DETAILED ACTION
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Responsive to the communication dated 9/1/2023
Claims 1 – 10 are presented for examination.
Priority
ADS dated 9/1/2023 claims foreign priority to CN 202110224539.5 dated 3/1/2021.
ADS dated 9/1/2023 also claims domestic benefit of PCT/CN2021/112947 dated 8/17/2021.
As of 9/29/2026 the USPTO Patent end-to-end system (PE2E) indicates “PCT/CN2021/112947 is currently unavailable in PE2E.” The Examiner is unable to verify the priority documents exist in the parent application or national stage file. No priority documents provided in the instant application. Accordingly, the Examiner cannot verify the priority.
Information Disclosure Statement
No IDS provided.
Request for Information
Under MPEP 704 and 37 CFR 1.105 the Examiner is requesting the following documents which the Examiner believes the Applicant has access to.
“基于强化学习的准分子激光器能量控制算法研究”,《中国激光》,孙 泽旭 等,第47卷第9期,第1-10页 2020-09-30.
Zexu, 孙泽旭 & Zebin, 冯泽斌 & Yi, 周翊 & Guangyi, 刘广义 & Xiaoquan, 韩晓泉. (2020). Energy Control of Excimer Laser Based on Reinforcement Learning. Chinese Journal of Lasers. 47. 0901002. 10.3788/CJL202047.0901002.
Drawings
The drawings dated 9/1/2023 have been reviewed. They are accepted.
Specification
The abstract dated 9/1/2023 has 140 words, 12 lines, and no legal phraseology. The abstract is accepted.
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 – 10 are rejected under 35 U.S.C. 101 because the claimed invention is directed to a judicial exception without significantly more.
Claim 1.
STEP 1: Yes. The claim recites “… method…”.
STEP 2A PRONG ONE: Yes.
The claim recites: “An excimer laser energy model identification method, comprising the following steps: Step S1: building a gated recurrent unit network for excimer laser energy model identification.
Step S2: setting energy harvesting conditions in a single laser pulse manner in a plurality of preset moments to harvest a training dataset for excimer laser energy model identification; and
Step S3: training the built gated recurrent unit network by using the training dataset, and ending the training to obtain an excimer laser energy model when reaching a training ending condition” which is the abstract idea of building a mathematical equation and then fitting the mathematical equation to data.
Evidence of this is provided by Setia_2004 which states, in section IV “… Neural networks (NNs) are… represented by a mathematical model…”. Section IV further teaches that “training” a neural network in the mathematical operation of fitting weights (i.e., coefficients) of the mathematical equation to data using gradient descent approach to minimize the root-mean-squared error (RMSE) which is the calculation of a Euclidean distance where the weights are calculated according to: w (n+1)=w (n)+ Δw (n)+ Δw (n+1).
Further evidence is provided by:
Pansera_2020 teaches that artificial intelligence is simply a mathematical formulation.
ACSE4403_2011 clearly illustrates the mathematical model commonly called a neural network.
Babs_2018 shows that neural networks is simply mathematics.
Browniee_2020 further shows how neural network are simply mathematical calculations that approximate a function. This indicates that a training a neural network is simply fitting the neural network equation to the mathematical function it is intended to approximate.
Therefore, the claim is directed towards a mathematical abstract idea.
STEP 2A PRONG TWO: No.
While the claim recites “An excimer laser energy model identification method” this is merely the name of the method and simply links the use of the mathematical calculations to the field of excimer lasers. As outlined above by the evidence cited in STEP 2A PRONG ONE, a method for neural network model identification is simply the act of training/fitting the equation known as a neural network to observed data (i.e., the function the equation is intended to approximate). Therefore, the name of the method simply means to a method for fitting an equation to data. The data is simply from the field of excimer lasers. Accordingly, this simply links the method to a filed of use.
While the claim recites “Step S1: building a gated recurrent unit network for excimer laser energy model identification” building a gated recurrent unit network is simply the act of writing/establishing the base equation including its coefficients. This is like building an equation of, for example, a line (Y = MX+B) where generic coefficients/weights of M and B are written down but their values have not been determined/fitted/trained to observed data. These elements further recite “for excimer laser energy” model. The elements simply link the establishment of a base equation to the field of excimer laser energy. This is not indicative of a practical application.
While the claim recites: “Step S2: setting energy harvesting conditions in a single laser pulse manner in a plurality of preset moments to harvest a training dataset for excimer laser energy model identification” these elements are simply necessary data gather steps. This step identifies the conditions for gathering data to which the mathematical equation is subsequently fitted. MPEP 2106.05(g) states that mere data gathering is not a practical application.
For example, performing clinical tests to obtain input for an equation, testing a system response, processing test data and gathering statistics based on the testing where the statistics are then used in a mathematical equation were all determined by the Court to be insignificant extra solution activity.
In the instant application, setting up “harvesting conditions” in order to “harvest a training dataset” that is subsequently used during training (i.e., mathematical fitting of the equation coefficients/weights to the data) is therefore found to be mere data gathering.
While the claim recites “and Step S3: training the built gated recurrent unit network by using the training dataset, and ending the training to obtain an excimer laser energy model when reaching a training ending condition” this is merely the act of executing mathematical calculations that fit an equation to gathered data. This is analogous to taking the generic equation of a line: Y = MX +B and fitting it to an observed line that has a slope of 2 and crosses the Y axis at 1 which would result in an equation of Y = 2X + 1. Fitting a mathematical equation to data is part of the abstract idea itself. Accordingly, these elements are not a practical application.
Therefore, the claim is simply linking the mathematical process of writing a mathematical equation, gathering data, and then fitting the mathematical equation to the observed data. The claim does not utilize the trained/fitted mathematical equation because there are not elements beyond the abstract idea that rely upon or use the trained/fitted mathematical equation in any meaningful way.
STEP 2B: No.
As outlined above, the claim is simply the mathematical process of writing a mathematical equation, gathering data, and then fitting the mathematical equation to the observed data. All these steps are directed towards fundamental mathematical actions of creating an equation that approximates observed data. This is fundamental mathematics. While the claim generally links the use of such mathematics to the field of excimer lasers, linking the use of a judicial exception to a field of use is not significantly more than the abstract idea itself.
Claim 10.
STEP 1: Yes. The claim recites “… apparatus…”
STEP 2A PRONG ONE. Yes. The claim recites substantially the same elements as claim 1 and is therefore found to be directed towards a mathematical abstract idea due to the same reasons as outlined above for claim 1.
STEP 2A PRONG TWO: No.
While the claim recites “…comprising a processor and a memory, wherein the processor reads a computer program or instruction in the memory, and is configured to execute the following operations…” this merely generally recites a general purpose computer invoked simply as a tool to perform the mathematical calculations. Such elements are not indicative of a practical application.
STEP 2B: No.
The claim is not significantly more due to the same reasons as outlined above for claim 1.
Claim 2 recites “wherein the gated recurrent unit network comprises gated recurrent units corresponding to a plurality of time sequences; each of the gated recurrent units comprises an input layer, a hidden layer and an output layer; and the input layer is connected with the hidden layer, the hidden layer is connected with the output layer, and the hidden layers of the adjacent gated recurrent units are connected” which merely articulates the logical/mathematical relationships of the variables in the mathematical equation called a gated recurrent network. Merely reciting a mathematical relationship is not a practical application nor is it significantly more than the abstract idea itself.
Claim 3 recites “wherein the energy harvesting conditions refer to the time interval of single laser pulses and a discharge high voltage value, and the time interval of the single laser pulses refers to the time interval from a current laser pulse to a former laser pulse” merely name the type of observed data. Such elements contextualize the observed data but these elements are not more than merely linking the mathematical calculations to named data in the field of use. Such elements are not a practical application nor are they significantly more than the abstract idea.
Claim 4 recites “wherein the hidden layer of each of the gated recurrent units comprises a reset gate r(t), a refresh gate z(t) and a candidate hidden layer state h̄(t); the refresh gate z(t) represents the information amount brought by the state of a former moment to a current moment, and is shown by: zt=σWz∙xt+Uz∙ht-1 in the formula, σ represents an activation function, the activation function is a sigmoid function, Wz represents an input weight matrix of the refresh gate, x(t) represents an input variable of the current gated recurrent unit network, Uz represents a transfer matrix of the hidden layer state of the refresh gate, and h(t-1) represents a hidden layer state of a former moment; the reset gate r(t) represents a degree of the current state ignoring the former moment state, and is shown by: r(t)=σ(Wr∙x(t)+Ur∙h(t-1)) in the formula, Wr represents an input weight matrix of the reset gate, and Ur represents a transfer matrix of the hidden layer state of the reset gate; the candidate hidden layer state h̄(t) is used for assisting the calculation of the hidden layer state h(t), and is shown by: h̄(t)=tanh(W∙x(t)+U∙(r(t)⊙h(t-1))) in the formula, W represents an input weight matrix of the candidate hidden layer state, U represents a transfer matrix of the candidate hidden layer state aiming at the hidden layer state of a former moment, and ⊙ represents a Hadamard product; and the hidden layer state h(t) of the current moment is shown by: h(t)=(1-z(t))⊙h(t-1)+z(t)⊙h̄(t)” however, this merely further recites mathematical relationships and calculation. This is part of the abstract idea itself and is not a practical application nor significantly more.
Claim 5 recites “wherein the output layer of the gated recurrent unit obtains the pulse energy E(t) of an excimer laser of a current moment according to the following formulas: y(t)=σ(Wy∙h(t)) E(t)=WE∙y(t) in the formulas, yt represents an energy factor of the pulse energy of the current moment, Wy represents a weight matrix of the hidden layer state to the output layer, and WE represents an output scale conversion coefficient” however, this merely further recites mathematical relationships and calculation. This is part of the abstract idea itself and is not a practical application nor significantly more.
Claim 6 recites “wherein the training dataset is a mean value of the actual laser pulse energy in the same position under each corresponding laser burst mode at each discharge high voltage, wherein at each discharge high voltage, all actual laser pulse energy under each corresponding laser burst mode is the actual laser pulse energy harvested after the energy harvesting conditions are set in the single laser pulse manner in a preset moment” however, this merely further recites mathematical relationships and calculation. This is part of the abstract idea itself and is not a practical application nor significantly more.
Claim 7 recites “wherein a loss function of the gated recurrent unit network is as follows: l=∑t=1n12(Et(t)-E(t))2 in the formula, Et(t) represents a mean value of the actual laser pulse energy in the same position under the laser burst mode at the discharge high voltage of a training sample at a current moment, E(t) represents a pulse energy sequence of the excimer laser output by the gated recurrent unit network, and n represents a specific moment” however, this merely further recites mathematical relationships and calculation. This is part of the abstract idea itself and is not a practical application nor significantly more.
Claim 8 recites wherein the training the built gated recurrent unit network by using the training dataset comprises the following steps: Step S31: randomly selecting one training sample from the training dataset, inputting the energy harvesting conditions corresponding to the mean value of all actual laser pulse energy in the training sample into the gated recurrent unit network one by one so that a pulse energy sequence under one burst mode is obtained through gated recurrent unit network training; Step S32: calculating a loss function of the gated recurrent unit network corresponding to the currently selected training sample, and updating training parameters of the gated recurrent unit network according to the loss function; Step S33: calculating an error between the pulse energy sequence under the burst mode currently output by the gated recurrent unit network and the current training sample; and Step S34: circularly executing Steps S31 to S33, and ending the training to obtain the excimer laser energy model till reaching the training ending condition” however, this merely further recites mathematical relationships and calculation. This is part of the abstract idea itself and is not a practical application nor significantly more.
Claim 9 recites “wherein the training ending condition of the gated recurrent unit network is the number of preset training times, or the number of preset times of circularly executing Steps S31 to S33, and the maximum error between each pulse energy in the pulse energy sequence under the burst mode output by the gated recurrent unit network in each time and the pulse energy in the same position in the training sample is smaller than 0.15 mJ” however, this merely further recites mathematical relationships and calculation. This is part of the abstract idea itself and is not a practical application nor significantly more.
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.
Claims 1, 2, 4, 10 are rejected under 35 U.S.C. 103 as being unpatentable over Setia_2004 (Sensitivity Analysis and Optimization of Excimer Laser Ablation for Microvia Formation Using Neural Networks and Genetic Algorithms, 2004 IEEE/SEMI Int’l Electronics Manufacturing Technology Symposium) in view of Yousef_2003 (Neural Network Modeling and Analysis of the Material Removal Process During Laser Machining, Int J Adv Manuf Technol (2003) 22: 41 – 53) in view of Ihlemann_1995 (Excimer Laser Ablation Patterning of Dielectric Layers, Applied Surface Sciences 86 (1995) 228-233) in view of Mozaffar_2018 (Data-driven prediction of the high-dimensional thermal history in directed energy depositions processes via recurrent neural networks, Manufacturing letters 18 (2018) 35 – 39).
Claim 1. Setia_2004 makes obvious “An excimer laser model identification method
(abstract: “… using an Anvik HexScan 2150 SXE excimer laser… the complex nonlinear interactions between process set points and responses are empirically modeled using the feed-forward neural network (NNS) employing the error back-propagation training algorithm. Neural networks encode the functional relationship between process conditions and responses…”; page 2 section II: “… several types of lasers, including… excimer, are commonly used… some of the advantages of excimer lasers over the other types of lasers include low thermal damage (or low heat-affected zone), high-resolution, ablation capacity, and low average power. In this study… excimer laser … was utilized…”), comprising the following steps: Step S1: building a network for excimer laser model identification” (page 4 section IV: “… in this study, the object-oriented neural network simulator… was used to construct the networks… Java-based software package developed by… Georgia Institute of Technology that allows the user to create, train, and test NN model… each NN starts with a random set of weights…”
PNG
media_image1.png
297
563
media_image1.png
Greyscale
) Step S2: setting harvesting conditions in alaser pulse manner in a plurality of preset moments to harvest a training dataset for excimer laser model identification”
((page 4 section IV: “… to build the laser… model, feed-forward NNs were trained… using the error back-propagation (BP) algorithms on… data acquired from the experimental trials…”; page 2 section III: “… experimental design was used [10]. The tool used for experimental design and analysis… the input factors varied were laser fluence, shot frequency, number of pulses, and the vertical and horizontal positions… the responses were the top via diameter, via wall angle… and ablated thickness… measurements were conducted… the measured data… the experimental input factors and their respective ranges are shown in Table 1…”); and Step S3: training the built network by using the training dataset, and ending the training to obtain an excimer lasermodel when reaching a training ending condition (page 4 section IV: “… allows the user to… train… to build the laser… model… NNs were trained… using… data acquired from the experimental data… using the initial weight matrix and compared with the experimental data (target) to generate an error value. The BP algorithm iteratively updates the weight matrix using the gradient descent approach to minimize the root-mean-squared error (RMSE)… the error is defined as the Euclidean distance… training concludes when the scaled RMSE converges to a user-defined value…”
PNG
media_image2.png
305
569
media_image2.png
Greyscale
).
While Setia_2004 teaches laser ablation is a material removal process that uses localized thermal energy caused by stimulated radiation where absorbance of light energy is translated into thermal energy (page 2 section II) and further teaches to build a neural network model, Setia_2004’s model calculates, for example, diameter and angle of ablation but does not explicitly calculate “energy.”
Further, while Setia_2004 teaches a neural network model of laser ablation (Figure 6, 7), Setia_2004 does not teach a “gated recurrent unit” network.
Also, while Setia_2004 teaches the “number of pulses” is a selectable choice and is further an input to the neural network model, Setia_2004 does not explicitly teach s “single” pulse.
Therefore, Setia_2004 does not explicitly recite an “energy” model nor a “gated recurrent unit” network nor a “single” laser pulse.
Yousef_2003, however, makes obvious an “energy” model (Fig. 6 shown below.
PNG
media_image3.png
595
1182
media_image3.png
Greyscale
).
Setia_2004 and Yousef_2003 are analogous art because they are from the same field of endeavor called lasers. Before the effective filing date, it would have been obvious to a person of ordinary skill in the art to combine Setia_2004 and Yousef_2003. The rationale for doing so would have been that Setia_2004 teaches to use a neural network to model an nonlinear lasers including the modeling of depth and diameter. See Figure 7 of Setial_2004. Yousef_2003 further teaches to model nonlinear behavior using neural networks and to combine an energy model with a model of depth and diameters. See Fig. 6 of Yousef_2003. Therefore, it would have been obvious to combine Setia_2004 and Yousef_2003 for the benefit of having a model to predict nonlinear energy, depth, and diameter behavior of a laser to obtain the invention as specified in the claims.
Setia_2004 and Yousef_2003 a “gated recurrent unit” network nor a “single” laser pulse.
Ihlemann_1995, however, makes obvious a “single” laser pulse from an excimer laser (introduction: “excimer laser ablation… for studying the ablation process… ablation depth per pulse… the ablation patterning of dielectric coatings by excimer laser radiation is investigated…”; section3.1 par 1: “… pulse numbers…”; Fig. 2 and Fig. 4 clearly illustrate the dept of ablation for 1 pulse.
PNG
media_image4.png
618
735
media_image4.png
Greyscale
PNG
media_image5.png
595
701
media_image5.png
Greyscale
Setia_2004 and Ihlemann_1995 are analogous art because they are from the same field of endeavor called lasers. Before the effective filing date, it would have been obvious to a person of ordinary skill in the art to combine Setia_2004 and Ihlemann_1995. The rationale for doing so would have been that Setia_2004 teaches to do experiment design and to vary experimental parameters that include the number of pulses of an excimer laser and to use the experimental data to train an excimer model. Ihlemann_1995 teaches that when performing experiments with excimer lasers to vary the number of pulses and to include a single laser pulse. Therefore, it would have been obvious to combine Setia_2004 and Ihlemann_1995 for the benefit of having training data for the model that include results from a single laser pulse so that the model has data allowing it to learn excimer behavior for a single pulse so that it can accurately predict that behavior in addition to multi-pulse behavior to obtain the invention as specified in the claims.
Setia_2004, Yousef_2003, and Ihlemann_1995 does not teach a “gated recurrent unit” network
Mozaffar_2018, however, makes obvious a “gated recurrent unit” network (abstract: “… in this work a recurrent neural network (RNN) structure with a Gated Recurrent Unit (GRU) formulation is proposed for predicting the high-dimensional thermal history… our results indicate that the model can accurately predict the thermal history… the model’s general applicability and ability to accurately predict thermal histories has been demonstrated…”; page 36 section 2: “a recurrent neural network (RNN) is a special type of artificial neural network… the long short term memory (LSTM)… and the gated recurrent unit (GRU)… are two successful formulations of RNN structures… properties of RNNs led to the hypothesis that the RNN structure can predict the temperature history of DED regardless of its highly nonlinear nature…”).
Setia_2004 and Mozaffar_2018 are analogous art because they are from the same field of endeavor called lasers. Before the effective filing date, it would have been obvious to a person of ordinary skill in the art to combine Setia_2004 and Mozaffar_2018. The rationale for doing so would have been that Setia_2004 teaches to model “complex nonlinear interactions between process set points and responses” and that such nonlinear situations can be modeled using neural networks. Mozaffar_2018 teaches that Gated Recurrent Unit (GRU) are a form of neural network that have “general applicability” and the ability to accurately predict thermal history’s have been demonstrated (abstract) and further that such networks can predict the temperature history regardless of high nonlinearity (page 36 section 2). Therefore, it would have been obvious to combine Setia_2004 and Mozaffar_2018 for the benefit of having a generally applicable neural network capable of accurate prediction regardless of high nonlinearity of the scenario being modeled to obtain the invention as specified in the claims.
Claim 10. The limitations of claim 10 are substantially the same as those of claim 1 and are rejected due to the same reasons as outlined above for claim 1. Additionally, Setia_2004 makes obvious “An excimer laser energy model identification apparatus, comprising a processor and a memory, wherein the processor reads a computer program or instruction in the memory, and is configured to execute the following operations” ( page 2 section III: “… the tool used for experimental design and analysis was the RS/Discover commercial software package…”; page 4 section IV: “… in this study, the object-oriented neural network simulation (ObOrNNS) was used to construct the networks [14]. ObOrNNS is a Java-based software package developed by Intelligent Semiconductor Manufacturing Group at the Georgia Institute of Technology…” EXAMINER NOTE: the teachings of commercial software makes obvious a computer apparatus comprising a processor and a memory that is configured to read and execute the software instructions).
Claim 2. Mozaffar_2018 makes obvious “wherein the gated recurrent unit network comprises gated recurrent units corresponding to a plurality of time sequences; each of the gated recurrent units comprises an input layer, a hidden layer and an output layer; and the input layer is connected with the hidden layer, the hidden layer is connected with the output layer, and the hidden layers of the adjacent gated recurrent units are connected” (Fig. 1
PNG
media_image6.png
614
1345
media_image6.png
Greyscale
).
Claim 4. Mozaffar_2018 makes obvious “wherein the hidden layer of each of the gated recurrent units comprises a reset gate r(t), a refresh gate z(t) and a candidate hidden layer state h̄(t); the refresh gate z(t) represents the information amount brought by the state of a former moment to a current moment, and is shown by: zt=σWz∙xt+Uz∙ht-1 in the formula, σ represents an activation function, the activation function is a sigmoid function, Wz represents an input weight matrix of the refresh gate, x(t) represents an input variable of the current gated recurrent unit network, Uz represents a transfer matrix of the hidden layer state of the refresh gate, and h(t-1) represents a hidden layer state of a former moment; the reset gate r(t) represents a degree of the current state ignoring the former moment state, and is shown by: r(t)=σ(Wr∙x(t)+Ur∙h(t-1)) in the formula, Wr represents an input weight matrix of the reset gate, and Ur represents a transfer matrix of the hidden layer state of the reset gate; the candidate hidden layer state h̄(t) is used for assisting the calculation of the hidden layer state h(t), and is shown by: h̄(t)=tanh(W∙x(t)+U∙(r(t)⊙h(t-1))) in the formula, W represents an input weight matrix of the candidate hidden layer state, U represents a transfer matrix of the candidate hidden layer state aiming at the hidden layer state of a former moment, and ⊙ represents a Hadamard product; and the hidden layer state h(t) of the current moment is shown by: h(t)=(1-z(t))⊙h(t-1)+z(t)⊙h̄(t)” ( Fig. 1 – see the equations shown in figure 1 below.
PNG
media_image7.png
529
546
media_image7.png
Greyscale
).
Claims 3 are rejected under 35 U.S.C. 103 as being unpatentable over Setia_2004 in view of Yousef_2003 in view of Ihlemann_1995 in view of Mozaffar_2018 in view of ITO_1995 (US 5,463,650).
Claim 3. Setia_2004 makes obvious “wherein the energy harvesting conditions refer to the time interval of single laser pulses, and the time interval of the single laser pulses refers to the time interval from a current laser pulse to a former laser pulse” (Table 1: pulse frequency. )
ITO_1995 makes obvious “and a discharge high voltage value” as an energy harvest condition (COL 2 lines 64 – COL 3 lines 1 – 10: “… controlling the output of an excimer laser device which repeats an operation of generating a predetermined number of successive laser beam pulses and suspending the pulse oscillation for a predetermined time subsequent to the successive laser beam pulse generation, and controls discharge voltage for exciting a laser gas in a laser chamber so that each of the laser beam pulses has a predetermined energy, in which parameters which contribute to the generation of spiking such a pulse oscillation suspension time are obtained and the discharge voltage for each of the successive laser beam pulses is changed on the basis of the parameters such that the successive laser beam pulse subsequent to the suspension time have the same energy…”).
Setia_2004 and ITO_1995 are analogous art because they are from the same field of endeavor called excimer laser. Before the effective filing date, it would have been obvious to a person of ordinary skill in the art to combine Setia_2004 and ITO_1995. The rationale for doing so would have been that Setia_2004 teaches to build a model of an excimer laser. ITO_1995 further teaches to control an excimer laser by controlling the discharge voltage. Therefore, it would have been obvious to combine Setia_2004 and ITO_1995 for the benefit of being able to use the neural network to not only predict the behavior of an excimer laser but also to control its behavior to obtain the invention as specified in the claims.
Claim Objections
Claims, 5, 6, 7, 8, and 9 objected to as being dependent upon a rejected base claim, but would be allowable if rewritten in independent form including all of the limitations of the base claim and any intervening claims and if all other rejections and objections are overcome.
Claim 5. recites “wherein the output layer of the gated recurrent unit obtains the pulse energy E(t) of an excimer laser of a current moment according to the following formulas: y(t)=σ(Wy∙h(t)) E(t)=WE∙y(t) in the formulas, yt represents an energy factor of the pulse energy of the current moment, Wy represents a weight matrix of the hidden layer state to the output layer, and WE represents an output scale conversion coefficient” and the mathematical equations are not taught by the cited prior art.
Claim 6 recites “wherein the training dataset is a mean value of the actual laser pulse energy in the same position under each corresponding laser burst mode at each discharge high voltage, wherein at each discharge high voltage, all actual laser pulse energy under each corresponding laser burst mode is the actual laser pulse energy harvested after the energy harvesting conditions are set in the single laser pulse manner in a preset moment” and the mathematical relationship is not taught by the cited prior art.
Claim 7 recites “wherein a loss function of the gated recurrent unit network is as follows: l=∑t=1n12(Et(t)-E(t))2 in the formula, Et(t) represents a mean value of the actual laser pulse energy in the same position under the laser burst mode at the discharge high voltage of a training sample at a current moment, E(t) represents a pulse energy sequence of the excimer laser output by the gated recurrent unit network, and n represents a specific moment” and the mathematical equation is not taught by the cited prior art.
Claim 8 recites “wherein the training the built gated recurrent unit network by using the training dataset comprises the following steps: Step S31: randomly selecting one training sample from the training dataset, inputting the energy harvesting conditions corresponding to the mean value of all actual laser pulse energy in the training sample into the gated recurrent unit network one by one so that a pulse energy sequence under one burst mode is obtained through gated recurrent unit network training; Step S32: calculating a loss function of the gated recurrent unit network corresponding to the currently selected training sample, and updating training parameters of the gated recurrent unit network according to the loss function; Step S33: calculating an error between the pulse energy sequence under the burst mode currently output by the gated recurrent unit network and the current training sample; and Step S34: circularly executing Steps S31 to S33, and ending the training to obtain the excimer laser energy model till reaching the training ending condition” and the mathematical relationships are not taught by the cited prior art.
Claim 9 recites “wherein the training ending condition of the gated recurrent unit network is the number of preset training times, or the number of preset times of circularly executing Steps S31 to S33, and the maximum error between each pulse energy in the pulse energy sequence under the burst mode output by the gated recurrent unit network in each time and the pulse energy in the same position in the training sample is smaller than 0.15 mJ” and the 0.15mj is not taught by the cited prior art.
Additional Prior art
Feng Zebin et al., Recognition of Energy Model of Excimer Laser by Gate Recurrent Unit, available online 11/23/2020. While this reference appears to anticipate all claims and while the authors are different than the inventive entity, the publication appears to involve Bejing RSLaser Opto-Electronics Technology Co. which is that assignee of the instant application. Therefore, this is prior art under 35 USC 102(a)(1) but it appears to be an exception under 35 USC 102(b)(1).
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to BRIAN S COOK whose telephone number is (571)272-4276. The examiner can normally be reached 8:00 AM - 5:00 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, Emerson Puente can be reached at 571-272-3652. 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.
/BRIAN S COOK/Primary Examiner, Art Unit 2187