An integrated photonic prediction method, apparatus, device, medium, and product

Abstract

The application discloses an integrated photon prediction method and device, equipment, medium and product, and relates to the technical field of integrated photon. The method comprises the following steps: after configuration data used for simulating the effect prediction of a target integrated photon object is acquired, sample data is collected from all historical simulation sample data discarded and matched with the target object by using a data collection algorithm, and the collection result is added to a sample data set; then, the data set is applied to calibrate and verify a machine learning model based on an artificial neural network to obtain a simulation effect prediction model or a reverse structure prediction model; finally, the prediction model is applied to obtain design parameters of the target integrated photon object and output the design parameters, so that all invalid simulation data generated in the design process can be fully utilized, simulation design time of different functions under the same integrated photon structure can be saved for users, and the application and popularization are facilitated.

Description

TECHNICAL FIELD

[0001] The present application belongs to the field of integrated photonics, and particularly relates to an integrated photonics prediction method, device, equipment, medium and product. BACKGROUND

[0002] In the post-Moore era (i.e., the era after the failure of Moore's Law), as integrated circuits approach physical and theoretical limits, integrated photonics technology (PIC, which is a technology system for integrating optical systems into a chip, similar to electronic integrated circuit technology; unlike electronic integrated circuit technology that integrates transistors, capacitors and / or resistors, etc., integrated photonics technology integrates various different optical or optoelectronic devices, such as lasers, electro-optical modulators, photodetectors, optical attenuators, optical multiplexers / demultiplexers, and optical amplifiers, etc.) is getting more and more attention as an alternative solution. However, whether it is integrated photonics device design or integrated optoelectronic system link design, a large amount of simulation time is required; at the same time, in the design process, the final design is not achieved in one step, and a large number of iterative simulations are required to design the final design structure that meets the requirements.

[0003] At present, a large amount of invalid simulation data generated from the initial idea to the design result is discarded, so that the existing integrated photonics design scheme has the problems of data waste and long design simulation time. Therefore, how to utilize these invalid simulation data to help designers save simulation time is a subject that those skilled in the art need to study. SUMMARY

[0004] The purpose of the present application is to provide an integrated photonics prediction method, device, computer equipment, computer readable storage medium and computer program product, to solve the problem of data waste and long design simulation time existing in the existing integrated photonics design scheme.

[0005] In order to achieve the above-mentioned purpose, the present application adopts the following technical scheme:

[0006] In a first aspect, an integrated photonics prediction method is provided, comprising:

[0007] Obtaining configuration data used for simulation effect prediction or reverse structure prediction of a target integrated photonics object, wherein the target integrated photonics object includes an integrated photonics device or an integrated optoelectronic system link to be designed, the configuration data contains design parameter configuration information of the target integrated photonics object, and the design parameter configuration information contains multi-dimensional design parameters to be optimized;

[0008] collecting the historical simulation sample data from all the historical simulation sample data that has been discarded and matched with the target integrated photon object by using a data collection algorithm and adding the collection result to a sample data set, wherein the historical simulation sample data contains historical simulation parameter values and historical simulation effect index values of the multi-dimensional design parameters as model input items and model output items;

[0009] applying the sample data set to calibrate and verify a machine learning model based on an artificial neural network to obtain a simulation effect prediction model for outputting a corresponding simulation effect index value after inputting a parameter value of the multi-dimensional design parameter or a reverse structure prediction model for outputting a corresponding parameter value of the multi-dimensional design parameter after inputting a simulation effect index value;

[0010] applying the simulation effect prediction model or the reverse structure prediction model to obtain and output the design parameter of the target integrated photon object.

[0011] Based on the above invention content, a new scheme for integrated photon design based on historical invalid simulation data is provided, that is, after obtaining configuration data for simulation effect prediction of a target integrated photon object, sample data is first collected from all historical simulation sample data that has been discarded and matched with the target object by using a data collection algorithm, and the collection result is added to a sample data set, then a data set is applied to calibrate and verify a machine learning model based on an artificial neural network to obtain a simulation effect prediction model or a reverse structure prediction model, finally the prediction model is applied to obtain and output the design parameter of the target integrated photon object, so that all invalid simulation data generated in the design process can be fully utilized to save simulation design time for different functions under the same integrated photon structure, facilitating practical application and promotion.

[0012] In one possible design, when the design parameter configuration information further contains a target function related to a simulation effect index of the target integrated photon object, the simulation effect prediction model is applied to obtain the design parameter of the target integrated photon object, including:

[0013] applying the simulation effect prediction model to optimize the multi-dimensional design parameter based on a group optimization algorithm to obtain an optimal search result of the multi-dimensional design parameter for optimizing the target function;

[0014] the optimal search result of the multi-dimensional design parameter is taken as an optimal design parameter of the target integrated photon object and imported into an integrated optoelectronic link simulation software for simulation to obtain a simulation effect index value corresponding to the optimal design parameter;

[0015] determining whether the simulation effect index value corresponding to the optimal design parameter meets a preset simulation effect expectation, and if so, outputting the optimal design parameter, otherwise, taking the optimal design parameter and the simulation effect index value corresponding to the optimal design parameter as a discarded historical simulation sample data.

[0016] In one possible design, a data collection algorithm is configured according to the sample data gap direction, and the historical simulation sample data is collected from all the discarded historical simulation sample data matching the target integrated photon object by using the data collection algorithm.

[0017] processing and analyzing the current sample data set to determine a sample data gap direction, wherein the sample data set contains a plurality of simulation sample data, and the simulation sample data contains simulation parameter values and simulation effect index values of the multi-dimensional design parameters as model input items and model output items;

[0018] configuring a data collection algorithm according to the sample data gap direction, and collecting the historical simulation sample data from all the discarded historical simulation sample data matching the target integrated photon object by using the data collection algorithm, wherein the historical simulation sample data contains historical simulation parameter values and historical simulation effect index values of the multi-dimensional design parameters as model input items and model output items;

[0019] determining whether the collected historical simulation sample data matches the sample data gap direction, and if so, adding the collected historical simulation sample data to the sample data set, otherwise, discarding the collected historical simulation sample data from being added to the sample data set.

[0020] In one possible design, a data collection algorithm is configured according to the sample data gap direction, and the historical simulation sample data is collected from all the discarded historical simulation sample data matching the target integrated photon object by using the data collection algorithm, including the following steps S221-S229:

[0021] S221. initializing an initial position corresponding to an initial parameter value of the multi-dimensional design parameter at a data gap periphery position according to the sample data gap direction, and then performing step S222, wherein the initial position and the data gap periphery position are respectively located in a multi-dimensional design parameter space constructed based on the multi-dimensional design parameter, and the data gap periphery position is determined according to the sample data gap direction;

[0022] S222. According to the initial parameter value of the multi-dimensional design parameter, a first historical simulation sample data matching the model input with the initial parameter value is searched from all historical simulation sample data discarded and matching the target integrated photon object by using a direct binary search algorithm, and then step S223 is executed, wherein the historical simulation sample data contains historical simulation parameter values and historical simulation effect index values of the multi-dimensional design parameter as model input and model output;

[0023] S223. A parameter serial number variable i is initialized as 1, and then step S224 is executed;

[0024] S224. Based on the current parameter value of the multi-dimensional design parameter, a parameter value of an i-th design parameter in the multi-dimensional design parameter is changed to obtain a new parameter value of the multi-dimensional design parameter, and then step S225 is executed;

[0025] S225. According to the new parameter value of the multi-dimensional design parameter, a second historical simulation sample data matching the model input with the new parameter value is searched from all the historical simulation sample data by using the direct binary search algorithm, and then step S226 is executed;

[0026] S226. Historical simulation effect index values as model output in the second historical simulation sample data and the first historical simulation sample data are compared, if the historical simulation effect index value in the second historical simulation sample data is better than the historical simulation effect index value in the first historical simulation sample data, step S227 is executed, otherwise step S228 is executed;

[0027] S227. The first historical simulation sample data is updated as the second historical simulation sample data, and if the parameter serial number variable i is less than the total dimension number of the multi-dimensional design parameter, the parameter serial number variable i is increased by 1, and then step S224 is executed, otherwise step S229 is executed;

[0028] S228. The parameter value of the i-th design parameter is restored to the value before the change, and if the parameter serial number variable i is less than the total dimension number of the multi-dimensional design parameter, the parameter serial number variable i is increased by 1, and then step S224 is executed, otherwise step S229 is executed;

[0029] S229. The first historical simulation sample data is taken as a collection result, and the collection is ended.

[0030] In one possible design, the current sample data set is processed and analyzed to determine the current sample data gap direction, including:

[0031] According to a definition domain of the simulation effect index value of the target integrated photon object, statistical analysis is performed on the current sample data set to determine a distribution interval of the simulation effect index value in the sample data set in the definition domain;

[0032] The distribution interval is eliminated in the definition domain to obtain a distribution missing interval in the definition domain;

[0033] According to the distribution missing interval, a current sample data gap direction is determined.

[0034] In one possible design, the artificial neural network adopts a residual network, wherein the number of layers of the residual network is 6-10 layers.

[0035] In a second aspect, an integrated photon prediction device is provided, which includes a configuration data acquisition unit, a sample data acquisition unit, a rate verification modeling unit, and a design parameter output unit.

[0036] The configuration data acquisition unit is configured to acquire configuration data for simulating the effect prediction or reverse structure prediction of a target integrated photon object, wherein the target integrated photon object includes an integrated photon device or an integrated optoelectronic system link to be designed, the configuration data contains design parameter configuration information of the target integrated photon object, and the design parameter configuration information contains multi-dimensional design parameters to be optimized.

[0037] The sample data acquisition unit is in communication connection with the configuration data acquisition unit and is configured to acquire historical simulation sample data from all historical simulation sample data that has been discarded and matched with the target integrated photon object by using a data acquisition algorithm, and add the acquisition result to a sample data set, wherein the historical simulation sample data contains historical simulation parameter values and historical simulation effect index values of the multi-dimensional design parameters used as model input items and model output items.

[0038] The rate verification modeling unit is in communication connection with the sample data acquisition unit and is configured to apply the sample data set to rate verification modeling of a machine learning model based on an artificial neural network to obtain a simulation effect prediction model for outputting corresponding simulation effect index values after inputting parameter values of the multi-dimensional design parameters or a reverse structure prediction model for outputting corresponding parameter values of the multi-dimensional design parameters after inputting simulation effect index values.

[0039] The design parameter output unit is in communication connection with the rate verification modeling unit and is configured to apply the simulation effect prediction model or the reverse structure prediction model to obtain and output design parameters of the target integrated photon object.

[0040] In a third aspect, the present application provides a computer device comprising a memory, a processor and a transceiver connected in sequence, wherein the memory is configured to store a computer program, the transceiver is configured to transmit and receive messages, and the processor is configured to read the computer program and execute the integrated photon prediction method according to any possible design of the first aspect.

[0041] In a fourth aspect, the present application provides a computer readable storage medium having instructions stored thereon, which, when executed on a computer, perform the integrated photon prediction method according to any possible design of the first aspect.

[0042] In a fifth aspect, the present application provides a computer program product comprising a computer program or instructions, which, when executed on a computer, implement the integrated photon prediction method according to any possible design of the first aspect.

[0043] Advantages of the above-mentioned scheme are as follows:

[0044] (1) The present application provides a new scheme for designing integrated photon based on historical invalid simulation data, that is, after obtaining configuration data for predicting the simulation effect of a target integrated photon object, sample data is collected from all historical simulation sample data that has been discarded and matched with the target object using a data collection algorithm, and the collection result is added to the sample data set, then the data set is applied to calibrate and verify the machine learning model based on artificial neural network to obtain a simulation effect prediction model or a reverse structure prediction model, finally the prediction model is applied to obtain the design parameters of the target integrated photon object and output them, thus the full use of all invalid simulation data generated in the design process can be achieved to save simulation design time for users to realize different functions under the same integrated photon structure;

[0045] (2) Through the use of invalid simulation data, users can further explore the functions and application scenarios of devices under the same integrated design structure;

[0046] (3) With the growth of data volume and the standardization of data sets, the boundaries between single devices (or systems) can be gradually smoothed by carefully designed data sets, thus realizing the simulation prediction of more devices and more systems, laying the foundation for supporting simulation prediction from single model to multi-topology structure, and facilitating practical application and promotion. BRIEF DESCRIPTION OF DRAWINGS

[0047] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or prior art description. Obviously, the drawings in the following description only constitute some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained from these drawings without creative effort.

[0048] Figure 1 The flowchart of the integrated photon prediction method provided by the embodiments of the present application.

[0049] Figure 2 The structural diagram of the integrated photon prediction device provided by the embodiments of the present application.

[0050] Figure 3 The structural diagram of the computer device provided by the embodiments of the present application. DETAILED DESCRIPTION

[0051] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or prior art description. Obviously, the drawings in the following description only constitute some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained from these drawings without creative effort. It should be noted that the description of these embodiment modes is used to help understand the present application, but does not constitute a limitation on the present application.

[0052] It should be understood that although the terms first and second, etc. may be used herein to describe various objects, these objects should not be limited by these terms. These terms are only used to distinguish one object from another. For example, the first object can be called the second object, and similarly, the second object can be called the first object, without departing from the scope of the example embodiments of the present application.

[0053] It should be understood that for the term "and / or" which may appear in the present application, it only describes the association relationship of the associated objects, which means that there can be three kinds of relationships, for example, A and / or B can mean that A exists alone, B exists alone or A and B exist together, etc. For example, A, B and / or C can mean that any one of A, B and C exists or any combination thereof; for the term " / and" which may appear in the present application, it describes another association object relationship, which means that there can be two kinds of relationships, for example, A / and B can mean that A exists alone or A and B exist together; in addition, for the character " / " which may appear in the present application, it generally means that the associated objects before and after are in an "or" relationship.

[0054] EMBODIMENTS

[0055] As Figure 1 shown in the first aspect, the integrated photonic prediction method can be executed by a computer device with certain computing resources, such as a cloud server, a personal computer (PC), a smart phone, a personal digital assistant (PDA), a wearable device, or the like. Figure 1 As Figure 1 shown, the integrated photonic prediction method can include the following steps S1-S4.

[0056] S1. Obtain configuration data for simulating the effect prediction or reverse structure prediction of the target integrated photonic object, wherein the target integrated photonic object includes but is not limited to an integrated photonic device or an integrated optoelectronic system link to be designed, and the configuration data includes but is not limited to design parameter configuration information of the target integrated photonic object, and the design parameter configuration information includes but is not limited to multi-dimensional design parameters to be optimized.

[0057] In the step S1, the integrated photonic device can be exemplified but not limited to a two-channel wavelength division multiplexing device of a photonic crystal-like structure, and the integrated optoelectronic system link can be exemplified but not limited to a four-channel wavelength division multiplexing link based on a micro-ring. The configuration data can be a default value or a specific value input by a person, so it can be obtained conventionally. The simulation effect prediction refers to predicting the simulation effect based on the design structure, and the reverse structure prediction refers to inversely predicting the design structure based on the simulation effect, which is completely opposite to the simulation effect prediction. In addition, the configuration data can also include but is not limited to a target function related to the simulation effect index of the target integrated photonic object, algorithm parameter configuration information such as population size, mutation rate, iteration number, classification operator, selection operator, crossover operator and / or mutation operator, and option configuration information such as recovery setting options (for example, if the recovery is selected, the recovery Json file needs to be selected) and / or record setting options (for example, record recovery file, optionally whether to configure; record recovery duration, default is 10 seconds; record parameters, other parameters are recorded together with the optimization parameters; record file, optionally whether to configure) and the like; the design parameter configuration information can also include but is not limited to design parameter dimension, parameter upper and lower boundary, parameter precision, equality constraint and / or inequality constraint and the like; and for example, if the target integrated photonic object is specifically a programmable photonic digital link composed of a plurality of programmable units, each programmable unit can work in the following three states through electrical control: straight-through state (i.e. the input light of the upper input port is output from the corresponding upper output port), cross state (also full coupling state, i.e. the input light of the upper input port is output from the lower output port) and partial coupling state (between the above two states, i.e. the proportion of output light from the output port changes with the change of coupling efficiency), then the multi-dimensional design parameter can be the state value / coupling efficiency of the plurality of programmable units.

[0058] S2. Collect the historical simulation sample data from all the historical simulation sample data that has been discarded and matched with the target integrated photonic object by using a data collection algorithm, and add the collection result to the sample data set, wherein the historical simulation sample data includes but is not limited to the historical simulation parameter value and the historical simulation effect index value of the multi-dimensional design parameter for the model input and the model output.

[0059] In the step S2, all the historical simulation sample data that have been discarded and matched with the target integrated photonic object are exemplified as follows: if the target integrated photonic object is a two-channel wavelength division multiplexing device of a photonic crystal structure, the historical simulation sample data must be sample data obtained by historical design and simulation for the two-channel wavelength division multiplexing device, and cannot be sample data obtained by historical design and simulation for other objects such as a four-channel wavelength division multiplexing link or a programmable photonic digital link. Considering different prediction directions, the historical simulation sample data will also be different as follows: if simulation effect prediction is required for the target integrated photonic object, the historical simulation sample data include but are not limited to historical simulation parameter values of the multi-dimensional design parameters used as model input items and historical simulation effect index values used as model output items; and if reverse structure prediction is required for the target integrated photonic object, the historical simulation sample data include but are not limited to historical simulation effect index values used as model input items and historical simulation parameter values of the multi-dimensional design parameters used as model output items. Generally speaking, the data collection algorithm can adopt a direct binary search method (DBS Method, also known as binary search algorithm, which is a method for efficiently searching for a specific element in an ordered array; its core idea is to locate the position of the target element by continuously narrowing the search range), but also considers that the data distribution for training a single model neural network at the present stage must meet certain distribution conditions, that is, the distribution of a certain result in all sample data must be relatively uniform, otherwise the trained model cannot realize the basic prediction function, therefore preferably, the data collection algorithm is used to collect the historical simulation sample data from all the historical simulation sample data that have been discarded and matched with the target integrated photonic object, and the collection result is added to the sample data set, including but not limited to the following steps S21-S23.

[0060] S21. Process and analyze the current sample data set to determine the current sample data gap direction, wherein the sample data set includes but is not limited to a plurality of simulation sample data, and the simulation sample data include but are not limited to simulation parameter values and simulation effect index values of the multi-dimensional design parameters used as model input items and model output items.

[0061] In step S21, the current sample dataset can be a manually imported dataset or a dataset obtained from historical data acquisition using the aforementioned data acquisition algorithm. The simulation effect index value is the simulation result corresponding to the simulation parameter values ​​of the multidimensional design parameters. Examples, but not limited to, are specific parameter values ​​for transmission efficiency or scattering parameters (which specifically include the following four parameters: S11—reflection coefficient of port 1, representing the reflection of the input signal at port 1; S21—forward transmission coefficient, representing the signal gain from port 1 to port 2; S12—reverse transmission coefficient, representing the reverse transmission from port 2 to port 1; S22—reflection coefficient of port 2, representing the reflection of the input signal at port 2). Specifically, the current sample dataset is processed and analyzed to determine the direction of the current sample data gap, including but not limited to the following steps S211 to S213.

[0062] S211. Based on the domain of the simulation effect index value of the target integrated photonic object, perform statistical analysis on the current sample dataset to determine the distribution range of the simulation effect index value in the domain within the sample dataset.

[0063] In step S211, for example, if the simulation effect index value is a scattering parameter value, then the domain can be (0,1). Then, through conventional statistical analysis, the distribution interval of the simulation effect index value in the domain can be determined to be (0,0.5) and (0.6,1).

[0064] S212. Remove the distribution interval from the defined domain to obtain the distribution missing interval in the defined domain.

[0065] In step S212, based on the specific example of step S211 above, the missing interval of the distribution can be obtained as (0.5, 0.6).

[0066] S213. Determine the direction of the current sample data gap based on the missing distribution interval.

[0067] S22. Configure a data acquisition algorithm according to the direction of the gap in the sample data, and use the data acquisition algorithm to acquire the historical simulation sample data from all historical simulation sample data that have been discarded and match the target integrated photonic object. The historical simulation sample data includes, but is not limited to, historical simulation parameter values ​​and historical simulation effect index values ​​of the multidimensional design parameters that are used as model inputs and model outputs.

[0068] In the step S22, specifically, a data collection algorithm is configured according to the sample data gap direction, and the data collection algorithm is used to collect the historical simulation sample data from all the historical simulation sample data that has been discarded and matched with the target integrated photon object, including the following steps S221-S229.

[0069] S221. An initial position corresponding to an initial parameter value of the multi-dimensional design parameter is initialized at a data gap periphery position according to the sample data gap direction, and then the step S222 is performed, wherein the initial position and the data gap periphery position are respectively located in a multi-dimensional design parameter space constructed based on the multi-dimensional design parameter, and the data gap periphery position is determined according to the sample data gap direction.

[0070] In the step S221, assuming that the multi-dimensional design parameter space is a three-dimensional space, and the sample data gap direction is a direction pointing to a coordinate position in the three-dimensional space, the multi-dimensional design parameter value closest to the coordinate position in the current sample data set can be taken as the data gap periphery position and the initial position.

[0071] S222. According to the initial parameter value of the multi-dimensional design parameter, a first historical simulation sample data in which a model input item matches the initial parameter value is searched from all the historical simulation sample data that has been discarded and matched with the target integrated photon object by using a direct binary search algorithm, and then the step S223 is performed, wherein the historical simulation sample data contains historical simulation parameter values and historical simulation effect index values of the multi-dimensional design parameter for model input items and model output items.

[0072] S223. A parameter serial number variable i is initialized to 1, and then the step S224 is performed.

[0073] S224. Based on a current parameter value of the multi-dimensional design parameter, a parameter value of an i-th design parameter in the multi-dimensional design parameter is changed to obtain a new parameter value of the multi-dimensional design parameter, and then the step S225 is performed.

[0074] S225. According to the new parameter value of the multi-dimensional design parameter, a second historical simulation sample data in which a model input item matches the new parameter value is searched from all the historical simulation sample data by using the direct binary search algorithm, and then the step S226 is performed.

[0075] S226. Compare the historical simulation performance indicator values in the second historical simulation sample data and the first historical simulation sample data as model output items, if the historical simulation performance indicator values in the second historical simulation sample data are better than those in the first historical simulation sample data, execute step S227, otherwise execute step S228.

[0076] S227. Update the first historical simulation sample data as the second historical simulation sample data, and if the parameter serial number variable i is less than the total number of dimensions of the multi-dimensional design parameters, make the parameter serial number variable i plus 1, and then return to execute step S224, otherwise execute step S229.

[0077] S228. Restore the parameter value of the i-th design parameter to the value before the change, and if the parameter serial number variable i is less than the total number of dimensions of the multi-dimensional design parameters, make the parameter serial number variable i plus 1, and then return to execute step S224, otherwise execute step S229.

[0078] S229. Take the first historical simulation sample data as the collection result, and end this collection.

[0079] S23. Determine whether the collected historical simulation sample data matches the sample data gap direction, if yes, add the collected historical simulation sample data to the sample data set, otherwise discard adding the collected historical simulation sample data to the sample data set.

[0080] In the step S23, determining whether the collected historical simulation sample data matches the sample data gap direction can include, but is not limited to: according to the definition domain of the simulation performance indicator value of the target integrated photon object, performing statistical analysis on the current sample data set and the collected historical simulation sample data again, determining a new distribution interval of the simulation performance indicator value in the definition domain, and then if the new distribution interval is expanded (at this time it means that the distribution missing interval is reduced), it can be determined that the collected historical simulation sample data matches the sample data gap direction, otherwise it is determined that the collected historical simulation sample data does not match the sample data gap direction.

[0081] S3. Apply the sample data set to calibrate and verify the machine learning model based on artificial neural network modeling to obtain a simulation performance prediction model for outputting corresponding simulation performance indicator values after inputting parameter values of the multi-dimensional design parameters, or a reverse structure prediction model for outputting corresponding parameter values of the multi-dimensional design parameters after inputting simulation performance indicator values.

[0082] In the step S3, the artificial neural network (ANN) is a mathematical model simulating the structure and function of biological neural networks for information processing and pattern recognition; the artificial neural network is composed of a large number of nodes (or called "neurons") connected to each other, each node represents a specific output function, called activation function; the connection between each two nodes represents a weighted value, called weight, which is equivalent to the memory of the artificial neural network; the output of the network depends on the connection mode, weight and activation function of the network. Therefore, based on a certain amount of historical simulation sample data, the simulation effect prediction model or the reverse structure prediction model can be trained through a conventional calibration verification modeling process (specifically including a model calibration process and a checking process, that is, a process of adjusting model parameters to make the simulation results consistent with the actual results by comparing the simulation results with the measured data). Specifically, the artificial neural network can but not limited to use a residual network, wherein the number of layers of the residual network is 6-10 layers, for example, 8 layers. In addition, the sample data set can be specifically divided into training set, validation set and test set, etc. in the calibration verification modeling process, and randomly selected according to the ratio of 3:1:1.

[0083] S4. Applying the simulation effect prediction model or the reverse structure prediction model to obtain the design parameters of the target integrated photonic object and output them.

[0084] In the step S4, since the model input of the reverse structure prediction model is the simulation effect index value, the target effect index value of the target integrated photonic object can be directly input into the reverse structure prediction model, and then the design parameters of the target integrated photonic object are directly obtained and output. Since the model output of the simulation effect prediction model is the simulation effect index value, the design parameters of the target integrated photonic object need to be continuously adjusted and imported into the simulation effect prediction model until the output simulation effect index value reaches the target effect index value, so that the model input data at this time can be used as the design parameters of the target integrated photonic object and output. For the latter, in order to automatically optimize the optimal design parameters of the target integrated photonic object, preferably, when the design parameter configuration information further contains a target function related to the simulation effect index of the target integrated photonic object, the simulation effect prediction model is applied to obtain the design parameters of the target integrated photonic object, including but not limited to the following steps S41-S43.

[0085] S41. Apply the simulation effect prediction model, and optimize the multi-dimensional design parameters based on a swarm optimization algorithm to obtain an optimal search result of the multi-dimensional design parameters and for optimizing the target function.

[0086] In the step S41, the swarm optimization algorithm can be, but is not limited to, a particle swarm optimization algorithm, a genetic algorithm, or a grey wolf algorithm, and a specific optimization process can be derived according to a conventional algorithm.

[0087] S42. The optimal search result of the multi-dimensional design parameters is taken as an optimal design parameter of the target integrated photon object and is imported into an integrated optoelectronic link simulation software for simulation to obtain a simulation effect index value corresponding to the optimal design parameter.

[0088] In the step S42, the integrated optoelectronic link simulation software is an existing software, for example, a pSim Plus software.

[0089] S43. It is judged whether the simulation effect index value corresponding to the optimal design parameter meets a preset simulation effect expectation, if yes, the optimal design parameter is output, otherwise, the optimal design parameter and the simulation effect index value corresponding to the optimal design parameter are taken as a discarded historical simulation sample data.

[0090] In the step S43, the simulation effect expectation can be specifically an example of exceeding a target effect index value of the target integrated photon object. In addition, after the optimal design parameter and the simulation effect index value corresponding to the optimal design parameter are discarded, they can be added to the sample data set through the foregoing step S2 so as to be utilized again.

[0091] Therefore, the integrated photon prediction method described in the foregoing steps S1-S4 provides a new scheme of integrated photon design based on historical invalid simulation data, that is, after configuration data used for simulation effect prediction of a target integrated photon object is obtained, sample data is collected from all historical simulation sample data that has been discarded and matched with the target object by using a data collection algorithm, and the collection result is added to the sample data set, then a machine learning model based on an artificial neural network is calibrated and verified modeling by using the data set to obtain a simulation effect prediction model or a reverse structure prediction model, finally, a prediction model is applied to obtain a design parameter of the target integrated photon object and output the design parameter, so that all invalid simulation data generated in the design process can be fully utilized to save simulation design time for users to realize different functions under the same integrated photon structure, and the method is convenient for practical application and promotion.

[0092] As Figure 2As shown, the second aspect of the embodiment provides a virtual device for implementing the integrated photon prediction method of the first aspect, including a configuration data acquisition unit, a sample data acquisition unit, a rate verification modeling unit and a design parameter output unit;

[0093] The configuration data acquisition unit is configured to acquire configuration data for simulating effect prediction or reverse structure prediction of a target integrated photon object, wherein the target integrated photon object includes an integrated photon device or an integrated optoelectronic system link to be designed, and the configuration data contains design parameter configuration information of the target integrated photon object, and the design parameter configuration information contains multi-dimensional design parameters to be optimized;

[0094] The sample data acquisition unit is in communication connection with the configuration data acquisition unit, configured to collect historical simulation sample data from all historical simulation sample data that has been discarded and matched with the target integrated photon object by using a data collection algorithm, and add the collection result to a sample data set, wherein the historical simulation sample data contains historical simulation parameter values and historical simulation effect index values of the multi-dimensional design parameters as model input items and model output items;

[0095] The rate verification modeling unit is in communication connection with the sample data acquisition unit, configured to apply the sample data set to rate verification modeling of a machine learning model based on an artificial neural network, to obtain a simulation effect prediction model for outputting corresponding simulation effect index values after inputting parameter values of the multi-dimensional design parameters, or a reverse structure prediction model for outputting corresponding parameter values of the multi-dimensional design parameters after inputting simulation effect index values;

[0096] The design parameter output unit is in communication connection with the rate verification modeling unit, configured to apply the simulation effect prediction model or the reverse structure prediction model to obtain and output design parameters of the target integrated photon object.

[0097] The working process, working details and technical effects of the foregoing device provided by the second aspect of the embodiment can be referred to the integrated photon prediction method of the first aspect, which will not be described here.

[0098] As Figure 3As shown, the third aspect of the embodiment provides a computer device for performing the integrated photonic prediction method according to the first aspect, which comprises a memory, a processor and a transceiver connected in sequence, wherein the memory is configured to store a computer program, the transceiver is configured to transmit and receive messages, and the processor is configured to read the computer program and perform the integrated photonic prediction method according to the first aspect. Specifically, the memory can include, but is not limited to, a random access memory (RAM), a read-only memory (ROM), a flash memory, a first-in first-out memory (FIFO) and / or a first-in last-out memory (FILO), etc.; and the processor can be, but is not limited to, a microprocessor of STM32F105 series. In addition, the computer device can further include, but is not limited to, a power module, a display screen and other necessary components.

[0099] The working process, working details and technical effects of the aforementioned computer device provided by the third aspect of the embodiment can be referred to the integrated photonic prediction method according to the first aspect, which will not be described here.

[0100] The fourth aspect of the embodiment provides a computer readable storage medium storing instructions of the integrated photonic prediction method according to the first aspect, i.e., the computer readable storage medium stores instructions, and when the instructions are executed on a computer, the integrated photonic prediction method according to the first aspect is performed. The computer readable storage medium refers to a carrier for storing data, which can include, but is not limited to, floppy disks, optical disks, hard disks, flash memories, USB flash drives and / or memory sticks, etc. The computer can be a general-purpose computer, a special-purpose computer, a computer network or other programmable devices.

[0101] The working process, working details and technical effects of the aforementioned computer readable storage medium provided by the fourth aspect of the embodiment can be referred to the integrated photonic prediction method according to the first aspect, which will not be described here.

[0102] The fifth aspect of the embodiment provides a computer program product comprising a computer program or instructions, which, when executed by a computer, implement the integrated photonic prediction method according to the first aspect. The computer can be a general-purpose computer, a special-purpose computer, a computer network or other programmable devices.

[0103] Finally, it should be noted that the above description is only the preferred embodiment of the present application, and is not intended to limit the protection scope of the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. An integrated photon prediction method, characterized in that, include: The configuration data is used to predict the simulation effect or reverse structure of a target integrated photonic object. The target integrated photonic object includes an integrated photonic device or integrated optoelectronic system link to be designed. The configuration data package contains the design parameter configuration information of the target integrated photonic object. The design parameter configuration information includes multi-dimensional design parameters to be optimized. A data acquisition algorithm is used to collect historical simulation sample data from all historical simulation sample data that have been discarded and match the target integrated photon object, and the acquisition results are added to the sample dataset. The historical simulation sample data contains historical simulation parameter values ​​and historical simulation effect index values ​​of the multidimensional design parameters that are mutually model input and model output. Using the sample dataset, the neural network-based machine learning model is calibrated and validated to obtain a simulation effect prediction model that outputs the corresponding simulation effect index value after inputting the parameter values ​​of the multidimensional design parameters, or a reverse structure prediction model that outputs the corresponding parameter values ​​of the multidimensional design parameters after inputting the simulation effect index value. The design parameters of the target integrated photonic object are obtained and output by applying the simulation effect prediction model or the reverse structure prediction model. Specifically, this includes: inputting the target effect index value of the target integrated photonic object into the reverse structure prediction model to obtain and output the design parameters of the target integrated photonic object; continuously adjusting the design parameters of the target integrated photonic object and importing them into the simulation effect prediction model until the output simulation effect index value reaches the target effect index value; and using the model input data at this time as the design parameters of the target integrated photonic object and outputting them.

2. The integrated photon prediction method according to claim 1, characterized in that, When the design parameter configuration information also includes an objective function related to the simulation effect index of the target integrated photonic object, the simulation effect prediction model is applied to obtain the design parameters of the target integrated photonic object, including: By applying the simulation effect prediction model, the multidimensional design parameters are optimized based on the swarm optimization algorithm to obtain the multidimensional design parameters and the optimal search results used to make the objective function optimal. The optimization search results of the multidimensional design parameters are used as the optimal design parameters of the target integrated photonic object and imported into the integrated optoelectronic link simulation software for simulation to obtain the simulation effect index value corresponding to the optimal design parameters. Determine whether the simulation effect index value corresponding to the optimal design parameter meets the preset simulation effect expectation. If so, output the optimal design parameter; otherwise, treat the optimal design parameter and the simulation effect index value corresponding to the optimal design parameter as a discarded historical simulation sample data.

3. The integrated photon prediction method according to claim 1, characterized in that, A data acquisition algorithm is used to collect historical simulation sample data from all discarded historical simulation sample data that match the target integrated photon object, and the acquisition results are added to the sample dataset, including: The current sample dataset is processed and analyzed to determine the direction of the current sample data gap. The sample dataset contains several simulation sample data. The simulation sample data contains simulation parameter values ​​and simulation effect index values ​​of the multidimensional design parameters that serve as model inputs and outputs. The data acquisition algorithm is configured according to the direction of the gap in the sample data, and the data acquisition algorithm is used to acquire the historical simulation sample data from all historical simulation sample data that have been discarded and match the target integrated photon object. The historical simulation sample data contains historical simulation parameter values ​​and historical simulation effect index values ​​of the multidimensional design parameters that are mutually model input and model output. Determine whether the collected historical simulation sample data matches the direction of the sample data gap. If yes, add the collected historical simulation sample data to the sample dataset; otherwise, abandon the addition of the collected historical simulation sample data to the sample dataset.

4. The integrated photon prediction method according to claim 3, characterized in that, Configure a data acquisition algorithm according to the direction of the gap in the sample data, and use the data acquisition algorithm to acquire the historical simulation sample data from all historical simulation sample data that have been discarded and match the target integrated photonic object, including the following steps S221 to S229: S221. Based on the direction of the sample data gap, initialize the initial position corresponding to the initial parameter value of the multidimensional design parameter at the position around the data gap, and then execute step S222, wherein the initial position and the position around the data gap are respectively located in the multidimensional design parameter space constructed based on the multidimensional design parameter, and the position around the data gap is determined according to the direction of the sample data gap; S222. Based on the initial parameter values ​​of the multidimensional design parameters, a direct binary search algorithm is used to search for the first historical simulation sample data that matches the initial parameter values ​​from all historical simulation sample data that have been discarded and match the target integrated photonic object. Then, step S223 is executed, wherein the historical simulation sample data contains historical simulation parameter values ​​and historical simulation effect index values ​​of the multidimensional design parameters that are mutually model input and model output items. S223. Change the parameter number variable Initialize to 1, then execute step S224; S224. Based on the current parameter value of the multidimensional design parameters, change the first parameter among the multidimensional design parameters. The parameter values ​​of each design parameter are obtained to obtain new parameter values ​​for the multidimensional design parameters, and then step S225 is executed; S225. Based on the new parameter values ​​of the multidimensional design parameters, the direct binary search algorithm is used to search for the second historical simulation sample data that matches the new parameter values ​​from all the historical simulation sample data, and then step S226 is executed. S226. Compare the historical simulation effect index values ​​in the second historical simulation sample data and the first historical simulation sample data that are used as model output items. If the historical simulation effect index value in the second historical simulation sample data is better than the historical simulation effect index value in the first historical simulation sample data, then proceed to step S227; otherwise, proceed to step S228. S227. Update the first historical simulation sample data to the second historical simulation sample data, and if the parameter sequence variable... If the number of dimensions is less than the total number of dimensions of the multidimensional design parameters, then the parameter index variable is set to... Increment by 1, then return to step S224; otherwise, proceed to step S229. S228. The first The parameter values ​​of each design parameter are restored to their original values, and if the parameter number variable... If the number of dimensions is less than the total number of dimensions of the multidimensional design parameters, then the parameter index variable is set to... Increment by 1, then return to step S224; otherwise, proceed to step S229. S229. Take the first historical simulation sample data as the acquisition result and end this acquisition.

5. The integrated photon prediction method according to claim 3, characterized in that, The current sample dataset is processed and analyzed to determine the direction of the current sample data gap, including: Based on the domain of the simulation effect index value of the target integrated photonic object, statistical analysis is performed on the current sample dataset to determine the distribution range of the simulation effect index value in the sample dataset within the domain. By removing the distribution interval from the defined domain, the distribution missing interval in the defined domain is obtained; Based on the missing intervals in the distribution, determine the direction of the current sample data gap.

6. The integrated photon prediction method according to claim 1, characterized in that, The neural network employs a residual network, wherein the residual network has 6 to 10 layers.

7. An integrated photon prediction device, characterized in that, It includes a configuration data acquisition unit, a sample data acquisition unit, a calibration and verification modeling unit, and a design parameter output unit; The configuration data acquisition unit is used to acquire configuration data for simulation effect prediction or reverse structure prediction of the target integrated photonic object. The target integrated photonic object includes an integrated photonic device or integrated optoelectronic system link to be designed. The configuration data package contains design parameter configuration information of the target integrated photonic object. The design parameter configuration information includes multi-dimensional design parameters to be optimized. The sample data acquisition unit is communicatively connected to the configuration data acquisition unit. It is used to acquire historical simulation sample data from all historical simulation sample data that have been discarded and match the target integrated photon object using a data acquisition algorithm, and add the acquisition results to the sample dataset. The historical simulation sample data package contains historical simulation parameter values ​​and historical simulation effect index values ​​of the multidimensional design parameters that are mutually model input and model output items. The calibration and verification modeling unit is communicatively connected to the sample data acquisition unit. It is used to apply the sample dataset to calibrate and verify the machine learning model based on the neural network, and to obtain a simulation effect prediction model that outputs the corresponding simulation effect index value after inputting the parameter value of the multidimensional design parameter or a reverse structure prediction model that outputs the corresponding parameter value of the multidimensional design parameter after inputting the simulation effect index value. The design parameter output unit is communicatively connected to the calibration and verification modeling unit. It is used to apply the simulation effect prediction model or the reverse structure prediction model to obtain and output the design parameters of the target integrated photonic object. Specifically, it includes: inputting the target effect index value of the target integrated photonic object into the reverse structure prediction model to obtain and output the design parameters of the target integrated photonic object; continuously adjusting the design parameters of the target integrated photonic object and importing them into the simulation effect prediction model until the output simulation effect index value reaches the target effect index value, and using the model input data at this time as the design parameters of the target integrated photonic object and outputting them.

8. A computer device, characterized in that, The device includes a memory, a processor, and a transceiver that are sequentially and communicatively connected, wherein the memory is used to store a computer program, the transceiver is used to send and receive messages, and the processor is used to read the computer program and execute the integrated photon prediction method as described in any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that... The computer-readable storage medium stores instructions that, when executed on a computer, perform the integrated photon prediction method as described in any one of claims 1 to 6.

10. A computer program product, comprising a computer program or instructions, characterized in that, When the computer program or the instructions are executed by a computer, they implement the integrated photon prediction method as described in any one of claims 1 to 6.

Images