DETAILED ACTION
This action is responsive to the claims filed on 07/08/2025. Claims 1, 6-9, 11, 16-18, and 22 are pending for examination.
This action is Final.
Response to Arguments
Applicant on pg. 12 argues:
In this regard, Applicant respectfully contends that amended independent claim 1, at least by incorporating the above subject matter is eligible, because it does not recite a judicial exception. Specifically, the data of "the plurality of fault events of the relay protection system" originates from devices (a relay protection apparatus, an intelligent terminal, and a combination unit) external to a risk evaluation and fault positioning apparatus for the relay protection system. The claimed method for risk evaluation and fault positioning for relay protection system cannot be performed or implemented in the human mind because a human mind cannot obtain relevant abnormality warnings from the relay protection system itself, the apparatuses within the relay protection system, and inside each of the apparatuses comprised in the relay protection system. Therefore, the claim 1 does not recite a mental process.
Applicant’s argument about the source of the data (fault events/warnings from field devices) does not resolve Step 2A, Prong One. Whether the claim recites a mental process is determined by the nature of the claimed analysis itself; obtaining or gathering data is evaluated as an additional element under Step 2A, Prong Two. Further limiting data-gathering to a particular type or source (e.g., relay-protection alarms) is generally insignificant extra-solution activity and does not by itself integrate a judicial exception into a practical application. See MPEP §2106.05(g) (mere data gathering/field-of-use limitations).
Applicant on pg. 13 argues:
In response, claim 1 is performed by a processor of the risk evaluation and fault positioning apparatus for the relay protection system, which achieves risk evaluation and precise fault positioning for the relay protection system. The claimed method is implemented "in conjunction with a particular machine or manufacture that is integral to the claim". For example, the processor is specifically recited. Therefore, as a whole, claim 1 cannot be interpreted as being reciting a judicial exception.
Reciting that the steps are “performed by a processor” does not, by itself, transform an abstract idea into eligible subject matter or show integration into a practical application. The claim describes use of a generic computer component at a high level of generality, without any specific improvement to computer functionality or any non-conventional architecture tied to the claimed analysis. Under MPEP §2106.05(f), such a recitation is treated as mere instructions to implement the exception on a computer and does not meaningfully limit the claim for eligibility purposes.
Furthermore, Applicant’s arguments concerning the Gao reference have been found persuasive. However, the arguments are moot in light of the new references used in the current rejection.
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, 6-9, 11, 16-18, and 22are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Statutory Categories
Claim 1 is directed to a method.
Claim 9 is directed to a system.
Claim 11 is directed to a CRM.
Independent Claims
Step 2A Prong 1: Does the claim recite an abstract idea, law of nature, or natural phenomenon?
Yes. Independent claims 1, 9, and 11 recites the following limitations that are abstract ideas in the form of mental processes:
Claim 1 recites:
dividing, by the processor, the plurality of fault events of the relay protection system into different hierarchy events, the different hierarchy events comprising a top event, bottom events and intermediate events, and constructing, by the processor, a fault tree of the relay protection system according to the different hierarchy events; (mental process of evaluation which can reasonably be performed in the human mind or with aid of pen and paper but for the recitation of generic computer components, see MPEP 2106.04(a)(2)(III) for more information on mental processes; it should also be noted that "[c]ourts have examined claims that required the use of a computer (the data processing device) and still found that the underlying, patent-ineligible invention could be performed via pen and paper or in a person’s mind." Versata Dev. Group v. SAP Am., Inc., 793 F.3d 1306, 1335, 115 USPQ2d 1681, 1702 (Fed. Cir.2015), furthermore the step of determining the discomfort level index value as a mathematical integer quantity may be performed by mental evaluation, judgment or opinion)
constructing, by the processor, a target Bayesian network according to a pre-built Bayesian network conditional probability distribution table and the plurality of statuses of each node (a mental process of evaluation which can reasonably be performed in the human mind or with aid of pen and paper but for the recitation of generic computer components).
transforming, by the processor, the different hierarchy events in the fault tree into different nodes of an initial Bayesian network, the different nodes comprising root nodes, a leaf node and intermediate nodes; (a mental process of transformation which can reasonably be performed in the human mind or with aid of pen and paper but for the recitation of generic computer components)
endowing, by the processor, each node of the initial Bayesian network with a plurality of statuses, the plurality of statuses comprising severe abnormality, abnormality and normality; (a mental process of evaluation which can reasonably be performed in the human mind or with aid of pen and paper but for the recitation of generic computer components)
and determining, by the processor, according to a prior probability that each of root nodes in the target Bayesian network is in different statuses, a probability that each of intermediate nodes in the target Bayesian network is in different statuses and a probability that a leaf node in the target Bayesian network is in different statuses to complete risk evaluation for the relay protection system; (this limitation is directed to mathematical ideas which involve mathematical calculations, formulas, or algorithms under Bayesian statistics but for the recitation of generic computer components, see MPEP 2106.04(a)(2)(I) for more information on mathematical concepts)
and determining, by the processor, according to a status of the leaf node in the target Bayesian network, a posterior probability of a status of one of the root node in the target Bayesian network by using the target Bayesian network to complete fault positioning for the relay protection system (this limitation is directed to mathematical ideas which involve mathematical calculations, formulas, or algorithms under Bayesian statistics but for the recitation of generic computer components);
wherein the dividing, by the processor, the plurality of fault events of the relay protection system into the different hierarchy events comprises: setting, by the processor, the abnormality warning event occurring in the relay protection system to be the top event; (a mental process of evaluation which can reasonably be performed by human mind or with additional aid of pen and paper but for the recitation of generic computer components)
setting, by the processor, each of the abnormality warning events for decomposing a fault cause to an apparatus comprised in the relay protection system to be a respective one of the intermediate events; (a mental process of evaluation which can reasonably be performed by human mind or with additional aid of pen and paper but for the recitation of generic computer components)
and setting, by the processor, each of the abnormality warning events for decomposing a fault cause of each of the apparatuses comprised in the relay protection system into each apparatus to be a respective one of the bottom events; (a mental process of evaluation which can reasonably be performed by human mind or with additional aid of pen and paper but for the recitation of generic computer components)
wherein the transforming, by the processor, the different hierarchy events in the fault tree into the different nodes of the initial Bayesian network comprises: respectively transforming, by the processor, the top event, the bottom events and the intermediate events of the fault tree into the leaf node, the root nodes and the intermediate nodes of the initial Bayesian network; (“transforming” encompasses a user manually assigning different hierarchy events in a fault tree into nodes of a Bayesian network and, therefore, constitutes a mental process of evaluation which can reasonably be performed by human mind or with additional aid of pen and paper but for the recitation of generic computer components)
and wherein the constructing, by the processor, the target Bayesian network according to the pre-built Bayesian network conditional probability distribution table and the plurality of statuses of each node comprises: removing, by the processor, part of the root nodes, that no abnormality warning occurs in history in apparatuses of a same model in a same manufacturer, from the initial Bayesian network to construct the target Bayesian network. (“removing” encompasses a user manually removing certain nodes from the initial Bayesian network and, therefore, constitutes a mental process of evaluation which can reasonably be performed by human mind or with additional aid of pen and paper but for the recitation of generic computer components)
Step 2A Prong 2: Does the claim recite additional elements that integrate the judicial exception into a practical application?No. The judicial exceptions recited in the above discussed claims is not integrated into a practical application.
The claims recite the following additional elements, but these additional elements are not sufficient to integrate the judicial exception into a practical application:
Claim 1 recites the following additional elements:
A risk evaluation and fault positioning method for a relay protection system, comprising (this limitation is merely generally linking the abstract ideas to a particular field of Bayesian networks. See MPEP 2106.05(h) for more information on field of use.).
by a/the processor (Mere instruction to apply exception using generic computer components recited at a high level of generality, see MPEP 2106.05(f).)
obtaining, by the processor, a plurality of fault events of the relay protection system, wherein the relay protection system comprises at least the following apparatuses: a relay protection apparatus, an intelligent terminal, and a combination unit, and the plurality of fault events comprise an abnormality warning event obtained from the relay protection system, abnormality warning events obtained from the apparatuses comprised in the relay protection system and abnormality warning events inside each of the apparatuses comprised in the relay protection system; (receiving information is considered insignificant extra-solution activity under MPEP 2106.05(g))
Therefore, the above limitations do not integrate the judicial exception into a practical application.
The additional limitations fail step 2A Prong 2 of the 101 analysis because they do not transform the claim into a practical application. These limitations are too abstract or lack technical improvement that would make the concept practically useful. Without clear utility or integration into a specific field, the claim does not relate to any particular application. It does not meet the requirements of Step 2A Prong 2, as it fails to make the concept meaningfully applicable in practice.
This claim recites the following additional elements for the purposes of Step 2B analysis:
A risk evaluation and fault positioning method for a relay protection system, comprising (this limitation is merely generally linking the abstract ideas to a particular field of Bayesian networks. See MPEP 2106.05(h) for more information on field of use.).
by a/the processor (Mere instruction to apply exception using generic computer components recited at a high level of generality, see MPEP 2106.05(f).)
Claims 9 and 11 recite limitations substantially similar to claim 1 and as such a similar analysis applies.
Claim 9 further recites additional limitations:
a processor; and a memory for storing instructions executable by the processor; wherein the processor is configured to: (computer components recited at this level of generality is merely invoking machinery as a tool to perform an existing process and is considered mere instructions to apply an exception. See MPEP 2106.05(f))
Dependent Claims
The remaining dependent claims do not recite additional elements, whether considered individually or in combination, that are sufficient to integrate the judicial exception into a practical application or amount to significantly more than the judicial exception.
The claims below recite additional limitations which fail step 2A Prong 2 of the 101 analysis because they do not transform the claim into a practical application. These limitations are too abstract or lack technical improvement that would make the concept practically useful. Without clear utility or integration into a specific field, the claim does not relate to any particular application. It does not meet the requirements of Step 2A Prong 2, as it fails to make the concept meaningfully applicable in practice.
The claims also fails Step 2B of the analysis because the additional limitations do not amount to significantly more than the abstract idea itself. The additional limitations do not enhance the claim in a way that would move it beyond its abstract ideas as they minimally elaborate on the core concept without adding any inventive or technical substance. The claims are unpatentable.
Claim 6 recites the further limitation of:
The method of claim 1 wherein logic of the Bayesian network conditional probability distribution table satisfies following formulas:
PNG
media_image1.png
105
392
media_image1.png
Greyscale
where P(N =N`) represents a probability that a node N is in a status i; P(M =M`) represents a probability that a father node M of the node N is in the status i, i=1, 2, 3; status 1 means normal, status 2 means abnormal, and status 3 means severely abnormal. (this limitation is directed to mathematical ideas which involve mathematical calculations, formulas, or algorithms under Bayesian statistics but for the recitation of generic computer components, see MPEP 2106.04(a)(2)(I) for more information on mathematical concepts)
Claim 7 recites the further limitation of:
The method of claim 1, wherein the determining, by the processor according to the prior probability that each of the root nodes in the target Bayesian network is in the different statuses, the probability that each of the intermediate nodes in the target Bayesian network is in the different statuses and the probability that the leaf node in the target Bayesian network is in the different statuses comprises: determining, by the processor by means of a following formula according to the prior probability that each of the root nodes in the target Bayesian network is in the different statuses, a probability that nodes taking the root node as a father node are in different statuses; (this limitation is directed to mathematical ideas which involve mathematical calculations, formulas, or algorithms under Bayesian statistics but for the recitation of generic computer components, see MPEP 2106.04(a)(2)(I) for more information on mathematical concepts)
repeatedly executing, by the processor, following steps until determining the probability that the leaf node in the target Bayesian network is in the different statuses: (this limitation is directed to mathematical ideas which involve mathematical calculations, formulas, or algorithms under Bayesian statistics but for the recitation of generic computer components, see MPEP 2106.04(a)(2)(I) for more information on mathematical concepts)
determining, by the processor by means of the following formula according to the probability that a target node is in different statuses, a probability that the nodes taking the target node as the father node are in different statuses, wherein the target node is a node that is determined last time and has a probability of different statuses;
PNG
media_image2.png
89
439
media_image2.png
Greyscale
where P(XX') is a probability that a node Xis in a status i; y is a farther node of the node X, j=1, 2,..., m; P(yj- is a probability that a node y is in a status kj;P(X=X'IyY'Y,Ym=Y is determined according to the Bayesian network conditional probability distribution table; i=1, 2, 3, k=1, 2, 3; state 1 means normal, status 2 means abnormal, and status 3 means severely abnormal; and k1, ..., km is a status permutation and combination of y1, ...,ym. (this limitation is directed to mathematical ideas which involve mathematical calculations, formulas, or algorithms under Bayesian statistics but for the recitation of generic computer components, see MPEP 2106.04(a)(2)(I) for more information on mathematical concepts)
Claim 8 recites the further limitation of:
The method of claim 5, wherein the determining, by the processor according to the status of the leaf node, the posterior probability of the status of the root node in the target Bayesian network by using the target Bayesian network comprises: under a known condition that a leaf node T is in a status i, calculating a posterior probability that a root node Z (=1, ..., n) in the target Bayesian network is in a status Sj by using Bayesian formulas:
PNG
media_image3.png
61
260
media_image3.png
Greyscale
where P(ZJ=ZiIT=T`) is a probability that the root node Z is in the status sj and the leaf node T is in the status i, i=1, 2, 3, sj=1, 2, 3; state 1 means normal, status 2 means abnormal, and status 3 means severely abnormal.
(this limitation is directed to mathematical ideas which involve mathematical calculations, formulas, or algorithms under Bayesian statistics but for the recitation of generic computer components, see MPEP 2106.04(a)(2)(I) for more information on mathematical concepts)
Claim 22 recites the further limitation of:
The method of claim 1, wherein when the plurality of statuses are used for modeling, in response to the one of the root nodes being in the severe abnormality or the abnormality is consistent with a severity of an abnormality warning for the one of the root nodes, a probability that the one of the root nodes is in the severe abnormality or the abnormality is the prior probability of the one of the root nodes; in response to the one of the root nodes being in the severe abnormality or the abnormality is not consistent with a severity of an abnormality warning for the one of the root nodes, the probability that the one of the root nodes is in the severe abnormality or the abnormality is 0; and a probability that the one of the root nodes is in normal is a probability that no abnormality warning occurs for the one of the root nodes. (this limitation is describing the conditions for modeling at a high level of generality and, therefore, constitutes mere instruction to apply the exception under MPEP 2106.05(f).)
Claims 16-18 recite limitations substantially similar to claims 6-8 and as such a similar analysis applies.
Claim Rejections - 35 USC § 103
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.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or non-obviousness.
This application currently names joint inventors. In considering patentability of the claims the examiner presumes that the subject matter of the various claims was commonly owned as of the effective filing date of the claimed invention(s) absent any evidence to the contrary. Applicant is advised of the obligation under 37 CFR 1.56 to point out the inventor and effective filing dates of each claim that was not commonly owned as of the effective filing date of the later invention in order for the examiner to consider the applicability of 35 U.S.C. 102(b)(2)(C) for any potential 35 U.S.C. 102(a)(2) prior art against the later invention.
Claims 1, 6 and 11, 16-17, 19-21 are rejected under 35 U.S.C. 103 as being unpatentable by Xiang (CN 109818335 A), hereafter referred to as Xiang, in view of Bobbio et al. (Bobbio, A., Portinale, L., Minichino, M., & Ciancamerla, E. (2001). Improving the analysis of dependable systems by mapping fault trees into Bayesian networks. Reliability Engineering & System Safety, 71(3), 249-260.), hereafter referred to as Bobbio and in further view of Kalet (Kalet, A. M. (2015). Bayesian networks from ontological formalisms in radiation oncology (Doctoral dissertation).), hereafter referred to as Kalet.
Claim 1: Xiang teaches:
A risk evaluation and fault positioning method for a relay protection system, performed by a processor of a risk evaluation and fault positioning apparatus for the relay protection system, the method comprising: (Xiang, page 1, paragraph 12, “Defect diagnosis module: establish a two-layer anomaly event model based on the fault tree analysis method, and perform layer-by-layer diagnostic evaluation of the two-layer anomaly model based on the abnormal event data set and the abnormal event defect level evaluation rule determined in the event building module to determine the protection system.”, Xiang’s method covers system-level risk evaluation and diagnostic (fault-positioning) in a relay protection context.)
obtaining, by the processor, a plurality of fault events of the relay protection system, wherein the relay protection system comprises at least the following apparatuses: a relay protection apparatus, an intelligent terminal, and a combination unit, and the plurality of fault events comprise an abnormality warning event obtained from the relay protection system, abnormality warning events obtained from the apparatuses comprised in the relay protection system and abnormality warning events inside each of the apparatuses comprised in the relay protection system; (Xiang, page 1, paragraph 4, “The protection system covers the merging unit, intelligent terminal, protection device and process layer switch, and sends the protection system real-time running status through SV, GOOSE and MMS messages.”
Xiang, page 4, paragraph 5, “The network communication module 3 acquires real-time message information of the protection system, and the real-time message information includes a protection system based on the IEC61850 standard. SV real-time messages, GOOSE real-time messages, and MMS realtime messages, such as packet interruptions, device hardware anomalies, and software alarms, that reflect the protection status. The SV real-time message refers to the message describing the current and voltage analog quantity; the GOOSE real-time message refers to the message describing the interlocking and tripping control commands between the IEDs and the signal timing characteristics; the MMS real-time message means : Adopt a publish-subscribe mechanism to describe messages such as alarms and action events.”, This explicitly lists the apparatus set (merging unit ≈ combination unit; intelligent terminal; protection device/relay) and the sources and types of abnormality warnings (system messages + device alarms) that are obtained.)
dividing, by the processor, a plurality of fault events of the relay protection system into different hierarchy events, (Xiang, page 1, paragraph 12, “Defect diagnosis module: establish a two-layer anomaly event model based on the fault tree analysis method, and perform layer-by-layer diagnostic evaluation of the two-layer anomaly model based on the abnormal event data set and the abnormal event defect level evaluation rule determined in the event building module to determine the protection system.”, Directly discloses hierarchical division of events and that the hierarchy is based on FTA, i.e., constructing a fault-tree-based model for diagnosis.)
PNG
media_image4.png
237
456
media_image4.png
Greyscale
Figure 2 of Xiang
the different hierarchy events comprising a top event, bottom events and intermediate events, and constructing, by the processor, a fault tree of the relay protection system according to the different hierarchy events;(Xiang, page 4, paragraph 5, “The first layer of the two-layer anomaly event model is the protection system link abnormal event layer, the second layer is the basic component abnormal event layer, and the protection system link abnormal event layer includes multiple protection system link abnormal events, including each protection system link abnormal event. Multiple basic component anomalies, all basic component exception events form the base component anomaly event layer. The two-layer anomaly event model is shown in Figure 2.”
Xiang, page 2, last paragraph, “Finally, determine the defect level of the protection system: the defect level with the highest defect level in all the protection system abnormal events is regarded as the protection system defect level; The level of the defect is proportional to the severity of the defect.”, it is interpreted that Top level events = overall “protection system defect level” (system outcome); Intermediate = “link abnormal event layer”; Bottom = “basic component abnormal event layer” (inside each apparatus). This is precisely the claimed three-tier hierarchy feeding the fault tree and is similarly reflected in figure 2 of Gao.)
Bobbio, in the same field of Bayesian fault analysis, teaches the following limitations which Xiang fails to teach:
transforming, by the processor, the different hierarchy events in the fault tree into different nodes of an initial Bayesian network, the different nodes comprising root nodes, a leaf node and an intermediate nodes; (Bobbio, page 252, col. 2, paragraph 2, “According to the translation rules for the basic gates, it is straightforward to map an FT [fault tree] into a binary BN, i.e. a BN [initial Bayesian network]…The conversion algorithm proceeds along the following steps: for each leaf node (i.e. primary event or system component) of the FT, create a root node [root nodes] in the BN; however, if more leaves of the FT represent the same primary event (i.e. the same component), create just one root node in the BN; • assign to root nodes in the BN the prior probability of the corresponding leaf node in the FT (computed at a given mission time t); • for each gate of the FT, create a corresponding node [intermediate nodes] in the BN; • label the node corresponding to the gate whose output is the TE of the FT as the Fault node [leaf node] in the BN; • connect nodes in the BN as corresponding gates are connected in the FT; • for each gate (OR, AND or k:n) in the FT assign the equivalent CPT to the corresponding node in the BN (see Figs. 1 and 2).”, Bobbio provides a method for transforming fault trees (FT) into a Bayesian Network (BN) mapping: FT bottom events → BN root nodes, FT internal gates → BN intermediate nodes, and FT Top Event → BN leaf/Fault node.)
Xiang further teaches:
endowing, by the processor, each node of the initial Bayesian network with a plurality of statuses, the plurality of statuses comprising severe abnormality, abnormality and normality; (Xiang, page 4, paragraph 1, “defect definition principle refers to: 1 general defect, the protection system has an abnormal phenomenon, but does not affect the realization of the protection function, such as protection against time anomaly, the analog quantity exceeds the limit; 2 serious defects, some protection functions are affected, but the protection can continue to operate. Such as the longitudinal protection channel warning, etc.; 3 critical defects, the protection can not operate normally, need to stop immediately, such as CPU abnormal alarm.”, Xiang provides three severity tiers (general/serious/critical) matching abnormality levels. The severity tiers of the fault tree would carry over to the Bayesian network taught by Bobbio)
Bobbio in the same field of Bayesian fault analysis, teaches the following limitations which Xiang fails to teach:
constructing, by the processor, a target Bayesian network according to a pre-built Bayesian network conditional probability distribution table and the plurality of statuses of each node; (Bobbio, page 250, col. 2, paragraph 5, “the quantitative part is completely specified by considering the probability of each value of a variable conditioned by every possible instantiation of its parents (i.e. by considering only local conditioning). These local conditional probabilities are specified by defining, for each node, a Conditional Probability Table (CPT). The CPT contains, for each possible value of the variables associated to a node, all the conditional probabilities with respect to all the combination of values of the variables associated to the parent nodes. Variables having no parents are called root variables and marginal prior probabilities are associated with them.”, Bobbio’s fault tree to Bayesian network transformation does assign CPTs (conditional probability tables, which includes multi-state entries) and root priors and is interpreted as the claimed “target BN” formed from a pre-built CPT table and node statuses.)
and determining, by the processor, according to a prior probability that each of root nodes in the target Bayesian network is in different statuses, (Bobbio, page 249, col. 2, paragraph 3, “A forward (or predictive) analysis, in which the probability of occurrence of any node of the network is calculated on the basis of the prior probabilities of the root nodes and the conditional dependence of each node.”, teaches that root nodes carry priors and forward inference uses those priors to propagate probabilities.)
a probability that each of intermediate nodes in the target Bayesian network is in different statuses (Bobbio, page 252, col. 2, paragraph 2, “connect nodes in the BN as corresponding gates are connected in the FT; for each gate (OR, AND or k:n) in the FT assign the equivalent CPT to the corresponding node in the BN (see Figs. 1 and 2).”, Once CPTs are assigned, each intermediate node’s state probabilities are determined by the CPT and upstream priors (i.e., the node’s “different statuses” probabilities).)
and a probability that a leaf node in the target Bayesian network is in different statuses to complete risk evaluation for the relay protection system; (Bobbio, page 252, paragraph 3, “Due to the very special nature of the gates appearing in a FT, non-root nodes of the BN are actually deterministic nodes and not random variables and the corresponding CPT can be assigned automatically. The prior probabilities on the root nodes are coincident with the corresponding probabilities assigned to the leaf nodes in the FT.”, similarly above for the root and intermediate nodes, leaf nodes are computed a probability using the CPT to complete (deterministic) risk evaluation for the system.)
and determining, by the processor, according to a status of the leaf node in the target Bayesian network, a posterior probability of a status of one of the root node in the target Bayesian network by using the target Bayesian network to complete fault positioning for the relay protection system. (Bobbio, page 252, paragraph 2, “assign to root nodes in the BN the prior probability of the corresponding leaf node in the FT (computed at a given mission time t);”)
Xiang further teaches:
wherein the dividing, by the processor, the plurality of fault events of the relay protection system into the different hierarchy events comprises: setting, by the processor, the abnormality warning event occurring in the relay protection system to be the top event; (Xiang, page 2, last paragraph, “Finally, determine the defect level of the protection system: the defect level with the highest defect level in all the protection system abnormal events is regarded as the protection system defect level; The level of the defect is proportional to the severity of the defect.”, The “protection system defect level” is the aggregated system outcome determined at the conclusion of the evaluation. Under broadest reasonable interpretation (BRI), this corresponds to the Top Event in a fault tree / BN (the system-level abnormality result).)
setting, by the processor, each of the abnormality warning events for decomposing a fault cause to an apparatus comprised in the relay protection system to be a respective one of the intermediate events; (Xiang, page 2, paragraph 2, “The protection system link abnormal event layer includes three protection system link abnormal events, namely: an abnormality event of the AC input link, an abnormal event of the IED device link, and an abnormal event of the data communication link.”, Xiang defines a link-level layer (AC input, IED device link, data communication link). These are between the system outcome and the basic component events, and is interpreted as the intermediate tier.)
and setting, by the processor, each of the abnormality warning events for decomposing a fault cause of each of the apparatuses comprised in the relay protection system into each apparatus to be a respective one of the bottom events; (Xiang, page 2, paragraph 1, “The system link abnormal event includes a plurality of basic component abnormal events, and all basic component abnormal events constitute the basic component abnormal event layer… The basic component abnormal events in the abnormal event of the AC input link include: an SV link abnormality alarm, an abnormality in the deviation of the homologous comparison evaluation, an abnormality in the current difference of the protection device, and an abnormality in the voltage input of the protection device.”, Xiang states that every link-level event decomposes into basic component anomalies and lists device-level items (merging unit, protection device, intelligent terminal). Under BRI, these are bottom events (inside each apparatus).)
Bobbio in the same field of Bayesian fault analysis, teaches the following limitations which Xiang fails to teach:
wherein the transforming, by the processor, the different hierarchy events in the fault tree into the different nodes of the initial Bayesian network comprises: respectively transforming, by the processor, the top event, the bottom events and the intermediate events of the fault tree into the leaf node, the root nodes and the intermediate nodes of the initial Bayesian network; (Bobbio, page 252, col. 2, paragraph 2, “According to the translation rules for the basic gates, it is straightforward to map an FT into a binary BN, i.e. a BN…The conversion algorithm proceeds along the following steps: for each leaf node (i.e. primary event or system component) of the FT, create a root node in the BN; however, if more leaves of the FT represent the same primary event (i.e. the same component), create just one root node in the BN; • assign to root nodes in the BN the prior probability of the corresponding leaf node in the FT (computed at a given mission time t); • for each gate of the FT, create a corresponding node in the BN; • label the node corresponding to the gate whose output is the TE of the FT as the Fault node in the BN; • connect nodes in the BN as corresponding gates are connected in the FT; • for each gate (OR, AND or k:n) in the FT assign the equivalent CPT to the corresponding node in the BN (see Figs. 1 and 2).”, Bobbio provides a method for transforming fault trees (FT) into a Bayesian Network (BN) mapping: FT bottom events → BN root nodes, FT internal gates → BN intermediate nodes, and FT Top Event → BN leaf/Fault node.)
It would have been obvious to a person of ordinary skill in the art (POSITA), before the effective filing date of the claimed invention, to apply Bobbio’s fault-tree-to-Bayesian-network translation to Xiang’s fault-tree model in order to obtain quantitative risk assessment and inference on a set of protection-system abnormality events. Xiang expressly organizes protection-system abnormality events—at system/link and device levels—using fault-tree analysis for diagnostic assessment and defect grading; Bobbio provides an algorithm for converting a FT to a BN and explains that BN forward (prior→leaf) and backward (evidence→root) analyses replicate and extend FT analyses. A motivation for the combination is leveraging well-established Bayesian network algorithms for further interpreting relationships found in fault tree analysis. (Bobbio, page 256, last paragraph, “One advantage is that there exist well-established algorithms that can compute the marginal posterior probability of each node”)
Kalet, in the same field of Bayesian networking, teaches the following which the above fails to teach:
and wherein the constructing, by the processor, the target Bayesian network according to the pre-built Bayesian network conditional probability distribution table and the plurality of statuses of each node comprises: removing, by the processor, part of the root nodes, that no abnormality warning occurs in history in apparatuses of a same model in a same manufacturer, from the initial Bayesian network to construct the target Bayesian network. (Kalet, page 31, paragraph 1, “Figure 3.2 shows the final network after pruning nodes that could not be populated from our data.”, Kalet teaches data-driven pruning of BN nodes that lack historical/populatable data prior to finalizing the model. The claimed rule (“no historical abnormality in same-model/same-manufacturer apparatus”) is a specific case of history-based pruning — i.e., removing a root-cause node when history provides no data/occurrence.)
It would have been obvious to a POSITA, before the effective filing date of the claimed invention, when implementing Bobbio’s BN (which requires priors and CPT entries) on real operational data, to prune nodes that cannot be parameterized from available historical records, as taught by Kalet. Bobbio’s BN construction assigns marginal priors to root nodes and CPTs to internal nodes; if the domain data contain no historical evidence for a node (e.g., no recorded abnormality history for a given root/device type), a POSITA would remove that node to produce a target BN that can actually be populated and inferred over.
Claim 11 recites limitations substantially similar to claim 1 and as such a similar analysis applies.
Claim 16 recites limitations substantially similar to claim 6 and as such a similar analysis applies.
Claims 7 and 17 are rejected under 35 U.S.C. 103 as being unpatentable over Xiang in view of Bobbio and Kalet as applied to claim 1 above, and in further view of Gao et al. (CN-106600127-A), hereafter referred to as Gao.
Claim 7: Gao teaches the limitations of claim 1, Gao further teaches:
The method of claim 1, wherein the determining, by the processor according to the prior probability that each of the root nodes in the target Bayesian network is in the different statuses, the probability that each of the intermediate nodes in the target Bayesian network is in the different statuses and the probability that the leaf node in the target Bayesian network is in the different statuses comprises:
determining, by the processor by means of a following formula according to the prior probability that each of the root nodes in the target Bayesian network is in the different statuses, a probability that nodes taking the root node as a father node are in different statuses; (Gao, page 6, table 4, “Table 4 -Node E, G probability of occurrence table”:
PNG
media_image5.png
235
668
media_image5.png
Greyscale
Table 4 represents the probability of occurrence (of fault events) for intermediate-level nodes (E,G) based on the conditional probabilities of their parent nodes. This teaches how prior probabilities for root level nodes (Node A, B, C, or D) propagate downward (decomposing a fault cause) to their child nodes.)
repeatedly executing, by the processor, following steps until determining the probability that the leaf node in the target Bayesian network is in the different statuses: (Gao, page 6, table 5, “Table 5- Node H probability Table":
PNG
media_image6.png
246
699
media_image6.png
Greyscale
the probability values in table 5 demonstrate that leaf node H is calculated based on the probabilities of its parent intermediate nodes (E,F,G) which was subsequently based on the root node probabilities (A,B,C,D), teaching that calculations of probability from root nodes are propagated and repeated downward to the leaf nodes.)
determining, by the processor, by means of the following formula according to the probability that a target node is in different statuses, a probability that the nodes taking the target node as the father node are in different statuses, wherein the target node is a node that is determined last time and has a probability of different statuses;
PNG
media_image2.png
89
439
media_image2.png
Greyscale
where P(XX') is a probability that a node Xis in a status i; y is a farther node of the node X, j=1, 2,..., m; P(yj- is a probability that a node y is in a status kj; P(X=X'IyY'Y,Ym=Y is determined according to the Bayesian network conditional probability distribution table; i=1, 2, 3, k=1, 2, 3; state 1 means normal, status 2 means abnormal, and status 3 means severely abnormal; and k1, ..., km is a status permutation and combination of y1, ...,ym. (Gao, page 6, table 5, the probability values in table 5 as shown above in this claim demonstrate that leaf node H is calculated based on the probabilities of its parent intermediate nodes (E,F,G). Since the parent probabilities were determined last time and has a probability of different statuses (the risk levels as disclosed above) and since it is used to determine the probabilities of subsequent child nodes, the functional teaching of this claim is met by this Bayesian calculation in Gao.)
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have incorporated the teachings disclosed in Xiang with the teachings disclosed in Gao in order to obtain quantified risk probabilities and graded outputs, rather than using Xiang’s non-Bayesian scoring alone. A motivation of which is to use Bayesian posterized fault probabilities to calculate risk grades, which Xiang lacks. (Gao, page 2, paragraph 7, “Step 4.2, draw the Bayesian network acyclic graph according to the father-son nodes found in the first step; Step 4.3, Determine the dependencies and independent relationships among nodes, and determine the conditional probability of each node; Step 4.4: Check the directed edges of each node to determine the correctness of the DAG plotted, and calculate the conditional probability according to the requirements… To assess the risk, we must first know the possibility of failure, that is, the failure rate of the equipment has been identified; followed by the consequences of failure to understand and evaluate the consequences of quantitative assessment of risk as a measure of risk . The function expression is as follows: R (t) = LE (t) × P (t) (5)”, Gao teaches how to build and quantify the Bayesian network to fit risk grading)
Claim 17 recites limitations substantially similar to claim 7 and as such a similar analysis applies.
Claims 8, 9, and 18 are rejected under 35 U.S.C. 103 as being unpatentable over Xiang in view of Bobbio and Kalet as applied to claim 1 above, and in further view of Tian et al. (Tian, J., & He, R. (2012). Computing posterior probabilities of structural features in Bayesian networks. arXiv preprint arXiv:1205.2612.), hereafter referred to as Tian.
Claim 8: Xiang, Bobbio, and Kalet teaches the limitations of claim 1, Tian, in the same field of Bayesian inference, further teaches:
The method of claim 1, wherein the determining, by the processor, according to the status of the leaf node, the posterior probability of the status of one of the root nodes in the target Bayesian network by using the target Bayesian network comprises: under a known condition that a leaf node T is in a status i, calculating, by the processor, a posterior probability that a root node Z (=1, ..., n) in the target Bayesian network is in a status Sj by using Bayesian formulas:
PNG
media_image3.png
61
260
media_image3.png
Greyscale
where P(ZJ=ZiIT=T`) is a probability that the root node Z is in the status sj and the leaf node T is in the status i, i=1, 2, 3, sj=1, 2, 3; state 1 means normal, status 2 means abnormal, and status 3 means severely abnormal.
(Tian, page 539, col. 1, last paragraph, “In the Bayesian approach to learn Bayesian networks from the training data D, we compute the posterior probability of a network G as
PNG
media_image7.png
45
258
media_image7.png
Greyscale
We can then compute the posterior probability of any hypothesis of interest by averaging over all possible networks. In this paper, we are interested in computing the posteriors of structural features. Let f be a structural feature represented by an indicator function such that f(G) is 1 if the feature is present in G and 0 otherwise. We have
PNG
media_image8.png
50
260
media_image8.png
Greyscale
”, this teaches that posterior probabilities in a Bayesian network are computed based on observed data denoted D, and since D consists of known states of nodes (including leaf nodes) this directly affects the calculated posterior probability of all nodes above such as root nodes. It is interpreted by the examiner that this effectively teaches determining the posterior probability of a root node based of a change in status (probability calculation) or a leaf node.)
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have incorporated the teachings disclosed in Xiang with the teachings disclosed in Tian (updating posterior probability calculations with aggregated state data sent upstream the network). A motivation for combination is gain additional posterior knowledge learned from subsequent nodes to affect the final inferred probability of the entire network (the root nodes).) (Tian, page 546, col. 1, paragraph 2, “The main advantage of our algorithm over the current state-of-the-art algorithms, the DP algorithm in [Koivisto, 2006] and the order MCMC in [Friedman and Koller, 2003], for computing the posterior probabilities of structural features is that those algorithms require special structure prior P(G) that is highly non uniform while we allow general prior P(G).”)
Claim 9 recites limitations substantially similar to claim 1 and as such a similar analysis applies.
Claim 9 further recites additional limitations which Tian teaches:
a processor; and a memory for storing instructions executable by the processor; wherein the processor is configured to: (Tian, page 543, col. 2, last paragraph, “All the experiments were run under Linux on an ordinary desktop PC with a 3.0GHz Intel Pentium processor and 2.0GB of memory.”)
The rationale for this combination is the same motivation as similarly applied above for claim 8.
Claim 18 recites limitations substantially similar to claim 8 and as such a similar analysis applies.
Claim 6 is rejected under 35 U.S.C. 103 as being unpatentable over Xiang view of Bobbio and Kalet as applied to claim 1 above, and in further view of Ali et al. (US 7774293 B2), hereafter referred to as Ali.
Claim 6: Xiang, Bobbio, and Kalet teaches the limitations of claim 1, Ali teaches:
The method of claim 1, wherein logic of the Bayesian network conditional probability distribution table satisfies following formulas:
PNG
media_image1.png
105
392
media_image1.png
Greyscale
where P(N =N`) represents a probability that a node N is in a status i; P(M =M`) represents a probability that a father node M of the node N is in the status i, i=1, 2, 3; status 1 means normal, status 2 means abnormal, and status 3 means severely abnormal. (Ali, col. 7, line 47, “The foregoing example illustrates a beneficial feature of the invention, i.e., f(.alpha.,L) is applied once at each node with the appropriate conditions L. The conditional marginal probabilities Pr( x), Pr(y| x), Pr(z|y x), Pr(x) and Pr(z|x) follow from the BBN.", this teaches that in a BBN (Bayesian Belief Network) each node has a conditional probability function that depends on the specific states of its parents nodes. It is interpreted by the examiner that this is analogous to the function of the claim language.)
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have incorporated the teachings disclosed in Xiang with the teachings disclosed in Ali (have conditional probability tables based on parent statuses/state). A motivation for combination is gain additional posterior knowledge learned from subsequent nodes to affect the final inferred probability of new node (the leaf nodes).) (Ali, col. 4, line 19, “The connections between the BBNs and logic models are defined by binary variables in the BBN that correspond to basic events in the FTs, or directly as pivotal events in the ESDs. The probability of those events are then determined by the BBN and propagated through each logic model. The framework of the present invention includes factors that have widespread influences, such as the previously mentioned training or quality of maintenance, and as such, BBNs may affect multiple basic events across multiple fault trees.”, allows for more accurate modeling of real world causal relationships)
Claim 22 is rejected under 35 U.S.C. 103 as being unpatentable over Xiang in view of Bobbio and Kalet as applied to claim 1 above, and in further view of (Manual, N. J. (2012). Version 4.18 and higher. Norsys Software Corp.), hereafter referred to as Norsys.
Claim 22: Xiang, Bobbio, and Kalet teaches the limitations of claim 1, Norsys teaches:
The method of claim 1, wherein when the plurality of statuses are used for modeling, (Norsys, page 34, paragraph 2, “For example, say the node 'Temperature' can take on the values cold, medium, and hot… A likelihood finding consists of one probability for each state of the node,”, Netica, a program for creating belief networks for probabilistic systems, explicitly models tri-state node statuses (normal/abnormal/severe maps to cold, medium, hot))
in response to the one of the root nodes being in the severe abnormality or the abnormality is consistent with a severity of an abnormality warning for the one of the root nodes, a probability that the one of the root nodes is in the severe abnormality or the abnormality is the prior probability of the one of the root nodes; (Norsys, page 21, paragraph 3, “After the Bayes net is constructed, it may be applied. For each variable we know the value of, we enter that value into its node as a finding (also known as "evidence")… The final beliefs are sometimes called posterior probabilities (with prior probabilities being the probabilities before any findings were entered). Probabilistic inference done within a Bayes net is called belief updating.” Netica explains the life cycle of inference in a Bayes net: once the network is built, you enter a finding (evidence) on a node; the network performs belief updating; the priors (before any findings) are transformed into posteriors (after the finding).
Page 45, paragraph 3, “If nothing is known regarding the value of this variable (i.e. missing data), then a question mark ? or an asterisk * should be used to indicate that. It is equivalent to ~{} which is a likelihood of all ones.”, In Netica’s neutral/complete-uncertainty likelihood form (the “likelihood of all ones”), because belief updating multiplies likelihood × prior for each state and normalizes, a uniform likelihood (equal across states) doesn’t change any state’s probability. Under broadest reasonable interpretation (BRI), a warning that is “consistent” (i.e., provides no preference among the relevant states) can be encoded as this neutral likelihood, so the state’s posterior remains equal to its prior)
in response to the one of the root nodes being in the severe abnormality or the abnormality is not consistent with a severity of an abnormality warning for the one of the root nodes, the probability that the one of the root nodes is in the severe abnormality or the abnormality is 0; (Norsys, page 36, paragraph 3, “In general, it is not possible to determine anything about what the belief of a node is going to be based just on its accumulated likelihood findings, except that states with a zero likelihood will have a zero belief.”, if likelihood (evidence findings) is zero (i.e., not consistent with the warning) then the probability is 0)
and a probability that the one of the root nodes is in normal is a probability that no abnormality warning occurs for the one of the root nodes. (Norsys, page 34, paragraph 3, “Say we have a thermosensor to measure 'Temperature', which is designed so that when the temperature is hot it is supposed to turn on. In actual practice we find that when the temperature is cold the sensor never goes on, when the temperature is medium it goes on 10% of time, and when it is hot it always goes on. If at a certain time we observe the sensor on, and want to enter this finding into the Temperature node, then we do so as a likelihood finding.", In the Netica Model, no warning (Temperature = cold) is the normal condition, so the probability of Normal aligns with the probability of cold for that node)
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have incorporated the teachings disclosed in Xiang with the teachings disclosed in Norsys. A motivation for combination is to encode warnings/alarms from the protection system as an application programming interface. (Norsys, page 22, paragraph 2, “Netica uses the fastest known algorithm for exact general probabilistic inference in a compiled Bayes net, which is message passing in a junction tree (or "join tree") of cliques... The quality of the compilation depends upon the elimination order used, which is a list of all the nodes in the net. Any order of the nodes will produce a successful compilation, but some do a much better job than others. You may specify an elimination order (perhaps from your own program, or by using Netica Application‟s “optimize compile”), or just let Netica API find a good one itself.”)
Conclusion
THIS ACTION IS MADE FINAL. Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to HYUNGJUN B YI whose telephone number is (703)756-4799. The examiner can normally be reached M-F 9-5.
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, Andrew Jung can be reached on (571) 270-3779. 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.
/H.B.Y./Examiner, Art Unit 2146
/SHAHID K KHAN/Primary Examiner, Art Unit 2146
Prosecution Insights