Simulation-deduction-oriented combat scheme quality-effect analysis method

Abstract

The invention provides a simulation deduction-oriented combat scheme quality and efficiency analysis method, which comprises the following steps of: establishing an efficiency index set directly related to a combat mission from the angle of a red party, and weighting the efficiency indexes; carrying out large sample deduction on all alternative combat schemes in the red party alternative combat scheme set based on a combat simulation system, quantifying an efficiency index set by using deduction result data to form an efficiency cuboid, and obtaining an index efficiency set of each alternative combat scheme; based on the Bayesian network, establishing a combat capability conversion mechanism model comprising a reference capability generation layer, a positive capability generation layer, a negative capability generation layer and an efficiency generation layer; sampling the combat capability conversion mechanism model to obtain a sample set, and deducing the negative capability of each efficiency index; and calculating the efficiency generation rate and the capability utilization rate of the alternative combat scheme based on the index efficiency set of the alternative combat scheme, the combat efficiency calculation function and the efficiency index negative capability, and further determining the combat scheme quality-efficiency level of the alternative combat scheme.

Description

Technical Field

[0001] The present invention relates to the technical field of military operations research, and in particular to a method for analyzing the quality and effectiveness of combat plans oriented to simulation deduction. Background Art

[0002] An operational plan is a specific operational plan and action plan formulated to achieve mission objectives during a military operation. For a specific combat mission, the staff department typically drafts multiple alternative operational plans and evaluates and analyzes them from multiple perspectives, such as damage, survivability, and cost-effectiveness, to assist commanders in their decision-making. With the advancement of computer technology, combat simulation has become a crucial support for scenario analysis. Combat simulation systems support highly realistic reproduction of complex battlefield environments and the efficient and low-cost collection of large-sample scenario simulation data, enabling lean analysis of operational plans, including quality and effectiveness analysis.

[0003] Quality and effectiveness analysis focuses on two core aspects of an operational plan: operational effectiveness and operational benefits. Operational effectiveness refers to the degree to which combat operations achieve their intended operational objectives in the actual battlefield environment. The higher the operational effectiveness, the more adaptable the operational plan is to the mission. Operational benefits, on the other hand, concern the efficiency with which operational capability is converted into operational effectiveness. Operational capability is the inherent level of achievement of the intended mission objectives within the operational plan and is determined by the various operational resources invested. This includes not only quantifiable, explicit elements such as personnel, equipment, and supplies, but also less quantifiable, implicit elements such as technical proficiency, logistical support, and tactical application. The relationship between operational capability and operational effectiveness is similar to the relationship between input and output. The higher the efficiency with which capability is converted into effectiveness, the greater the operational benefits and the correspondingly more cost-effectiveness of the operational plan. Scientific and effective quality and effectiveness analysis has important application value in guiding operational resource allocation and optimizing operational plans.

[0004] However, the current level of quality and effectiveness analysis of combat plans is only at the level of combat effectiveness assessment, and the analysis of combat benefits is not yet in-depth. Generally, only the cost-effectiveness of the plan can be calculated, but this cannot take into account the implicit factors of combat capability, and it is difficult to answer deep-seated questions related to the efficiency of combat capability conversion, such as whether combat capability has been fully converted or whether combat effectiveness is generated in the most economical way. This greatly restricts the analysis and optimization of combat plans. Summary of the Invention

[0005] In view of this, the present invention provides a method for analyzing the quality and effectiveness of combat plans based on simulation and deduction, comprising:

[0006] Step 1: From the perspective of the Red Army, based on each link and element of the combat process, establish a set of effectiveness indicators I that are directly related to the combat mission, and assign weights W to the established effectiveness indicators; where I = {I m |m=1~N I},W={ωm |m=1~N I}, where N I is the number of performance indicators, I m is the mth performance indicator, ω m Performance Index I m The weight of

[0007] Step 2: Based on the combat simulation system, conduct a large-sample simulation of all the alternative combat plans in the Red side's alternative combat plan set C, and use the simulation result data to quantify the effectiveness index set I to form an effectiveness cuboid S; where C = {C n |n=1~N C}, N C is the number of alternative combat options, C n is the nth alternative combat plan; the dimension of the efficiency cuboid S is N C ×N I ×N S , N S The number of independent deductions for each alternative combat plan, any element S in S nmk Representative Effectiveness Index I m In Alternative Operations Plan C n The effectiveness value obtained in the kth deduction is used to obtain the alternative combat plan C n The performance set U of the indicator n ;

[0008] Step 3: Based on the Bayesian network, establish a combat capability conversion mechanism model that includes a baseline capability generation layer, a positive capability generation layer, a negative capability generation layer, and an effectiveness generation layer. The baseline capability in the baseline capability generation layer is used to obtain the corresponding positive capability in the positive capability generation layer, the negative capability generation layer forms the negative capability, and the effectiveness value in the effectiveness generation layer is obtained based on the positive and negative capabilities.

[0009] Step 4: Sampling the combat capability transformation mechanism model established by the predetermined number of times to obtain the sample set Based on the obtained sampling samples, it is inferred that each performance indicator has the negative capacity {β m |m=1~N I}; where L is the number of sampling times, For all β m The posterior distribution P(β m |S)’s i-th sample;

[0010] Step 5, based on alternative operation plan C n The performance set U of the indicator n , combat effectiveness calculation function, and the inferred negative capacity of each effectiveness indicator, calculate the alternative combat plan C nThe efficiency generation rate θ S and capacity utilization θ α ; Based on the calculated effectiveness generation rate and capacity utilization rate, determine the alternative combat plan C n The quality and effectiveness level of the combat plan is V.

[0011] Furthermore, in step 1, the effectiveness indicator set I established includes: emergency response effectiveness, support effectiveness, reconnaissance effectiveness, mobility effectiveness, strike effectiveness, survival effectiveness, and coordination effectiveness.

[0012] Furthermore, in step 3, the combat capability conversion mechanism model is specifically as follows:

[0013] The benchmark capability generation layer contains the node set {α n |n=1~N C}, where node α n Indicates alternative combat plan C n Intrinsic overall combat capability:

[0014]

[0015] The positive capacity generation layer contains a set of nodes The nodes Indicates alternative combat plan C n In the performance index I m Actual combat capabilities demonstrated:

[0016]

[0017] The negative capacity generation layer contains the node set {β m |m=1~N I}, where node β m Indicates the blue team and the complex battlefield environment factors in the effectiveness index I m The ability to resist the alignment ability shown above:

[0018]

[0019] The performance generation layer includes a node set {S nmk |n=1~N C ,m=1~N I ,k=1~N S}, is the performance cuboid S:

[0020]

[0021] in, is the truncated normal distribution on the interval [0,+∞), LNormal(·) is the lognormal distribution, Beta(·) is the Beta distribution, μ α , τα ,τ,μ β , τ β are the parameters of the corresponding distributions, and ln(·) is the natural logarithm function.

[0022] Furthermore, in step 4, β m Approximately the sample mean

[0023]

[0024] Where avg(·) is the average value function.

[0025] Furthermore, in step 5, the combat effectiveness calculation function Specifically:

[0026]

[0027] in, and Are any indicator performance set and corresponding weight set, satisfying

[0028] Furthermore, in step 5, alternative combat plan C n The performance set U of the indicator n Specifically:

[0029] U n ={u nm |m=1~N I}

[0030] Among them, u nm For indicator I m Based on alternative combat plan C n The mean performance of all deduction samples:

[0031] u nm =avg({S nmk |k=1~N S}).

[0032] Furthermore, in step 5, for alternative combat plan C n , efficiency generation rate θ S Determined by the following formula:

[0033]

[0034] Among them, U S is the solution to the following optimization problem:

[0035]

[0036] Furthermore, in step 5, for alternative combat plan C n , capacity utilization θ α Determined by the following formula:

[0037]

[0038] in, is the solution to the following optimization problem:

[0039]

[0040] Further, in step 5, the operational plan C to be analyzed n The quality and effectiveness level V of the combat plan is determined by the following formula:

[0041]

[0042] The beneficial effects of the present invention are: first, although the quality and efficiency level in the present invention is a single numerical value, it is a comprehensive indicator of combat effectiveness and combat benefit, and has a clear and intuitive graphical form, which can effectively represent the degree of optimization (adaptability and economy) of the combat plan; second, the combat capability conversion mechanism model proposed in the present invention, on the one hand, establishes a probabilistic correlation between capability and effectiveness, and on the other hand, can measure the explicit and implicit elements of combat capability at the abstract level, which is conducive to more objective and comprehensive quantification of the generation basis of combat effectiveness. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the technical solutions of the specific embodiments of the present invention, the following is a brief introduction to the drawings required for use in the description of the specific embodiments. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without paying any creative work.

[0044] Figure 1 The implementation steps of the method of the present invention;

[0045] Figure 2 The Bayesian network of the combat capability conversion mechanism model in the present invention;

[0046] Figure 3 It is a graphic representation of the quality and effectiveness levels in the present invention;

[0047] Figure 4 Graph showing the quality and efficiency levels in an embodiment of the present invention. DETAILED DESCRIPTION

[0048] To make the purpose, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without making creative efforts shall fall within the scope of protection of the present invention.

[0049] The implementation steps of the method of the present invention are shown in Figure 1 .

[0050] The simulation-based combat plan quality and effectiveness analysis method includes the following steps:

[0051] (1) Constructing operational effectiveness evaluation indicators

[0052] From the perspective of the Red Team, we should adopt the idea of ​​combat effectiveness evaluation, fully consider all aspects and elements of the combat process, establish a set of indicators directly related to the combat mission, ensure the systematicity, completeness, distinguishability and quantifiability of the indicators, reasonably control the scale of indicators, highlight the main indicators, and use existing technologies (including but not limited to Delphi method, priority diagram method, hierarchical analysis method, etc.) to empower indicators.

[0053] Formally, let the index set and weight set be I = {I m |m=1~N I} and W={ω m |m=1~N I}, where N I is the number of indicators, I m is the mth indicator, ω m For indicator I m The weight of

[0054] In this example, a typical joint combat scenario is used as the background. The Red Army command organization determines the combat mission and combat determination at this level by understanding the combat intention of the superior, and jointly formulates the combat effectiveness evaluation indicators with the combat evaluation personnel. For simplicity, only N is listed here. I = 7 key indicators, see Table 1. The Delphi method was used to assign weights to indicators.

[0055] Table 1

[0056]

[0057]

[0058] (2) Scheme deduction and indicator quantification

[0059] For all the alternative combat plans of the Red Army, large-sample deductions are conducted based on the combat simulation system, and the deduction result data is used to quantify the indicator set to form an effectiveness rectangle.

[0060] Formally, the set of equipment selection schemes is C = {C n |n=1~N C}, where N C is the number of options, C n is the nth solution. The efficiency cuboid is S, and its dimension is N C ×N I ×N S , where N S The number of independent deductions for each solution is recommended to be N S ≥50. Any element S in S nmk Representative Index I m In Plan C n The effectiveness value obtained in the kth deduction of nmk They are all benefit-type (i.e. the larger the value, the better) and dimensionless, and belong to the same interval [0,1].

[0061] In the embodiment, the Red Army staff department formulates N by adjusting factors such as task force, equipment allocation, attack target, and action sequence. C = 3 alternative combat plans {C1, C2, C3}, which are used as the assumption input of a certain combat simulation system, and each plan is tested N times. S = 100 independent deductions, ultimately forming a 3 × 7 × 100-dimensional performance cuboid S. The calculation method for the performance value of each indicator is shown in the indicator definition in Table 1.

[0062] (3) Establishing a combat capability transformation mechanism model

[0063] Based on the Bayesian network, we establish Figure 2 The combat capability transformation mechanism model is shown in Figure 1, where circular nodes represent random variables, directed edges represent conditional probabilistic dependencies between random variables, and rectangular boxes represent the replication of their internal structure. The transformation mechanism is abstracted into four levels in the network:

[0064] ① Benchmark capability generation layer: contains node set {α n |n=1~N C}, where node α n Indicates plan C n The inherent overall combat capability, called baseline capability, is a comprehensive measure of both explicit and implicit elements of combat capability;

[0065] ②Positive Capacity Generation Layer: Contains a set of nodes The nodes Indicates plan C n In indicator Im The actual combat capability demonstrated is called positive capability, which has an improving effect on the effectiveness of indicators;

[0066] ③ Negative capacity generation layer: contains the node set {β m |m=1~N I}, where node β m Indicates that factors such as the blue team and the complex battlefield environment are in the index I m The ability to counteract the positive ability shown above is called negative ability, which has a weakening effect on the effectiveness of indicators;

[0067] ④Effectiveness generation layer: contains node set {S nmk |n=1~N C ,m=1~N I ,k=1~N S}, which is the performance cuboid in step (2).

[0068] Let the random variable satisfy the following probability distribution:

[0069]

[0070] in, is the truncated normal distribution on the interval [0,+∞), LNormal(·) is the lognormal distribution, Beta(·) is the Beta distribution, μ α , τ α 、μ β , τ β and τ are the parameters of the corresponding distributions, and ln(·) is the natural logarithm function.

[0071] In the embodiment, let the parameter μ α =μ β =0, τ α =τ β =0.1,τ=1.

[0072] (4) Inferring the negative capacity of indicators

[0073] Inference indicator negative ability {β m |m=1~N I}, we need to first infer all β m The posterior distribution P(β m |S). However, it is usually difficult to give the exact analytical expressions of these posterior distributions, so an approximate method is used here to calculate P(β m |S) Perform MCMC (Markov Chain Monte Carlo) sampling (existing technology) to generate a sample set Where L is the number of sampling times, and it is recommended to take L≥1000. is the posterior distribution P(βm |S) is the i-th sample. Thus, β m Can be approximated as the sample mean

[0074]

[0075] Where avg(·) is the average value function.

[0076] In the embodiment, in MultiBUGS Bayesian network modeling software 1 In step (3), the model is programmed and the sampling times L=10000 are set to start the P(β m |S) and then find the negative capacity of the indicator (in an approximate sense) based on the sampled samples, as shown in the last column of Table 1.

[0077] (5) Analysis of the quality and effectiveness of operational plans

[0078] To implement this step, you first need to:

[0079] ① Select the solution C to be analyzed n , calculate its index performance set U n :

[0080] U n ={u nm |m=1~N I}

[0081] where u nm For indicator I m Based on Scheme C n The mean performance of all deduction samples is:

[0082] u nm =avg({S nmk |k=1~N S})

[0083] ②Select the combat effectiveness calculation function in and Are any indicator performance set and corresponding weight set, satisfying The functional form of Γ(·) includes but is not limited to weighted sum, weighted product, etc.

[0084] In the embodiment, C3 is selected as the scheme to be analyzed, and U3 is calculated as shown in Table 2. The combat effectiveness calculation function uses the weighted product, that is:

[0085]

[0086] On this basis, operational effectiveness is divided into two indicators: capability generation rate and capability utilization rate. The quality and effectiveness analysis of Plan C3 is carried out according to the following steps:

[0087] (5.1) Calculate the efficiency generation rate

[0088] Given the weight set W in step (1), solve the problem of the performance set of indicators is a constrained optimization problem with decision variables:

[0089] maximize U Γ(U,W)

[0090]

[0091] The optimization problem aims to answer how much combat effectiveness a given combat capability can generate at most. (1) is an isocapability surface, and all points on the surface have the same n The same summed positive power on the index set I. Using the SLSQP method (existing technology) to solve the optimization problem, we get the solution U S , then the efficiency generation rate θ S Defined as:

[0092]

[0093] According to this definition and optimization characteristics, on the one hand, θ S ∈[0,1], on the other hand, when keeping the total positive capacity unchanged, there is an optimal solution whose combat effectiveness is better than solution C. n higher Obviously θ S The larger it is, the higher the rate of combat effectiveness generation.

[0094] In the embodiment, the minimize function under the optimize module in the Python language scipy library is used to solve the optimization problem, and the solver adopts the SLSQP method. S See Table 2, the corresponding θ S =0.788, which shows that by fully tapping the potential of existing combat capabilities and keeping the total positive capacity unchanged, the combat effectiveness can be improved by about 27% relative to Plan C3.

[0095] (5.2) Computing capacity utilization

[0096] Given the weight set W in step (1), solve the problem of the performance set of indicators is a constrained optimization problem with decision variables:

[0097]

[0098] Γ(U,W)=Γ(Un ,W)

[0099] The optimization problem aims to answer the minimum combat capability required to generate a given combat effectiveness. (2) is the iso-efficiency surface, and all points on the surface have the same n The same combat effectiveness value. Using the SLSQP method (existing technology) to solve the optimization problem, the solution is obtained. Then the capacity utilization rate θ α Defined as:

[0100]

[0101] According to this definition and optimization characteristics, on the one hand, θ α ∈[0,1], on the other hand, when keeping the combat effectiveness unchanged, there is an optimal solution whose combat capability requirement is greater than that of solution C. n Lower than (1-θ α )×100%, which essentially means saving of combat resources. α The larger it is, the higher the utilization rate of combat capability.

[0102] In the embodiment, the minimize function under the optimize module in the Python language scipy library is used to solve the optimization problem, and the solver adopts the SLSQP method. α See Table 2, the corresponding θ α =0.647, which shows that by re-planning the allocation of combat resources, the combat capability requirements can be reduced by about 35% relative to Plan C3 while maintaining the same combat effectiveness.

[0103] Table 2

[0104]

[0105] (5.3) Calculating the quality and effectiveness of combat plans

[0106] Calculate the quality and effectiveness of the combat plan. Let Plan C n The quality and efficiency level V is:

[0107]

[0108] Where (θ S +θ α ) / 2 is called the operational benefit factor. The larger its value is, the better the economic efficiency of the plan is. In particular, when (θ S +θ α ) / 2=1, the plan has reached the theoretical optimum in terms of economic efficiency: it is impossible to improve combat effectiveness without increasing combat capability, nor is it possible to reduce combat capability requirements without weakening combat effectiveness. The geometric meaning of the quality and efficiency level V is Figure 3 The area of ​​the shaded triangle is obviously V∈[0,1]. The larger the area, the higher the quality and efficiency level.

[0109] In the embodiment, the operational effectiveness of scheme C3 is Γ(U3,W) = 0.392, and its quality efficiency level is V = 0.281. The corresponding area is shown in Figure 4 Obviously, the quality and efficiency level of Plan C3 is low.

[0110] From the above introduction, it can be seen that the present invention provides a simulation-based combat plan quality and effectiveness analysis method. Within the framework of Bayesian networks, based on the Beta regression concept, a probabilistic mechanism model for the conversion of combat capability to combat effectiveness is established. Relying on large-sample simulation and deduction data of combat plans, the level of countermeasures implicit in the combat effectiveness index can be inferred. Then, along the equal capability surface and the equal effectiveness surface, the effectiveness generation rate and capability utilization rate of the combat plan are obtained by solving the constrained optimization problem. Based on this, a formula and a graphical representation of the combat plan quality and effectiveness level are given. Although the quality and effectiveness level in the present invention is a single numerical value, it is a comprehensive indicator of combat effectiveness and combat benefit, and has a clear and intuitive graphical form, which can effectively represent the optimization degree (adaptability and economy) of the combat plan. The combat capability conversion mechanism model proposed in the present invention not only establishes a probabilistic correlation between capability and effectiveness, but also can measure both the explicit and implicit elements of combat capability at an abstract level, which is conducive to more objective and comprehensive quantification of the generation basis of combat effectiveness.

[0111] Please note that the various technical features of the above embodiments can be combined arbitrarily. In order to make the description concise, not all possible combinations of the various technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification. The above embodiments only express several implementation methods of the present application, and their descriptions are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that for those skilled in the art, without departing from the concept of the present application, several variations and improvements can be made, which all fall within the scope of protection of the present application. Therefore, the scope of protection of the patent in this application shall be based on the attached claims.

Claims

1. A method for analyzing the quality and effectiveness of combat plans based on simulation and deduction, characterized by: The method includes: Step 1: From the perspective of the Red Army, based on each link and element of the combat process, establish a set of effectiveness indicators I that are directly related to the combat mission, and assign weights W to the established effectiveness indicators; where I = {I m |m=1~N I },W={ω m |m=1~N I }, where N I is the number of performance indicators, I m is the mth performance indicator, ω m Performance Index I m The weight of Step 2: Based on the combat simulation system, conduct a large-sample simulation of all the alternative combat plans in the Red side's alternative combat plan set C, and use the simulation result data to quantify the effectiveness index set I to form an effectiveness cuboid S; where C = {C n |n=1~N C }, N C is the number of alternative combat options, C n is the nth alternative combat plan; the dimension of the efficiency cuboid S is N C ×N I ×N S , N S The number of independent deductions for each alternative combat plan, any element S in S nmk Representative Effectiveness Index I m In Alternative Operations Plan C n The effectiveness value obtained in the kth deduction is used to obtain the alternative combat plan C n The performance set U of the indicator n ; Step 3: Based on the Bayesian network, establish a combat capability conversion mechanism model that includes a baseline capability generation layer, a positive capability generation layer, a negative capability generation layer, and an effectiveness generation layer. The baseline capability in the baseline capability generation layer is used to obtain the corresponding positive capability in the positive capability generation layer, the negative capability generation layer forms the negative capability, and the effectiveness value in the effectiveness generation layer is obtained based on the positive and negative capabilities. Step 4: Sampling the combat capability transformation mechanism model established by the predetermined number of times to obtain the sample set Based on the obtained sampling samples, it is inferred that each performance indicator has the negative capacity {β m |m=1~N I }; where L is the number of sampling times, For all β m The posterior distribution P(β m |S)’s i-th sample; Step 5, based on alternative operation plan C n The performance set U of the indicator n , combat effectiveness calculation function, and the inferred negative capacity of each effectiveness indicator, calculate the alternative combat plan C n The efficiency generation rate θ S and capacity utilization θ α ; Based on the calculated effectiveness generation rate and capacity utilization rate, determine the alternative combat plan C n The quality and effectiveness level of the combat plan is V.

2. The method according to claim 1, wherein In step 1, the effectiveness index set I established includes: emergency response effectiveness, support effectiveness, reconnaissance effectiveness, mobility effectiveness, strike effectiveness, survivability effectiveness, and coordination effectiveness.

3. The method according to claim 1, wherein In step 3, the combat capability conversion mechanism model is as follows: The benchmark capability generation layer contains the node set {α n |n=1~N C }, where node α n Indicates alternative combat plan C n Intrinsic overall combat capability: The positive capacity generation layer contains a set of nodes The nodes Indicates alternative combat plan C n In the performance index I m Actual combat capabilities demonstrated: The negative capacity generation layer contains the node set {β m |m=1~N I }, where node β m Indicates the blue team and the complex battlefield environment factors in the effectiveness index I m The ability to resist the alignment ability shown above: The performance generation layer includes a node set {S nmk |n=1~N C ,m=1~N I ,k=1~N S }, is the performance cuboid S: in, is the truncated normal distribution on the interval [0,+∞), LNormal(·) is the lognormal distribution, Beta(·) is the Beta distribution, μ α , τ α ,τ,μ β , τ β are the parameters of the corresponding distributions, and ln(·) is the natural logarithm function.

4. The method according to claim 3, wherein In step 4, β m Approximately the sample mean Where avg(·) is the average value function.

5. The method according to claim 1, wherein In step 5, the combat effectiveness calculation function Specifically: in, and Are any indicator performance set and corresponding weight set, satisfying 6. The method according to claim 5, wherein In step 5, alternative action plan C n The performance set U of the indicator n Specifically: U n ={u nm |m=1~N I } Among them, u nm For indicator I m Based on alternative combat plan C n The mean performance of all deduction samples: u nm =avg({S nmk |k=1~N S })。 7. The method according to claim 6, wherein In step 5, for alternative action plan C n , efficiency generation rate θ S Determined by the following formula: Among them, U S is the solution to the following optimization problem:

8. The method according to claim 6, wherein In step 5, for alternative action plan C n , capacity utilization θ α Determined by the following formula: in, is the solution to the following optimization problem:

9. The method according to claim 7 or 8, wherein In step 5, the operational plan C to be analyzed n The quality and effectiveness level V of the combat plan is determined by the following formula:

Images