CN113807040B - Optimized design method for microwave circuit - Google Patents
Optimized design method for microwave circuit Download PDFInfo
- Publication number
- CN113807040B CN113807040B CN202111112523.1A CN202111112523A CN113807040B CN 113807040 B CN113807040 B CN 113807040B CN 202111112523 A CN202111112523 A CN 202111112523A CN 113807040 B CN113807040 B CN 113807040B
- Authority
- CN
- China
- Prior art keywords
- sample
- training
- design parameters
- microwave circuit
- output
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/30—Circuit design
- G06F30/32—Circuit design at the digital level
- G06F30/337—Design optimisation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/30—Circuit design
- G06F30/32—Circuit design at the digital level
- G06F30/33—Design verification, e.g. functional simulation or model checking
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/004—Artificial life, i.e. computing arrangements simulating life
- G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2111/00—Details relating to CAD techniques
- G06F2111/06—Multi-objective optimisation, e.g. Pareto optimisation using simulated annealing [SA], ant colony algorithms or genetic algorithms [GA]
Abstract
The invention discloses an optimal design method for a microwave circuit, which comprises the following steps: obtaining sample model design parameters by utilizing LHS, and obtaining corresponding sample response by utilizing Matlab-HFSS joint simulation technology; calculating the correlation coefficients of all sample responses and target responses, and selecting the sample with the largest correlation coefficient as an optimized sample and other samples as training samples; training the ELM by using a training sample, predicting design parameters of response of an optimized sample, and optimizing input weights and threshold values of the ELM by using BSO; and establishing a mapping relation between the design parameters of the microwave circuit model and the response by using the ELM after optimizing the input weight and the threshold value, training by using all training samples in the training process, and predicting the model design parameters corresponding to the target response in the prediction process. The invention improves the training and predicting quality of the neural network, reduces the number of training samples required and the time required for optimally designing the microwave circuit, realizes the automation of the optimal design of the microwave circuit, and improves the optimal design efficiency.
Description
Technical Field
The invention relates to the technical field of wireless communication, in particular to an optimal design method for a microwave circuit.
Background
With rapid development of modern wireless communication technology, the requirements of a communication system on a microwave circuit are higher and higher, and in order to meet different requirements of the communication system, the microwave circuit needs to be continuously optimally designed. In order to improve the design efficiency of the microwave circuit and reduce the time cost of the design, the design of the microwave circuit by using an algorithm has become a necessary trend.
With the development of computers in recent years, the computing power of the computers has been greatly improved, and the artificial neural network (Artificial Neural Network, ANN) reenters the field of view of people, and is paid attention to by more and more students, so that the computers are widely applied. ANN is a mathematical model for simulating the actual neural network of the human brain, which can establish a certain mapping relation between the input and the output of the network, and by selecting an appropriate algorithm and utilizing a manually given training sample, the weight and the threshold value for minimizing the error between the network output and the actual sample output of each iteration are searched. When the neural network is trained, the input to a given network can be mapped to the output of the network.
The most commonly used microwave circuit optimization design methods at present are a space mapping algorithm, a group intelligent optimization algorithm and a machine learning algorithm.
The space mapping algorithm expresses the microwave circuit by two models, one is a fine model with low simulation speed and accurate simulation result; the other is a coarse model with high simulation speed but inaccurate simulation results. Firstly, optimizing in a coarse model to obtain ideal coarse model parameters, then obtaining fine model parameters through a mapping relation, and finally substituting the fine model parameters into a fine model for verification. The space mapping algorithm is utilized, tedious and time-consuming optimization work can be carried out in the coarse model, and the fine model only carries out simple verification work, so that the complex fine model is prevented from being directly optimized, and the optimization time is shortened. However, most of the fine models of microwave circuits cannot be represented by coarse models, and thus there are certain limitations in optimizing designs using spatial mapping algorithms.
The group intelligent optimization algorithm generally searches a local optimal solution of a problem by utilizing a specific algorithm, then determines a global optimal solution through multiple iterations, and generally adopts a variation idea to increase the diversity of the algorithm in the optimization process, so that the algorithm can accurately search the global optimal solution while ensuring convergence. In the actual optimization design process, a large number of candidate solutions are needed to be given out, full-wave electromagnetic simulation is carried out for a large number of times, and the fitness value of the full-wave electromagnetic simulation is calculated, so that the time cost required by the algorithm is high, and the randomness is high.
Among machine learning algorithms, ANN is one of the most commonly used methods for establishing a mapping relationship, and mainly includes two processes. In the training process, a certain number of samples are utilized to establish the mapping relation between the design parameters of the microwave circuit model and the response thereof; in the prediction process, the target response is taken as input, and ideal model design parameters can be obtained through network output.
However, when the conventional ANN method is used for optimizing design, a large number of iterations are required to be performed by using a gradient descent method to optimize network parameters, and the problems of poor learning ability, low training speed, low prediction accuracy and the like generally exist, so that the time complexity is high in practical application, and a large amount of time and labor cost are wasted.
Disclosure of Invention
Aiming at the technical problems, the invention provides a novel general optimal design method for a microwave circuit, which obtains model design parameters of target response with lower time and labor cost.
In order to achieve the above object, the present invention provides the following technical solutions:
an optimized design method for a microwave circuit comprises the following steps:
s1, designing ideal target response according to design indexes;
s2, obtaining a certain number of sample model design parameters by utilizing Latin hypercube sampling (Latin hypercube sampling, LHS), and obtaining corresponding sample responses by utilizing Matlab-HFSS joint simulation technology;
s3, calculating correlation coefficients of all sample responses and target responses, and selecting the response with the largest correlation coefficient and design parameters thereof as an optimized sample, and other responses and design parameters as training samples;
s4, training an extreme learning machine (Extreme Learning Machine, ELM) by using a training sample, predicting design parameters of response of an optimized sample, and optimizing input weights and threshold values of the ELM by using a brain storm optimization algorithm (Brain Storm Optimization Algorithm, BSO);
s5, establishing a mapping relation between the design parameters of the microwave circuit model and the response by using the ELM after optimizing the input weight and the threshold value, training by using all training samples including the optimized sample in the training process, and predicting the model design parameters corresponding to the target response in the prediction process;
s6, substituting the predicted design parameters into the microwave circuit for simulation verification, and if the response does not accord with the design index, adding the design parameters and the response into the original training sample set, and turning to the step S3 until the response and the design parameters meeting the design index are obtained.
Further, the ELM network structure includes an input layer, an hidden layer, and an output layer, each of which is provided with a plurality of neurons.
Further, the ELM training process in step S4 is:
is provided with N arbitrary training samples (X i ,Y i ) Wherein X is input i In n dimensions, i.e. X i ＝[x i1 ,x i2 ,…,x in ] T Output Y i In m dimension, i.e. Y i ＝[y i1 ,y i2 ,…,y im ] T The output of the extreme learning machine with an hidden layer of L neurons is expressed as:
wherein g (x) is an activation function, W j ＝[w j1 ,w j2 ,…,w jn ] T To input weights, θ j Threshold, beta, for the j-th neuron of the hidden layer j ＝[β j1 ,β j2 ,…,β jm ] T For output weights, "·" represents inner product;
the goal of web learning is to minimize the output error of the web from the output of the samples, expressed as:
i.e. W is present j 、θ j And beta j So that
The above is represented by a matrix:
Hβ＝Y (4)
where H is the output of the hidden layer neuron, beta is the output weight, Y is the sample output, and
by inputting X through a given network, obtain
And->
So that
Equivalent to minimizing the loss function:
further, if the weight W is input j And a threshold value theta j Randomly or manually set, the output weight β is calculated by:
β＝H + Y (10)
wherein H is + Is the generalized inverse of matrix H.
Further, the ELM prediction process in step S4 is: given the input weight ω, the threshold θ, and the output weight β, the predicted sample output Y is found according to equations (4) - (7) given the predicted sample input X.
Further, in step S4, the parameter optimization process is performed by using a BSO algorithm based on a reverse learning initialization policy, where the reverse learning process is as follows: and randomly generating an initial population, generating a reverse population according to the initial population, calculating fitness values of the initial population and the reverse population, and finally selecting optimal solutions of fitness values in all the populations as final initial population.
Further, in step S4, the process of optimizing the input weight and the threshold value of the ELM using BSO is as follows: calculating the correlation coefficients of all sample responses and target responses, and selecting the sample response and the design parameters thereof with the largest correlation coefficient with the target responses as optimized samples, and other responses and the design parameters thereof as training samples; and training the ELM by using a training sample, and obtaining the input weight and the threshold value of the extreme learning machine which minimize the fitness function through BSO optimization.
Further, the optimization objective of BSO is:
min e (11)
wherein e is an fitness function, and the expression is as follows:
wherein the method comprises the steps of
Design parameters, X, of optimized samples predicted for extreme learning machine Optimizing samples The design parameters of the sample are optimized for practical knowledge.
The BSO optimization variables are the input weights omega and the threshold value theta of the ELM, namely, the values omega and theta when the optimization targets are met are obtained.
Compared with the prior art, the invention has the beneficial effects that:
compared with the traditional optimal design method, the optimal design method for the microwave circuit can realize automatic operation of all processes, can effectively reduce the time and labor cost of optimal design and improve the optimal design efficiency; compared with a space mapping method, the method has more universality; compared with a group intelligent optimization algorithm, the method does not need to perform full-wave electromagnetic simulation for a large number of times, so that the time cost is reduced; compared with a general machine learning method, the method adopts BSO to optimize ELM network parameters, so that the ELM learning capacity is higher, and the prediction accuracy is higher.
In conclusion, the optimal design method for the microwave circuit is inspired from the ANN, the proper neural network is selected, and the network parameters are optimized by matching with a certain algorithm, so that the training and prediction quality of the neural network can be improved, the number of required training samples is reduced, and the time cost required by optimally designing the microwave circuit is reduced; meanwhile, an algorithm and a joint simulation technology are utilized to realize the automatic operation of the microwave circuit optimization design process, so that the actual problem is solved, and the optimization design efficiency is improved.
Drawings
In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings that are needed in the embodiments will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments described in the present invention, and other drawings may be obtained according to these drawings for a person having ordinary skill in the art.
Fig. 1 is a flowchart of an optimized design method for a microwave circuit according to an embodiment of the present invention.
Fig. 2 is a schematic diagram of an ELM network structure according to an embodiment of the present invention.
Fig. 3 is a reverse learning flow chart provided in an embodiment of the present invention.
FIG. 4 is a flow chart of input weights and thresholds for a BSO optimized ELM provided by an embodiment of the present invention.
Detailed Description
For a better understanding of the present technical solution, the method of the present invention is described in detail below with reference to the accompanying drawings.
As shown in fig. 1, the invention provides an optimized design method for a microwave circuit, which comprises the following steps:
(1) Designing an ideal target response according to the design index;
(2) Obtaining a certain number of sample model design parameters by utilizing LHS, and obtaining corresponding sample responses by utilizing Matlab-HFSS joint simulation technology;
(3) Calculating the correlation coefficients of all sample responses and target responses, and selecting the response with the largest correlation coefficient and the design parameters thereof as optimized samples, and other responses and design parameters as training samples;
(4) Training the ELM by using a training sample, predicting design parameters of response of an optimized sample, and optimizing input weights and threshold values of the ELM by using BSO;
(5) Establishing a mapping relation between the design parameters of the microwave circuit model and the response by using the ELM after optimizing the input weight and the threshold value, training by using all training samples including the optimized sample in the training process, and predicting the model design parameters corresponding to the target response in the prediction process;
(6) Substituting the predicted design parameters into the microwave circuit for simulation verification, and if the response does not accord with the design index, adding the design parameters and the response into the original training sample set, and turning to the step (3) until the response and the design parameters meeting the design index are obtained.
The technical scheme of the invention is based on an Extreme Learning Machine (ELM), adopts a brain storm optimization algorithm (BSO) to improve the ELM, and adopts Latin Hypercube Sampling (LHS) and a joint simulation technology to obtain training samples. The method mainly comprises three parts of obtaining training samples, training prediction and optimizing network parameters.
1. Acquisition of training sample portions
The acquiring a training sample portion includes acquiring model design parameters of the training sample and acquiring a response of the training sample.
The model design parameter part for obtaining the training sample is realized by LHS. The LHS is a layered sampling technology, higher sampling precision can be obtained with fewer sampling times, design parameters of each model obtained through the LHS are mutually independent and are uniformly generated in each interval, and the quality of training samples can be effectively improved.
The response part for obtaining the training sample is realized by Matlab-HFSS joint simulation, and instructions such as HFSS modeling, design parameter modification, response output and operation are written in the script by utilizing Matlab software and HFSS-Matlab-Api library functions, so that the simulation can be realized by utilizing Matlab software to control HFSS electromagnetic simulation software. The Matlab-HFSS joint simulation has remarkable advantages for building a complex structure model and repeatedly modifying specified design parameters, and the optimized design of the microwave circuit can be automatically completed in the Matlab by calling HFSS electromagnetic simulation software for simulation, so that the design parameters are not required to be brought into HFSS software, and the labor cost is saved.
2. Training a predictive part
The training prediction part is realized by adopting ELM, the ELM is a neural network based on forward propagation, and compared with the traditional algorithm, the training prediction part has the advantages of higher learning speed and higher generalization capability while ensuring the precision, and the network structure is shown in figure 2. Wherein omega ij For the input weight of the i-th neuron of the input layer to the j-th neuron of the hidden layer, theta L Threshold, beta, for the hidden layer L-th neuron jk The output weights from the jth neuron of the hidden layer to the kth neuron of the output layer.
In actual operation, the extreme learning machine can be divided into two processes of training and predicting:
(1) Training process: the input and output of training samples for extreme learning machines are known, and the mapping relationship between the training samples is found.
Is provided with N arbitrary training samples (X i ,Y i ) Wherein X is input i In n dimensions, i.e. X i ＝[x i1 ,x i2 ,…,x in ] T Output Y i In m dimension, i.e. Y i ＝[y i1 ,y i2 ,…,y im ] T The output of an extreme learning machine with an hidden layer of L neurons can be expressed as:
wherein g (x) is an activation function, W j ＝[w j1 ,w j2 ,…,w jn ] T To input weights, θ j Threshold, beta, for the j-th neuron of the hidden layer j ＝[β j1 ,β j2 ,…,β jm ] T For output weights, "·" represents inner product.
The goal of web learning is to minimize the output error of the web from the output of the samples, expressed as:
i.e. W is present j 、θ j And beta j So that
The above is represented by a matrix:
Hβ＝Y (4)
where H is the output of the hidden layer neuron, beta is the output weight, Y is the sample output, and
by inputting X through a given network, it is desirable to obtain
And->
So that
Equivalent to minimizing the loss function:
the traditional learning algorithm adopts a gradient descent method to solve, but the gradient descent method needs to be used for carrying out multiple iterations, and all parameters are required to be adjusted in the iteration process, so that the process is complex and time-consuming. In the extreme learning machine, if the weight W is input j And a threshold value theta j If randomly or manually set, the matrix H in the formula (4) is uniquely determined, and the output weight β can be calculated by the following formula:
β＝H + Y (10)
wherein H is + Is the generalized inverse of matrix H.
(2) The prediction process comprises the following steps: the input of the extreme learning machine and the mapping relation between the input and the output are known, and the output of the extreme learning machine is calculated.
In the prediction process, the mapping relation between the training sample input and the training sample output is known, namely the input weight omega, the threshold value theta and the output weight beta of the extreme learning machine are known, the prediction sample input X is given, and the prediction sample output Y can be obtained according to the formulas (4) - (7).
The ELM features that the input weights from the input layer to the hidden layer and the threshold values of the hidden layer units are set randomly or artificially, and then the updating is not needed, and the output weights from the hidden layer to the output layer are determined once by solving the equation set.
The ELM has the advantages of simple network structure, high convergence speed, strong generalization capability and the like, and is very suitable for solving the problem of optimal design of a microwave circuit.
3. Optimizing network parameter parts
The network parameters in ELM mainly include an input weight ω, a threshold θ, and an output weight β, and since β is calculated from ω and θ, which are randomly or artificially set, ω and θ are selected to be important. To improve the quality of the network, this section optimizes ω and θ with BSO.
BSO is proposed based on the elicitation of a brainstorming conference by humans at group decision time. Firstly, initializing a random strategy to generate a certain number of ideas, and uniformly distributing all the ideas in a search space as much as possible; then, according to the similarity, dividing all ideas into a certain number of classes by using a K-means clustering method; secondly, calculating the fitness value of all ideas according to the given fitness function, and selecting the idea with the optimal fitness value in each class, wherein the fitness value represents the quality of the idea; and finally, selecting the overall optimal idea through iteration.
The population initialization process of the original BSO algorithm adopts a purely random strategy, so that the quality of the final optimization result and the convergence rate of the population are poor. The scheme of the invention adopts an initialization strategy of reverse learning to improve the BSO algorithm, firstly searches a reverse population of an initial population, and finally selects an optimal solution to carry out a subsequent iterative optimization process through competition between the current population and the reverse population, wherein a reverse learning flow chart is shown in figure 3.
The BSO can search the global optimal solution while ensuring the convergence, is very suitable for solving the optimization problem of the multi-peak high-dimensional function, and is very suitable for optimizing ELM multi-dimensional network parameters.
A flowchart of BSO optimized extreme learning machine input weights and thresholds is shown in fig. 4 below. Firstly, designing ideal target response according to design indexes; then calculating the correlation coefficients of the responses of all training samples and ideal target responses, and selecting the training sample with the largest correlation coefficient as an optimized sample; secondly, training the ELM by using other training samples, and predicting model design parameters corresponding to response of the optimized samples; and finally, obtaining the input weight and the threshold value of the extreme learning machine which minimize the fitness function by using a brain storm optimization algorithm.
The optimization objective of the brain storm algorithm is:
min e (11)
wherein e is an fitness function, and the expression is as follows:
wherein the method comprises the steps of
Design parameters, X, of optimized samples predicted for extreme learning machine Optimizing samples The design parameters of the sample are optimized for practical knowledge.
In summary, the invention provides an optimal design method for a microwave circuit, which has the key points that:
1) The LHS is utilized to obtain a specified number of model design parameters as training samples, so that the samples are uniformly distributed in the space, and the quality of the samples is improved;
2) Obtaining the response of design parameters by Matlab-HFSS joint simulation so as to realize the full-flow automatic operation;
3) Improving a BSO algorithm by utilizing a population initialization strategy of reverse learning so as to improve the optimization capability of the BSO;
4) The ELM is improved by using BSO, and the input weight and threshold of the ELM are optimized to improve learning ability, accuracy and precision of the ELM.
The above embodiments are only for illustrating the technical solution of the present invention, and are not limiting; although the invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may be modified or some technical features may be replaced with others, which may not depart from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims (8)
1. The optimized design method for the microwave circuit is characterized by comprising the following steps of:
s1, designing ideal target response according to design indexes;
s2, obtaining a certain number of sample model design parameters by utilizing LHS, and obtaining corresponding sample responses by utilizing Matlab-HFSS joint simulation technology;
s3, calculating correlation coefficients of all sample responses and target responses, and selecting the response with the largest correlation coefficient and design parameters thereof as an optimized sample, and other responses and design parameters as training samples;
s4, training the ELM by using a training sample, predicting design parameters of response of an optimized sample, and optimizing input weights and threshold values of the ELM by adopting BSO;
s5, establishing a mapping relation between the design parameters of the microwave circuit model and the response by using the ELM after optimizing the input weight and the threshold value, training by using all training samples including the optimized sample in the training process, and predicting the model design parameters corresponding to the target response in the prediction process;
s6, substituting the predicted design parameters into the microwave circuit for simulation verification, and if the response does not accord with the design index, adding the design parameters and the response into the original training sample set, and turning to the step S3 until the response and the design parameters meeting the design index are obtained.
2. The optimal design method for a microwave circuit according to claim 1, wherein the ELM network structure comprises an input layer, an hidden layer and an output layer, and the input layer, the hidden layer and the output layer are all provided with a plurality of neurons.
3. The optimal design method for a microwave circuit according to claim 1, wherein the training process of ELM in step S4 is:
is provided with N arbitrary training samples (X i ,Y i ) Wherein X is input i In n dimensions, i.e. X i ＝[x i1 ,x i2 ,…,x in ] T Output Y i In m dimension, i.e. Y i ＝[y i1 ,y i2 ,…,y im ] T The output of the extreme learning machine with an hidden layer of L neurons is expressed as:
wherein g (x) is an activation function, W j ＝[w j1 ,w j2 ,…,w jn ] T To input weights, θ j Threshold, beta, for the j-th neuron of the hidden layer j ＝[β j1 ,β j2 ,…,β jm ] T For output weights, "·" represents inner product;
the goal of web learning is to minimize the output error of the web from the output of the samples, expressed as:
i.e. W is present j 、θ j And beta j So that
The above is represented by a matrix:
Hβ＝Y (4)
where H is the output of the hidden layer neuron, beta is the output weight, Y is the sample output, and
by inputting X through a given network, obtain
And->
So that
Equivalent to minimizing the loss function:
4. the method of optimizing design for a microwave circuit according to claim 3, wherein if a weight W is inputted j And a threshold value theta j Randomly or manually set, the output weight β is calculated by:
β＝H + Y (10)
wherein H is + Is the generalized inverse of matrix H.
5. The optimal design method for a microwave circuit according to claim 3, wherein the ELM prediction process in step S4 is as follows: given the input weight ω, the threshold θ, and the output weight β, the predicted sample output Y is found according to equations (4) - (7) given the predicted sample input X.
6. The optimal design method for the microwave circuit according to claim 1, wherein in step S4, the parameter optimization process is performed by adopting a BSO algorithm based on a reverse learning initialization strategy, and the reverse learning flow is as follows: and randomly generating an initial population, generating a reverse population according to the initial population, calculating fitness values of the initial population and the reverse population, and finally selecting optimal solutions of fitness values in all the populations as final initial population.
7. The optimized design method for microwave circuit according to claim 1, wherein the process of optimizing the input weight and threshold value of ELM using BSO in step S4 is as follows: calculating the correlation coefficients of all sample responses and target responses, and selecting the sample response and the design parameters thereof with the largest correlation coefficient with the target responses as optimized samples, and other responses and the design parameters thereof as training samples; and training the ELM by using a training sample, and obtaining the input weight and the threshold value of the extreme learning machine which minimize the fitness function through BSO optimization.
8. The optimal design method for a microwave circuit according to claim 7, wherein the optimization objective of BSO is:
min e (11)
wherein e is an fitness function, and the expression is as follows:
wherein the method comprises the steps of
Design parameters, X, of optimized samples predicted for extreme learning machine Optimizing samples The design parameters of the sample are optimized for practical knowledge. />
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111112523.1A CN113807040B (en) | 2021-09-23 | 2021-09-23 | Optimized design method for microwave circuit |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111112523.1A CN113807040B (en) | 2021-09-23 | 2021-09-23 | Optimized design method for microwave circuit |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113807040A CN113807040A (en) | 2021-12-17 |
| CN113807040B true CN113807040B (en) | 2023-06-09 |
Family
ID=78940206
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111112523.1A Active CN113807040B (en) | 2021-09-23 | 2021-09-23 | Optimized design method for microwave circuit |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113807040B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114492253A (en) * | 2022-01-20 | 2022-05-13 | 电子科技大学 | Rapid optimization method for half-space mapping of microstrip circuit in combination with Gaussian process |
| CN116111984A (en) * | 2022-12-06 | 2023-05-12 | 中国电子科技集团公司信息科学研究院 | Filter design optimization method and device, filter, equipment and medium |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101976294A (en) * | 2010-11-18 | 2011-02-16 | 南开大学 | Space mapping algorithm-based automatic computer-aid tuning design method for microwave device |
| CN112766608A (en) * | 2021-02-03 | 2021-05-07 | 燕山大学 | Cement mill system power consumption index prediction method based on extreme learning machine |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9038005B2 (en) * | 2013-03-15 | 2015-05-19 | Resonant Inc. | Network synthesis design of microwave acoustic wave filters |
| CN109933881A (en) * | 2019-03-06 | 2019-06-25 | 武汉大学 | A kind of Fault Diagnosis of Power Electronic Circuits method based on optimization deepness belief network |
-
2021
- 2021-09-23 CN CN202111112523.1A patent/CN113807040B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101976294A (en) * | 2010-11-18 | 2011-02-16 | 南开大学 | Space mapping algorithm-based automatic computer-aid tuning design method for microwave device |
| CN112766608A (en) * | 2021-02-03 | 2021-05-07 | 燕山大学 | Cement mill system power consumption index prediction method based on extreme learning machine |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113807040A (en) | 2021-12-17 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110472778A (en) | A kind of short-term load forecasting method based on Blending integrated study | |
| CN106600059B (en) | Intelligent power grid short-term load prediction method based on improved RBF neural network | |
| CN113807040B (en) | Optimized design method for microwave circuit | |
| CN109376848A (en) | A kind of door control unit neural network of simplification | |
| CN109034034A (en) | A kind of vein identification method based on nitrification enhancement optimization convolutional neural networks | |
| CN111191785B (en) | Structure searching method based on expansion search space for named entity recognition | |
| CN113762603B (en) | Photovoltaic base station short-term photovoltaic power prediction method based on improved sparrow algorithm optimization | |
| CN111416797B (en) | Intrusion detection method for optimizing regularization extreme learning machine by improving longicorn herd algorithm | |
| CN105427241B (en) | Distortion correction method for large-view-field display equipment | |
| CN112231839B (en) | Intelligent exploration method for peak-tracking sampling of aircraft aiming at high time-consuming constraint | |
| CN116402002B (en) | Multi-target layered reinforcement learning method for chip layout problem | |
| CN106296434B (en) | Grain yield prediction method based on PSO-LSSVM algorithm | |
| CN112149883A (en) | Photovoltaic power prediction method based on FWA-BP neural network | |
| CN105913078A (en) | Multi-mode soft measurement method for improving adaptive affine propagation clustering | |
| CN112232577A (en) | Power load probability prediction system and method for multi-core intelligent meter | |
| Alber et al. | Backprop evolution | |
| CN112766603A (en) | Traffic flow prediction method, system, computer device and storage medium | |
| CN116526450A (en) | Error compensation-based two-stage short-term power load combination prediction method | |
| CN115689070A (en) | Energy prediction method for optimizing BP neural network model based on imperial butterfly algorithm | |
| CN115906303A (en) | Planar microwave filter design method and device based on machine learning | |
| CN116307211A (en) | Wind power digestion capability prediction and optimization method and system | |
| CN111159489A (en) | Searching method | |
| CN114564787A (en) | Bayesian optimization method, device and storage medium for target-related airfoil design | |
| CN109408896B (en) | Multi-element intelligent real-time monitoring method for anaerobic sewage treatment gas production | |
| CN116415177A (en) | Classifier parameter identification method based on extreme learning machine |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |