System program model calibration method and device based on optimization algorithm

Abstract

The invention provides a system program model calibration method and device based on an optimization algorithm, and relates to the technical field of computers. The method comprises the steps of establishing a system program model; a card reading program reads an input card and records and arranges parameter information; judging whether a parameter optimization instruction exists or not; automatic calibration of the system program model is achieved through result evaluation, key component parameter selection and optimization algorithm parameter optimization, and finally the calculation accuracy of the system analysis program model (called the system program model for short) is improved.

Description

Technical Field

[0001] The present application relates to the field of computer technology, and in particular to a method and device for calibrating a system program model based on an optimization algorithm, an electronic device, and a storage medium. Background Art

[0002] Reactor accident reproduction requires the use of system analysis programs to accurately model the system, and the quality of modeling determines the accuracy of simulation. Using existing data to correct the modeling of the system is one of the ideas for optimizing modeling. Nuclear reactors need to optimize parameters in various aspects such as design and simulation. For example, in core design, parameter optimization technology is needed to optimize the core layout plan, refueling plan and system layout; in reactor simulation calculations, parameter optimization technology can be used to optimize the modeling method so that the program can obtain the most accurate solution possible. This application is similar to parameter sensitivity analysis; in addition, the use of parameter optimization technology can also improve the correlation coefficient of the empirical relationship used in the existing program to improve its prediction accuracy for related phenomena.

[0003] Parameter optimization is a method to achieve design goals. By parameterizing the design goals and using optimization methods, the design variables are continuously adjusted so that the design results continue to approach the parameterized target values. Summary of the invention

[0004] In view of the above problems, the present application is proposed to provide a method and device for calibrating a system program model based on an optimization algorithm, an electronic device and a storage medium to overcome the above problems or at least partially solve the above problems, and to realize automatic calibration of the system program model through result evaluation, key component parameter selection, and optimization algorithm optimization parameters, thereby ultimately improving the calculation accuracy of the system analysis program model (referred to as the system program model). The technical solution is as follows:

[0005] In a first aspect, a system program model calibration method based on an optimization algorithm is provided, the method comprising:

[0006] Step 1: Establish system program model; proceed to step 2;

[0007] Step 2: The card reader reads the input card and records and organizes the parameter information; then proceed to step 3;

[0008] Step 3: Determine whether there is a parameter optimization instruction;

[0009] If not, proceed to step seven;

[0010] If yes, proceed to step 4;

[0011] Step 4: Select key component parameters; proceed to step 5;

[0012] Step 5: Use the parameter optimization program and select the optimization algorithm to optimize the parameters; proceed to step 6;

[0013] Step 6: The optimized parameters are automatically written into the input card; proceed to step 7;

[0014] Step 7: System program runs; if there is no parameter optimization instruction, go to step 10; if there is a parameter optimization instruction, go to step 8;

[0015] Step 8: Determine whether the system program reports an error and has no calculation results;

[0016] If not, proceed to step 10;

[0017] If yes, go to step nine;

[0018] Step 9: The parameter combination that cannot converge is given a low fitness, and the optimized parameters are optimized continuously, and then return to step 5;

[0019] Step 10: Read the output file;

[0020] Step 11: Result evaluation;

[0021] Conduct result evaluation based on the predicted data and experimental data or design data in the output file;

[0022] Step 12: Determine whether the acceptance criteria are met;

[0023] If not, proceed to step 13;

[0024] If yes, proceed to step 14;

[0025] Step 13: Based on the evaluation results, form parameter optimization instructions and give parameter optimization directions; return to step 3;

[0026] Step 14: End of analysis.

[0027] In one possible implementation, when performing result evaluation in steps 11 to 14, firstly, a steady-state parameter optimization calibration analysis is performed. After the steady-state acceptance criteria are met, a transient parameter optimization calibration analysis is then performed. After the transient acceptance criteria are met, steady-state and transient parameter optimization calibration analyses are then performed iteratively in sequence until the parameter optimization meets both steady-state and transient acceptance criteria, and the analysis is finally terminated.

[0028] In a possible implementation, the method of evaluating the results of steps 11 to 14 is as follows:

[0029] Read the output file to obtain the predicted data, and then compare it with the design data or experimental data to determine whether the data time series are of equal length;

[0030] If the data time series are of equal length, the data are fast Fourier transformed to obtain the average amplitude and weighted frequency, and then it is determined whether the acceptance criteria are met;

[0031] If the data time series are not of equal length, the data are dynamically time-warped to obtain the Euclidean distance, and then it is determined whether the acceptable standard is met;

[0032] If the acceptance criteria are met, the analysis ends; if the acceptance criteria are not met, parameter optimization instructions are formed and the parameter optimization direction is given.

[0033] In a possible implementation, the key component parameters are selected in step 4, specifically including:

[0034] Before parameter optimization, component parameters are selected, and the principal component analysis method is used to reduce the dimension of component parameter data involved in the calculation object to select key component parameters.

[0035] In a possible implementation, in step 5, the optimization algorithm optimization parameters are selected, specifically including:

[0036] The machine learning optimization algorithm based on neural network finds the predicted value that satisfies the objective function, and reversely infers the corresponding input parameters based on the predicted value to realize automatic adjustment of the modeling parameters. At the same time, it combines the simulated annealing algorithm or genetic algorithm, uses the model trained by the neural network as the prediction formula, optimizes and solves the objective function, obtains the optimal solution of the objective function globally, and then obtains the optimal modeling parameters of the system program.

[0037] In a possible implementation, the parameter combination that cannot converge in step nine is assigned a low fitness, and the optimized parameters are further optimized, specifically including:

[0038] The optimized parameter combination causes the system program calculation to be unable to converge. By utilizing the characteristics of the genetic algorithm, a low fitness is assigned to the parameter combination that cannot converge. For the optimized parameters, the genetic algorithm is used again to re-optimize the parameters.

[0039] In a possible implementation, a parameter optimization direction is given, including:

[0040] When fast Fourier transform is selected, the optimization direction is to minimize the average amplitude; when dynamic time warping is selected, the optimization direction is to minimize the distance between the predicted data and the design data or experimental data.

[0041] In a second aspect, a system program model calibration device based on an optimization algorithm is provided, the device comprising:

[0042] Establishing unit, used to carry out step 1: establishing system program model; entering step 2 of calibration unit;

[0043] The calibration unit is used to perform step 2: the card reader program reads the input card, records and organizes the parameter information; and proceeds to step 3;

[0044] Step 3: Determine whether there is a parameter optimization instruction;

[0045] If not, proceed to step seven;

[0046] If yes, proceed to step 4;

[0047] Step 4: Select key component parameters; proceed to step 5;

[0048] Step 5: Use the parameter optimization program and select the optimization algorithm to optimize the parameters; proceed to step 6;

[0049] Step 6: The optimized parameters are automatically written into the input card; proceed to step 7;

[0050] Step 7: System program runs; if there is no parameter optimization instruction, go to step 10; if there is a parameter optimization instruction, go to step 8;

[0051] Step 8: Determine whether the system program reports an error and has no calculation results;

[0052] If not, proceed to step 10;

[0053] If yes, go to step nine;

[0054] Step 9: The parameter combination that cannot converge is given a low fitness, and the optimized parameters are optimized continuously, and then return to step 5;

[0055] Step 10: Read the output file;

[0056] Step 11: Result evaluation;

[0057] Conduct result evaluation based on the predicted data and experimental data or design data in the output file;

[0058] Step 12: Determine whether the acceptance criteria are met;

[0059] If not, proceed to step 13;

[0060] If yes, proceed to step 14;

[0061] Step 13: Based on the evaluation results, form parameter optimization instructions and give parameter optimization directions; return to step 3;

[0062] Step 14: End of analysis.

[0063] According to a third aspect, an electronic device is provided, comprising a processor and a memory, wherein a computer program is stored in the memory, and the processor is configured to run the computer program to execute any one of the above-mentioned system program model calibration methods based on the optimization algorithm.

[0064] In a fourth aspect, a storage medium is provided, wherein the storage medium stores a computer program, wherein the computer program is configured to execute any of the above-mentioned system program model calibration methods based on the optimization algorithm when running.

[0065] Through the above-mentioned technical scheme, the embodiments of the present application provide a system program model calibration method and device based on an optimization algorithm, an electronic device and a storage medium. The method realizes automatic calibration of the system program model through result evaluation, key component parameter selection, and optimization algorithm optimization parameters, thereby ultimately improving the calculation accuracy of the system analysis program model (referred to as the system program model).

[0066] Furthermore, in this embodiment, the result evaluation compares the predicted data with the design data or the experimental data, and gives the direction of parameter optimization; the key component parameters are selected by the principal component analysis method to reduce the data dimension; the optimization algorithm is based on machine learning optimization algorithms such as neural networks, combined with simulated annealing algorithms or genetic algorithms, to optimize and solve the objective function; finally, through the iterative calculation of result evaluation, key component parameter selection, and optimization algorithm optimization parameters, the system program model parameters are automatically calibrated in steady-state and transient calculations, thereby ultimately improving the calculation accuracy of the system analysis program model. BRIEF DESCRIPTION OF THE DRAWINGS

[0067] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in describing the embodiments of the present application are briefly introduced below.

[0068] Figure 1 A flow chart of a system program model calibration method based on an optimization algorithm provided in an embodiment of the present application is shown;

[0069] Figure 2 It is a schematic diagram of the result evaluation process provided in the embodiment of the present application;

[0070] Figure 3 is a flow chart of a simulated annealing algorithm provided in an embodiment of the present application;

[0071] Figure 4 is a flow chart of a genetic algorithm provided in an embodiment of the present application;

[0072] Figure 5 A structural diagram of a system program model calibration device based on an optimization algorithm provided in an embodiment of the present application is shown. DETAILED DESCRIPTION

[0073] The exemplary embodiments of the present application will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present application are shown in the accompanying drawings, it should be understood that the present application can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided in order to enable a more thorough understanding of the present application and to fully convey the scope of the present application to those skilled in the art.

[0074] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that such use is interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the term "including" and its variants are to be interpreted as open-ended terms meaning "including but not limited to".

[0075] Embodiment 1:

[0076] In order to solve the above technical problems, the embodiment of the present application provides a system program model calibration method based on an optimization algorithm, such as Figure 1 As shown, the system program model calibration method based on the optimization algorithm may include the following steps:

[0077] Step 1: Establish system program model; proceed to step 2;

[0078] Step 2: The card reader reads the input card and records and organizes the parameter information; then proceed to step 3;

[0079] Step 3: Determine whether there is a parameter optimization instruction;

[0080] If not, proceed to step seven;

[0081] If yes, proceed to step 4;

[0082] Step 4: Select key component parameters; proceed to step 5;

[0083] Step 5: Use the parameter optimization program and select the optimization algorithm to optimize the parameters; proceed to step 6;

[0084] Step 6: The optimized parameters are automatically written into the input card; proceed to step 7;

[0085] Step 7: System program runs; if there is no parameter optimization instruction, go to step 10; if there is a parameter optimization instruction, go to step 8;

[0086] Step 8: Determine whether the system program reports an error and has no calculation results;

[0087] If not, proceed to step 10;

[0088] If yes, go to step nine;

[0089] Step 9: The parameter combination that cannot converge is given a low fitness, and the optimized parameters are optimized continuously, and then return to step 5;

[0090] Step 10: Read the output file;

[0091] Step 11: Result evaluation;

[0092] Conduct result evaluation based on the predicted data and experimental data or design data in the output file;

[0093] Step 12: Determine whether the acceptance criteria are met;

[0094] If not, proceed to step 13;

[0095] If yes, proceed to step 14;

[0096] Step 13: Based on the evaluation results, form parameter optimization instructions and give parameter optimization directions; return to step 3;

[0097] Step 14: End of analysis.

[0098] This embodiment realizes automatic calibration of the system program model through result evaluation, key component parameter selection, and optimization algorithm optimization parameters, and ultimately improves the calculation accuracy of the system analysis program model (referred to as the system program model).

[0099] The system program model established in the above step 1 may be a system program model such as RELAP (Reactor Excursion and Leak Analysis Program), which is not limited in this embodiment.

[0100] In the above step 2, the card reader program reads the input card, which can be read by a card reader program such as Python (a widely used high-level programming language that supports multiple programming paradigms, including object-oriented, imperative, functional and procedural programming), and this embodiment does not limit this.

[0101] The parameter optimization program used in the above step 5 may be a parameter optimization program such as GEATPY (Genetic and Evolutionary Algorithm Toolbox for Python), which is not limited in this embodiment.

[0102] Embodiment 2:

[0103] A possible implementation method is provided in an embodiment of the present application. When performing result evaluation in steps eleven to fourteen, a steady-state parameter optimization calibration analysis is first performed. After the steady-state acceptance standard is met, a transient parameter optimization calibration analysis is then performed. After the transient acceptance standard is met, steady-state and transient parameter optimization calibration analyses are iteratively performed in sequence until the parameter optimization meets both steady-state and transient acceptance standards, and the analysis is finally terminated.

[0104] Embodiment 3:

[0105] A possible implementation method is provided in the embodiment of the present application. Figure 2 This is a schematic diagram of the result evaluation process provided by the embodiment of the present application, that is, the method of result evaluation carried out in steps 11 to 14 above is specifically as follows:

[0106] Read the output file to obtain the predicted data, and then compare it with the design data or experimental data to determine whether the data time series are of equal length;

[0107] If the data time series are of equal length, the data are fast Fourier transformed to obtain the average amplitude and weighted frequency, and then it is determined whether the acceptance criteria are met;

[0108] If the data time series are not of equal length, the data are dynamically time-warped to obtain the Euclidean distance, and then it is determined whether the acceptable standard is met;

[0109] If the acceptance criteria are met, the analysis ends; if the acceptance criteria are not met, parameter optimization instructions are formed and the parameter optimization direction is given.

[0110] Here, when the parameter optimization direction is given, when the fast Fourier transform is selected, the optimization direction is to minimize the average amplitude; when the dynamic time warping is selected, the optimization direction is to minimize the distance between the predicted data and the design data or experimental data.

[0111] This embodiment compares the predicted data with the design data or the experimental data through result evaluation, forms parameter optimization instructions and gives parameter optimization directions, reduces the number of optimization iterations, and improves the optimization speed.

[0112] Embodiment 4:

[0113] The present application provides a possible implementation method, in which the key component parameters are selected in step 4, which may specifically include:

[0114] Before parameter optimization, component parameters are selected, and the principal component analysis method is used to reduce the dimension of component parameter data involved in the calculation object to select key component parameters.

[0115] The calculation steps for selecting key component parameters are as follows:

[0116] (1) Obtain a component parameter data set m×n (for example, 20 components, 6 parameters for each component, that is, 20×6);

[0117] (2) Subtract the mean value to obtain DataAdjust (here is the data set after adjusting the original data);

[0118] (3) Calculate the covariance matrix of DataAdjust;

[0119] (4) Calculate the eigenvectors and eigenvalues ​​of the covariance matrix;

[0120] (5) Sort the eigenvalues ​​from large to small;

[0121] (6) retain the top N (N is a positive integer) eigenvectors so that the cumulative variance contribution rate reaches 80%;

[0122] (7) Project the sample points onto the selected feature vector to obtain a new data set;

[0123] (8) Analyze the raw data loadings in each principal component;

[0124] (9) Select parameters with high load as the optimization object.

[0125] This embodiment selects component parameters before parameter optimization, uses the principal component analysis method to reduce the dimension of component parameter data involved in the calculation object, and selects key component parameters, which can reduce the program's requirements on computer performance.

[0126] A possible implementation method is provided in the embodiment of the present application. In step 5, the optimization algorithm optimization parameters are selected, which may specifically include:

[0127] The machine learning optimization algorithm based on neural network finds the predicted value that satisfies the objective function, and reversely infers the corresponding input parameters based on the predicted value to realize automatic adjustment of the modeling parameters. At the same time, it combines the simulated annealing algorithm or genetic algorithm, uses the model trained by the neural network as the prediction formula, optimizes and solves the objective function, obtains the optimal solution of the objective function globally, and then obtains the optimal modeling parameters of the system program.

[0128] This embodiment is based on machine learning optimization algorithms such as neural networks, combined with simulated annealing algorithms or genetic algorithms, and uses a prediction model trained by a neural network to optimize and solve the objective function, obtain the optimal solution of the objective function globally, and then obtain the optimal modeling parameters of the system analysis program.

[0129] Embodiment 5:

[0130] The execution process of the simulated annealing algorithm mentioned above can be found in Figure 3 , specifically including the following steps:

[0131] Generate an initial solution using the example input card and calculate the objective function;

[0132] Perturbations produce new solutions;

[0133] Determine whether the difference between the new solution and the initial solution of the objective function is less than or equal to 0. If so, accept the new solution; if not, accept the new solution according to the Metropolis criterion. Here, the Metropolis criterion is a strategy used to decide whether to accept a new solution in an optimization problem. Specifically, if the energy of the new state is lower than the current state (that is, the new solution is better than the current solution), accept the new state; if the energy of the new state is higher than the current state, accept the new state with a certain probability. Continue to determine whether the number of iterations has been reached. If not, continue to perturb to generate a new solution; if the number of iterations has been reached, continue to determine whether the termination condition is met. If the termination condition is met, the operation ends and the optimal solution is returned; if the termination condition is not met, slowly reduce the temperature and reset the number of iterations.

[0134] Embodiment 6:

[0135] The execution flow chart of the genetic algorithm mentioned above can be found in Figure 4 , specifically including the following steps:

[0136] (1) Population initialization;

[0137] (2) Individual fitness evaluation;

[0138] (3) Select operation;

[0139] (4) Crossover operation;

[0140] (5) Mutation operation;

[0141] (6) Produce a new generation of population;

[0142] (7) Determine whether the new generation of population meets the conditions. If so, improve or solve the practical problems; if not, continue to evaluate individual fitness.

[0143] This embodiment combines a simulated annealing algorithm or a genetic algorithm, uses a prediction model trained by a neural network, performs an optimization solution to the objective function, obtains the optimal solution of the objective function globally, and then obtains the optimal modeling parameters of the system analysis program.

[0144] A possible implementation method is provided in the embodiment of the present application. In step nine, the parameter combination that cannot converge is given a low fitness, and the optimized parameters are further optimized, specifically including:

[0145] The optimized parameter combination causes the system program calculation to be unable to converge. By utilizing the characteristics of the genetic algorithm, a low fitness is assigned to the parameter combination that cannot converge. For the optimized parameters, the genetic algorithm is used again to re-optimize the parameters.

[0146] It should be noted that the size of the sequence number of each step in the above embodiment does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiment of the present application. In practical applications, all the above possible implementation methods can be combined in any way to form possible embodiments of the present application, which will not be described one by one here.

[0147] Based on the system program model calibration method based on the optimization algorithm provided in the above embodiments, and based on the same inventive concept, an embodiment of the present application also provides a system program model calibration device based on the optimization algorithm.

[0148] Figure 5 is a structural diagram of a system program model calibration device based on an optimization algorithm provided in an embodiment of the present application. Figure 5 As shown, the system program model calibration device based on the optimization algorithm may specifically include an establishment unit 510 and a calibration unit 520 .

[0149] The establishment unit 510 is used to perform step 1: establish a system program model; and enter step 2 of the calibration unit;

[0150] The calibration unit 520 is used to perform step 2: the card reading program reads the input card and records and organizes the parameter information; and then proceeds to step 3;

[0151] Step 3: Determine whether there is a parameter optimization instruction;

[0152] If not, proceed to step seven;

[0153] If yes, proceed to step 4;

[0154] Step 4: Select key component parameters; proceed to step 5;

[0155] Step 5: Use the parameter optimization program and select the optimization algorithm to optimize the parameters; proceed to step 6;

[0156] Step 6: The optimized parameters are automatically written into the input card; proceed to step 7;

[0157] Step 7: System program runs; if there is no parameter optimization instruction, go to step 10; if there is a parameter optimization instruction, go to step 8;

[0158] Step 8: Determine whether the system program reports an error and has no calculation results;

[0159] If not, proceed to step 10;

[0160] If yes, go to step nine;

[0161] Step 9: The parameter combination that cannot converge is given a low fitness, and the optimized parameters are optimized continuously, and then return to step 5;

[0162] Step 10: Read the output file;

[0163] Step 11: Result evaluation;

[0164] Conduct result evaluation based on the predicted data and experimental data or design data in the output file;

[0165] Step 12: Determine whether the acceptance criteria are met;

[0166] If not, proceed to step 13;

[0167] If yes, proceed to step 14;

[0168] Step 13: Based on the evaluation results, form parameter optimization instructions and give parameter optimization directions; return to step 3;

[0169] Step 14: End of analysis.

[0170] A possible implementation manner is provided in an embodiment of the present application, where the calibration unit 520 is further configured to:

[0171] When evaluating the results of steps 11 to 14, first perform a steady-state parameter optimization calibration analysis. Once the steady-state acceptance criteria are met, then perform a transient parameter optimization calibration analysis. Once the transient acceptance criteria are met, iterate and perform steady-state and transient parameter optimization calibration analyses in sequence until the parameter optimization meets both steady-state and transient acceptance criteria, and the analysis is finally completed.

[0172] A possible implementation manner is provided in an embodiment of the present application, where the calibration unit 520 is further configured to:

[0173] The method of result evaluation carried out from step 11 to step 14 is as follows:

[0174] Read the output file to obtain the predicted data, and then compare it with the design data or experimental data to determine whether the data time series are of equal length;

[0175] If the data time series are of equal length, the data are fast Fourier transformed to obtain the average amplitude and weighted frequency, and then it is determined whether the acceptance criteria are met;

[0176] If the data time series are not of equal length, the data are dynamically time-warped to obtain the Euclidean distance, and then it is determined whether the acceptable standard is met;

[0177] If the acceptance criteria are met, the analysis ends; if the acceptance criteria are not met, parameter optimization instructions are formed and the parameter optimization direction is given.

[0178] A possible implementation manner is provided in an embodiment of the present application, where the calibration unit 520 is further configured to:

[0179] In step 4, key component parameters are selected, including:

[0180] Before parameter optimization, component parameters are selected, and the principal component analysis method is used to reduce the dimension of component parameter data involved in the calculation object to select key component parameters.

[0181] A possible implementation manner is provided in an embodiment of the present application, where the calibration unit 520 is further configured to:

[0182] In step 5, select the optimization algorithm optimization parameters, including:

[0183] The machine learning optimization algorithm based on neural network finds the predicted value that satisfies the objective function, and reversely infers the corresponding input parameters based on the predicted value to realize automatic adjustment of the modeling parameters. At the same time, it combines the simulated annealing algorithm or genetic algorithm, uses the model trained by the neural network as the prediction formula, optimizes and solves the objective function, obtains the optimal solution of the objective function globally, and then obtains the optimal modeling parameters of the system program.

[0184] A possible implementation manner is provided in an embodiment of the present application, where the calibration unit 520 is further configured to:

[0185] In step 9, the parameter combination that cannot converge is given a low fitness, and the optimized parameters are further optimized, including:

[0186] The optimized parameter combination causes the system program calculation to be unable to converge. By utilizing the characteristics of the genetic algorithm, a low fitness is assigned to the parameter combination that cannot converge. For the optimized parameters, the genetic algorithm is used again to re-optimize the parameters.

[0187] A possible implementation manner is provided in an embodiment of the present application, where the calibration unit 520 is further configured to:

[0188] Provide parameter optimization directions, including:

[0189] When fast Fourier transform is selected, the optimization direction is to minimize the average amplitude; when dynamic time warping is selected, the optimization direction is to minimize the distance between the predicted data and the design data or experimental data.

[0190] Based on the same inventive concept, an embodiment of the present application also provides an electronic device, including a processor and a memory, wherein a computer program is stored in the memory, and the processor is configured to run the computer program to execute the system program model calibration method based on the optimization algorithm of any one of the above embodiments.

[0191] Based on the same inventive concept, an embodiment of the present application further provides a storage medium, in which a computer program is stored, wherein the computer program is configured to execute the system program model calibration method based on the optimization algorithm of any of the above embodiments when running.

[0192] Those skilled in the art can clearly understand that the specific working processes of the systems, devices, and modules described above can refer to the corresponding processes in the aforementioned method embodiments, and for the sake of brevity, they are not further described here.

[0193] Those skilled in the art can understand that the technical solution of the present application can be essentially or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, which includes a number of program instructions, so that an electronic device (such as a personal computer, a server, or a network device, etc.) executes all or part of the steps of the method described in each embodiment of the present application when running the program instructions. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk, etc., various media that can store program codes.

[0194] Alternatively, all or part of the steps of implementing the aforementioned method embodiments may be accomplished by hardware associated with program instructions (such as electronic devices such as personal computers, servers, or network devices), and the program instructions may be stored in a computer-readable storage medium. When the program instructions are executed by a processor of an electronic device, the electronic device executes all or part of the steps of the methods described in the embodiments of the present application.

[0195] The above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that within the spirit and principles of the present application, they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not deviate from the protection scope of the present application.

Claims

1. A system program model calibration method based on an optimization algorithm, characterized in that: The method comprises: Step 1: Establish system program model; proceed to step 2; Step 2: The card reader reads the input card and records and organizes the parameter information; then proceed to step 3; Step 3: Determine whether there is a parameter optimization instruction; If not, proceed to step seven; If yes, proceed to step 4; Step 4: Select key component parameters; proceed to step 5; Step 5: Use the parameter optimization program and select the optimization algorithm to optimize the parameters; proceed to step 6; Step 6: The optimized parameters are automatically written into the input card; proceed to step 7; Step 7: System program runs; if there is no parameter optimization instruction, go to step 10; if there is a parameter optimization instruction, go to step 8; Step 8: Determine whether the system program reports an error and has no calculation results; If not, proceed to step 10; If yes, go to step nine; Step 9: The parameter combination that cannot converge is given a low fitness, and the optimized parameters are optimized continuously, and then return to step 5; Step 10: Read the output file; Step 11: Result evaluation; Conduct result evaluation based on the predicted data and experimental data or design data in the output file; Step 12: Determine whether the acceptance criteria are met; If not, proceed to step 13; If yes, proceed to step 14; Step 13: Based on the evaluation results, form parameter optimization instructions and give parameter optimization directions; return to step 3; Step 14: End of analysis.

2. A system program model calibration method based on optimization algorithm as claimed in claim 1, characterized in that: When evaluating the results of steps 11 to 14, first perform a steady-state parameter optimization calibration analysis. Once the steady-state acceptance criteria are met, then perform a transient parameter optimization calibration analysis. Once the transient acceptance criteria are met, iterate and perform steady-state and transient parameter optimization calibration analyses in sequence until the parameter optimization meets both steady-state and transient acceptance criteria, and the analysis is finally completed.

3. The system program model calibration method based on optimization algorithm according to claim 1 or 2, characterized in that: The method of result evaluation carried out from step 11 to step 14 is as follows: Read the output file to obtain the predicted data, and then compare it with the design data or experimental data to determine whether the data time series are of equal length; If the data time series are of equal length, the data are fast Fourier transformed to obtain the average amplitude and weighted frequency, and then it is determined whether the acceptance criteria are met; If the data time series are not of equal length, the data are dynamically time-warped to obtain the Euclidean distance, and then it is determined whether the acceptable standard is met; If the acceptance criteria are met, the analysis ends; if the acceptance criteria are not met, parameter optimization instructions are formed and the parameter optimization direction is given.

4. The system program model calibration method based on optimization algorithm according to claim 1 or 2, characterized in that: In step 4, key component parameters are selected, including: Before parameter optimization, component parameters are selected, and the principal component analysis method is used to reduce the dimension of component parameter data involved in the calculation object to select key component parameters.

5. The system program model calibration method based on optimization algorithm according to claim 1 or 2, characterized in that: In step 5, select the optimization algorithm optimization parameters, including: The machine learning optimization algorithm based on neural network finds the predicted value that satisfies the objective function, and reversely infers the corresponding input parameters based on the predicted value to realize automatic adjustment of the modeling parameters. At the same time, it combines the simulated annealing algorithm or genetic algorithm, uses the model trained by the neural network as the prediction formula, optimizes and solves the objective function, obtains the optimal solution of the objective function globally, and then obtains the optimal modeling parameters of the system program.

6. The system program model calibration method based on optimization algorithm according to claim 1 or 2, characterized in that: In step 9, the parameter combination that cannot converge is given a low fitness, and the optimized parameters are further optimized, including: The optimized parameter combination causes the system program calculation to be unable to converge. By utilizing the characteristics of the genetic algorithm, a low fitness is assigned to the parameter combination that cannot converge. For the optimized parameters, the genetic algorithm is used again to re-optimize the parameters.

7. The system program model calibration method based on optimization algorithm as claimed in claim 3, characterized in that: Provide parameter optimization directions, including: When fast Fourier transform is selected, the optimization direction is to minimize the average amplitude; when dynamic time warping is selected, the optimization direction is to minimize the distance between the predicted data and the design data or experimental data.

8. A system program model calibration device based on an optimization algorithm, characterized in that: The device comprises: Establishing unit, used to carry out step 1: establishing system program model; entering step 2 of calibration unit; The calibration unit is used to perform step 2: the card reader program reads the input card, records and organizes the parameter information; and proceeds to step 3; Step 3: Determine whether there is a parameter optimization instruction; If not, proceed to step seven; If yes, proceed to step 4; Step 4: Select key component parameters; proceed to step 5; Step 5: Use the parameter optimization program and select the optimization algorithm to optimize the parameters; proceed to step 6; Step 6: The optimized parameters are automatically written into the input card; proceed to step 7; Step 7: System program runs; if there is no parameter optimization instruction, go to step 10; if there is a parameter optimization instruction, go to step 8; Step 8: Determine whether the system program reports an error and has no calculation results; If not, proceed to step 10; If yes, go to step nine; Step 9: The parameter combination that cannot converge is given a low fitness, and the optimized parameters are optimized continuously, and then return to step 5; Step 10: Read the output file; Step 11: Result evaluation; Conduct result evaluation based on the predicted data and experimental data or design data in the output file; Step 12: Determine whether the acceptance criteria are met; If not, proceed to step 13; If yes, proceed to step 14; Step 13: Based on the evaluation results, form parameter optimization instructions and give parameter optimization directions; return to step 3; Step 14: End of analysis.

9. An electronic device, characterized in that: The system comprises a processor and a memory, wherein a computer program is stored in the memory, and the processor is configured to run the computer program to execute the system program model calibration method based on the optimization algorithm according to any one of claims 1 to 7.

10. A storage medium, characterized in that: The storage medium stores a computer program, wherein the computer program is configured to execute the system program model calibration method based on the optimization algorithm according to any one of claims 1 to 7 when running.

Images