Detailed Action
The following NON-FINAL office action is in response to application 18/223756 filed on 7/19/2023. This communication is the first action on the merits.
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Status of Claims
Claims 1-20 are currently pending and have been rejected as follows.
Information Disclosure Statement
The information disclosure statements (IDS) submitted on 9/1/23, 6/6/23, 4/25/25, 7/11/25, 11/20/25, 1/16/26, and 2/19/26 comply with the provisions of 37 CFR 1.97 and are being considered.
Specification
The disclosure is objected to because of the following informality:
Paragraph [0121] refers to “graph 420 of Fig. 3D,” where applicant likely means to reference Fig. 4D.
Appropriate correction is required.
Claim Rejections - 35 USC § 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or
composition of matter, or any new and useful improvement thereof, may obtain a patent therefor,
subject to the conditions and requirements of this title.
Claims 1-20 are rejected under 35 U.S.C. 101 because the claimed invention is directed to a judicial exception without significantly more. A subject matter eligibility analysis is set forth below. See MPEP 2106.
Specifically, representative Claim 1 recites:
A system, comprising:
a metering system, comprising one or more processors and memory, located on a utility grid downstream from a substation to:
detect, via a sensor of the metering system, current associated with electricity consumed in the utility grid during a first time interval;
determine, based on the current detected by the sensor, a current harmonic metric and a power metric associated with the electricity delivered over the utility grid in the first time interval;
input the current harmonic metric and the power metric into a model trained with machine learning and deployed on the metering system to determine a likelihood that at least a portion of the electricity delivered over the utility grid in the first time interval is used to charge an electric vehicle; and
execute an action associated with performance of the utility grid responsive to the likelihood satisfying a threshold.
Step 1:
Under Step 1 of the analysis, Claim 1 belongs to a statutory category, namely it is a system claim. Likewise, Claim 11 is a method claim and Claim 20 is a product claim.
Step 2A – Prong I:
Under Step 2A, prong 1: This part of the eligibility analysis evaluates whether the claim recites a judicial exception. As explained in MPEP 2106.04, subsection II, a claim “recites” a judicial exception when the judicial exception is “set forth” or “described” in the claim.
In the instant case, Claim 1 is found to recite at least one judicial exception (i.e. abstract idea), and is a mental concept and mathematical calculation. This can be seen in the following claim limitations: “determine, based on the current detected by the sensor, a current harmonic metric and a power metric” and “input the current harmonic metric and the power metric into a model trained with machine learning and deployed on the metering system to determine a likelihood that at least a portion”. Determining a harmonic metric can be accomplished by mentally reviewing sensor data and selecting a metric, while a power metric can be determined via Fourier analysis or other statistical techniques [See instant spec. [0062]- [0063], module 206]. Similarly, inputting the harmonic metric and power metric into a model trained with machine learning to determine a likelihood is a use of a generic machine learning model [See inst. Spec. [0082]-[0085]], which has been determined to be abstract [See Recentive Analytics, Inc. v. Fox Corp. (2025). From Page 18 – “Today, we hold only that patents that do no more than claim the application of generic machine learning to new data environments, without disclosing improvements to the machine learning models to be applied, are patent ineligible under § 101.]
Similar limitations comprise the abstract ideas of Claims 11 and 20.
Step 2A – Prong II:
Step 2A, prong 2 of the eligibility analysis evaluates whether the claim as a whole integrates the recited judicial exception(s) into a practical application of the exception. This evaluation is performed by (a) identifying whether there are any additional elements recited in the claim beyond the judicial exception, and (b) evaluating those additional elements individually and in combination to determine whether the claim as a whole integrates the exception into a practical application.
In addition to the abstract ideas recited in Claim 1, the claimed system recites additional elements including “a system, comprising: a metering system, comprising one or more processors and memory, located on a utility grid downstream from a substation,” “detect, via a sensor of the metering system, current associated with electricity consumed in the utility grid during a first time interval,” “the electricity delivered over the utility grid in the first time interval is used to charge an electric vehicle,” “execute an action associated with performance of the utility grid responsive to the likelihood satisfying a threshold.” The “system, comprising a metering system, comprising one or more processors and memory…” of Claim 1, the “metering system comprising one or more processors and memory” of Claim 11, and the “non-transitory computer-readable medium that stores processor-executable instructions that, when executed by one or more processors, cause the one or more processors” of Claim 20 merely recite generic computer components. See MPEP 2106.05(f). Detecting, “via a sensor…” is broadly and generically recited, and is merely routine data gathering necessary for the abstract determining step. “The electricity delivered over a grid…” is an attempt to generally link the abstract idea of determining and inputting the current harmonic and power metric to the technological environment of electricity delivered over a power grid. See MPEP 2106.05(h). Executing “an action associated with performance…” is broadly and generically recited and can be interpreted as merely providing an alert, which is insignificant post-solution activity. See MPEP 2106.05(g) “Insignificant Extra-Solution Activity.”
Thus, under Step 2A, prong 2 of the analysis, even when viewed in combination, these additional elements do not integrate the recited judicial exception into a practical application and the claims (1, 11, and 20) are directed to the judicial exception.
Step 2B:
Under Step 2B, the claims do not include additional elements that are sufficient to amount to significantly more than the judicial exception because the additional elements, as described above with respect to Step 2A Prong 2, merely amount to a general purpose computer system that attempts to apply the abstract idea in a technological environment, limiting the abstract idea to a particular field of use, and/or merely performs insignificant extra-solution activit(ies).
The generic data gathering, processing, and output steps, are recited at such a high level of generality (e.g. “processor”, “memory”) that they represent no more than mere instructions to apply the judicial exceptions on a computer. It can also be viewed as nothing more than an attempt to generally link the use of the judicial exceptions to the technological environment of a computer. Noting MPEP 2106.04(d)(I): “It is notable that mere physicality or tangibility of an additional element or elements is not a relevant consideration in Step 2A Prong Two. As the Supreme Court explained in Alice Corp., mere physical or tangible implementation of an exception does not guarantee eligibility. Alice Corp. Pty. Ltd. v. CLS Bank Int’l, 573 U.S. 208, 224, 110 USPQ2d 1976, 1983-84 (2014) ("The fact that a computer ‘necessarily exist[s] in the physical, rather than purely conceptual, realm,’ is beside the point")”.
Such insignificant extra-solution activity (“execute an action…”), amounts to no more than a data output step and, when re-evaluated under Step 2B is further found to be well-understood, routine, and conventional as evidenced by MPEP 2106.05(d)(II) (describing conventional activities that include transmitting and receiving data over a network, electronic recordkeeping, storing and retrieving information from memory, and electronically scanning or extracting data from a physical document).
Therefore, similarly the combination and arrangement of the above identified additional elements when analyzed under Step 2B also fails to necessitate a conclusion that Claim 1, as well as Claims 11 and 20, amount to significantly more than the abstract idea.
With regards to the dependent claims, Claims 2-10 and 12-19 merely further expand upon the algorithm/abstract idea and do not set forth further additional elements that integrate the recited abstract idea into a practical application or amount to significantly more. Therefore, these claims are found ineligible for the reasons described for parent claims 1 and 11.
Specifically, Claims 2 and 12 generically recite a routine data transfer and training of the machine learning model. Training the machine learning model is within the abstract idea, while receiving the data is merely routine data gathering.
Similarly, Claims 3-6 and 13-16 expand on the machine learning model training and merely provide specifics on data parameters, sampling rates, labeling, and the type of model. These are no more than data manipulations and gathering steps, and the generically recited convolution neural network long short-term memory network is generically recited and subject to no improvements by the claimed system.
Specifically, for Claims 5 and 15, note that the use of a Support Vector Machine as a classifier is well-understood, conventional, and routine (See Recentive Analytics, Inc. v. Fox Corp. (2025). From Page 5 – “The machine learning technology described in the patents is conventional, as the patents’ specifications demonstrate. See, e.g., ’367 patent, col. 6 ll. 1–5 (requiring “any suitable machine learning technology . . . such as, for example: a gradient boosted random forest, a regression, a neural network, a decision tree, a support vector machine, a Bayesian network, [or] other type of technique”); ’811 patent, col. 3 l. 23 (requiring the application of “any suitable machine learning technique.”).” A such, Claims 2-6 and 12-16 do not render the claimed invention into practical application, nor do they amount to significantly more than the judicial exception.
Claims 7 and 17, which are substantively similar to Claim 1, contains the same abstract ideas found in Claim 1 and its dependent claims. Claim 7 introduces the following additional element: “a data processing system comprising one or more processors coupled with memory,” which is broadly and generically recited and thus does not render the claimed system into practical application or amount to significantly more. Claim 8 provides a further numerical limitation to the sampling resolution, but the recited data gathering at a sample rate of 1 Hz is well-understood, routine, and conventional. See US 20210247424 A1, Paragraph [0076] – “Since the order of magnitude for the available data transmission speeds on these digital bus 2 is typically above megabits per second (Mbps), they are totally suitable for sampling rates of 1 kHz and even above.” and Paragraph [0078] – “According to the preferred embodiment illustrated on FIG. 1, reports E.sub.m on power/energy metered are periodically sent back at a lower frequency, typically ranging between 1 Hz and 10 Hz, in order not to create too much traffic load on the digital bus 2.” See also US 20210172920 A1, Paragraph [0084] – “Typically, the period of each cycle may range from 1 microsecond to 100 microseconds…The period of data collection is much longer and depends on the sampling frequency needed for testing. This sampling frequency may range from 10 Hz to 1 Hz but is typically chosen to match with the frequency of local alternating currents”.
Claims 9 and 18 recite updating the model via a network. The model update is no more than generically recited machine learning algorithm training, while the receipt via a network is merely a data transfer step necessary to perform the abstract idea. Claims 10 and 19 merely provide further limitations to the “execute an action” limitation of Claim 1, and also amount to insignificant post-solution activity that is also well understood, routine, and conventional in the art. The triggering of an alarm has been identified by courts as a well-understood, routine, and conventional practice and does not amount to significantly more than the abstract idea itself (See Parker v. Flook, 437 U.S. 584 (1978). From Page 595 – “Here it is absolutely clear that respondent's application contains no claim of patentable invention. The chemical processes involved in catalytic conversion of hydrocarbons are well known, as are the practice of monitoring the chemical process variables, the use of alarm limits to trigger alarms, the notion that alarm limit values must be recomputed and readjusted, and the use of computers for "automatic monitoring alarming.”). None of the above-listed claims are sufficient to integrate the abstract ideas of Claim 1 into practical application, nor do they amount to significantly more.
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claims 1, 10, 11, 19, and 20 are rejected under Decker et. al (US 20240149736 A1) in view of Woodman et. Al, Modeling Harmonic Impacts of Electric Vehicle Chargers on Distribution Networks, IEEE, 2018 [hereinafter “Woodman”].
Regarding Claim 1, Decker discloses a system, comprising: a metering system, comprising one or more processors and memory, located on a utility grid downstream from a substation [Paragraph [0018] – “FIG. 1 illustrates an exemplary physical topology of a power distribution network showing devices at various points, or nodes, on the network. FIG. 1 depicts a premises monitoring system 100 and a power distribution network 110. In an example, the premises monitoring system 100 receives information from endpoint meters in the power distribution network 110 and determines premises on the power distribution network 110 that are performing electric vehicle charging operations.” – Fig. [1] shows meters 130, 131… to be downstream from the substation; Paragraph [0019] – “The premises monitoring system 100 includes a premises monitoring application 101 and a headend system 102. The premises monitoring application 101 executes on a computing device as depicted in FIG. 7.” – see utility grid in Fig. [1] and contents of computing system (memory and processor) in Fig. [7]] to: detect, via a sensor of the metering system, power consumption values associated with electricity consumed in the utility grid during a first time interval [Paragraph [0019] – “The premises monitoring application 101 can receive metering data such as voltage, load, power consumption, etc., from meters that are installed at customers' premises.”].
Decker does not disclose that the power measurements are current measurements or determining, based on the current detected by the sensor, a current harmonic metric and a power metric associated with the electricity delivered over the utility grid in the first time interval.
However, Woodman discloses measuring current measurements as a power consumption parameter [Abstract - “Voltage and current data collected from in-service EV charging stations were used to create harmonic profiles of the EV charging units.”]
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to measure current, as disclosed by Woodman, as one of the power consumption metrics disclosed by Decker in order to determine the electric load contributions on the utility grid.
Woodman further discloses determining, based on the current detected by the sensor, a current harmonic metric and a power metric associated with the electricity delivered over the utility grid [Woodman 2018, pp. 3 – “Equation 1 shows the mathematical definition of THD. Note THD is a ratio of a load’s harmonic current content with respect to the load’s fundamental current, I1; THD is a metric of the harmonic content of a single load, and not a metric of the harmonic content within a feeder.” - current harmonics from a single meter; “Total Demand Distortion (TDD) considers distortion of all the loads on a feeder with respect to the size of the distribution feeder…EV charging draws high current, and the TDD changes throughout a charging event.” – total power requirements, signature current and current harmonics during EV charging] in a first time interval [Woodman 2018, pp. 4 – “Current magnitudes and phase angles were selected from representative ranges of time throughout the charging cycle and represent the changing levels of THD presented to the system by the charger.”].
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to use the current harmonic metric of Woodman along with the power metric disclosed by Decker with the metering system of Decker to better distinguish the electric load from an EV charger on the utility grid from other loads in a time interval.
Decker, as modified, would disclose inputting the current harmonic metric and the power metric into a model trained with machine learning and deployed on the metering system [Paragraph [0022] – “In an example, the premises monitoring application 101 may apply one or more machine-learning models to the data obtained from the meters 130 and 131, the power distribution network 110, other relevant data sources, or a combination thereof to generate predicted indications of premises conditions.”] to determine a likelihood that at least a portion of the electricity delivered over the utility grid in the first time interval is used to charge an electric vehicle [Paragraph [0022] – “For example, the premises monitoring application 101 may leverage data obtained from the meters 130 and 131, and other data sources in the power distribution network 110…In an example, the premises monitoring application 101 may determine a condition (e.g., no electric vehicle charging, Level 1 electric vehicle charging, Level 2 electric vehicle charging, etc.) of a particular target premises.” – determines charging or no charging condition; Paragraph [0032] – “…if the aggregate number of new premises performing electric vehicle charging operations is expected to result in the power distribution network 110 consuming a threshold percentage more power…” – expected power consumption data that indicates electric vehicle charging is a likelihood];
and executing an action associated with performance of the utility grid [Paragraph [0032] – “In such an example, if the aggregate number of new premises performing electric vehicle charging operations is expected to result in the power distribution network 110 consuming a threshold percentage more power (e.g., 1% more, 5% more, etc.) than after the previous adjustment, then the power supply may be adjusted.” – executing the action of adjusting the power supply after detecting that some threshold percentage more power is being used over the power distribution network] responsive to the likelihood satisfying a threshold [Paragraph [0032] – “…if the aggregate number of new premises performing electric vehicle charging operations is expected to result in the power distribution network 110 consuming a threshold percentage more power…” – expected power consumption is a likelihood].
Regarding Claim 11, Decker discloses a method, comprising: detecting, by a metering system comprising one or more processors and memory located on a utility grid downstream from a substation [Paragraph [0018] – “FIG. 1 illustrates an exemplary physical topology of a power distribution network showing devices at various points, or nodes, on the network. FIG. 1 depicts a premises monitoring system 100 and a power distribution network 110. In an example, the premises monitoring system 100 receives information from endpoint meters in the power distribution network 110 and determines premises on the power distribution network 110 that are performing electric vehicle charging operations.” – Fig. [1] shows meters 130, 131… to be downstream from the substation; Paragraph [0019] – “The premises monitoring system 100 includes a premises monitoring application 101 and a headend system 102. The premises monitoring application 101 executes on a computing device as depicted in FIG. 7.” – see utility grid in Fig. [1] and contents of computing system (memory and processor) in Fig. [7]], via a sensor of the metering system, power consumption values associated with electricity consumed in the utility grid during a first time interval [Paragraph [0019] – “The premises monitoring application 101 can receive metering data such as voltage, load, power consumption, etc., from meters that are installed at customers' premises.”].
Decker does not disclose that the power measurements are current measurements or determining, by the metering system, based on the current detected by the sensor, a current harmonic metric and a power metric associated with the electricity delivered over the utility grid in the first time interval.
However, Woodman discloses measuring current measurements as a power consumption parameter [Abstract - “Voltage and current data collected from in-service EV charging stations were used to create harmonic profiles of the EV charging units.”]
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to measure current, as disclosed by Woodman, as one of the power consumption metrics disclosed by Decker in order to determine the electric load contributions on the utility grid.
Woodman further discloses determining, by the metering system [Woodman 2018, pp. 4, para. [0002] – “A Tektronix PA4000 three-phase power analyzer was use to collect data.”], based on the current detected by the sensor, a current harmonic metric and a power metric associated with the electricity delivered over the utility grid [Woodman pp. 3, para. [0001] – “Equation 1 shows the mathematical definition of THD. Note THD is a ratio of a load’s harmonic current content with respect to the load’s fundamental current, I1”; pp. 3, para. [0001] – “THD is a metric of the harmonic content of a single load, and not a metric of the harmonic content within a feeder.” - current harmonics from a single meter; pp. 3, para. [0002] - Total Demand Distortion (TDD) considers distortion of all the loads on a feeder with respect to the size of the distribution feeder…EV charging draws high current, and the TDD changes throughout a charging event.” – total power requirements, signature current and current harmonics during EV charging] in a first time interval [Woodman pp. 4, para. [0007] – “Current magnitudes and phase angles were selected from representative ranges of time throughout the charging cycle and represent the changing levels of THD presented to the system by the charger.”].
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to use the current harmonic metric of Woodman along with the power metric disclosed by Decker with the metering system of Decker to better distinguish the electric load from an EV charger on the utility grid from other loads in a time interval.
The combination of Decker and Woodman discloses inputting, by the metering system, the current harmonic metric and the power metric into a model trained with machine learning and deployed on the metering system [Paragraph [0022] – “For example, the meters 130 and 131 may transmit data relating to the power consumption at the premises 124 and 125 to the headend system 102. In an example, the premises monitoring application 101 may determine a condition (e.g., no electric vehicle charging, Level 1 electric vehicle charging, Level 2 electric vehicle charging, etc.) of a particular target premises.”] to determine a likelihood that at least a portion of the electricity delivered over the utility grid in the first time interval is used to charge an electric vehicle [Paragraph [0022] – “For example, the premises monitoring application 101 may leverage data obtained from the meters 130 and 131, and other data sources in the power distribution network 110…In an example, the premises monitoring application 101 may determine a condition (e.g., no electric vehicle charging, Level 1 electric vehicle charging, Level 2 electric vehicle charging, etc.) of a particular target premises.” – determines charging or no charging condition; Paragraph [0032] – “…if the aggregate number of new premises performing electric vehicle charging operations is expected to result in the power distribution network 110 consuming a threshold percentage more power…” – expected power consumption data that indicates electric vehicle charging is a likelihood];
and executing, by the metering system [Paragraph [0032] – “In an example, the power supply of the power distribution network 110 may be adjusted contemporaneously with detecting a new premises performing electric vehicle charging operations.”], an action associated with performance of the utility grid [Paragraph [0032] – “In such an example, if the aggregate number of new premises performing electric vehicle charging operations is expected to result in the power distribution network 110 consuming a threshold percentage more power (e.g., 1% more, 5% more, etc.) than after the previous adjustment, then the power supply may be adjusted.” – executing the action of adjusting the power supply after detecting that some threshold percentage more power is being used over the power distribution network] responsive to the likelihood satisfying a threshold [Paragraph [0032] – “…if the aggregate number of new premises performing electric vehicle charging operations is expected to result in the power distribution network 110 consuming a threshold percentage more power…” – expected power consumption is a likelihood].
Regarding Claim 10, Decker, as modified, would disclose the system of claim 1, comprising: the metering system, to execute the action, provides an alert to a data processing system indicating electric vehicle charging in the first time interval [Paragraph [0028] – “FIG. 2 is a flowchart of a process 200 for implementing the premises monitoring application 101 to control power supplies of the power distribution network 110.”; Paragraph [0032] – “In such an example, if the aggregate number of new premises performing electric vehicle charging operations is expected to result in the power distribution network 110 consuming a threshold percentage more power (e.g., 1% more, 5% more, etc.) than after the previous adjustment, then the power supply may be adjusted. Other threshold percentages may also be used, and other triggering events associated with the addition of premises performing electric vehicle charging operations may also trigger adjustment of the power supply at block 208.” – See Fig. [2]; premises monitoring application is alerted to adjust power supply].
Regarding Claim 19, Decker, as modified, would disclose the method of claim 11, comprising: executing, by the metering system, the action to provide an alert to a data processing system indicating electric vehicle charging in the first time interval. [Paragraph [0028] – “FIG. 2 is a flowchart of a process 200 for implementing the premises monitoring application 101 to control power supplies of the power distribution network 110.”; Paragraph [0032] – “In such an example, if the aggregate number of new premises performing electric vehicle charging operations is expected to result in the power distribution network 110 consuming a threshold percentage more power (e.g., 1% more, 5% more, etc.) than after the previous adjustment, then the power supply may be adjusted. Other threshold percentages may also be used, and other triggering events associated with the addition of premises performing electric vehicle charging operations may also trigger adjustment of the power supply at block 208.” – See Fig. [2]; premises monitoring application is alerted to adjust power supply].
Regarding Claim 20, Decker discloses a non-transitory computer-readable medium that stores processor-executable instructions [Paragraph [0053] – “The memory device 704 includes any suitable non-transitory computer-readable medium for storing data, program code, or both.”] that, when executed by one or more processors [Paragraph [0052] – “The processor 702 executes computer-executable program code 730 stored in a memory device 704, accesses data 720 stored in the memory device 704, or both.”], cause the one or more processors to [Paragraph [0055] – “The computing device 700 executes program code 730 that configures the processor 702 to perform one or more of the operations described herein.”]: detect, via a sensor of a metering system, power consumption values associated with electricity consumed in a utility grid during a first time interval [Paragraph [0019] – “The premises monitoring application 101 can receive metering data such as voltage, load, power consumption, etc., from meters that are installed at customers' premises.”].
Decker does not disclose that the power measurements are current measurements or determining, based on the current detected by the sensor, a current harmonic metric and a power metric associated with the electricity delivered over the utility grid in the first time interval.
However, Woodman discloses that the power measurements are current measurements [Abstract - “Voltage and current data collected from in-service EV charging stations were used to create harmonic profiles of the EV charging units.”]
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to measure current, as disclosed by Woodman, as one of the power consumption metrics disclosed by Decker in order to determine the electric load contributions on the utility grid.
Woodman further discloses determining, based on the current detected by the sensor, a current harmonic metric and a power metric associated with the electricity delivered over the utility grid [Woodman 2018, pp. 3 – “Equation 1 shows the mathematical definition of THD. Note THD is a ratio of a load’s harmonic current content with respect to the load’s fundamental current, I1; THD is a metric of the harmonic content of a single load, and not a metric of the harmonic content within a feeder.” - current harmonics from a single meter; “Total Demand Distortion (TDD) considers distortion of all the loads on a feeder with respect to the size of the distribution feeder…EV charging draws high current, and the TDD changes throughout a charging event.” – total power requirements, signature current and current harmonics during EV charging] in a first time interval [Woodman 2018, pp. 4 – “Current magnitudes and phase angles were selected from representative ranges of time throughout the charging cycle and represent the changing levels of THD presented to the system by the charger.”].
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to use the current harmonic metric of Woodman along with the power metric disclosed by Decker with the metering system of Decker to better distinguish the electric load from an EV charger on the utility grid from other loads in a time interval.
Decker, as modified would disclose inputting the current harmonic metric and the power metric into a model trained with machine learning and deployed on the metering system [Paragraph [0022] – “In an example, the premises monitoring application 101 may apply one or more machine-learning models to the data obtained from the meters 130 and 131, the power distribution network 110, other relevant data sources, or a combination thereof to generate predicted indications of premises conditions.”] to determine a likelihood that at least a portion of the electricity delivered over the utility grid in the first time interval is used to charge an electric vehicle [Paragraph [0022] – “For example, the premises monitoring application 101 may leverage data obtained from the meters 130 and 131, and other data sources in the power distribution network 110…In an example, the premises monitoring application 101 may determine a condition (e.g., no electric vehicle charging, Level 1 electric vehicle charging, Level 2 electric vehicle charging, etc.) of a particular target premises.” – determines charging or no charging condition; Paragraph [0032] – “…if the aggregate number of new premises performing electric vehicle charging operations is expected to result in the power distribution network 110 consuming a threshold percentage more power…” – expected power consumption data that indicates electric vehicle charging is a likelihood];
and executing an action associated with performance of the utility grid [Paragraph [0032] – “In such an example, if the aggregate number of new premises performing electric vehicle charging operations is expected to result in the power distribution network 110 consuming a threshold percentage more power (e.g., 1% more, 5% more, etc.) than after the previous adjustment, then the power supply may be adjusted.” – executing the action of adjusting the power supply after detecting that some threshold percentage more power is being used over the power distribution network] responsive to the likelihood satisfying a threshold [Paragraph [0032] – “…if the aggregate number of new premises performing electric vehicle charging operations is expected to result in the power distribution network 110 consuming a threshold percentage more power…” – expected power consumption is a likelihood].
Claims 2 and 12, 9 and 18 are rejected under Decker et. al., in view of Woodman et. al., in further view of Abaas et. al. (US 20220416541 A1)
Regarding Claim 2, Decker, as modified, would disclose the system of claim 1.
Decker, as modified, does not disclose that the system comprises the metering system to receive, via a network, the model from a data processing system remote from the metering system, wherein the data processing system trains the model via machine learning.
Abaas, however discloses the metering system comprising: the metering system to receive, via a network, the model [Paragraph [0051] – “In one or more cases, each smart meter 104a-104c is configured to implement a trained machine learning model 110a-110c to predict power consumption, outages, and/or other operational parameters of the associated microgrid 108a-108c.” – machine learning model is on meters 108a-108c, which comprise he metering system].
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, for the smart meters of Abaas to implement the trained machine learning model received via a network from the metering system of Decker and Woodman in order to implement the machine learning model locally, allowing individual smart meters of the subsystem to better identify periods of EV charging.
The combination of Decker, Woodman, and Abaas would disclose that the model comes from a data processing system remote from the metering system [Decker, Paragraph [0022] – “In an example, the premises monitoring application 101 may apply one or more machine-learning models to the data obtained from the meters 130 and 131, the power distribution network 110, other relevant data sources, or a combination thereof to generate predicted indications of premises conditions” – the premises monitoring system 100 is remote from the metering system 110 and the premises monitoring application 101 is the data processing system],
and wherein the data processing system trains the model via machine learning [Decker, Paragraph [0019] – “The premises monitoring application 101 executes on a computing device as depicted in FIG. 7.” – training the model occurs on the computing device, see Fig. [7]; Paragraph [0025] – “In an example, the machine-learning models of the premises monitoring application 101 may be trained…”].
Regarding Claim 12, Decker, as modified, would disclose the method of claim 11.
Decker, as modified, does not disclose that the method comprises: receiving, by the metering system via a network, the model from a data processing system remote from the metering system, wherein the data processing system trains the model via machine learning.
However, Abaas discloses receiving, by the metering system via a network, the model [Paragraph [0051] – “In one or more cases, each smart meter 104a-104c is configured to implement a trained machine learning model 110a-110c to predict power consumption, outages, and/or other operational parameters of the associated microgrid 108a-108c.” – machine learning model is on meters 108a-108c, which comprise he metering system].
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, for the smart meters of Abaas to implement the trained machine learning model received via a network from the metering system of Decker and Woodman in order to implement the machine learning model locally, allowing individual smart meters of the subsystem to better identify periods of EV charging.
The combination of Decker, Woodman, and Abaas would disclose that the model comes from a data processing system remote from the metering system [Decker, Paragraph [0022] – “In an example, the premises monitoring application 101 may apply one or more machine-learning models to the data obtained from the meters 130 and 131, the power distribution network 110, other relevant data sources, or a combination thereof to generate predicted indications of premises conditions” – the premises monitoring system 100 is remote from the metering system 110 and the premises monitoring application 101 is the data processing system],
and wherein the data processing system trains the model via machine learning [Decker, Paragraph [0019] – “The premises monitoring application 101 executes on a computing device as depicted in FIG. 7.” – training the model occurs on the computing device, see Fig. [7]; Paragraph [0025] – “In an example, the machine-learning models of the premises monitoring application 101 may be trained…”].
Regarding Claim 9, Decker, as modified, would disclose the system of claim 1.
Decker, as modified, does not disclose the metering system to receive, via a network, an update to the model from a data processing system that re-trains the model based on additional data to generate the update to the model.
However, Abaas discloses the metering system to receive, via a network, an update to the model [Paragraph [0051] – “In one or more cases, each smart meter 104a-104c is configured to implement a trained machine learning model 110a-110c to predict power consumption, outages, and/or other operational parameters of the associated microgrid 108a-108c.” – machine learning model is on meters 108a-108c, which comprise the metering system] .
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, for the smart meters of Abaas to receive updates to the machine learning model received via a network from the metering system of Decker and Woodman in order to implement the machine learning model locally, allowing individual smart meters of the subsystem to better identify periods of EV charging.
The combination of Decker, Woodman, and Abaas discloses that the update to the model is received from a data processing system [Decker, Paragraph [0022] – “In an example, the premises monitoring application 101 may apply one or more machine-learning models to the data obtained from the meters 130 and 131, the power distribution network 110, other relevant data sources, or a combination thereof to generate predicted indications of premises conditions.” – refer to Fig. [1]] that re-trains the model based on additional data to generate the update to the model [Paragraph [0043] – “At block 410, the process 400 involves updating the trained machine-learning model using additional real-world consumption data.” – see also Fig. [4] for overview of training of machine learning model].
Regarding Claim 18, Decker, as modified, would disclose the method of claim 11.
Decker, as modified, does not disclose receiving, by the metering system the metering system via a network, an update to the model from a data processing system that re-trains the model based on additional data to generate the update to the model.
However, Abaas discloses receiving, by the metering system via a network, an update to the model [Paragraph [0051] – “In one or more cases, each smart meter 104a-104c is configured to implement a trained machine learning model 110a-110c to predict power consumption, outages, and/or other operational parameters of the associated microgrid 108a-108c.” – machine learning model is on meters 108a-108c, which comprise the metering system] .
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, for the smart meters of Abaas to receive updates to the machine learning model received via a network from the metering system of Decker and Woodman in order to implement the machine learning model locally, allowing individual smart meters of the subsystem to better identify periods of EV charging.
The combination of Decker, Woodman, and Abaas discloses receiving the update to the model from a data processing system [Decker, Paragraph [0022] – “In an example, the premises monitoring application 101 may apply one or more machine-learning models to the data obtained from the meters 130 and 131, the power distribution network 110, other relevant data sources, or a combination thereof to generate predicted indications of premises conditions.” – refer to Fig. [1]] that re-trains the model based on additional data to generate the update to the model [Paragraph [0043] – “At block 410, the process 400 involves updating the trained machine-learning model using additional real-world consumption data.” – see also Fig. [4] for overview of training of machine learning model].
Claims 3 and 13 are rejected under Decker et. al in view of Woodman et. al, in further view of Naido et. al. (US 20210165032 A1).
Regarding Claim 3, Decker, as modified, would disclose the system of claim 1 wherein the model is trained with first training data [Decker, Paragraph [0025] – “…the machine-learning models of the premises monitoring application 101 may be trained based on a historical corpus of data obtained from one or more power distribution networks.” – historical corpus is first training data].
Decker, as modified, does not disclose that the first training data is sampled at a first resolution and second training data sampled at a second resolution that is greater than the first resolution.
However, Naido discloses that the first training data is sampled at a first resolution [Paragraph [0023]-[0024] – “In an embodiment, a first power system device and second power system device measure data in response to a condition in a corresponding segment of the power transmission network. Here, the first device may be at a first end and sample data at a first sampling rate…the first power system device processes the measured data. Here, the first power system device may sample data at a low sampling rate (e.g., <4 KHz).”].
and a second training data set sampled at a second resolution that is greater than the first resolution [Paragraph [0023]-[0024] –“…while the second device may be at a second end and sample data at a second sampling rate…The fault zone information can be used for locating the fault with the disturbance data received from the second power system device (which may only be sampling data (i.e., not doing any processing) at a high sampling rate (e.g., >1 MHz)”].
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to sample the second training set at a greater resolution than the first training set, as disclosed by Naido, to sample the high-resolution data received locally by the metering system disclosed by Decker and Woodman to capture the current harmonics and power waveform.
Regarding Claim 13, Decker, as modified, discloses the method of claim 1 wherein the model is trained with first training data [Decker, Paragraph [0025] – “…the machine-learning models of the premises monitoring application 101 may be trained based on a historical corpus of data obtained from one or more power distribution networks.” – historical corpus is first training data].
Decker, as modified, does not disclose that the first training data is sampled at a first resolution and second training data sampled at a second resolution that is greater than the first resolution.
However, Naido discloses that the first training data is sampled at a first resolution [Paragraph [0023]-[0024] – “In an embodiment, a first power system device and second power system device measure data in response to a condition in a corresponding segment of the power transmission network. Here, the first device may be at a first end and sample data at a first sampling rate…the first power system device processes the measured data. Here, the first power system device may sample data at a low sampling rate (e.g., <4 KHz).”].
and a second training data set sampled at a second resolution that is greater than the first resolution [Paragraph [0023]-[0024] –“…while the second device may be at a second end and sample data at a second sampling rate…The fault zone information can be used for locating the fault with the disturbance data received from the second power system device (which may only be sampling data (i.e., not doing any processing) at a high sampling rate (e.g., >1 MHz)”].
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to sample the second training set at a greater resolution than the first training set, as disclosed by Naido, to sample the high-resolution data received locally by the metering system disclosed by Decker and Woodman to capture the current harmonics and power waveform.
Claims 4 and 14 are rejected under Decker et. al in view of Woodman et. al, in further view of Lin et. al. (US 20240054322 A1).
Regarding Claim 4, Decker as modified would disclose the system of claim 1, wherein the model is trained with first training data comprising historical load metrics [Paragraph [0039] – “At block 402, the process 400 involves accessing a corpus of training and validation power consumption data...The corpus of training data may be a National Renewable Energy Laboratory (NREL) dataset labeled with electric vehicle charging operation information for 600 premises taken over 52 weeks with readings taken at 10-minute intervals.”].
Decker, as modified, does not disclose that the training is done for a plurality of time stamps, wherein each of the plurality of time stamps is assigned one of a first label that indicates electric vehicle charging or a second label that indicates no electric vehicle charging during the corresponding plurality of time stamps.
Lin, however discloses that the training is done for a plurality of time stamps [Paragraph [0064]-[0065] – “Accordingly, surveyed energy usage data can be combined with labeled energy usage data to bolster the performance of the training machine learning model...In some embodiments, input 302 and/or training data 308 can include information other than energy usage information. For example, weather information relative to the energy usage data…a time stamp relative to the energy usage data, and other relevant information can be included in input 302 and/or training data 308.”].
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, assign the timestamps and energy usage labels of Lin to the training data disclosed by the combination of Decker and Woodman in order to clearly define time intervals in the training data.
The combination of Decker, Woodman, and Lin would disclose wherein each of the plurality of time stamps is assigned one of a first label that indicates electric vehicle charging or a second label that indicates no electric vehicle charging during the corresponding plurality of time stamps [Decker, Paragraph [0037]-[0038] – “Temporal sequencing 308 of the power consumption data 302 may also be used by the machine-learning model. As the power consumption data 302 is provided in the time domain, various features of the power consumption data 302 may be relevant to the temporal sequence of the power consumption data 302. For example, the machine-learning model may use the temporal sequencing 308 to identify features of the power consumption data 302 that are likely to occur in series with particular events or at certain times during a particular day. Electric vehicle charging occurring at particular times (e.g., overnight) may be a significant indicator used by the machine-learning model to identify electric vehicle charging operations at a premises… Classifications 310 may be generated for each of the premises represented in the power consumption data 302 based on the low-level feature detection 304, the high-level feature detection 306, and the temporal sequencing 308. In an example, the classifications can include an indication that the premises does not perform electric vehicle charging operations, an indication that the premises performs Level 1 electric vehicle charging operations, or an indication that the premises performs Level 2 electric vehicle charging operations.” – temporal sequencing classifies no charging and charging intervals (level 1 and level 2), which can receive their corresponding labels according to the aforementioned methods].
Regarding Claim 14, Decker as modified would disclose the system of claim 1, wherein the model is trained with first training data comprising historical load metrics [Paragraph [0039] – “At block 402, the process 400 involves accessing a corpus of training and validation power consumption data...The corpus of training data may be a National Renewable Energy Laboratory (NREL) dataset labeled with electric vehicle charging operation information for 600 premises taken over 52 weeks with readings taken at 10-minute intervals.”].
Decker, as modified, does not disclose that the training is done for a plurality of time stamps, wherein each of the plurality of time stamps is assigned one of a first label that indicates electric vehicle charging or a second label that indicates no electric vehicle charging during the corresponding plurality of time stamps.
Lin, however discloses that the training is done for a plurality of time stamps [Paragraph [0064]-[0065] – “Accordingly, surveyed energy usage data can be combined with labeled energy usage data to bolster the performance of the training machine learning model...In some embodiments, input 302 and/or training data 308 can include information other than energy usage information. For example, weather information relative to the energy usage data…a time stamp relative to the energy usage data, and other relevant information can be included in input 302 and/or training data 308.”].
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, assign the timestamps and energy usage labels of Lin to the training data disclosed by the combination of Decker and Woodman in order to clearly define time intervals in the training data.
The combination of Decker, Woodman, and Lin would disclose wherein each of the plurality of time stamps is assigned one of a first label that indicates electric vehicle charging or a second label that indicates no electric vehicle charging during the corresponding plurality of time stamps [Decker, Paragraph [0037]-[0038] – “Temporal sequencing 308 of the power consumption data 302 may also be used by the machine-learning model. As the power consumption data 302 is provided in the time domain, various features of the power consumption data 302 may be relevant to the temporal sequence of the power consumption data 302. For example, the machine-learning model may use the temporal sequencing 308 to identify features of the power consumption data 302 that are likely to occur in series with particular events or at certain times during a particular day. Electric vehicle charging occurring at particular times (e.g., overnight) may be a significant indicator used by the machine-learning model to identify electric vehicle charging operations at a premises… Classifications 310 may be generated for each of the premises represented in the power consumption data 302 based on the low-level feature detection 304, the high-level feature detection 306, and the temporal sequencing 308. In an example, the classifications can include an indication that the premises does not perform electric vehicle charging operations, an indication that the premises performs Level 1 electric vehicle charging operations, or an indication that the premises performs Level 2 electric vehicle charging operations.” – temporal sequencing classifies no charging and charging intervals (level 1 and level 2), which can receive their corresponding labels according to the aforementioned methods].
Claims 5 and 15 are rejected under Decker et. al., in view of Woodman et. al., in further view of Lu et. al. (US 20200122598 A1).
Regarding Claim 5, Decker, as modified discloses the system of claim 1.
Decker, as modified, does not disclose wherein the training data used to train the model is filtered based on a comparison of an output of a support vector machine trained to classify a harmonics and power data set as corresponding to electric vehicle charging or no electric vehicle charging with a second threshold.
Lu, however, discloses wherein data is filtered based on a comparison of an output of a support vector machine trained to classify a harmonics and power data set [Paragraph [0026] – “Account time series data: Data that represents the electric consumption in kWh units measured and collected from a house meter for an account under consideration as a whole over a period of time. The measured data is typically collected in time intervals.” – See Fig. [2], 210 and 210a] as corresponding to electric vehicle charging or no electric vehicle charging with a second threshold [Paragraph [0084] – “In one embodiment, the system implements and uses a logistic classifier, and a decision-tree based algorithm (e.g., Support Vector Machine SVM) to identify EV charges. In the case of the logistic classifier, suspected EV events with an assigned probability of greater than 0.5 are classified as EV charges. Of course, other values may be used.” – classifying EV charges based on an assigned probability is filtering; probability of 0.5 is the second threshold].
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to use the SVM and second threshold of Lu to perform an initial classification (EV charging or no EV charging) of the training data disclosed by Decker and Woodman in order to improve the machine learning training process.
The combination of Decker, Woodman, and Lu discloses that the data is training data used to train the model [Decker, Paragraph [0039] – “In an example, a corpus of data may be divided into a set of training data and a set of validation data that is used to validate the machine-learning model that is trained using the set of training data.”].
Regarding Claim 15, Decker, as modified, discloses the method of claim 11.
Decker, as modified, does not disclose wherein training data used to train the model is filtered based on a comparison of an output of a support vector machine trained to classify a harmonics and power data set as corresponding to electric vehicle charging or no electric vehicle charging with a second threshold.
Lu, however, discloses wherein data is filtered based on a comparison of an output of a support vector machine trained to classify a harmonics and power data set as corresponding to electric vehicle charging or no electric vehicle charging with a second threshold [Paragraph [0084] – “In one embodiment, the system implements and uses a logistic classifier, and a decision-tree based algorithm (e.g., Support Vector Machine SVM) to identify EV charges. In the case of the logistic classifier, suspected EV events with an assigned probability of greater than 0.5 are classified as EV charges. Of course, other values may be used.” – classifying EV charges based on an assigned probability is filtering; probability of 0.5 is the second threshold].
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to use the SVM and second threshold of Lu to perform an initial classification (EV charging or no EV charging) of the training data disclosed by Decker and Woodman in order to improve the machine learning training process.
The combination of Decker, Woodman, and Lu discloses that the data is training data used to train the model [Decker, Paragraph [0039] – “In an example, a corpus of data may be divided into a set of training data and a set of validation data that is used to validate the machine-learning model that is trained using the set of training data.”].
Claims 6 and 16 are rejected under Decker et. al, in view of Woodman et. al, in further view of Anderson et. al. (US 11322976 B1).
Regarding Claim 6, Decker, as modified, discloses the system of claim 1.
The combination does not disclose wherein the model comprises a convolution neural network long short- term memory network.
However, Anderson discloses wherein the model comprises a convolution neural network long short-term memory network [Col 28, Ln. 29-60 – “FIG. 11 is a flow chart of an example of a process for generating and using a machine-learning model according to some aspects…can classify input data among two or more classes; cluster input data among two or more groups; predict a result based on input data; identify patterns or trends in input data; identify a distribution of input data in a space; or any combination of these. Examples of machine-learning models can include…ensembles or other combinations of machine-learning models. In some examples, neural networks can include deep neural networks, feed-forward neural networks, recurrent neural networks, convolutional neural networks, radial basis function (RBF) neural networks, echo state neural networks, long short-term memory neural networks…or any combination of these.”; Col.29, Ln. 42-43 – “In block 1106, a machine-learning model is trained using the training data. The machine-learning model can be trained in a supervised, unsupervised, or semi-supervised manner.” – Fig. [11]].
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to use the combination of convolutional neural networks and long short-term memory neural networks of Anderson as the machine learning model disclosed in the combination of Decker and Woodman to detect periods of EV charging and no charging because these are known, effective machine-learning techniques.
Regarding Claim 16, Decker, as modified discloses the method of claim 11.
The combination does not disclose wherein the model comprises a convolution neural network long short- term memory network.
However, Anderson discloses wherein the model comprises a convolution neural network long short-term memory network [Col 28, Ln. 29-60 – “FIG. 11 is a flow chart of an example of a process for generating and using a machine-learning model according to some aspects…can classify input data among two or more classes; cluster input data among two or more groups; predict a result based on input data; identify patterns or trends in input data; identify a distribution of input data in a space; or any combination of these. Examples of machine-learning models can include…ensembles or other combinations of machine-learning models. In some examples, neural networks can include deep neural networks, feed-forward neural networks, recurrent neural networks, convolutional neural networks, radial basis function (RBF) neural networks, echo state neural networks, long short-term memory neural networks…or any combination of these.”; Col.29, Ln. 42-43 – “In block 1106, a machine-learning model is trained using the training data. The machine-learning model can be trained in a supervised, unsupervised, or semi-supervised manner.” – Fig. [11]].
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to use the combination of convolutional neural networks and long short-term memory neural networks of Anderson as the machine learning model disclosed in the combination of Decker and Woodman to detect periods of EV charging and no charging.
Claims 7 and 17 are rejected in view of Decker et. al., in view of Woodman et. al., in view of Lu et. al., in view of Anderson et. al., in further view of Naido et. al. (US 20210165032 A1).
Regarding Claim 7, Decker, as modified, discloses the system of claim 1, comprising: a data processing system comprising one or more processors coupled with memory [Paragraph [0052] – “FIG. 7 illustrates an exemplary computing device used for detecting electric vehicle charging operations…The depicted example of a computing device 700 includes a processor 702 communicatively coupled to one or more memory devices 704. The processor 702 executes computer-executable program code 730 stored in a memory device 704, accesses data 720 stored in the memory device 704, or both.”], the data processing system to: receive time-series data sampled at a first resolution corresponding to residential electric load [Paragraph [0055] – “The computing device 700 executes program code 730 that configures the processor 702 to perform one or more of the operations described herein. For example, the program code 730 causes the processor to perform the operations described in FIGS. 1-6.”; Paragraph [0019] – “The premises monitoring application 101 can receive metering data such as voltage, load, power consumption, etc., from meters that are installed at customers' premises.”];
Decker, as modified does not disclose detecting, based on a second threshold, one or more time intervals of electric vehicle charging in the time-series data.
Lu, however, discloses detecting, based on a second threshold, one or more time intervals of electric vehicle charging in the time-series data [Paragraph [0073] – “The classifier can be trained and executed a number of times on account time series data. The classifier is configured to generate a list of detected EV events that the classifier identified…” – SVM is the classifier; Paragraph [0084] – “…In one embodiment, the system implements and uses a logistic classifier, and a decision-tree based algorithm (e.g., Support Vector Machine SVM) to identify EV charges. In the case of the logistic classifier, suspected EV events with an assigned probability of greater than 0.5 are classified as EV charges. Of course, other values may be used.” – a time series consists of multiple time intervals; probability of 0.5 is the second threshold; refer to Fig. [2]];
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to identify periods of EV charging using the probability threshold disclosed by Lu using the metering system disclosed by Decker and Woodman to make an initial classification of the time series data.
The combination of Decker, Woodman, and Lu would disclose labeling, based on the detection, the time-series data to indicate the one or more time intervals of electric vehicle charging to generate labeled time-series data [Decker, Paragraph [0030] – “At block 206, the process 200 involves generating classification information for the set of premises based on an output of the trained machine-learning model. In an example, the trained machine-learning model may be trained to output an indication of the premises performing electric vehicle charging operations or not performing electric vehicle charging operations.” – see labels in Fig. [3], 312];
The combination does not disclose receiving a harmonics and power data set sampled at a second resolution greater than the first resolution.
However, Naido discloses receiving a harmonics and power data set sampled at a second resolution greater than the first resolution [Paragraph [0023]-[0024] – In an embodiment, a first power system device and second power system device measure data in response to a condition in a corresponding segment of the power transmission network. Here, the first device may be at a first end and sample data at a first sampling rate, while the second device may be at a second end and sample data at a second sampling rate, and the devices can measure the data at the corresponding ends…Here, the first power system device may sample data at a low sampling rate (e.g., <4 KHz)…The fault zone information can be used for locating the fault with the disturbance data received from the second power system device (which may only be sampling data (i.e., not doing any processing) at a high sampling rate (e.g., >1 MHz).”].
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to sample the second training set at a greater resolution than the first training set, as disclosed by Naido, to sample the high-resolution data received locally by the metering system disclosed by Decker, Woodman, and Lu to capture the current harmonics and power waveform.
The combination of Decker, Woodman, Lu, and Naido does not disclose inputting the labeled time-series data and the harmonics and power data set into a support vector machine to train the support vector machine to classify the harmonics and power data set to indicate electric vehicle charging or no electric vehicle charging or using the trained support vector machine to filter a second harmonics and power data set to generate a filtered second harmonics and power data set.
However, Lu discloses inputting the labeled time-series data [See Fig. 2 blocks 210, “Account time Series Data (From Known EVs)” and 210a “Account Time Series Data (House with Known Non-EV)”, both of which train classifier at block 270] and the harmonics and power data set into a support vector machine to train the support vector machine to classify the harmonics and power data set to indicate electric vehicle charging or no electric vehicle charging [Paragraph [0072] – “At block 270, the EV charge patterns and their associated charge features and input to one or more machine learning classifiers (the classifier). This data trains the classifier to learn from the charge features in order to identify and detect an EV charge event from an unknown data set. Likewise, the non-EV charge patterns and their associated charge features are input to train the classifier to identify non-EV charge patterns and distinguish them from EV charge events. – the classifier is the SVM, the unknown data set is the harmonics and power data set (as opposed to the labeled time-series data set); Paragraph [0083]-[0084] – “(3) Given a time series for an account, is an EV being charged? In one embodiment, the system implements and uses a logistic classifier, and a decision-tree based algorithm (e.g., Support Vector Machine SVM) to identify EV charges. In the case of the logistic classifier, suspected EV events with an assigned probability of greater than 0.5 are classified as EV charges. Of course, other values may be used.” - probability of 0.5 is the threshold; see also Fig. [2], 210 and 210a, which shows data from known EV and known non EV sources training the SVM at block 270]
and using the trained support vector machine to filter a second harmonics and power data set to generate a filtered second harmonics and power data set [Paragraph [0072]-[0073] – “At block 270, the EV charge patterns and their associated charge features and input to one or more machine learning classifiers (the classifier)…The classifier can be trained and executed a number of times on account time series data.” – the classifier is the SVM, SVM is trained after first training and execution, second dataset is generated with successive training and execution; Paragraph [0084] – “In one embodiment, the system implements and uses a logistic classifier, and a decision-tree based algorithm (e.g., Support Vector Machine SVM) to identify EV charges. In the case of the logistic classifier, suspected EV events with an assigned probability of greater than 0.5 are classified as EV charges. Of course, other values may be used.” – classifying EV charges based on an assigned probability is filtering].
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to use the SVM and second threshold of Lu to generate a data set filtered according to an initial classification (EV charging or no EV charging) of the training data disclosed by Decker, Woodman, Lu, and Naido in order to improve the machine learning training process.
The combination of Decker, Woodman, Lu, and Naido does not disclose training, via a convolution neural network long short-term memory network, the model with the filtered second harmonics and power data set.
However, Anderson discloses training, via a convolution neural network long short-term memory network [Anderson, Col 28, Ln. 29-60 – “FIG. 11 is a flow chart of an example of a process for generating and using a machine-learning model according to some aspects…can classify input data among two or more classes; cluster input data among two or more groups; predict a result based on input data; identify patterns or trends in input data; identify a distribution of input data in a space; or any combination of these. Examples of machine-learning models can include…ensembles or other combinations of machine-learning models. In some examples, neural networks can include deep neural networks, feed-forward neural networks, recurrent neural networks, convolutional neural networks, radial basis function (RBF) neural networks, echo state neural networks, long short-term memory neural networks…or any combination of these.”; Col.29, Ln. 42-43 – “In block 1106, a machine-learning model is trained using the training data. The machine-learning model can be trained in a supervised, unsupervised, or semi-supervised manner.” – Fig. [11]], the model with the filtered second harmonics and power data set [Paragraph [0043] – “At block 410, the process 400 involves updating the trained machine-learning model using additional real-world consumption data.”].
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to use the combination of convolutional neural networks and long short-term memory neural networks of Anderson as the machine learning model disclosed in the combination of Decker, Woodman, Lu, and Naido to detect periods of EV charging and no charging.
Regarding Claim 17, Decker, as modified, discloses the method of claim 11, comprising: receiving, by a data processing system comprising one or more processors coupled with memory [Paragraph [0052] – “FIG. 7 illustrates an exemplary computing device used for detecting electric vehicle charging operations…The depicted example of a computing device 700 includes a processor 702 communicatively coupled to one or more memory devices 704. The processor 702 executes computer-executable program code 730 stored in a memory device 704, accesses data 720 stored in the memory device 704, or both.”], time-series data sampled at a first resolution corresponding to residential electric load [Paragraph [0055] – “The computing device 700 executes program code 730 that configures the processor 702 to perform one or more of the operations described herein. For example, the program code 730 causes the processor to perform the operations described in FIGS. 1-6.”; Paragraph [0019] – “The premises monitoring application 101 can receive metering data such as voltage, load, power consumption, etc., from meters that are installed at customers' premises.”];
Decker, as modified, does not disclose detecting, by the data processing system based on a second threshold, one or more time intervals of electric vehicle charging in the time-series data.
Lu, however, discloses detecting, by the data processing system [Paragraph [0035] – “With reference to FIG. 2, one embodiment of an EV detection process 200 is shown, which is implemented and performed by the EV detection system 100 of FIG. 1. The process 200 is a computer-implemented process performed by a computing system. The described actions and functions are performed at least by a processer that accesses data from a memory or storage device and generates data read from and/or written to the memory.”] based on a second threshold, one or more time intervals of electric vehicle charging in the time-series data [Paragraph [0073] – “The classifier can be trained and executed a number of times on account time series data. The classifier is configured to generate a list of detected EV events that the classifier identified…” – SVM is the classifier; Paragraph [0084] – “…In one embodiment, the system implements and uses a logistic classifier, and a decision-tree based algorithm (e.g., Support Vector Machine SVM) to identify EV charges. In the case of the logistic classifier, suspected EV events with an assigned probability of greater than 0.5 are classified as EV charges. Of course, other values may be used.” – a time series consists of multiple time intervals; - probability of 0.5 is the threshold; refer to Fig. [2]];
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to identify periods of EV charging using the probability threshold disclosed by Lu using the metering system disclosed by Decker and Woodman to make an initial classification of the time series data.
The combination of Decker, Woodman, and Lu would disclose labeling, by the data processing system [Paragraph [0052] – “FIG. 7 illustrates an exemplary computing device used for detecting electric vehicle charging operations…The depicted example of a computing device 700 includes a processor 702 communicatively coupled to one or more memory devices 704. The processor 702 executes computer-executable program code 730 stored in a memory device 704, accesses data 720 stored in the memory device 704, or both.”] based on the detection, the time-series data to indicate the one or more time intervals of electric vehicle charging to generate labeled time- series data [Paragraph [0030] – “At block 206, the process 200 involves generating classification information for the set of premises based on an output of the trained machine-learning model. In an example, the trained machine-learning model may be trained to output an indication of the premises performing electric vehicle charging operations or not performing electric vehicle charging operations.” – see labels in Fig. [3], 312];
The combination of decker, Woodman, and Lu does not disclose receiving, by the data processing system, a harmonics and power data set sampled at a second resolution greater than the first resolution.
However, Naido discloses receiving, by the data processing system [Paragraph [0025] – “In accordance with the invention, each power system device can communicate with a remote device. Such communication is enabled with a communication hardware, software and/or firmware of the power system device…controller, wherein the gateway device or controller has such communication capabilities.”], a harmonics and power data set sampled at a second resolution greater than the first resolution [Paragraph [0023]-[0024] – “In an embodiment, a first power system device and second power system device measure data in response to a condition in a corresponding segment of the power transmission network. Here, the first device may be at a first end and sample data at a first sampling rate, while the second device may be at a second end and sample data at a second sampling rate, and the devices can measure the data at the corresponding ends…Here, the first power system device may sample data at a low sampling rate (e.g., <4 KHz)…The fault zone information can be used for locating the fault with the disturbance data received from the second power system device (which may only be sampling data (i.e., not doing any processing) at a high sampling rate (e.g., >1 MHz).”].
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to sample the second training set at a greater resolution than the first training set, as disclosed by Naido, to sample the high-resolution data received locally by the metering system disclosed by Decker, Woodman, and Lu to capture the current harmonics and power waveform.
The combination of Decker, Woodman, Lu, and Naido does not explicitly disclose inputting, by the data processing system, the labeled time-series data and the harmonics and power data set into a support vector machine to train the support vector machine to classify the harmonics and power data set to indicate electric vehicle charging or no electric vehicle charging; using, by the data processing system, the trained support vector machine to filter a second harmonics and power data set to generate a filtered second harmonics and power data set; or training, by the data processing system via a convolution neural network long short-term memory network, the model with the filtered second harmonics and power data set.
However, Lu discloses inputting, by the data processing system [Paragraph [0035] – “With reference to FIG. 2, one embodiment of an EV detection process 200 is shown, which is implemented and performed by the EV detection system 100 of FIG. 1. The process 200 is a computer-implemented process performed by a computing system. The described actions and functions are performed at least by a processer that accesses data from a memory or storage device and generates data read from and/or written to the memory.”] the labeled time-series data [See Fig. 2 blocks 210, “Account time Series Data (From Known EVs)” and 210a “Account Time Series Data (House with Known Non-EV)”, both of which train classifier at block 270] and the harmonics and power data set into a support vector machine to train the support vector machine to classify the harmonics and power data set to indicate electric vehicle charging or no electric vehicle charging [Paragraph [0072] – “At block 270, the EV charge patterns and their associated charge features and input to one or more machine learning classifiers (the classifier). This data trains the classifier to learn from the charge features in order to identify and detect an EV charge event from an unknown data set. Likewise, the non-EV charge patterns and their associated charge features are input to train the classifier to identify non-EV charge patterns and distinguish them from EV charge events. – the classifier is the SVM, the unknown data set is the harmonics and power data set (as opposed to the labeled time-series data set); Paragraph [0083]-[0084] – “(3) Given a time series for an account, is an EV being charged? In one embodiment, the system implements and uses a logistic classifier, and a decision-tree based algorithm (e.g., Support Vector Machine SVM) to identify EV charges. In the case of the logistic classifier, suspected EV events with an assigned probability of greater than 0.5 are classified as EV charges. Of course, other values may be used.” - probability of 0.5 is the threshold; see also Fig. [2], 210 and 210a, which shows data from known EV and known non EV sources training the SVM at block 270]
and using, by the data processing system [Paragraph [0035] – “With reference to FIG. 2, one embodiment of an EV detection process 200 is shown, which is implemented and performed by the EV detection system 100 of FIG. 1. The process 200 is a computer-implemented process performed by a computing system. The described actions and functions are performed at least by a processer that accesses data from a memory or storage device and generates data read from and/or written to the memory.”], the trained support vector machine to filter a second harmonics and power data set to generate a filtered second harmonics and power data set [Paragraph [0072]-[0073] – “At block 270, the EV charge patterns and their associated charge features and input to one or more machine learning classifiers (the classifier)…The classifier can be trained and executed a number of times on account time series data.” – the classifier is the SVM, SVM is trained after first training and execution, second dataset is generated with successive training and execution; Paragraph [0084] – “In one embodiment, the system implements and uses a logistic classifier, and a decision-tree based algorithm (e.g., Support Vector Machine SVM) to identify EV charges. In the case of the logistic classifier, suspected EV events with an assigned probability of greater than 0.5 are classified as EV charges. Of course, other values may be used.” – classifying EV charges based on an assigned probability is filtering].
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to use the SVM and second threshold of Lu to generate a data set filtered according to an initial classification (EV charging or no EV charging) of the training data disclosed by Decker, Woodman, Lu, and Naido in order to improve the machine learning training process.
The combination of Decker, Woodman, Lu, and Naido does not disclose training, by the data processing system via a convolution neural network long short-term memory network, the model with the filtered second harmonics and power data set.
However, Anderson discloses training, by the data processing system via a convolution neural network long short-term memory network [Col 28, Ln. 29-60 – “FIG. 11 is a flow chart of an example of a process for generating and using a machine-learning model according to some aspects…can classify input data among two or more classes; cluster input data among two or more groups; predict a result based on input data; identify patterns or trends in input data; identify a distribution of input data in a space; or any combination of these. Examples of machine-learning models can include…ensembles or other combinations of machine-learning models. In some examples, neural networks can include deep neural networks, feed-forward neural networks, recurrent neural networks, convolutional neural networks, radial basis function (RBF) neural networks, echo state neural networks, long short-term memory neural networks…or any combination of these.”; Col.29, Ln. 42-43 – “In block 1106, a machine-learning model is trained using the training data. The machine-learning model can be trained in a supervised, unsupervised, or semi-supervised manner.” – Fig. [11]], the model with the filtered second harmonics and power data set [Paragraph [0043] – “At block 410, the process 400 involves updating the trained machine-learning model using additional real-world consumption data.”].
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to use the combination of convolutional neural networks and long short-term memory neural networks of Anderson as the machine learning model disclosed in the combination of Decker, Woodman, Lu, and Naido to detect periods of EV charging and no charging.
Claim 8 is rejected under Decker et. al., in view of Woodman et. al., in view of Lu et. al., in view of Anderson et. al., in further view of Naido et. al., in further view of Gupta et. al. (US 20130289788 A1).
Regarding Claim 8, the combination of Decker, Woodman, Lu, Naido, and Anderson discloses the system of claim 7.
The combination does not disclose wherein the first resolution corresponds to 1 Hz, the second resolution corresponds to 10 kHz, and the current detected by the sensor of the metering system is sampled at 10 kHz.
However, Gupta discloses wherein the first resolution corresponds to 1 Hz [Paragraph [0031] – “Medium resolution data 220 may be sampled at a frequency of around every few seconds.” – 1 Hz is on this scale], the second resolution corresponds to 10 kHz [Paragraph [0031] – “High resolution data 210 may be sampled at a higher frequency, for example every millisecond or microsecond.” – 10 kHz falls within this range], and the current detected by the sensor of the metering system is sampled at 10 kHz [Paragraph [0031] – “Medium resolution data 220 may be sampled at a frequency of around every few seconds.” – see also the waveforms Fig. [2A]-[2C], noting that the current harmonics have already been addressed in view of Woodman].
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to use sampling rates on the order of those disclosed by Gupta to sample the first and second training data, as disclosed by the combination of Decker, Woodman, Lu, Naido, and Anderson in order to capture the waveforms associated with the sampled electric load.
Pertinent Prior Art
US 20220147035 A1, Rodemann, T., Method and System for Detecting Faults in a Charging Infrastructure System for Electric Vehicles, 2022.
US 8140283 B2, Benmouyal, G., Independent Frequency Measurement and Tracking, 2012.
US 20190081476 A1, Konya, M.J., Electric Power Grid Supply and Load prediction, 2019.
US-20220414484-A1, Garza, C. I., Service Location Anomalies, 2022.
US 20210110313 A1, Jones, R.B., Computer-Based systems, Computing Components and Computing Objects Configured to Implement Dynamic Outlier Bias Reduction in Machine Learning Models, 2021.
US 20160055419 A1, Fischer, B., Computer-Implemented method for Identifying Owners of Electric Automobile from Non-owners Using Power Consumption Data in Entity, Involves Determining Probabilistic Classification for each of Set of Users by Analyzing Load-curve Data, 2016.
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to JANELLE A HOLMES whose telephone number is (571)272-4336. The examiner can normally be reached Monday - Friday 8:00 m - 5 pm.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Arleen M Vazquez can be reached at (571) 272-2619. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of published or unpublished applications may be obtained from Patent Center. Unpublished application information in Patent Center is available to registered users. To file and manage patent submissions in Patent Center, visit: https://patentcenter.uspto.gov. Visit https://www.uspto.gov/patents/apply/patent-center for more information about Patent Center and https://www.uspto.gov/patents/docx for information about filing in DOCX format. For additional questions, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/J.A.H./Examiner, Art Unit 2857
/ARLEEN M VAZQUEZ/Supervisory Patent Examiner, Art Unit 2857