A method for evaluating air defense combat effectiveness based on ABMS and orthogonal test

Abstract

The present invention discloses an air defense combat effectiveness evaluation method based on ABMS and orthogonal testing, comprising the following steps: 1. establishing an air defense combat simulation model based on ABMS; 2. performing air defense combat effectiveness simulation based on orthogonal testing to obtain significant influencing factors of air defense combat; 3. performing air defense combat effectiveness simulation based on the significant influencing factors based on orthogonal testing; 4. obtaining optimal training data and worst-case training data for air defense combat effectiveness evaluation; 5. combining M′ sets of air defense combat effectiveness evaluation training sets, optimal training data, and worst-case training data to obtain an air defense combat effectiveness evaluation training set; 6. training a neural network evaluation model; and 7. performing air defense combat effectiveness evaluation in real scenarios based on the trained neural network evaluation model. The method of the present invention has simple steps, a reasonable design, is more practical, reduces the problem of large simulation workload and time consumption, and improves the accuracy of air defense combat effectiveness evaluation.

Description

Technical Field

[0001] The present invention belongs to the technical field of air defense combat effectiveness evaluation, and in particular relates to an air defense combat effectiveness evaluation method based on ABMS and orthogonal test. Background Art

[0002] Combat effectiveness primarily supports system demonstration, optimizes combat plans, and measures combat effectiveness. It is a crucial indicator for evaluating the combat capability of an equipment system. Currently, four common methods for evaluating air defense combat effectiveness include modeling based on queuing theory, ADC methods, comprehensive evaluation methods, and simulation platforms. The first three methods are largely based on expert-provided weights or prior probabilities for each system, making the evaluation results significantly influenced by subjective factors. The fourth method, based on simulation platforms, suffers from the problem that the number of experiments performed on the platform increases exponentially with the number of factors studied, making simulations computationally intensive and time-consuming. Summary of the Invention

[0003] The technical problem to be solved by the present invention is to address the deficiencies in the above-mentioned prior art and provide an air defense combat effectiveness evaluation method based on ABMS and orthogonal testing. The method has simple steps and a reasonable design. An air defense combat simulation model is established to simulate real air defense combat, which is more in line with reality. An air defense combat effectiveness evaluation training set is obtained through orthogonal testing, which reduces the problem of large simulation volume and time consumption, increases the input of optimal training data and worst training data, and improves the accuracy of air defense combat effectiveness evaluation.

[0004] To solve the above technical problems, the present invention adopts a technical solution: an air defense combat effectiveness evaluation method based on ABMS and orthogonal test, characterized in that the method comprises the following steps:

[0005] The method comprises the following steps:

[0006] Step 1: Establish an air defense combat simulation model based on ABMS:

[0007] Step 101: Using a computer to establish an air defense combat simulation model based on the ABMS; wherein the air defense combat simulation model includes an agent model of the combined missile and artillery air defense system, an agent model of the air defense missile, an agent model of the anti-aircraft artillery shell, and an agent model of the enemy aircraft;

[0008] Step 102: Set the effectiveness evaluation index of the air defense combat simulation model to include the first index and the second index; set the L factor set that affects the evaluation index of the air defense combat simulation model to {X1,...,X l ,...,X L}; where X l represents the lth air defense operation influencing factor, l and L are both positive integers, and 1≤l≤L, and L is a positive integer not less than 3;

[0009] Step 2: Conduct air defense combat effectiveness simulation based on orthogonal experiments to obtain significant influencing factors of air defense combat:

[0010] Step 201: From the L factor set {X1,,...,X l ,...,X L} and record them as A sets of influencing factors to be studied in air defense operations, namely {X1',...,X' a ,...,X' A}, where X' a represents the ath influencing factor of air defense operations to be studied, a and A are both positive integers, and 1≤a≤A, A is less than L;

[0011] Step 202: Perform air defense combat effectiveness simulation based on an orthogonal test to obtain M first indices and M second indices corresponding to the M air defense combat influencing factor design schemes;

[0012] Step 203: Use a computer to perform correlation analysis on the M air defense operation influencing factor design schemes, the M first indicators, and the M second indicators, and obtain air defense operation significant influencing factors from the A set of air defense operation influencing factors to be studied; wherein, each obtained air defense operation significant influencing factor is recorded as B air defense operation significant influencing factor sets, namely {X1",...,X b ”,...,X'' B}, where X b " represents the bth significant influencing factor of air defense operations, b and B are both positive integers, and 1≤b≤B, B is less than A;

[0013] Step 3: Conduct air defense combat effectiveness simulation based on orthogonal experiments according to significant influencing factors:

[0014] According to the significant influencing factors, air defense combat effectiveness simulation is carried out based on orthogonal experiments to obtain M′ groups of air defense combat effectiveness evaluation training sets;

[0015] Step 4: Obtaining the best and worst training data for air defense combat effectiveness evaluation:

[0016] A computer is used to perform variance analysis on M air defense combat influencing factor design schemes, M first indicators, and M second indicators to obtain the optimal training data and the worst training data;

[0017] Step 5: Combine the M′ group of air defense combat effectiveness evaluation training sets, the optimal training data, and the worst training data to obtain M′+2 groups of air defense combat effectiveness evaluation training sets;

[0018] Step 6: Training of the neural network evaluation model:

[0019] Step 601: Construct a neural network evaluation model;

[0020] Step 602: Using the significant influencing factors of air defense operations as the input layer and the first and second indicators as the output layer, the air defense operations effectiveness evaluation training set obtained in step 5 is input into the neural network evaluation model in step 601 for training, thereby obtaining a trained neural network evaluation model.

[0021] Step 7: Evaluate the air defense combat effectiveness in real scenarios based on the trained neural network evaluation model.

[0022] The above-mentioned air defense combat effectiveness evaluation method based on ABMS and orthogonal test is characterized in that: in step 101, the missile-artillery combined air defense system agent model includes a radar system agent model, an air defense command and control system agent model and an anti-aircraft artillery system agent model.

[0023] The above-mentioned air defense combat effectiveness evaluation method based on ABMS and orthogonal testing is characterized in that: in step 202, air defense combat effectiveness simulation is performed based on the orthogonal testing to obtain M first indicators and M second indicators corresponding to the M air defense combat influencing factor design schemes. The specific process is as follows:

[0024] Step 2021: Set the ath air defense operation influencing factor X' to be studied a The minimum value is recorded as X' a,min , the ath air defense operation influencing factor to be studied is X' a The maximum value is recorded as X' a,max , set the ath air defense operation influencing factor X' to be studied a The value is located at X' a,min ~X' a,max The step length when the range is set is recorded as X' a,s ;

[0025] Step 2022: Input A air defense operation influencing factors to be studied, and use a computer to obtain design schemes for the air defense operation influencing factors using orthogonal experimental design software; wherein the total number of design schemes for the air defense operation influencing factors is M, where M is a positive integer;

[0026] Step 2023: Using a computer, input the mth air defense combat influencing factor design scheme into the air defense combat simulation model for simulation, and obtain the mth first index and the mth second index corresponding to the mth air defense combat influencing factor design scheme; wherein m is a positive integer and 1≤m≤M;

[0027] Step 2024: Repeat step 2023 multiple times to complete the simulation of M air defense combat influencing factor design schemes, and obtain M first indicators and M second indicators;

[0028] In step 203, a computer is used to perform correlation analysis on the M air defense operation influencing factor design schemes, the M first indicators, and the M second indicators, and obtain significant air defense operation influencing factors from the A set of air defense operation influencing factors to be studied. The specific process is as follows:

[0029] Step 2031: Using a computer and orthogonal experimental design software, perform a correlation analysis on the M air defense operation influencing factor design schemes, the M first indicators, and the M second indicators to obtain a correlation value of each air defense operation influencing factor with the first indicator and the second indicator.

[0030] Step 2032: Using a computer, based on the correlation values ​​of the various air defense operation influencing factors with the first indicator and the second indicator, select the air defense operation influencing factors corresponding to the correlation values ​​greater than 0.1 and record them as significant air defense operation influencing factors.

[0031] The above-mentioned air defense combat effectiveness evaluation method based on ABMS and orthogonal test is characterized by: in step 3, air defense combat effectiveness simulation is performed based on orthogonal test according to significant influencing factors to obtain M′ groups of air defense combat effectiveness evaluation training sets. The specific process is as follows:

[0032] Step 301: Set the bth significant influencing factor X for air defense operations b "The minimum value is recorded as X b ” ,min , the bth significant influencing factor of air defense operations is X b The maximum value of " is recorded as X b ” ,max , set the bth significant influencing factor X for air defense operations b "The value is located at X b ” ,min ~X b ” ,max The step length when in range is recorded as X b ” ,s ;

[0033] Step 302: According to the method described in step 2022, B significant influencing factors of air defense operations are input to obtain M′ design schemes of significant influencing factors of air defense operations; wherein M′ is a positive integer;

[0034] Step 303: According to the method described in step 2023, the m′th design scheme for the significant influencing factor of air defense operations is input for simulation, and the m′th first index and the m′th second index corresponding to the m′th design scheme for the significant influencing factor of air defense operations are obtained; wherein m′ is a positive integer and 1≤m′≤M′;

[0035] Step 304: Set the m′th design scheme of the significant influencing factor of air defense operations as the input layer, the m′th first indicator and the m′th second indicator as the output layer, and obtain M′ groups of air defense operations effectiveness evaluation training sets.

[0036] The above-mentioned air defense combat effectiveness evaluation method based on ABMS and orthogonal test is characterized in that: in step 4, a computer is used to perform variance analysis on M air defense combat influencing factor design schemes, M first indicators and M second indicators to obtain optimal training data and worst training data. The specific process is as follows:

[0037] Step 401: Using a computer, perform variance analysis on the M air defense combat influencing factor design schemes, the M first indices, and the M second indices to obtain a main effects plot of each air defense combat influencing factor and the average value of the first indices, as well as a main effects plot of each air defense combat influencing factor and the average value of the second indices; wherein the abscissa of the main effects plot of each air defense combat influencing factor and the average value of the first indices and the main effects plot of each air defense combat influencing factor and the average value of the second indices are the number of factor levels on the abscissa and the average value on the ordinate.

[0038] Step 402: Using a computer, based on the main effects plots of each air defense operation influencing factor and the average value of the first index, and the main effects plots of each air defense operation influencing factor and the average value of the second index, the optimal combination solution is recorded as the factor level number corresponding to the maximum average value of the first index and the minimum average value of the second index under each air defense operation influencing factor, and the worst combination solution is recorded as the factor level number corresponding to the minimum average value of the first index and the maximum average value of the second index under each air defense operation influencing factor;

[0039] The data corresponding to the significant influencing factors of air defense operations are selected from the optimal combination scheme and the worst combination scheme and recorded as the optimal training data and the worst training data.

[0040] Compared with the prior art, the present invention has the following advantages:

[0041] 1. The method of the present invention has simple steps and reasonable design, and solves the problem that the evaluation results are greatly affected by subjective factors and the simulation calculation is large and time-consuming.

[0042] 2. The present invention establishes an air defense combat simulation model based on ABMS, including an agent model of a combined missile and artillery air defense system, an agent model of an air defense missile, an agent model of an anti-aircraft artillery shell, and an agent model of an enemy aircraft, simulating real air defense combat and being more realistic.

[0043] 3. The present invention performs air defense combat effectiveness simulation based on orthogonal experiments, first obtaining significant influencing factors of air defense combat, and then performing air defense combat effectiveness simulation based on orthogonal experiments according to the significant influencing factors to obtain an air defense combat effectiveness evaluation training set, and adding optimal training data and worst training data to the air defense combat effectiveness evaluation training set can improve the fitting accuracy of the neural network evaluation model, thereby improving the accuracy of the air defense combat effectiveness evaluation.

[0044] 4. The present invention has strong practicality and adaptability, and can be used to study different air defense combat influencing factors according to different air defense combat scenarios, thereby achieving the purpose of air defense combat safety protection to the greatest extent.

[0045] 5. The present invention simulates air defense combat effectiveness based on orthogonal experiments to obtain significant influencing factors of air defense operations, which makes it easier for control and prevention commanders to adjust according to the significant influencing factors of air defense operations to obtain better air defense combat effectiveness, thereby providing great help for control and prevention commanders to select the optimal air defense combat strategy.

[0046] 6. The present invention can not only obtain the optimal input of the neural network evaluation model based on orthogonal experiments, but also greatly reduce the number of training times required to train the neural network evaluation model, so that the trained neural network can meet the evaluation requirements.

[0047] In summary, the method of the present invention has simple steps and reasonable design. It establishes an air defense combat simulation model to simulate real air defense combat, which is more in line with reality. It obtains an air defense combat effectiveness evaluation training set through orthogonal experiments, reduces the problem of large simulation volume and time consumption, increases the input of optimal training data and worst training data, and improves the accuracy of air defense combat effectiveness evaluation.

[0048] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 It is a flowchart of the method of the present invention.

[0050] Figure 2 This is the state diagram of the Agent model of the missile and artillery combined air defense system of the present invention.

[0051] Figure 3 This is the state diagram of the air defense missile Agent model of the present invention.

[0052] Figure 4This is the state diagram of the antiaircraft artillery shell Agent model of the present invention.

[0053] Figure 5 This is the state diagram of the enemy aircraft Agent model of the present invention.

[0054] Figure 6 Schematic diagram of the enemy aircraft track of the present invention. DETAILED DESCRIPTION

[0055] like Figure 1 As shown, the air defense combat effectiveness evaluation method based on ABMS and orthogonal test of the present invention includes the following steps:

[0056] Step 1: Establish an air defense combat simulation model based on ABMS:

[0057] Step 101: Using a computer to establish an air defense combat simulation model based on the ABMS; wherein the air defense combat simulation model includes an agent model of the combined missile and artillery air defense system, an agent model of the air defense missile, an agent model of the anti-aircraft artillery shell, and an agent model of the enemy aircraft;

[0058] Step 102: Set the effectiveness evaluation index of the air defense combat simulation model to include the first index and the second index; set the L factor set that affects the evaluation index of the air defense combat simulation model to {X1,...,X l ,...,X L}; where X l represents the lth air defense operation influencing factor, l and L are both positive integers, and 1≤l≤L, and L is a positive integer not less than 3;

[0059] Step 2: Conduct air defense combat effectiveness simulation based on orthogonal experiments to obtain significant influencing factors of air defense combat:

[0060] Step 201: From the L factor set {X1,,...,X l ,...,X L} and record them as A sets of influencing factors to be studied in air defense operations, namely {X1',...,X' a ,...,X' A}, where X' a represents the ath influencing factor of air defense operations to be studied, a and A are both positive integers, and 1≤a≤A, A is less than L;

[0061] Step 202: Perform air defense combat effectiveness simulation based on an orthogonal test to obtain M first indices and M second indices corresponding to the M air defense combat influencing factor design schemes;

[0062] Step 203: Use a computer to perform correlation analysis on the M air defense operation influencing factor design schemes, the M first indicators, and the M second indicators, and obtain air defense operation significant influencing factors from the A set of air defense operation influencing factors to be studied; wherein, each obtained air defense operation significant influencing factor is recorded as B air defense operation significant influencing factor sets, namely {X1",...,X b ”,...,X'' B}, where X b " represents the bth significant influencing factor of air defense operations, b and B are both positive integers, and 1≤b≤B, B is less than A;

[0063] Step 3: Conduct air defense combat effectiveness simulation based on orthogonal experiments according to significant influencing factors:

[0064] According to the significant influencing factors, air defense combat effectiveness simulation is carried out based on orthogonal experiments to obtain M′ groups of air defense combat effectiveness evaluation training sets;

[0065] Step 4: Obtaining the best and worst training data for air defense combat effectiveness evaluation:

[0066] A computer is used to perform variance analysis on M air defense combat influencing factor design schemes, M first indicators, and M second indicators to obtain the optimal training data and the worst training data;

[0067] Step 5: Combine the M′ group of air defense combat effectiveness evaluation training sets, the optimal training data, and the worst training data to obtain M′+2 groups of air defense combat effectiveness evaluation training sets;

[0068] Step 6: Training of the neural network evaluation model:

[0069] Step 601: Construct a neural network evaluation model;

[0070] Step 602: Using the significant influencing factors of air defense operations as the input layer and the first and second indicators as the output layer, the air defense operations effectiveness evaluation training set obtained in step 5 is input into the neural network evaluation model in step 601 for training, thereby obtaining a trained neural network evaluation model.

[0071] Step 7: Evaluate the air defense combat effectiveness in real scenarios based on the trained neural network evaluation model.

[0072] In this embodiment, the missile-artillery combined air defense system Agent model in step 101 includes a radar system Agent model, an air defense command and control system Agent model, and an antiaircraft artillery system Agent model.

[0073] In this embodiment, in step 202, air defense combat effectiveness simulation is performed based on an orthogonal test to obtain M first indicators and M second indicators corresponding to the M air defense combat influencing factor design schemes. The specific process is as follows:

[0074] Step 2021: Set the ath air defense operation influencing factor X' to be studied a The minimum value is recorded as X' a,min , the ath air defense operation influencing factor to be studied is X' a The maximum value is recorded as X' a,max , set the ath air defense operation influencing factor X' to be studied a The value is located at X' a,min ~X' a,max The step length when the range is set is recorded as X' a,s ;

[0075] Step 2022: Input A air defense operation influencing factors to be studied, and use a computer to obtain design schemes for the air defense operation influencing factors using orthogonal experimental design software; wherein the total number of design schemes for the air defense operation influencing factors is M, where M is a positive integer;

[0076] Step 2023: Using a computer, input the mth air defense combat influencing factor design scheme into the air defense combat simulation model for simulation, and obtain the mth first index and the mth second index corresponding to the mth air defense combat influencing factor design scheme; wherein m is a positive integer and 1≤m≤M;

[0077] Step 2024: Repeat step 2023 multiple times to complete the simulation of M air defense combat influencing factor design schemes, and obtain M first indicators and M second indicators;

[0078] In step 203, a computer is used to perform correlation analysis on the M air defense operation influencing factor design schemes, the M first indicators, and the M second indicators, and obtain significant air defense operation influencing factors from the A set of air defense operation influencing factors to be studied. The specific process is as follows:

[0079] Step 2031: Using a computer and orthogonal experimental design software, perform a correlation analysis on the M air defense operation influencing factor design schemes, the M first indicators, and the M second indicators to obtain a correlation value of each air defense operation influencing factor with the first indicator and the second indicator.

[0080] Step 2032: Using a computer, based on the correlation values ​​of the various air defense operation influencing factors with the first indicator and the second indicator, select the air defense operation influencing factors corresponding to the correlation values ​​greater than 0.1 and record them as significant air defense operation influencing factors.

[0081] In this embodiment, in step 3, air defense combat effectiveness simulation is performed based on orthogonal experiments according to significant influencing factors to obtain M′ sets of air defense combat effectiveness evaluation training sets. The specific process is as follows:

[0082] Step 301: Set the bth significant influencing factor X for air defense operations b "The minimum value is recorded as X b ” ,min , the bth significant influencing factor of air defense operations is X b The maximum value of " is recorded as X b ” ,max , set the bth significant influencing factor X for air defense operations b "The value is located at X b ” ,min ~X b ” ,max The step length when in range is recorded as X b ” ,s ;

[0083] Step 302: According to the method described in step 2022, B significant influencing factors of air defense operations are input to obtain M′ design schemes of significant influencing factors of air defense operations; wherein M′ is a positive integer;

[0084] Step 303: According to the method described in step 2023, the m′th design scheme for the significant influencing factor of air defense operations is input for simulation, and the m′th first index and the m′th second index corresponding to the m′th design scheme for the significant influencing factor of air defense operations are obtained; wherein m′ is a positive integer and 1≤m′≤M′;

[0085] Step 304: Set the m′th design scheme of the significant influencing factor of air defense operations as the input layer, the m′th first indicator and the m′th second indicator as the output layer, and obtain M′ groups of air defense operations effectiveness evaluation training sets.

[0086] In this embodiment, in step 4, a computer is used to perform variance analysis on the M air defense combat influencing factor design schemes, the M first indicators, and the M second indicators to obtain the optimal training data and the worst training data. The specific process is as follows:

[0087] Step 401: Using a computer, perform variance analysis on the M air defense combat influencing factor design schemes, the M first indices, and the M second indices to obtain a main effects plot of each air defense combat influencing factor and the average value of the first indices, as well as a main effects plot of each air defense combat influencing factor and the average value of the second indices; wherein the abscissa of the main effects plot of each air defense combat influencing factor and the average value of the first indices and the main effects plot of each air defense combat influencing factor and the average value of the second indices are the number of factor levels on the abscissa and the average value on the ordinate.

[0088] Step 402: Using a computer, based on the main effects plots of each air defense operation influencing factor and the average value of the first index, and the main effects plots of each air defense operation influencing factor and the average value of the second index, the optimal combination solution is recorded as the factor level number corresponding to the maximum average value of the first index and the minimum average value of the second index under each air defense operation influencing factor, and the worst combination solution is recorded as the factor level number corresponding to the minimum average value of the first index and the maximum average value of the second index under each air defense operation influencing factor;

[0089] The data corresponding to the significant influencing factors of air defense operations are selected from the optimal combination scheme and the worst combination scheme and recorded as the optimal training data and the worst training data.

[0090] In this embodiment, in step 101, a computer is used to establish an air defense combat simulation model based on ABMS using AnyLogic software.

[0091] In this embodiment, in step 401, a computer is used to perform variance analysis on the M air defense combat influencing factor design schemes, the M first indicators, and the M second indicators using miniTab software.

[0092] In this embodiment, in step 2023, a computer using AnyLogic software is used to input the mth air defense combat influencing factor design scheme into the air defense combat simulation model for simulation, thereby obtaining the mth first indicator and the mth second indicator corresponding to the mth air defense combat influencing factor design scheme.

[0093] In this embodiment, a neural network evaluation model is constructed in step 601, specifically: the number of nodes in the input layer is B, the number of nodes in the output layer is n and n is equal to 2, the number of nodes in the hidden layer is N, N = B + n + a′; where a′ is an adjustment coefficient, and its value ranges from 0 to 10.

[0094] In this embodiment, it is further preferred to set the number of input layer nodes to 5, the number of output layer nodes to 2, and the number of hidden layer nodes to between 3 and 13. After comparison, the best overall performance was achieved when the number of hidden layer nodes was 8. The training method used was Bayesian regularization, and the final neural network evaluation model structure was 5×8×2.

[0095] In this embodiment, it is further preferred that the first indicator is the interception rate and the second indicator is the enemy aircraft penetration distance. In actual use, other performance evaluation indicators can also be selected.

[0096] In this embodiment, the missile and artillery combined air defense system Agent model is as follows: Figure 2 As shown, Figure 2 The left side of the figure is the radar system agent model. Figure 2 The middle part is the Agent model of the control and command system. Figure 2The right side of the figure is the antiaircraft gun system agent model; the air defense missile agent model is as follows Figure 3 As shown, the antiaircraft artillery shell Agent model is as follows Figure 4 As shown, the enemy agent model is as follows Figure 5 shown.

[0097] In this embodiment, it should be noted that the agent model of the combined missile and artillery air defense system is the same as the defense vehicle in the real scene, and the agent model of the air defense missile, the agent model of the anti-aircraft artillery shell, and the agent model of the enemy aircraft are respectively the same as the air defense missile, anti-aircraft artillery shell, and the enemy aircraft in the real scene, thereby realizing the simulation of real air defense operations.

[0098] In this embodiment, the simulation of the missile and artillery combined air defense system Agent model is set as a fixed-position missile and artillery combined air defense system. A missile and artillery combined air defense system C is located in the south direction at a distance r from the defended target E. The enemy aircraft's dispatch direction, i.e., the heading angle θ and the route shortcut d, are visible. Figure 6 The frequency of enemy aircraft sorties follows the Poisson distribution law, and air defense missiles are launched in a one-to-one interception mode.

[0099] In this embodiment, ABMS is agent-based modeling and simulation.

[0100] In this embodiment, the value of L is 35.

[0101] In this embodiment, the value of A is 9, and the A air defense combat influencing factors to be studied are the enemy aircraft flight altitude, radar detection range, enemy aircraft sortie frequency, missile reloading time, enemy aircraft flight speed, artillery shell reloading time, enemy aircraft heading angle, the distance between the combined missile and artillery air defense system and the defended target, and route shortcuts.

[0102] In this embodiment, the orthogonal experiment design software in step 203 is SPSSAU software.

[0103] In this embodiment, the ath air defense operation influencing factor X' to be studied is used. a The value is located at X' a,min ~X' a,max The step length of the range is used to obtain the ath air defense combat influencing factor X' a The number of levels.

[0104] In this embodiment, the level numbers of the A air defense combat influencing factors to be studied are all 9.

[0105] In this embodiment, the design scheme for air defense combat influencing factors is 9 factors and 9 level values, so M is 81.

[0106] In this embodiment, variance analysis shows that the values ​​of the four factors of enemy aircraft flight altitude, enemy aircraft heading angle, route shortcut, and the distance between the combined missile and artillery air defense system and the defended target have little effect on the first indicator and the second indicator of the system, and there is no obvious correlation.

[0107] In this embodiment, the value of B is 5, and the B significant influencing factors of air defense operations are radar detection range, enemy aircraft sortie frequency, missile reloading time, enemy aircraft flight speed, and artillery shell reloading time.

[0108] In this embodiment, the correlation analysis in step 206 is Pearson correlation analysis or Spearman correlation analysis.

[0109] In this embodiment, B is 5, and the design scheme of air defense combat influencing factors is 5 factors and 9 level values, so M′ is 81, and M′+2 is 83.

[0110] In this embodiment, the neural network evaluation model is a BP neural network.

[0111] In this embodiment, it is further preferred that the number of nodes in the hidden layer is 8.

[0112] In this embodiment, when used specifically, a computer is used to utilize the NSGAⅡ algorithm in the PlatEMO platform, the number of decision variables is set to 5, objective function 1 is the penetration distance, objective function 2 is 1-the first index, the lower limits of the 5 decision variables are [2, 122, 10, 5, 100, 3], the upper limits of the 5 decision variables are [18, 298, 90, 45, 900, 27], the initial population size is 100, the number of evaluations is 10,000, and the Pareto frontier and the corresponding Pareto solution set are obtained.

[0113] In this embodiment, the five decision variables are radar detection range, enemy aircraft sortie frequency, missile reloading time, enemy aircraft flight speed, and artillery shell reloading time.

[0114] In this example, the Pareto solution set was compared with the M′ training set for air defense combat effectiveness evaluation. Within a certain error range, the Pareto solution set of the NSGA II algorithm was found to contain the orthogonal experiment design scheme. In this example, as shown in Table 1, the maximum error was 6.22% for the first indicator, demonstrating the credibility of the scheme obtained using the orthogonal experiment method.

[0115] Table 1 Comparison of design results between NSGAⅡ and orthogonal experiment

[0116]

[0117] In summary, the method of the present invention has simple steps and reasonable design. It establishes an air defense combat simulation model to simulate real air defense combat, which is more in line with reality. It obtains an air defense combat effectiveness evaluation training set through orthogonal experiments, reduces the problem of large simulation volume and time consumption, increases the input of optimal training data and worst training data, and improves the accuracy of air defense combat effectiveness evaluation.

[0118] The above description is only a preferred embodiment of the present invention and does not limit the present invention in any way. Any simple modification, change and equivalent structural change made to the above embodiment based on the technical essence of the present invention shall still fall within the scope of protection of the technical solution of the present invention.

Claims

1. A method for evaluating air defense combat effectiveness based on ABMS and orthogonal test, characterized in that: The method comprises the following steps: Step 1: Establish an air defense combat simulation model based on ABMS: Step 101: Using a computer to establish an air defense combat simulation model based on the ABMS; wherein the air defense combat simulation model includes an agent model of the combined missile and artillery air defense system, an agent model of the air defense missile, an agent model of the anti-aircraft artillery shell, and an agent model of the enemy aircraft; Step 102: Set the effectiveness evaluation index of the air defense combat simulation model to include the first index and the second index; set the L factor set that affects the evaluation index of the air defense combat simulation model to {X1,...,X l ,...,X L }; where X l represents the lth air defense operation influencing factor, l and L are both positive integers, and 1≤l≤L, and L is a positive integer not less than 3; Step 2: Conduct air defense combat effectiveness simulation based on orthogonal experiments to obtain significant influencing factors of air defense combat: Step 201: From the L factor set {X1,...,X l ,...,X L } and record them as A sets of influencing factors to be studied in air defense operations, namely {X1',...,X' a ,...,X' A }, where X' a represents the ath influencing factor of air defense operations to be studied, a and A are both positive integers, and 1≤a≤A, A is less than L; Step 202: Perform air defense combat effectiveness simulation based on an orthogonal test to obtain M first indices and M second indices corresponding to the M air defense combat influencing factor design schemes; Step 203: Use a computer to perform correlation analysis on the M air defense operation influencing factor design schemes, the M first indicators, and the M second indicators, and obtain air defense operation significant influencing factors from the A set of air defense operation influencing factors to be studied; wherein, each obtained air defense operation significant influencing factor is recorded as B air defense operation significant influencing factor sets, namely {X1",...,X b ”,...,X B ''}, where X b " represents the bth significant influencing factor of air defense operations, b and B are both positive integers, and 1≤b≤B, B is less than A; Step 3: Conduct air defense combat effectiveness simulation based on orthogonal experiments according to significant influencing factors: According to the significant influencing factors, air defense combat effectiveness simulation is carried out based on orthogonal experiments to obtain M′ groups of air defense combat effectiveness evaluation training sets; Step 4: Obtaining the best and worst training data for air defense combat effectiveness evaluation: A computer is used to perform variance analysis on M air defense combat influencing factor design schemes, M first indicators, and M second indicators to obtain the optimal training data and the worst training data; Step 5: Combine the M′ group of air defense combat effectiveness evaluation training sets, the optimal training data, and the worst training data to obtain M′+2 groups of air defense combat effectiveness evaluation training sets; Step 6: Training of the neural network evaluation model: Step 601: Construct a neural network evaluation model; Step 602: Using the significant influencing factors of air defense operations as the input layer and the first and second indicators as the output layer, the air defense operations effectiveness evaluation training set obtained in step 5 is input into the neural network evaluation model in step 601 for training, thereby obtaining a trained neural network evaluation model. Step 7: Evaluate the air defense combat effectiveness in real scenarios based on the trained neural network evaluation model.

2. The air defense combat effectiveness evaluation method based on ABMS and orthogonal testing according to claim 1 is characterized by: The missile-artillery combined air defense system agent model in step 101 includes a radar system agent model, an air defense command and control system agent model, and an antiaircraft gun system agent model.

3. The air defense combat effectiveness evaluation method based on ABMS and orthogonal test according to claim 1 or 2, characterized in that: In step 202, air defense combat effectiveness simulation is performed based on an orthogonal test to obtain M first indices and M second indices corresponding to the M air defense combat influencing factor design schemes. The specific process is as follows: Step 2021: Set the ath air defense operation influencing factor X' to be studied a The minimum value is recorded as X' a,min , the ath air defense operation influencing factor to be studied is X' a The maximum value is recorded as X' a,max , set the ath air defense operation influencing factor X' to be studied a The value is located at X' a,min ~X' a,max The step length when the range is set is recorded as X' a,s ; Step 2022: Input A air defense operation influencing factors to be studied, and use a computer to obtain design schemes for the air defense operation influencing factors using orthogonal experimental design software; wherein the total number of design schemes for the air defense operation influencing factors is M, where M is a positive integer; Step 2023: Using a computer, input the mth air defense combat influencing factor design scheme into the air defense combat simulation model for simulation, and obtain the mth first index and the mth second index corresponding to the mth air defense combat influencing factor design scheme; wherein m is a positive integer and 1≤m≤M; Step 2024: Repeat step 2023 multiple times to complete the simulation of M air defense combat influencing factor design schemes, and obtain M first indicators and M second indicators; In step 203, a computer is used to perform correlation analysis on the M air defense operation influencing factor design schemes, the M first indicators, and the M second indicators, and obtain significant air defense operation influencing factors from the A set of air defense operation influencing factors to be studied. The specific process is as follows: Step 2031: Using a computer and orthogonal experimental design software, perform a correlation analysis on the M air defense operation influencing factor design schemes, the M first indicators, and the M second indicators to obtain a correlation value of each air defense operation influencing factor with the first indicator and the second indicator. Step 2032: Using a computer, based on the correlation values ​​of the various air defense operation influencing factors with the first indicator and the second indicator, select the air defense operation influencing factors corresponding to the correlation values ​​greater than 0.1 and record them as significant air defense operation influencing factors.

4. The air defense combat effectiveness evaluation method based on ABMS and orthogonal testing according to claim 3 is characterized by: In step 3, the air defense combat effectiveness simulation is carried out based on the orthogonal test according to the significant influencing factors to obtain M′ groups of air defense combat effectiveness evaluation training sets. The specific process is as follows: Step 301: Set the bth significant influencing factor X for air defense operations b "The minimum value is recorded as X b ” ,min , the bth significant influencing factor of air defense operations is X b The maximum value of " is recorded as X b ” ,max , set the bth significant influencing factor X for air defense operations b "The value is located at X b ” ,min ~X b ” ,max The step length when in range is recorded as X b ” ,s ; Step 302: According to the method described in step 2022, B significant influencing factors of air defense operations are input to obtain M′ design schemes of significant influencing factors of air defense operations; wherein M′ is a positive integer; Step 303: According to the method described in step 2023, the m′th design scheme for the significant influencing factor of air defense operations is input for simulation, and the m′th first index and the m′th second index corresponding to the m′th design scheme for the significant influencing factor of air defense operations are obtained; wherein m′ is a positive integer and 1≤m′≤M′; Step 304: Set the m′th design scheme of the significant influencing factor of air defense operations as the input layer, the m′th first indicator and the m′th second indicator as the output layer, and obtain M′ groups of air defense operations effectiveness evaluation training sets.

5. The air defense combat effectiveness evaluation method based on ABMS and orthogonal testing according to claim 3 is characterized by: In step 4, a computer is used to perform variance analysis on the M air defense combat influencing factor design schemes, the M first indicators, and the M second indicators to obtain the optimal training data and the worst training data. The specific process is as follows: Step 401: Using a computer, perform variance analysis on the M air defense combat influencing factor design schemes, the M first indices, and the M second indices to obtain a main effects plot of each air defense combat influencing factor and the average value of the first indices, as well as a main effects plot of each air defense combat influencing factor and the average value of the second indices; wherein the abscissa of the main effects plot of each air defense combat influencing factor and the average value of the first indices and the main effects plot of each air defense combat influencing factor and the average value of the second indices are the number of factor levels on the abscissa and the average value on the ordinate. Step 402: Using a computer, based on the main effects plots of each air defense operation influencing factor and the average value of the first index, and the main effects plots of each air defense operation influencing factor and the average value of the second index, the optimal combination solution is recorded as the factor level number corresponding to the maximum average value of the first index and the minimum average value of the second index under each air defense operation influencing factor, and the worst combination solution is recorded as the factor level number corresponding to the minimum average value of the first index and the maximum average value of the second index under each air defense operation influencing factor; The data corresponding to the significant influencing factors of air defense operations are selected from the optimal combination scheme and the worst combination scheme and recorded as the optimal training data and the worst training data.

Images