LkAN neural network optimization method based on swarm algorithm, equipment support system evaluation method and electronic equipment

Abstract

The embodiment of the application provides the technical field of data processing based on a specific model, in particular to a LKAN neural network optimization method based on a swarm algorithm, an equipment support system evaluation method and an electronic device. The method steps include: determining individual candidate solutions and a current optimal solution, updating the individual candidate solutions through the current optimal solution; performing iterative operation, evaluating the current optimal solution through a target function, generating an improved swarm algorithm and returning a global optimal solution; constructing a Legendre polynomial neural network, and setting an input layer of the Legendre polynomial neural network. The equipment support system performance evaluation method provided by the embodiment of the application is based on the improved LKAN neural network, realizes the equipment support system performance evaluation, can evaluate the actual effect of the current equipment support system under the existing equipment support capability when completing the equipment support task, and is a comprehensive embodiment of the effective degree of the equipment support activity completing the specified task.

Description

TECHNICAL FIELD

[0001] Embodiments of the present application relate to the technical field of data processing based on specific models, and in particular to an LKAN neural network optimization method based on a swarm algorithm, an equipment support system evaluation method, and an electronic device. BACKGROUND

[0002] An equipment support system is an important logistics support for troops, and can provide sufficient combat resources for the troops to ensure the smooth progress of combat. The effectiveness evaluation of the equipment support system can effectively find problems in the system and timely develop improvement programs to improve the effectiveness of the equipment support system, which is one of the key technologies in the field of equipment support. Therefore, how to effectively and accurately evaluate the effectiveness of the equipment support system has become a research hotspot. At present, researchers at home and abroad mainly use the analytic hierarchy process, the ADC model, the fuzzy comprehensive evaluation method, and the neural network model to evaluate the effectiveness. The analytic hierarchy process realizes the quantitative analysis of the decision results through the judgment matrix according to the experience of experts, but this method follows the principles of top-down and layer-by-layer transmission, which is not conducive to reflecting the feedback of the index layer to the target layer and the mutual influence between the indexes at different levels, and the subjective factor is strong. The ADC model uses the function of the system availability vector A, the credibility matrix D, and the inherent ability matrix C to represent the system effectiveness, but when the evaluation object contains multiple sub-objects and each sub-object contains multiple initial states, the complexity of determining the initial state and the complexity of calculating the state transition probability will increase exponentially. The fuzzy comprehensive evaluation method is based on fuzzy mathematics, which improves the accuracy and reliability of the evaluation results, but when the number of system effectiveness evaluation indexes increases, the relative membership degree coefficient will be small, which may cause the weight vector to be unable to match the fuzzy matrix, resulting in evaluation failure. The use of neural networks for effectiveness evaluation can effectively solve the influence of subjective factors in the evaluation process and can generate accurate and objective intelligent evaluation models. Based on the above analysis, it can be seen that the related art equipment support system effectiveness evaluation method has problems such as dependence on expert experience and strong subjectivity, which leads to the subjectivity of the evaluation results.

[0003] In view of the above problems, the related art has not yet proposed an effective technical solution, which has not met people's requirements and needs to be improved. SUMMARY

[0004] The main purpose of the embodiments of the present application is to provide an LKAN neural network optimization method based on a swarm algorithm, an equipment support system evaluation method, and an electronic device to solve the problem of dependence on expert experience and strong subjectivity in the related art.

[0005] According to an aspect of the embodiments of the present application, an LKAN neural network optimization method based on a swarm algorithm is provided, which comprises: determining an individual candidate solution and a current optimal solution, updating the individual candidate solution through the current optimal solution; performing iterative operation, evaluating the current optimal solution through a target function, adjusting the individual candidate solution and the current optimal solution according to the relationship between the individual candidate solution and a search space, generating an improved swarm algorithm and returning a global optimal solution; constructing a Legendre polynomial neural network, and setting an input layer of the Legendre polynomial neural network according to the global optimal solution.

[0006] According to at least one embodiment of the embodiments of the present application, the determining an individual candidate solution and a current optimal solution, and updating the individual candidate solution through the current optimal solution, further comprises: initializing the swarm algorithm, setting a current individual position, a current best position, a maximum iteration number, an exploration parameter and a development parameter in a position updating formula; balancing the exploration and development mechanisms of the swarm algorithm through the exploration parameter and the development parameter; and obtaining a next individual position by taking the absolute value of the difference between the current individual position and the current best position.

[0007] According to at least one embodiment of the embodiments of the present application, the performing iterative operation, evaluating the current optimal solution through a target function, adjusting the individual candidate solution and the current optimal solution according to the relationship between the individual candidate solution and a search space, generating an improved swarm algorithm and returning a global optimal solution, further comprises: initializing parameters of a position updating formula in the swarm algorithm, the parameters comprising a population size, a maximum iteration number and a current global optimal solution; detecting a current individual position of the position updating formula in the swarm algorithm, and reassigning the current individual position back to a previous position if the current individual position exceeds the search space of the swarm algorithm.

[0008] According to at least one embodiment of the embodiments of the present application, in the process of reassigning the current individual position back to the previous position, evaluating the individual candidate solution through the target function comprises: initializing a candidate solution set corresponding to the individual candidate solution and the target function, calculating the fitness value of the individual candidate solution through the target function, and updating the position of the individual candidate solution; dynamically adjusting the parameters in the position updating formula according to the value of the target function, so as to keep the individual candidate solution within a preset search space; and repeating the above steps until the iteration number reaches a preset maximum iteration number or the target function satisfies a convergence condition.

[0009] According to at least one embodiment of the present application, the Legendre polynomial neural network is constructed, and an input layer of the Legendre polynomial neural network is set according to the global optimal solution, specifically, the Legendre polynomial neural network comprises an input layer, a hidden layer and an output layer, the input layer is used to receive the global optimal solution, the hidden layer uses a Legendre polynomial as a base function, and the output layer is used to output an optimization result.

[0010] According to another aspect of the present application, an equipment support system evaluation method is provided, comprising the LKAN neural network optimization method based on the improved swarm algorithm, and further comprising: obtaining data samples, preprocessing the data samples to generate preprocessed data; setting the Legendre polynomial neural network, corresponding the number of input layers of the Legendre polynomial neural network to the number of equipment support system performance evaluation indexes; training the Legendre polynomial neural network by using the improved swarm algorithm, and evaluating the equipment support system performance by using the trained Legendre neural network to obtain an equipment support system performance evaluation value.

[0011] According to at least one embodiment of the present application, the data samples are obtained, the data samples are preprocessed to generate preprocessed data, and specifically, the data samples are normalized to generate normalized data, and the normalized data is used to eliminate the influence of index units and numerical orders of magnitude on the data samples.

[0012] According to at least one embodiment of the present application, the Legendre polynomial neural network is set, and further comprising: using a Legendre polynomial as a base function in an output layer of the Legendre polynomial neural network; training the Legendre polynomial neural network based on the improved swarm algorithm, initializing individual positions of a position updating formula in the swarm algorithm, making the individual positions correspond to combinations of parameters of the Legendre polynomial neural network; evaluating the individual positions of the position updating formula according to a target function, updating the corresponding individual positions according to an evaluation result; and repeating the above steps until the number of iterations reaches a maximum number of iterations.

[0013] According to still another aspect of the present application, an electronic device is provided, comprising: at least one processor; and a memory connected with the at least one processor in communication; wherein the memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor to make the at least one processor execute the method.

[0014] According to still another aspect of the present application, a computer program product is provided, comprising a computer program, and the computer program is executed by a processor to implement the method.

[0015] The embodiment of the present application has the beneficial technical effects that:

[0016] The embodiment of the present application introduces a new parameter adjustment mechanism, improves the dynamic balance of the swarm algorithm in the search process, and improves the search efficiency and global optimization ability of the algorithm. The improved swarm algorithm is applied to the parameter adjustment of the neural network, and does not use the standard basis function and the spline function of the related technology, but uses the Legendre polynomial, which is a series of orthogonal polynomials and plays a crucial role in approximation theory and numerical analysis. The embodiment of the present application uses the improved swarm algorithm to dynamically adjust the network parameters of the Legendre neural network (LKAN), optimizes the neural network, and uses the improved swarm algorithm to adjust the LKAN neural network parameters. The position of the mouse corresponds to a combination of LKAN neural network parameters.

[0017] The equipment support system effectiveness evaluation method provided by the embodiment of the present application is based on the improved LKAN neural network, realizes the equipment support system effectiveness evaluation, can evaluate the actual effect of the current equipment support system under the existing equipment support capability to complete the equipment support task, and is a comprehensive embodiment of the effective degree of the completion of the designated task of the equipment support activity. BRIEF DESCRIPTION OF DRAWINGS

[0018] In order to more clearly illustrate the technical solutions in the specific embodiments or related technologies of the present application, the drawings needed to be used in the specific embodiments or related technology description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0019] Figure 1 is a flowchart of the LKAN neural network optimization method based on the swarm algorithm.

[0020] Figure 1A is a system architecture diagram of the LKAN neural network optimization system based on the swarm algorithm.

[0021] Figure 2 is a flowchart of the optimization technical solution provided by steps S11 to S13.

[0022] Figure 3 is a flowchart of the optimization technical solution provided by steps S21 to S22.

[0023] Figure 4 is a flowchart of the optimization technical solution provided by steps S221 to S223.

[0024] Figure 5is a flow chart of an equipment support system evaluation method.

[0025] Figure 5A is an architecture diagram of an equipment support system evaluation system.

[0026] Figure 6 is a method flow chart for optimizing an LKAN neural network using an improved swarm algorithm.

[0027] Figure 7 is a method flow chart for evaluating the effectiveness of an equipment support system based on LKAN neural network optimization.

[0028] Figure 8 is a structural schematic diagram of an electronic device. DETAILED DESCRIPTION

[0029] In order for those skilled in the art to better understand the embodiments of the present application, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all. Based on the embodiments in the embodiments of the present application, all other embodiments obtained by those skilled in the art without creative labor should fall within the scope of protection of the embodiments of the present application.

[0030] It should be noted that the embodiments and features in the embodiments of the present application can be combined with each other without conflict. The embodiments of the present application will be described in detail below with reference to the drawings and in conjunction with the embodiments

[0031] The related art equipment system effectiveness evaluation method is highly subjective and relies on expert experience, making it difficult to accurately evaluate the effectiveness of the equipment support system. The LKAN neural network optimized by the improved swarm algorithm (MRSO) provided in the embodiments of the present application can handle the difficult-to-analyze regularity within the system, has strong generalization ability and fast learning convergence speed, and can well avoid human factor interference, and more objectively and accurately evaluate the effectiveness of the equipment support system.

[0032] As shown in Figure 1 The embodiments of the present application provide a LKAN neural network optimization method based on a swarm algorithm, applied to a LKAN neural network optimization system based on a swarm algorithm. The LKAN neural network optimization method here can be applied to the evaluation of the effectiveness of the equipment support system. The method steps include:

[0033] Step S1, determine individual candidate solutions and current optimal solutions, and update the individual candidate solutions through the current optimal solutions.

[0034] Step S2, iterative operation is performed, the current optimal solution is evaluated through the objective function, the individual candidate solution and the current optimal solution are adjusted according to the relationship between the individual candidate solution and the search space, an improved swarm algorithm is generated and the global optimal solution is returned.

[0035] Step S3, the Legendre polynomial neural network is constructed, and the input layer of the Legendre polynomial neural network is set according to the global optimal solution.

[0036] For the method steps disclosed in the above embodiments, the method steps are described as a series of action combinations for the purpose of simple description, but those skilled in the art should know that the embodiments of the present application are not limited by the order of the described actions, because according to the embodiments of the present application, certain steps can be performed in other order or simultaneously. Secondly, those skilled in the art should know that the embodiments described in the specification are all preferred embodiments, and the actions involved are not necessarily required by the embodiments of the present application.

[0037] Swarm algorithm principle:

[0038] (1) Pursuit of prey: the pursuit behavior of rats is a typical social activity. The most effective search agent is determined as the rat who knows the location of the prey. The rest of the people adjust their positions according to the position of this best rat, as follows:

[0039]

[0040] where, represents the position of the ith rat (solution), (t) represents the current iteration number. (A) represents the position of the best candidate solution found so far. The calculation of (A) is as follows:

[0041]

[0042] R and C are random values, where R ranges between [1, 5] and C ranges between [0, 2]. These values are used as parameters for the exploration and exploitation mechanisms in the algorithm: R = rand(1, 5), C = rand(0, 2).

[0043] (2) Fighting with prey

[0044] The fighting behavior is mathematically represented as follows:

[0045]

[0046] The next position of rat number i is represented as Parameters A and C are crucial for balancing the exploration and exploitation mechanisms. A small A value (e.g., 1) combined with a medium C value emphasizes exploitation, while other values can shift the focus towards exploration.

[0047] (3) Improved Mouse Swarm Optimization Algorithm (MRSO)

[0048] The Improved Mouse Swarm Optimization (MRSO) algorithm aims to improve RSO performance by balancing exploration and exploitation. This balance is achieved through a function in the "chase prey" part, which is static and depends on a single parameter C. In MRSO, this function is restated as follows:

[0049]

[0050] F3=2×rand(0,1)-1×rand(0,1)

[0051] A Modified =F1×F2×F3,t=0,1,2,…,Max iteration

[0052] The formula has been updated as follows:

[0053]

[0054] Initialize parameter A Modified After C and R, the objective function is used to evaluate the results, and the optimal solution is saved as... Then the rat's position is updated using the equation, and the parameters R, A, and A' are updated as well. Modified If a mouse's position exceeds the search space, it is reassigned to a previously selected center for adjustment. The position of each mouse is then tested and evaluated using an objective function. If a position is found that is better than the previous one... A better solution would be Update to this new optimal position. This process will continue with the maximum number of iterations. Finally, use the identified optimal position. To select a location. These modifications improve performance by obtaining the optimal fitness function.

[0055] 2. Kolmogorov-Arnold Network:

[0056] The LKAN architecture improves the accuracy and efficiency of predictions. KAN follows the Kolmogorov-Arnold representation theorem, a fundamental result in approximation theory that states any continuous multivariate function on a bounded region can be represented as a superposition of a univariate function and linear operations. Mathematically, for a continuous function on an n-dimensional unit hypercube… This theorem guarantees the continuous univariate function Φ j and φ ij The existence of

[0057]

[0058] Where x = (x1, x2, ..., x) n ( ) is the input vector. Based on this theorem, KAN has been developed as a novel neural network architecture that approximates multivariable functions by combining univariate functions across layers. A typical KAN consists of L layers. Given an input vector... The output of KAN can be expressed as:

[0059]

[0060] Each Φ l Represents having n l dimensional input and n l+1 A KAN layer with 1D output, and can be defined as a matrix of 1D functions:

[0061] Φ l =) l,ij}, l=1,…,L,i=1,…,n l j = 1, ..., n l+1 ……(3)

[0062] In the initial implementation of KAN, φ was defined as a weighted combination of basis functions b and B-spline functions.

[0063] φ(x)=ω b b(x)+ω s spline(x)……(4)

[0064] Where b(x) and spline(x) are defined as:

[0065]

[0066] Among them, c i ω b and ω s These are trainable parameters, represented by θ. KAN ={c i ,ω b ,ω s} represents spline B. i The characteristics are the polynomial order k and the number of grid points g.

[0067] In this embodiment, KAN uses Legendre polynomials instead of standard basis functions and spline functions. Legendre polynomials are a series of orthogonal polynomials that play a crucial role in approximation theory and numerical analysis. They can be defined on the interval [1, 1] satisfying a recursive relation:

[0068]

[0069] Therefore, for LKAN, we have:

[0070] φ(x) = ∑ i c i T i (x) … … (7)

[0071] where c i is a learnable parameter. KAN and LKAN have similar structures in approximating complex functions, but they differ in the number of parameters and dependence on grid points. In KAN, the total number of trainable parameters is affected by the input size, the number of hidden layers, the number of neurons in each layer, the grid size, and the polynomial order, denoted by |θ| KAN ~ O(n l H 2 (k+g)), where g is the grid size and k is the polynomial order. In contrast, LKAN eliminates the need for grid points, reducing the number of parameters to |θ| lKAN ~ O(n l H 2 k), which simplifies the model and improves efficiency by focusing only on the polynomial order k.

[0072] In the present invention, LKAN is used to predict the effectiveness evaluation value of the equipment support system. The network takes relevant variables as input and predicts the effectiveness evaluation value of the equipment support system. To facilitate the training of the network, the total cost function CF is defined as the individual cost functions of the network, that is:

[0073] CF = MSE(y, y * ) … … (8)

[0074] where MSE represents the mean square error, y * represents the true data. The LKAN architecture utilizes Legendre polynomials as basis functions, providing a more flexible and adaptive approach to representing the solution space. This flexibility enables LKAN to capture complex patterns with higher accuracy and efficiency.

[0075] The technical solutions provided by steps S1 to S3 optimize the Legendre polynomial neural network based on the swarm algorithm, and the main application scenario can be used for evaluating the effectiveness of the equipment support system. The neural network can more efficiently approximate complex function relationships, improve the accuracy and objectivity of the effectiveness evaluation of the equipment support system, dynamically adjust the LKAN network parameters through the improved swarm algorithm, enable the network to more efficiently approximate complex function relationships, improve the prediction accuracy of the model, avoid the algorithm from falling into local optimization, enhance the global search capability of the algorithm, improve the optimization performance of the neural network, find the global optimal solution in a smaller number of iteration times through reasonable parameter setting and optimization strategy, improve the convergence speed, use the Legendre polynomial as the base function of the neural network, effectively process complex nonlinear relationships, and enable the LKAN network to perform well when processing complex problems. In summary, the technical solutions provided by steps S1 to S3 optimize the LKAN neural network parameters through the improved swarm algorithm, not only improve the optimization accuracy and convergence speed of the network, but also enhance the global search capability and the ability to adapt to complex problems of the algorithm, and have a good improvement effect on the effectiveness evaluation results of the equipment support system.

[0076] As shown in Figure 1A The embodiment of the present application provides a LKAN neural network optimization system based on a swarm algorithm, which is used for implementing a LKAN neural network optimization method based on a swarm algorithm, and comprises the following modules.

[0077] An individual candidate solution updating module is configured to update the individual candidate solution based on the current optimal solution.

[0078] A global optimal solution generating module is configured to perform iterative operation, evaluate the current optimal solution through a target function, adjust the individual candidate solution and the current optimal solution based on the relationship between the individual candidate solution and the search space, generate an improved swarm algorithm, and return the global optimal solution.

[0079] A neural network input layer setting module is configured to construct a Legendre polynomial neural network, and set the input layer of the Legendre polynomial neural network based on the global optimal solution.

[0080] The above-described system embodiments are only illustrative, for example: each functional module, unit or subsystem in the system can or can not be physically separated, or can or can not be a physical unit, i.e., can be located in the same place or distributed to multiple different systems and their subsystems or modules. Those skilled in the art can select part or all of the functional modules, units or subsystems to achieve the purpose of the embodiments of the present application according to actual needs, and those skilled in the art can understand and implement the above-mentioned cases without creative labor.

[0081] AsFigure 2 As shown, in step S1, determining individual candidate solutions and the current optimal solution, and updating the individual candidate solutions using the current optimal solution, further includes:

[0082] Step S11: Initialize the mouse swarm algorithm by setting the current individual position, current best position, maximum number of iterations, exploration parameters, and development parameters in the position update formula.

[0083] Step S12: Balance the exploration and development mechanisms of the mouse swarm algorithm by exploring parameters and developing parameters.

[0084] Step S13: Subtract the current individual position from the current best position, and take the absolute value of the difference to obtain the next individual position.

[0085] In the optimization schemes provided in steps S11 to S13, the LKAN network parameters are dynamically adjusted using an improved mouse swarm algorithm, enabling the network to more efficiently approximate complex functional relationships and improve the model's prediction accuracy. By reasonably setting parameters and optimization strategies, the algorithm can find the global optimum within a fewer iterations, improving the convergence speed. Legendre polynomials, as basis functions, can effectively handle complex nonlinear relationships, making the LKAN network perform well in handling complex problems. The optimization algorithm reduces unnecessary computation, improves computational efficiency, and lowers computational costs. In summary, the optimization schemes provided in steps S11 to S13 utilize an improved mouse swarm algorithm to optimize the LKAN neural network parameters, not only improving the network's optimization accuracy and convergence speed but also enhancing the algorithm's global search capability and adaptability to complex problems.

[0086] like Figure 3 As shown, in step S2, iterative calculations are performed to evaluate the current optimal solution using the objective function. The individual candidate solutions and the current optimal solution are adjusted based on the relationship between the individual candidate solutions and the search space to generate an improved mouse swarm algorithm and return the global optimal solution. This further includes:

[0087] Step S21: Initialize the parameters of the position update formula in the mouse swarm algorithm. The parameters include the population size, the maximum number of iterations, and the current global optimal solution.

[0088] Step S22 involves detecting the current individual position in the position update formula of the mouse swarm algorithm. If the current individual position exceeds the search space of the mouse swarm algorithm, the current individual position is reassigned back to its previous position. The purpose of step S22 is to check whether the current individual position exceeds the search space, which refers to the set of all possible solutions explored by the algorithm during the optimization process.

[0089] The optimization technical solution provided in steps S21 to S22 improves the prediction accuracy of the swarm algorithm model, avoids the algorithm from falling into local optimization, enhances the global search ability of the algorithm, improves the optimization performance, and through reasonable setting of parameters and optimization strategies, the algorithm can find the global optimal solution in a smaller number of iteration times, improves the convergence speed, not only improves the optimization accuracy and convergence speed of the network, but also enhances the global search ability and the ability to adapt to complex problems of the algorithm. As can be seen, the Legendre polynomial as the basis function can effectively handle complex nonlinear relationships, so that the LKAN network performs well in handling complex problems.

[0090] As shown in Figure 4 In step S22, the current individual position is re-assigned back to the previous position, and the individual candidate solution is evaluated by the objective function, including:

[0091] Step S221, initialize the candidate solution set corresponding to the individual candidate solution and the objective function, calculate the fitness value of the individual candidate solution by the objective function, and update the position of the individual candidate solution.

[0092] Step S222, dynamically adjust the parameters in the position update formula according to the value of the objective function, so that the individual candidate solution remains within the preset search space.

[0093] Step S223, repeat the above steps until the number of iterations reaches the preset maximum number of iterations or the objective function meets the convergence condition.

[0094] In the optimization technical solution provided in steps S221 to S223, the candidate solution set refers to the set of all possible solutions, which are evaluated and updated in the optimization process. The initialization and update of the candidate solution set is the core part of the algorithm, which directly affects the performance and optimization results of the algorithm, and the objective function is a function for evaluating the quality of each candidate solution. In the optimization process, the value of the objective function determines the fitness of the candidate solution, thereby guiding the search direction of the algorithm. By evaluating the fitness value of the candidate solution through the objective function, the algorithm can determine which solution is closer to the global optimal solution. In the optimization technical solution provided in steps S221 to S223, by reasonably initializing the candidate solution set and the objective function, and dynamically adjusting the parameters in the position update formula, not only the optimization accuracy and convergence speed are improved, but also the global search ability and the ability to adapt to complex problems of the algorithm are enhanced, providing powerful data processing ability and algorithm support for the evaluation of the effectiveness of the equipment support system.

[0095] In step S3, the Legendre polynomial neural network is constructed, and the input layer of the Legendre polynomial neural network is set according to the global optimal solution, specifically:

[0096] Legendre polynomial neural networks consist of an input layer, hidden layers, and an output layer. The input layer receives the global optimal solution, the hidden layers use Legendre polynomials as basis functions, and the output layer outputs the optimization result.

[0097] In step S3, the single-layer structure of the LKAN network reduces the computational complexity of the algorithm, making it suitable for various optimization problems. Based on the improved mouse swarm algorithm, the LKAN network parameters are dynamically adjusted so that the individual mouse position in the position update formula corresponds to a combination of LKAN neural network parameters. By optimizing the algorithm to adjust the LKAN neural network parameters, the prediction accuracy and efficiency of the model are improved.

[0098] like Figure 5 The equipment support system evaluation method shown is based on the LKAN neural network optimization method of the mouse swarm algorithm, and includes:

[0099] Step T1: Obtain data samples, preprocess the data samples, and generate preprocessed data.

[0100] In step T1, the data samples refer to data samples related to the equipment support system. Preprocessing the data can involve normalized data processing, which is part of standardized data processing. Other standardized data processing methods include Z-Score standardization, Min-Max standardization, Log transformation, and Box-Cox transformation, etc. Preprocessing the data samples related to the equipment support system makes the data more suitable for subsequent analysis, improving the model's performance and effectiveness. This involves cleaning, extracting, selecting, and transforming features, and reducing data dimensionality through methods such as principal component analysis.

[0101] Step T2 involves configuring the Legendre polynomial neural network by matching the number of input layers of the network with the number of performance evaluation indicators for the equipment support system.

[0102] Step T3: Train the Legendre multinomial neural network using the improved mouse swarm algorithm, and use the trained Legendre neural network to evaluate the effectiveness of the equipment support system, thereby obtaining the equipment support system effectiveness evaluation value.

[0103] The technical solutions provided by steps T1 to T3 realize efficient evaluation of the equipment support system efficiency by preprocessing, constructing and training the neural network model on the related data of the equipment support system, make the data more suitable for subsequent analysis by normalizing, cleaning, feature extraction, selection and dimension reduction (such as principal component analysis) of the data, improve the data quality, correspond the input layer of the Legendre polynomial neural network to the number of equipment support system efficiency evaluation indexes, ensure that the model can accurately process the data related to the equipment support system, improve the adaptability of the model, train the neural network by using the improved swarm algorithm, improve the efficiency and convergence speed of the training process, and finally evaluate the equipment support system efficiency by using the trained Legendre polynomial neural network, so that an accurate efficiency evaluation value can be obtained, and a reliable basis can be provided for optimization and improvement of the equipment support system.

[0104] The technical solutions provided by steps T1 to T3 can be combined with the technical solutions provided by steps S1 to S3, or combined with the optimization technical solutions of any one or more steps in the technical solutions of steps S1 to S3, to form various technical solutions. Steps S1 to S3 optimize the LKAN neural network based on the improved swarm algorithm, and apply the optimized results to steps T1 to T3, so that the evaluation of the equipment support system efficiency is realized.

[0105] As shown in Figure 5A The embodiment of the present application provides an equipment support system evaluation system for realizing the equipment support system evaluation method, which comprises:

[0106] The preprocessing data generation module is configured to obtain data samples, preprocess the data samples, and generate preprocessing data.

[0107] The neural network setting module is configured to set the Legendre polynomial neural network, and correspond the number of input layers of the Legendre polynomial neural network to the number of equipment support system efficiency evaluation indexes.

[0108] The equipment support system efficiency evaluation module is configured to train the Legendre polynomial neural network by using the improved swarm algorithm, evaluate the equipment support system efficiency by using the trained Legendre neural network, and obtain an equipment support system efficiency evaluation value.

[0109] The embodiments of the system described above are only illustrative, for example: wherein each functional module, unit or subsystem in the system can or can not be physically separated, or can or can not be a physical unit, that is, can be located in the same place, or can be distributed to multiple different systems and their subsystems or modules. Those skilled in the art can select part or all of the functional modules, units or subsystems to achieve the purpose of the embodiments of the present application according to the actual needs, and those skilled in the art can understand and implement without creative labor.

[0110] Preferably, in step T1, the data samples are obtained, the data samples are preprocessed, and the preprocessed data is generated. Specifically, the data samples are normalized to generate normalized data, and the normalized data is used to eliminate the influence of index units and numerical orders of magnitude on the data samples.

[0111] Preferably, in step T2, the Legendre polynomial neural network is set, and further includes:

[0112] Step T21, using Legendre polynomials as base functions in the output layer of the Legendre polynomial neural network;

[0113] Step T22, training the Legendre polynomial neural network based on the improved swarm algorithm, initializing the individual position of the position update formula in the swarm algorithm, and making the individual position correspond to the combination of parameters of the Legendre polynomial neural network;

[0114] Step T23, evaluating the individual position of the position update formula according to the objective function, and updating the corresponding individual position according to the evaluation result;

[0115] Step T24, repeating the above steps until the number of iterations reaches the maximum number of iterations.

[0116] In the optimization technical solution provided in steps T21 to T24, the setting and training process of Legendre polynomial neural network are further refined. By using Legendre polynomial as the base function in the output layer and combining the improved swarm algorithm for parameter optimization, efficient training and optimization of the equipment support system effectiveness evaluation model are realized. Finally, by optimizing the neural network structure and training algorithm, the performance and evaluation accuracy of the model are improved. The optimization technical solution provided in steps T21 to T24 introduces Legendre polynomial as the base function in the output layer of the Legendre polynomial neural network. By utilizing its orthogonality and good approximation ability, the fitting ability and generalization performance of the model for complex data are enhanced. By initializing the positions of individuals in the swarm algorithm and corresponding them to the combination of neural network parameters, a reasonable starting point is provided for the optimization process. Based on the evaluation of the pros and cons of individual positions, the individual positions are dynamically updated according to the evaluation results to ensure that the algorithm can efficiently search for the optimal solution. By repeating the above evaluation and updating process until the maximum iteration number is reached, the integrity and convergence of the training process are guaranteed. In summary, the optimization technical solution provided in steps T21 to T24 combines Legendre polynomial with neural network and introduces the improved swarm algorithm for training, which is an innovative modeling and optimization method that can effectively solve the shortcomings of traditional methods in complex system effectiveness evaluation, realize efficient modeling and optimization of equipment support system effectiveness evaluation, and improve evaluation accuracy and model performance.

[0117] As shown in Figure 6 , Figure 6 a specific embodiment of optimizing LKAN neural network using improved swarm algorithm is shown, in Figure 6 which the LKAN neural network parameters are adjusted using the improved swarm algorithm, and the position of the mouse corresponds to a combination of LKAN neural network parameters. The LKAN neural network parameter optimization process is optimized using the improved swarm algorithm.

[0118] As shown in Figure 7 , Figure 7 an equipment support system effectiveness evaluation process based on LKAN neural network optimization is shown. The actual effect of the equipment support system in completing the equipment support task under the existing equipment support capability is reflected, and the effectiveness evaluation comprehensively reflects the effectiveness of the equipment support activity in completing the specified task. By introducing a new parameter adjustment mechanism, the dynamic balance of the improved swarm algorithm in the search process is improved, and the search efficiency and global optimization ability of the algorithm are improved

[0119] In the equipment support system effectiveness evaluation process based on LKAN neural network optimization shown in Figure 7 , at least the following steps are included:

[0120] According to the influence of modern war by high-tech and the development trend of equipment support system, on the basis of the investigation of related troops and expert opinions, the main factors in the operation of equipment support system are comprehensively considered to establish the equipment support system effectiveness evaluation index system.

[0121] On the basis of consulting relevant information and literature, the data samples for simulation experiment are determined through relevant investigation.

[0122] In order to make the equipment support system effectiveness evaluation results more in line with the actual situation and truly reflect the equipment support system effectiveness, the data samples need to be normalized. The normalized data can eliminate the influence of index unit and numerical order of magnitude, reduce network prediction error, and accelerate network training speed.

[0123] The LKAN network structure is determined. For example, the number of input layer nodes of LKAN neural network is the same as the number of equipment support system effectiveness evaluation indexes, and the equipment support system effectiveness evaluation value of output layer is only 1. The LKAN neural network is trained through the improved swarm algorithm and training sample data.

[0124] The embodiment of the application also provides an alternative technical solution of the equipment support system effectiveness evaluation process based on LKAN neural network optimization. Any step in the alternative technical solution can be used as any specific embodiment of the embodiment of the application. For example, the steps of index system construction and data preprocessing can be applied to the technical solutions in steps S1 to S3, and the determination of neural network structure can also be applied to any technical solution of steps S1 to S3. The specific steps of the alternative technical solution are as follows:

[0125] Index system construction and data preprocessing: the equipment support system effectiveness evaluation index system is constructed, the data samples for simulation experiment are determined through investigation, and the data samples are normalized to eliminate the influence of index unit and numerical order of magnitude, reduce network prediction error, and accelerate network training speed.

[0126] Neural network structure determination: the LKAN neural network structure is determined, the number of input layer nodes is consistent with the number of equipment support system effectiveness evaluation indexes, the output layer is the evaluation value of equipment support system effectiveness, and the number of hidden layer nodes is determined according to the empirical formula. The empirical formula can adopt n=m+k+α, wherein n is the number of hidden layer, m is the number of input layer nodes, k is the number of output layer nodes, and α is a constant, and the value range is [1, 10].

[0127] Encoding operation: the encoding operation in the algorithm is executed by using binary encoding, and each value size of radial basis function center, width and output weight of LKAN neural network is represented by a multi-bit binary string. The size of individual is determined according to the number of input layer, output layer and hidden layer, that is, the length of binary encoding string.

[0128] Population initialization and fitness calculation: set the population size and the maximum iteration number, randomly initialize the population, and calculate the fitness value of the current individual according to the fitness function.

[0129] In this step, the fitness function can adopt Wherein: x' represents the normalized sample data value, x represents the original sample data value, x max x min respectively represent the maximum value and the minimum value of the same index in the sample data.

[0130] Selection operation: sort the last generation individuals and offspring individuals according to the individual fitness value, and select the individuals with large fitness value to form a new generation population.

[0131] Determine whether the current iteration number is equal to the maximum iteration number, if yes, output the best individual of the population, and use it as the parameter of the LKAN neural network structure, if not, continue to perform the sorting operation.

[0132] Crossing and mutation operation: perform complete crossing operation on the last generation population individuals, and calculate the fitness value of the candidate offspring individuals.

[0133] Pre-mutate the last generation population individuals and the candidate offspring individuals, then calculate the fitness value of the pre-mutated individuals, if the fitness value of the pre-mutated individuals is greater than the fitness value of the current individuals, mutation occurs; otherwise, no mutation operation is performed, and the step of the fitness function is skipped to continue execution until mutation operation occurs.

[0134] Result output: output the best individual of the population, and use it as the parameter of the LKAN neural network structure, and apply it to the effectiveness evaluation of the equipment support system.

[0135] The alternative technical solution provided by the embodiment of the application improves the evaluation accuracy of the equipment support system, enables the neural network to more efficiently approximate a complex function relationship, improves the prediction accuracy of the model, enhances the global search capability, introduces random disturbance and dynamic adjustment parameters, avoids the algorithm from falling into a local optimum, enhances the global search capability of the algorithm, improves the optimization performance, and through reasonable setting of parameters and optimization strategies, the algorithm can find a global optimal solution in a smaller number of iteration times, and improves the convergence speed. Compared with the defect that the equipment support system effectiveness evaluation method in the related art relies on expert experience and has a strong subjective factor, the embodiment of the application automatically adjusts the network parameters through the optimization algorithm, reduces the interference of human factors, ensures that the evaluation result is more objective and accurate, reduces unnecessary calculation, improves the calculation efficiency, and reduces the calculation cost. At the same time, the improved algorithm can more quickly find an optimal solution, further improves the evaluation efficiency, and also enhances the global search capability and the ability to adapt to complex problems of the algorithm, thereby providing an efficient and accurate method for evaluation of the effectiveness of the equipment support system.

[0136] As shown in Figure 8 the person skilled in the art provides corresponding electronic devices and storage media based on the LKAN neural network optimization method based on the ant colony algorithm and the equipment support system evaluation method:

[0137] The electronic device includes one or more processors 51 and a memory 52, and one processor 51 is taken as an example in the figure.

[0138] The controller can also include an input device 53 and an output device 54.

[0139] The processor 51, the memory 52, the input device 53 and the output device 54 can be connected through a bus or other means, and the connection through the bus is taken as an example in the figure.

[0140] The processor 51 can be a central processing unit (CPU), and the processor 51 can also be other general-purpose processors, digital signal processors (DSP), application specific integrated circuits (ASIC), field programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. chips or combinations of the above chips, and the general-purpose processor can be a microprocessor or any conventional processor.

[0141] The memory 52, as a non-transitory computer readable storage medium, can be used to store non-transitory software programs, non-transitory computer executable programs and modules, such as program instructions / modules corresponding to the control method in the embodiments of the present application. The processor 51 executes various functions and data processing of the server by running the non-transitory software programs, instructions and modules stored in the memory 52, that is, implements the method of the above method embodiments.

[0142] The memory 52 can include a program storage area and a data storage area, wherein the program storage area can store an operating system and application programs required by at least one function; the data storage area can store data created according to the use of the processing device of the server operation, etc. In addition, the memory 52 can include a high-speed random access memory, and can also include a non-transitory memory, such as at least one magnetic disk storage device, a flash memory device, or other non-transitory solid-state memory device. In some embodiments, the memory 52 can optionally include a memory disposed remotely with respect to the processor 51, and these remote memories can be connected to the network connection device through a network. Examples of the above network include but are not limited to the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0143] The input device 53 can receive input digital or character information, and generate key signal inputs related to user settings and function controls of the processing device of the server. The output device 54 can include a display device such as a display screen.

[0144] One or more modules are stored in the memory 52, and when executed by the one or more processors 51, the method shown in any of the specific embodiments of the present application is executed.

[0145] Those skilled in the art can understand that the implementation of all or part of the above method embodiments can be completed by a computer program instructing related hardware, and the program can be stored in a computer readable storage medium. When the program is executed, it can include the processes of the above method embodiments. The storage medium can be a magnetic disk, an optical disk, a read-only memory (ROM), a random access memory (RAM), a flash memory (FM), a hard disk drive (HDD) or a solid-state drive (SSD), etc. The storage medium can also include a combination of the above types of memories.

[0146] In the description of the embodiments of the present application in the specification, the description of the terms "one embodiment", "an example", "a specific example” and the like is intended to indicate that the specific feature, structure, material or the like described is included in at least one embodiment or example of the present application. In the specification, illustrative representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or the like described can be combined in any appropriate manner in one or more embodiments or examples.

[0147] In addition, the technical solutions among the various embodiments of the present application can be combined with each other, but it must be based on the fact that a person skilled in the art can realize it, and when the combination of technical solutions appears contradictory or cannot be realized, it should be considered that the combination of technical solutions does not exist, nor is it within the protection scope claimed by the embodiments of the present application.

[0148] All features disclosed in the embodiments of the present application, or steps in all disclosed methods or processes, can be combined in any manner, except for mutually exclusive features and / or steps. Any feature disclosed in the specification of the embodiments of the present application can be replaced by other equivalent or similar features with the same purpose, unless specifically stated. That is, each feature is only an example of a series of equivalent or similar features, unless specifically stated. Throughout the specification, the same reference signs indicate the same elements.

[0149] Those skilled in the art can understand that the modules in the device in the embodiments can be adaptively changed and arranged in one or more devices different from the embodiments. The modules or units or components in the embodiments can be combined into one module or unit or component, and furthermore can be divided into multiple sub-modules or sub-units or sub-components. All features disclosed in the specification of the embodiments of the present application (including the corresponding claims, abstract and drawings) and all processes or units of any method or device disclosed in the same can be combined in any combination, except that at least some of such features and / or processes or units are mutually exclusive. Unless explicitly stated otherwise, each feature disclosed in the specification of the embodiments of the present application (including the corresponding claims, abstract and drawings) can be replaced by an alternative feature providing the same, equivalent or similar purpose.

[0150] Although the embodiments of the present application are described in conjunction with the accompanying drawings, various modifications and changes can be made by those skilled in the art without departing from the spirit and scope of the present application, and such modifications and changes fall within the scope defined by the appended claims.

Claims

1. A method for optimizing an LKAN neural network based on a mouse swarm optimization algorithm, characterized in that, include: Determine individual candidate solutions and the current optimal solution, and update the individual candidate solutions using the current optimal solution; Perform iterative calculations, evaluate the current optimal solution using the objective function, adjust the individual candidate solutions and the current optimal solution according to the relationship between individual candidate solutions and the search space, generate an improved mouse swarm algorithm, and return the global optimal solution, further including: Initialize the parameters of the position update formula in the mouse swarm algorithm, including the population size, maximum number of iterations, and the current global optimum; The current individual position in the position update formula of the mouse swarm algorithm is detected. If the current individual position exceeds the search space of the mouse swarm algorithm, the current individual position is reassigned back to the previous position. Construct a Legendre polynomial neural network, and configure the input layer of the Legendre polynomial neural network according to the global optimal solution.

2. The LKAN neural network optimization method based on the mouse swarm algorithm according to claim 1, characterized in that, The process of determining individual candidate solutions and the current optimal solution, and updating the individual candidate solutions using the current optimal solution, further includes: The mouse swarm algorithm is initialized by setting the current individual position, current best position, maximum number of iterations, exploration parameters, and development parameters in the position update formula; The exploration and development mechanisms of the mouse swarm algorithm are balanced using the exploration parameters and the development parameters. The next individual position is obtained by subtracting the current individual position from the current best position and taking the absolute value of the difference.

3. The LKAN neural network optimization method based on the mouse swarm algorithm according to claim 1, characterized in that, In the process of reassigning the current individual position back to the previous position, candidate solutions for the individual are evaluated through an objective function, including: The candidate solution set and objective function corresponding to the individual candidate solution are initialized, the fitness value of the individual candidate solution is calculated through the objective function, and the position of the individual candidate solution is updated. The parameters in the position update formula are dynamically adjusted according to the value of the objective function, so that the individual candidate solutions are kept within the preset search space; Repeat the above steps until the number of iterations reaches the preset maximum number of iterations or the objective function meets the convergence condition.

4. The LKAN neural network optimization method based on the mouse swarm algorithm according to claim 1, characterized in that, The construction of the Legendre polynomial neural network, specifically the setting of the input layer of the Legendre polynomial neural network based on the global optimal solution, is as follows: The Legendre polynomial neural network includes an input layer, a hidden layer, and an output layer. The input layer is used to receive the global optimal solution, the hidden layer uses Legendre polynomials as basis functions, and the output layer is used to output the optimization result.

5. A method for evaluating an equipment support system, characterized in that, The LKAN neural network optimization method based on the mouse swarm algorithm according to any one of claims 1 to 4 further includes: Obtain data samples, preprocess the data samples, and generate preprocessed data; The Legendre polynomial neural network is configured such that the number of input layers of the Legendre polynomial neural network corresponds to the number of performance evaluation indicators of the equipment support system. An improved mouse swarm algorithm was used to train a Legendre multinomial neural network, and the trained Legendre neural network was used to evaluate the effectiveness of the equipment support system, thus obtaining the equipment support system effectiveness evaluation value.

6. The equipment support system evaluation method according to claim 5, characterized in that, The process of acquiring data samples and preprocessing them to generate preprocessed data specifically involves: The data sample is normalized to generate normalized data, which is used to eliminate the influence of the indicator unit and its numerical order of magnitude on the data sample.

7. The equipment support system evaluation method according to claim 5, characterized in that, The configuration of the Legendre polynomial neural network further includes: In the output layer of the Legendre polynomial neural network, Legendre polynomials are used as basis functions. The Legendre multinomial neural network is trained based on the improved mouse swarm algorithm. The individual positions in the position update formula of the mouse swarm algorithm are initialized so that the individual positions correspond to the combination of parameters of the Legendre multinomial neural network. The individual position is evaluated based on the objective function and the position update formula is used to update the corresponding individual position based on the evaluation result. Repeat the above steps until the maximum number of iterations is reached.

8. An electronic device, characterized in that, The electronic device includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores a computer program executable by the at least one processor, the computer program being executed by the at least one processor to cause the at least one processor to perform the method according to any one of claims 1 to 7.

9. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the method described in any one of claims 1 to 7.

Images