System intelligent horizontal driving combat effectiveness prediction method based on small samples

Abstract

The invention relates to a method for predicting system intelligent horizontal driving combat effectiveness based on small samples, belongs to the technical field of combat system simulation and evaluation, and aims to solve the problem of data scarcity and technical bottleneck existing in combat system effectiveness prediction. A correlation degree matrix is obtained through mechanism analysis under the condition that simulation data samples are few, and prediction quantity of combat effectiveness change caused by intelligent horizontal change is superposed on simulation experiment samples, so that the prediction quantity of combat effectiveness is formed. Therefore, a set of system intelligent level combat effectiveness prediction method with good feasibility and interpretability under the condition of small simulation experiment samples is established, and effective theoretical and technical support is provided for intelligent upgrading of a combat system.

Description

Technical Field

[0001] The present invention relates to the technical field of combat system simulation and evaluation, and in particular to a method for predicting combat effectiveness driven by system intelligence level based on a small sample. Background Art

[0002] Modern warfare is characterized by a large number of equipment, high degree of informatization, deep coupling of combat elements, and rapid changes in battlefield situations. Uncertainty and complexity increase exponentially with the increase in the number of equipment, and systemized warfare has become the main mode of warfare. Future system confrontation warfare requires rapidity, automation, intelligence, and autonomy. Intelligent technology is urgently needed to assist commanders and soldiers in completing various combat tasks and improve the intelligence level of combat equipment in the system to adapt to the high-speed, complex, and changing battlefield environment. The intelligence level of the combat system affects combat effectiveness to a certain extent. Researching the correlation between intelligence level and combat effectiveness and using intelligence level to predict effectiveness can improve the intelligence level in a targeted manner and improve the effectiveness of the combat system.

[0003] At present, there are few studies on the correlation between the intelligence level of the combat system and its combat effectiveness. In addition, due to the high cost of building a combat simulation system, it is difficult to collect enough effectiveness data under different intelligence level configuration scenarios, which cannot meet the data volume requirements for predictive model training. Therefore, data-based prediction methods are difficult to implement.

[0004] Therefore, establishing a method for predicting combat effectiveness based on the system intelligence level with good feasibility and explainability under small simulation experimental sample conditions is an important topic in the field of combat system intelligence level and combat effectiveness evaluation, and provides effective theoretical and technical support for the intelligent upgrade of combat systems. Summary of the invention

[0005] In view of the above problems, the present invention provides a prediction method for the combat effectiveness driven by the system intelligence level based on a small sample. The present invention aims at the data scarcity problem and technical bottleneck in the prediction of combat system effectiveness, studies the acquisition of the correlation degree matrix through mechanism analysis when there are few simulation data samples, and superimposes the predicted amount of combat effectiveness changes caused by changes in intelligence level on the simulation experiment samples to form a prediction amount of combat effectiveness. Thus, a method for predicting combat effectiveness by the system intelligence level with good feasibility and explainability under the conditions of small simulation experiment samples is established, providing effective theoretical and technical support for the intelligent upgrade of the combat system.

[0006] The present invention provides a method for predicting the combat effectiveness of system intelligence level driving based on a small sample, comprising:

[0007] The first step is to build multiple intelligence level indicators and multiple combat effectiveness indicators of the combat system;

[0008] Preferably, the specific steps of constructing the intelligence level index and combat effectiveness index of the combat system in step 1 include:

[0009] Decomposing the combat mission into a plurality of basic mission links 1 and a plurality of basic mission links 2;

[0010] Decompose the intelligence level of the combat system according to the basic task links and construct corresponding intelligence level indicators;

[0011] Decomposing the combat effectiveness of the combat system according to the multiple basic task links 2, and constructing corresponding combat effectiveness indicators;

[0012] The second step is to analyze the correlation between each intelligence level indicator and each combat effectiveness indicator through mechanism;

[0013] Preferably, the specific steps of analyzing the correlation between each intelligence level indicator and each combat effectiveness indicator through mechanism in step 2 are:

[0014] Step 201: simultaneously determine the degree of dependence of the basic task link of each intelligence level indicator on human participation and decision-making and whether the basic task link directly affects each combat effectiveness indicator, obtain a determination result, and obtain the degree of correlation between each intelligence level indicator and each combat effectiveness indicator based on the determination result;

[0015] Step 202: Determine the correlation matrix W according to the correlation between each intelligence level indicator and each combat effectiveness indicator. m×n ; Where m is the number of intelligence level indicators, and n is the number of combat effectiveness indicators.

[0016] Furthermore, the degree of association includes five levels: very strong, relatively strong, medium, relatively weak and very weak;

[0017] Specifically, the specific steps of obtaining the correlation between each intelligence level indicator and each combat effectiveness indicator in step 201 include:

[0018] Determine the degree of dependence of basic task links at each intelligence level on human participation and decision-making. If the degree of dependence is small, it means that the corresponding intelligence level plays an important role and has a significant impact on combat effectiveness; if the degree of dependence is large, it means that the corresponding intelligence level plays a supporting role and will not have a significant impact on combat effectiveness;

[0019] Determine whether the basic task links of each intelligence level have a direct impact on each combat effectiveness indicator; if so, it means that the corresponding intelligence level is highly correlated with the combat effectiveness indicator, such as the early warning detection intelligence level has a high correlation with the target detection rate; if not, it means that the corresponding intelligence level will not have a significant impact on combat effectiveness.

[0020] The third step is to determine the small sample size required for the simulation experiment; based on the correlation between each intelligence level indicator and each combat effectiveness indicator, obtain the predicted change amount of the combat effectiveness indicator caused by the change amount of the intelligence level indicator corresponding to each sample;

[0021] On the basis of the combat effectiveness index of each sample, the predicted change amount of the combat effectiveness index caused by the change amount of the corresponding intelligence level index is superimposed to obtain multiple predicted combat effectiveness; taking the average of the multiple predicted combat effectiveness to obtain the final predicted combat effectiveness;

[0022] It can be understood that the small sample required for the simulation experiment is a sample when the amount of simulation experiment sample data is small;

[0023] Preferably, the specific steps of realizing the combat effectiveness prediction driven by the intelligence level of the combat system based on the final prediction value in step 3 are:

[0024] Step 301, determine the small sample size required for the simulation experiment; when i=1, it is the first sample;

[0025] Step 302: Obtain the intelligence level index X corresponding to the i-th sample i The change in ΔX i ;

[0026] Furthermore, in step 302, the intelligence level index X corresponding to the i-th sample i The change in ΔX i The expression is:

[0027] ΔX i =XX i

[0028] Among them, X is the intelligence level to be predicted;

[0029] Step 303: Obtain the intelligence level index X corresponding to the i-th sample based on the correlation between each intelligence level and each combat effectiveness. i The predicted change in combat effectiveness index ΔY caused by the change in i ;

[0030] Furthermore, in step 303, the intelligence level index X corresponding to the i-th sample i The predicted change in combat effectiveness index ΔY caused by the change ini The expression is:

[0031] ΔY i =W m×n T ΔX i

[0032] Step 304: The combat effectiveness index Y corresponding to the i-th sample i On the basis of superimposing the intelligence level index X corresponding to the i-th sample i The predicted change in combat effectiveness index caused by the change in i ′;

[0033] Furthermore, the corresponding predicted combat effectiveness Y in step 304 i The expression of ′ is:

[0034] Y i ′=Y i +ΔY i =W m×n T ·(XX i )+Y i

[0035] Step 305: Set the combat effectiveness threshold; determine the corresponding predicted combat effectiveness Y i ′Whether it exceeds the combat effectiveness threshold, if it exceeds, the corresponding predicted combat effectiveness Y i 'Truncate; if not exceeded, proceed to the next step;

[0036] Step 306: Repeat steps 302-305, traverse all samples, obtain the predicted combat effectiveness corresponding to each sample, take the average, and obtain the final predicted combat effectiveness Y.

[0037] Furthermore, the expression of the final predicted combat effectiveness Y is:

[0038]

[0039] Wherein, k is the small sample size described in the simulation experiment.

[0040] The fourth step is to realize the combat effectiveness prediction driven by the intelligence level of the combat system based on the final predicted combat effectiveness.

[0041] Compared with the prior art, the present invention has at least the following beneficial effects:

[0042] (1) The prediction method of the present invention has low requirements on the number of simulation experiment samples, and prediction can also be performed under single sample conditions;

[0043] (2) The prediction method of the present invention has a clear mathematical expression and good computability and interpretability;

[0044] (3) The prediction method of the present invention has concise and clear steps, is simple to implement, has low complexity, and is easy to integrate into a simulation system or evaluation software. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] The drawings are only for the purpose of illustrating particular embodiments and are not to be construed as limiting the invention.

[0046] Figure 1 It is a schematic diagram of a flow chart of the operational effectiveness driven by the intelligence level of a small sample prediction system in an embodiment of the present invention;

[0047] Figure 2 Schematic diagram of a correlation degree matrix in an embodiment of the present invention;

[0048] Figure 3 A schematic diagram of the intelligence level index of the combat system in an embodiment of the present invention;

[0049] Figure 4 Schematic diagram of combat effectiveness indicators in an embodiment of the present invention. DETAILED DESCRIPTION

[0050] In order to more clearly understand the above-mentioned purpose, features and advantages of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that, in the absence of conflict, the embodiments of the present invention and the features in the embodiments can be combined with each other. In addition, the present invention can also be implemented in other ways different from those described herein, and therefore, the protection scope of the present invention is not limited by the specific embodiments disclosed below.

[0051] A specific embodiment of the present invention, as Figure 1-4 , discloses a prediction method for the combat effectiveness driven by the system intelligence level based on a small sample. In order to illustrate the effectiveness of the method proposed by the present invention, the above technical solution of the present invention is described in detail through a specific embodiment below. The specific implementation steps are as follows:

[0052] The present invention provides a method for predicting the combat effectiveness of system intelligence level driving based on a small sample, comprising:

[0053] The first step is to build multiple intelligence level indicators and multiple combat effectiveness indicators of the combat system;

[0054] Preferably, the specific steps of constructing the intelligence level index and combat effectiveness index of the combat system in step 1 include:

[0055] Decomposing the combat mission into a plurality of basic mission links 1 and a plurality of basic mission links 2;

[0056] Decompose the intelligence level of the combat system according to the basic task links and construct corresponding intelligence level indicators;

[0057] Decomposing the combat effectiveness of the combat system according to the multiple basic task links 2, and constructing corresponding combat effectiveness indicators;

[0058] The second step is to analyze the correlation between each intelligence level indicator and each combat effectiveness indicator through mechanism;

[0059] Preferably, the specific steps of analyzing the correlation between each intelligence level indicator and each combat effectiveness indicator through mechanism in step 2 are:

[0060] Step 201: simultaneously determine the degree of dependence of the basic task link of each intelligence level indicator on human participation and decision-making and whether the basic task link directly affects each combat effectiveness indicator, obtain a determination result, and obtain the degree of correlation between each intelligence level indicator and each combat effectiveness indicator based on the determination result;

[0061] Step 202: Determine the correlation matrix W according to the correlation between each intelligence level indicator and each combat effectiveness indicator. m×n ; Where m is the number of intelligence level indicators, and n is the number of combat effectiveness indicators.

[0062] Furthermore, the degree of association includes five levels: very strong, relatively strong, medium, relatively weak and very weak;

[0063] Table 1 Correlation degree and its quantitative value

[0064] Degree of association Quantization value Very strong 1 Strong 0.75 medium 0.5 Weaker 0.25 Very weak 0

[0065] Specifically, the specific steps of obtaining the correlation between each intelligence level indicator and each combat effectiveness indicator in step 201 include:

[0066] Determine the degree of dependence of basic task links at each intelligence level on human participation and decision-making. If the degree of dependence is small, it means that the corresponding intelligence level plays an important role and has a significant impact on combat effectiveness; if the degree of dependence is large, it means that the corresponding intelligence level plays a supporting role and will not have a significant impact on combat effectiveness;

[0067] Determine whether the basic task links of each intelligence level have a direct impact on each combat effectiveness indicator; if so, it means that the corresponding intelligence level is highly correlated with the combat effectiveness indicator, such as the early warning detection intelligence level has a high correlation with the target detection rate; if not, it means that the corresponding intelligence level will not have a significant impact on combat effectiveness.

[0068] The third step is to determine the small sample size required for the simulation experiment; based on the correlation between each intelligence level indicator and each combat effectiveness indicator, obtain the predicted change amount of the combat effectiveness indicator caused by the change amount of the intelligence level indicator corresponding to each sample;

[0069] On the basis of the combat effectiveness index of each sample, the predicted change amount of the combat effectiveness index caused by the change amount of the corresponding intelligence level index is superimposed to obtain multiple predicted combat effectiveness; taking the average of the multiple predicted combat effectiveness to obtain the final predicted combat effectiveness;

[0070] It can be understood that the small sample required for the simulation experiment is a sample when the amount of simulation experiment sample data is small;

[0071] Preferably, the specific steps of realizing the combat effectiveness prediction driven by the intelligence level of the combat system based on the final prediction value in step 3 are:

[0072] Step 301, determine the small sample size required for the simulation experiment; when i=1, it is the first sample;

[0073] Step 302: Obtain the intelligence level index X corresponding to the i-th sample i The change in ΔX i ;

[0074] Furthermore, in step 302, the intelligence level index X corresponding to the i-th sample i The change in ΔX i The expression is:

[0075] ΔX i =XX i

[0076] Among them, X is the intelligence level to be predicted;

[0077] Step 303: Obtain the intelligence level index X corresponding to the i-th sample based on the correlation between each intelligence level and each combat effectiveness. i The predicted change in combat effectiveness index ΔY caused by the change in i ;

[0078] Furthermore, in step 303, the intelligence level index X corresponding to the i-th sample i The predicted change in combat effectiveness index ΔY caused by the change in i The expression is:

[0079] ΔY i =W m×n T ΔX i

[0080] Step 304: The combat effectiveness index Y corresponding to the i-th sample i On the basis of superimposing the intelligence level index X corresponding to the i-th sample i The predicted change in combat effectiveness index caused by the change in i ′;

[0081] Furthermore, the corresponding predicted combat effectiveness Y in step 304 i The expression of ′ is:

[0082] Y i ′=Y i +ΔY i =W m×n T ·(XX i )+Y i

[0083] Step 305: Set the combat effectiveness threshold; determine the corresponding predicted combat effectiveness Y i ′Whether it exceeds the combat effectiveness threshold, if it exceeds, the corresponding predicted combat effectiveness Y i 'Truncate; if not exceeded, proceed to the next step;

[0084] Step 306: Repeat steps 302-305, traverse all samples, obtain the predicted combat effectiveness corresponding to each sample, take the average, and obtain the final predicted combat effectiveness Y.

[0085] Furthermore, the expression of the final predicted combat effectiveness Y is:

[0086]

[0087] Wherein, k is the small sample size described in the simulation experiment.

[0088] Step 4: Based on the final predicted combat effectiveness, the combat effectiveness prediction driven by the intelligence level of the combat system is realized.

[0089] First, construct the indicators of intelligence level and combat effectiveness of the combat system;

[0090] like Figure 3 As shown in the figure, the intelligence level of the combat system is taken as the root node, and the intelligence level of combat planning, early warning detection, command and control, interception and attack, combat assessment and comprehensive support are taken as indicators; Figure 4 As shown in the figure, the combat effectiveness indicators are decomposed into effectiveness indicators such as target detection probability, target identification probability, target tracking probability, situation construction coverage, situation analysis accuracy, decision-making plan generation rate, guidance success rate, damage rate, mission completion rate, and cost-effectiveness ratio.

[0091] Then, the correlation matrix between intelligence level and combat effectiveness is obtained through mechanism analysis.

[0092] As shown in Table 2, the correlation between each intelligence level indicator and each performance indicator can be obtained, and the correlation degree matrix can be quantified and converted into the following:

[0093]

[0094] Table 1 Correlation degree table

[0095]

[0096]

[0097]

[0098] Finally, the predicted value of performance change is superimposed on the existing small simulation experiment samples to achieve performance prediction driven by intelligent level.

[0099] Suppose through simulation we get an intelligence level sample X1 = [0.4 0.8 0.5 0.7 0.5 0.3] T And the corresponding performance data Y1 = [0.8 0.7 0.6 0.6 0.6 0.7 0.8 0.7 0.6 0.7] T Then the intelligence level can be obtained as X = [0.3 0.7 0.4 0.6 0.4 0.2] T , the corresponding performance prediction value

[0100] Y=[0.675 0.55 0.45 0.475 0.475 0.5 0.55 0.45 0.375 0.475] T .

[0101] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by any technician familiar with the technical field within the technical scope disclosed by the present invention should be covered within the protection scope of the present invention.

Claims

1. A prediction method for system intelligence level driven combat effectiveness based on small samples, characterized in that: include: The first step is to build multiple intelligence level indicators and multiple combat effectiveness indicators of the combat system; The second step is to analyze the correlation between each intelligence level indicator and each combat effectiveness indicator through mechanism; The third step is to determine the small sample size required for the simulation experiment; based on the correlation between each intelligence level indicator and each combat effectiveness indicator, obtain the predicted change amount of the combat effectiveness indicator caused by the change amount of the intelligence level indicator corresponding to each sample; Superimposing the predicted change amount of the combat effectiveness index caused by the change amount of the corresponding intelligence level index on the combat effectiveness index of each sample to obtain multiple predicted combat effectiveness; taking the average of the multiple predicted combat effectiveness to obtain the final predicted combat effectiveness; The fourth step is to realize the combat effectiveness prediction driven by the intelligence level of the combat system based on the final predicted combat effectiveness.

2. The method for predicting the combat effectiveness driven by the system intelligence level according to claim 1 is characterized in that: The specific steps of constructing the intelligence level index and combat effectiveness index of the combat system in step 1 include: Decomposing the combat mission into a plurality of basic mission links 1 and a plurality of basic mission links 2; Decompose the intelligence level of the combat system according to the basic task links and construct corresponding intelligence level indicators; According to the multiple basic task links, the combat effectiveness of the combat system is decomposed and corresponding combat effectiveness indicators are constructed.

3. The method for predicting the combat effectiveness driven by the system intelligence level according to claim 1 is characterized in that: The specific steps of analyzing the correlation between each intelligence level indicator and each combat effectiveness indicator through mechanism in step 2 are as follows: Step 201: simultaneously determine the degree of dependence of the basic task link of each intelligence level indicator on human participation and decision-making and whether the basic task link directly affects each combat effectiveness indicator, obtain a determination result, and obtain the degree of correlation between each intelligence level indicator and each combat effectiveness indicator based on the determination result; Step 202: Determine the correlation matrix W according to the correlation between each intelligence level indicator and each combat effectiveness indicator. m×u ; Where m is the number of intelligence level indicators, and n is the number of combat effectiveness indicators.

4. The method for predicting the combat effectiveness driven by the system intelligence level according to claim 3 is characterized in that: The specific steps of obtaining the correlation between each intelligence level indicator and each combat effectiveness indicator in step 201 include: Determine the degree of dependence of basic task links at each intelligence level on human participation and decision-making. If the degree of dependence is small, it means that the corresponding intelligence level plays an important role and has a significant impact on combat effectiveness; if the degree of dependence is large, it means that the corresponding intelligence level plays a supporting role and will not have a significant impact on combat effectiveness; Determine whether the basic task links of each intelligence level have a direct impact on each combat effectiveness indicator; if so, it means that the corresponding intelligence level is highly correlated with the combat effectiveness indicator, such as the early warning detection intelligence level has a high correlation with the target detection rate; if not, it means that the corresponding intelligence level will not have a significant impact on combat effectiveness.

5. The method for predicting the combat effectiveness driven by the system intelligence level according to claim 1 is characterized in that: The small sample required for the simulation experiment is a sample when the amount of simulation experiment sample data is small.

6. The method for predicting the combat effectiveness driven by the system intelligence level according to claim 3 is characterized in that: The specific steps of realizing the combat effectiveness prediction driven by the intelligence level of the combat system based on the final prediction value in step 3 are as follows: Step 301, determine the small sample size required for the simulation experiment; when i=0, it is the first sample; Step 302: Obtain the intelligence level index X corresponding to the i-th sample i The change in ΔX i ; Step 303: Obtain the intelligence level index X corresponding to the i-th sample based on the correlation between each intelligence level and each combat effectiveness. i The predicted change in combat effectiveness index ΔY caused by the change in i ; Step 304: The combat effectiveness index Y corresponding to the i-th sample i The intelligence level index X corresponding to the i-th sample is superimposed i The predicted change in combat effectiveness index caused by the change in i ′; Step 305: Set the combat effectiveness threshold; determine the corresponding predicted combat effectiveness Y i ′Whether it exceeds the combat effectiveness threshold, if it exceeds, the corresponding predicted combat effectiveness Y i 'Truncate; if not exceeded, proceed to the next step; Step 306: Repeat steps 302-305, traverse all samples, obtain the predicted combat effectiveness corresponding to each sample, take the average, and obtain the final predicted combat effectiveness Y.

7. The method for predicting the combat effectiveness driven by the system intelligence level according to claim 6 is characterized in that: The corresponding predicted combat effectiveness Y in step 304 i The expression of ′ is: Y i ′=Y i +ΔY i =W m×n T ·(XX i )+Y i 。 8. The method for predicting the combat effectiveness driven by the system intelligence level according to claim 7 is characterized in that: The final predicted combat effectiveness Y is expressed as: Wherein, k is the small sample size described in the simulation experiment.

Images