A dynamic target assignment method and system based on an improved DPSO algorithm

Abstract

The application provides a dynamic target allocation method and system based on an improved DPSO algorithm, and belongs to the field of dynamic target allocation. In order to solve the problem that the convergence and solution accuracy of a swarm intelligence optimization algorithm are poor in the process of existing multi-missile cooperative target allocation, the traditional PSO algorithm is discretized based on a decimal coding and speed integerization mode, the population updating process is ensured to always meet resource constraints and integer constraints through modular initialization and boundary value processing, and solutions that violate the strike effectiveness constraint are constantly eliminated by introducing a penalty function into a target function. In order to further improve the performance of the algorithm, field expert knowledge is introduced in the solution process, and fast and high-quality solution of the multi-missile cooperative target allocation problem is realized.

Description

TECHNICAL FIELD

[0001] The application relates to the field of dynamic target allocation, in particular to a dynamic target allocation method and system based on an improved DPSO algorithm. BACKGROUND

[0002] With the increasing complexity of modern battlefield environment, the traditional mode has been difficult to meet the demand for rapid and efficient. Especially in the face of multiple targets and complex situations, the single missile's strike effect is limited, and it is difficult to achieve rapid response and efficient attack. Therefore, how to improve the efficiency of the missile has become the key. The multi-missile cooperative strategy emerges as the times require, which breaks through the limitations of single missile system through information exchange and cooperation between missiles, significantly improves the striking accuracy and flexibility. This strategy enables missiles to select and destroy targets more quickly and accurately, achieving maximum effect at a lower cost. In modern battlefield, the challenge of target allocation is particularly prominent. The demand for multi-target and multi-task attack requires the system to quickly and accurately select and allocate targets, but in the existing multi-missile cooperative target allocation process, the convergence and solution accuracy of swarm intelligence optimization algorithm are poor. Therefore, researchers have proposed a multi-missile cooperative strategy based on a dynamic target allocation model, combined with an improved discrete particle swarm optimization algorithm, considering factors such as resource constraints and strike efficiency, which greatly improves the efficiency and accuracy of target allocation. SUMMARY

[0003] The technical problem to be solved by the application is:

[0004] To solve the problem of poor convergence and solution accuracy of swarm intelligence optimization algorithm in the existing multi-missile cooperative target allocation process.

[0005] The technical scheme adopted by the application to solve the above technical problem is:

[0006] The application provides a dynamic target allocation method based on an improved DPSO algorithm, comprising the following steps:

[0007] S100, a dynamic target allocation model is established for a task, aiming to complete the strike task while saving ammunition;

[0008] S200, the traditional particle swarm optimization algorithm is discretized through a decimal encoding and speed integerization method, and modular initialization and boundary value processing are used to ensure that resource constraints and integer constraints are always met when the population is updated;

[0009] S300, an improved discrete particle swarm optimization algorithm is used to solve the model obtained in step S200, including introducing a penalty function into the objective function to gradually eliminate solutions that violate the strike efficiency constraint, and introducing domain expert knowledge to reduce the randomness and uncertainty of the initial solution.

[0010] Furthermore, in step S100, it specifically includes: according to the OODA observation-shooting-observation-shooting cycle, in a single engagement model, assuming that the red side has Types of missiles, denoted as The number of missiles of each type that can be used in this round is , the weight of each type of missile in this round of planning is ;

[0011] Define the red team's planning matrix as:

[0012]

[0013] Among them, the matrix Line and The column distribution represents the number of missiles and targets to be attacked. The variable Indicates the Targets to be hit The allocated Types of missiles the number of

[0014] The lethality matrix of the red team against the blue team is defined as:

[0015]

[0016] Among them, the variable Each missile Can damage enemy targets to be attacked ;

[0017] The blue side's loss is equivalent to the sum of the red side's attack effectiveness against it. If all the red side's missiles hit the target, the blue side will get The loss model of a target to be hit is:

[0018]

[0019] Considering resource constraints, strike effectiveness constraints and integer constraints, the multi-missile coordinated target allocation optimization model is established as follows:

[0020] Maximize

[0021] st

[0022]

[0023]

[0024]

[0025] wherein, is a sustained damage factor;

[0026] The constraint set is used to limit the number of each type of missile dispatched by the red side to no more than the available number of missiles after deployment; the constraint set introduces a hyperparameter to limit the maximum strike effectiveness , limiting the maximum number of strikes that the red side can dispatch to the target number times; the constraint set is used to limit the decision variable of the red side to be a non-negative integer;

[0027] An external penalty function method is used to define the penalty for violating the th constraint in the constraint set as:

[0028]

[0029] The constraint set is introduced into the objective function in the form of a penalty, and solutions that violate the constraints are eliminated in iterations, and the new objective function is:

[0030]

[0031] wherein, represents a penalty factor, and ; when a solution does not meet the th constraint in the constraint set, it can be known that , since , therefore , the corresponding objective function value is reduced, and thus is eliminated in the process of continuous iteration;

[0032] The multi-missile cooperative target allocation model based on the penalty function is obtained as:

[0033] Maximize

[0034] s.t.

[0035]

[0036] .

[0037] Further, in step S200, specifically comprising: in the multi-missile cooperative target allocation model established in step S100, the allocation scheme of the red side missiles is represented as an m-row n-column matrix , wherein the i-th row represents the number of missiles of the i-th type of missile allocated to each target respectively;

[0038] A modular coding method is used, and a flattening process is performed to obtain a coding sequence;

[0039] wherein the coding sequence has a total of Each variable constitutes a module, and represents a type of missile Each variable constitutes a module, and represents a type of missile The allocation scheme for the target to be attacked The allocation scheme for the target to be attacked The allocation scheme for the target to be attacked The allocation scheme for the target to be attacked .

[0040] Further, in step S300, the updated speed is rounded according to the constraint set established in step S100, so as to ensure that the updated decision variable is an integer.

[0041] Further, in order to prevent the boundary constraint of the search space, the lower bound of a single decision variable is absorbed by the boundary, and the lower bound of a single decision variable is modularized and reflected based on the entire module, and the reflection rule is re-random allocation.

[0042] Further, in the process of introducing expert knowledge in the WTA field, the missile is allocated to the target with the maximum damage effectiveness, including sorting the damage effectiveness of different targets, that is, arranging the elements in the first row of the damage matrix in descending order, and then allocating the missile to the target corresponding to each element in the form of a random number, and updating the number of remaining missiles after each allocation.

[0043]

[0044] The fitness function is consistent with the objective function in the multi-missile cooperative target allocation model:

[0045] .

[0046] The dynamic target allocation system based on the improved DPSO algorithm has program modules corresponding to the above steps, and executes the steps in the dynamic target allocation method based on the improved DPSO algorithm.

[0047] The computer readable storage medium stores a computer program, and the computer program is configured to realize the steps of the dynamic target allocation method based on the improved DPSO algorithm when called by the processor.

[0048] Compared with the prior art, the beneficial effects of the present application are:

[0049] ​​​​​1) Multi-missile coordination strategy can significantly improve the effectiveness and flexibility of missiles by information exchange and coordinated action among missiles. This strategy enables missiles to select and destroy targets more quickly and accurately, achieving maximum effect at a lower cost.

[0050] 2) Multi-missile coordination strategy can consider battlefield situation and quickly and accurately allocate attack targets according to certain rules, which adapts to the complexity and dynamics of modern environment, improving the accuracy and efficiency of missile attack.

[0051] 3) By establishing a dynamic target allocation model that maximizes enemy losses and minimizes own consumption, and proposing an improved discrete particle swarm optimization algorithm, the multi-missile coordination target allocation problem is quickly and high-quality solved. This algorithm not only considers weapon resources and attack effectiveness constraints, but also improves the performance and practicality of the algorithm by introducing penalty function and domain expert knowledge.

[0052] 4) At the action level, multi-missile coordination strategy focuses on coordinated guidance of missiles to ensure saturation attack on targets. Through coordinated research on each stage of missile flight, better initial conditions are provided for the terminal guidance segment, thus improving the coordinated attack capability and hitting accuracy of missiles.

[0053] 5) The research and development of multi-missile coordination strategy and related technologies promote the innovation and progress of technology. This strategy not only improves the effectiveness and flexibility of missile attack, but also provides new ideas and means for the future. BRIEF DESCRIPTION OF DRAWINGS

[0054] Figure 1 Flowchart of a dynamic target allocation method based on an improved DPSO algorithm in an embodiment of the present application;

[0055] Figure 2 Red "shooting-observation-shooting" logic diagram in an embodiment of the present application;

[0056] Figure 3 Update process diagram of PSO algorithm in two-dimensional search space in an embodiment of the present application;

[0057] Figure 4 Decimal integer encoding sequence diagram in an embodiment of the present application;

[0058] Figure 5 Optimal fitness value scatter plot obtained by running four algorithms 100 times in an embodiment of the present application. DETAILED DESCRIPTION

[0059] In order to make the above objectives, features and advantages of the present application more obvious and easy to understand, specific embodiments of the present application are described in detail below with reference to the accompanying drawings.

[0060] In order to realize multi-target positioning in a low-computing-power and low-power-consumption platform such as a micro robot, the present application proposes a dynamic target allocation method based on an improved DPSO algorithm. Firstly, for the target of completing a strike task while saving ammunition, a dynamic target allocation model is established by considering weapon resources and strike effectiveness constraints, and maximizing enemy loss and minimizing own consumption. An improved discrete particle swarm optimization algorithm is proposed for model solving. Based on decimal encoding and integer speed, the traditional PSO algorithm is discretized, and the population update process is ensured to always meet the resource constraints and integer constraints through modular initialization and boundary value processing. The solution that violates the strike effectiveness constraint is constantly eliminated by introducing a penalty function into the objective function. In order to further improve the performance of the algorithm, domain expert knowledge is introduced in the solving process, and fast and high-quality solution of the multi-missile cooperative target allocation problem is realized.

[0061] Specific implementation scheme one: in combination with Figures 1 to 4 As shown in the figure, the present application provides a dynamic target allocation method based on an improved DPSO algorithm, comprising the following steps:

[0062] S100, for a task, a dynamic target allocation model is established for the target of completing a strike task while saving ammunition;

[0063] The red and blue sides attack and defense is usually a multi-round attack and defense process, and the dynamic weapon target allocation (DWTA) problem of the process belongs to a multi-stage decision problem; according to the classic OODA (observation, positioning, decision, action) cycle, the logic of the red side is "observation-shooting-observation-shooting", as shown in the figure; Figure 2 ;

[0064] In the above logic, the single "observation-shooting" process in each round of combat belongs to the static weapon target allocation (SWTA) problem, so the DWTA problem can be divided into multiple SWTA problems for modeling and solving;

[0065] In the single combat model, it is assumed that the red side has types of missiles, denoted as , the number of each type of missile available in this round is , and the weight of each type of missile in the plan for this round is , the weight represents the value of the missile, which depends on the cost of the missile and the launch cost factor;

[0066] It is assumed that the blue side has types of targets to be attacked, denoted as The number of each type of target to be attacked in the current round is The weight of each type of target in the current round is The weight represents the importance of the target to be attacked;

[0067] In each round of combat, the red side needs to optimize the target allocation model and send corresponding missiles to attack the target to be attacked, so the planning matrix of the red side is defined as:

[0068]

[0069] Where, the matrix Row and Column distribution represents the number of missiles and target types, and the variable represents the number of the first target to be attacked allocates the first type of missile ;

[0070] Since each type of missile has different effects on enemy targets to be attacked, the red side needs to define the damage matrix of the blue side as:

[0071]

[0072] Where, the variable represents that each missile can damage enemy targets to be attacked .

[0073] The loss of the blue side is equivalent to the sum of the attack effectiveness of the red side, if the red side missiles hit the target, then the loss model of the first target to be attacked of the blue side can be obtained as:

[0074]

[0075] Considering the resource constraints, attack effectiveness constraints and integer constraints, the multi-missile cooperative target allocation optimization model is established as:

[0076] Maximize

[0077] s.t.

[0078]

[0079]

[0080]

[0081] Where, to continuously damage the factor;

[0082] In the above model, the constraint set is used to limit the number of each type of missile dispatched by the red side to not exceed the available number of missiles after deployment; the constraint set introduces a hyperparameter to limit the maximum attack effectiveness , limiting the red side to a maximum of target number times the attack effectiveness, avoiding the waste of resources caused by the red side dispatching excessive forces; the constraint set is used to limit the value of the red side decision variable to be a non-negative integer; however, considering the difficulty of solving multi-constraint optimization problems, an external penalty function method is used, and the penalty for violating the th constraint in the constraint set is defined as:

[0083]

[0084] Thus, the constraint set is introduced into the objective function in the form of a penalty, and solutions that violate the constraints are eliminated in the iteration; therefore, the new objective function is:

[0085]

[0086] where is the penalty factor, and ; when a solution does not meet the th constraint in the constraint set, it can be known that , since , , the corresponding objective function value is reduced, and thus is eliminated in the process of continuous iteration;

[0087] Thus, the multi-missile cooperative target allocation model based on the penalty function is obtained as:

[0088] Maximize

[0089] s.t.

[0090]

[0091]

[0092] This kind of WTA problem is a classic combinatorial optimization problem in the field of mathematics, and the problem has the following characteristics: 1, WTA problem belongs to NP-complete problem, and the optimal solution can only be solved by complete enumeration method, and when facing large-scale problem; 2, the dimension of the decision variable depends on the product of the equipment type and the target type, and the scale is huge; 3, with nonlinear objective function; 4, due to the change of battlefield, the constraint condition and the objective function have certain randomness; 5, its decision variable is discrete and discontinuous; The solution of this kind of nonlinear integer programming problem is very complex, and it is still feasible to obtain the optimal solution of the problem when the problem size is small; However, once the problem size is expanded, there is almost no deterministic optimization algorithm in the traditional branch and bound method, dynamic programming method and other exact algorithms can calculate the optimal distribution scheme of the problem within the polynomial time; Therefore, it is more suitable for WTA problem to solve the suboptimal solution with less time by using intelligent optimization algorithm; In view of this, the application proposes an improved discrete particle swarm optimization algorithm (IDPSO) for model solving;

[0093] S200, by decimal encoding and speed integerization method, the traditional particle swarm optimization algorithm (PSO) is discretized, and through modularization initialization and boundary value processing, it is guaranteed that the resource constraint and integer constraint are always satisfied when the population is updated;

[0094] The particle swarm optimization algorithm is initially inspired by the regularity of bird flocking activities, and then through simulating the process of bird flocking and fish flocking in nature, a simplified model is established by using group intelligence, and the model optimization and problem solving are realized through information sharing and mutual cooperation between each individual in the group. Each bird in the bird flock has a memory function and can remember the position closest to the food (individual extreme value), and through information sharing between the flocks, it also knows the position of the bird closest to the food in the whole flock (global extreme value); Therefore, in the process of searching for food, the bird flies to the position closest to the food (individual extreme value) and the position of the bird closest to the food in the whole flock (global extreme value). In the PSO algorithm, each feasible solution is regarded as a bird, the optimal solution to be solved is regarded as food, and the distance between each bird and food is regarded as fitness. Each solution is guided by and to approach the optimal solution, which has the following steps:

[0095] Suppose there are particles in a dimensional search space to form a colony, wherein the particle can be expressed as:

[0096]

[0097] No. The speed of a particle is also a dimensional vector, expressed as:

[0098]

[0099] Each particle has a storage function. The optimal position searched by a particle so far is called an individual extreme value, resulting in a D-dimensional vector:

[0100]

[0101] The entire particle swarm obtains a global extreme value through information sharing:

[0102]

[0103] After finding these two optimal values, each particle updates its own velocity and adjusts its search direction using the following formula:

[0104]

[0105] The speed update process of the PSO algorithm in the two-dimensional search space is as follows: Figure 3 As shown. The update speed formula is divided into three parts, where is the inertia term, which represents the influence of the magnitude and direction of the velocity at the previous moment on the velocity at the current moment. Since the initial velocity of each particle is randomly initialized, this part has a certain degree of exploration randomness. It is the self-cognition term, which represents the influence of one's historical experience on speed. It is the vector from the current position to one's optimal position. is the social cognitive term, which represents the influence of the social history experience of the particle group on the speed, and is the vector pointing from the current position to the optimal position of the group; is the learning factor, which measures the weight of the influence of self-cognition and social cognition on the direction of particle exploration. , indicating that personal experience is as important as social experience; for Random numbers within a certain range add randomness to each particle during the experience learning process;

[0106] Afterwards, each particle updates its own position according to the following formula:

[0107]

[0108] in, is the number of iterations, and respectively represent the first , the first iteration, the position of the first particle in the first dimension, represent the first , the first iteration, the first dimension optimized speed of the first particle; in order to enhance the optimization ability of the particle, the particle speed is generally updated by the following formula:

[0109]

[0110] wherein, is an inertia factor, used to measure the influence of the flight inertia on the particle exploration direction, since the inertia term has random exploration, the inertia factor is greater, the particle exploration is more random, and the global optimization ability is stronger, the inertia factor is smaller, the particle local optimization ability is stronger; therefore, the linear decreasing weight search strategy is currently adopted:

[0111]

[0112] wherein, represents the current iteration number, represents the maximum number of iterations, represents the minimum value of the inertia factor, represents the maximum value of the inertia factor; in the initial iteration, the inertia weight is larger, the particle swarm exploration is more random, and it quickly converges to the global optimal point, with the increase of the iteration number, the inertia weight decreases, and the particle swarm performs more detailed local search, and finally converges to the global optimum.

[0113] Since the WTA problem to be solved in this step belongs to a combinatorial optimization problem, the solution space is discrete, and the PSO algorithm is only applicable to optimization problems with continuous solution space; therefore, the improved discrete PSO algorithm for the WAT problem is proposed by referring to the optimization mechanism of the traditional PSO algorithm;

[0114] For the WTA problem, the decision variables in the combinatorial optimization model do not need to be transformed complexly, but are directly mapped to the solution sequence required by the algorithm in a modular form, the coding is convenient and easy to understand, and decoding is not required, the fitness of each particle can be directly solved, and the algorithm solving speed is effectively accelerated;

[0115] In the multi-missile cooperative target allocation model established in step S100, the allocation scheme of the red missile is represented as an m-row n-column matrix , where the i-th row represents the number of missiles assigned to each target by the i-th missile. Considering that the number of missiles launched per round must not exceed the total number of missiles, the decision variables in each row of the matrix affect each other during the optimization process. Therefore, a modular coding method is adopted and flattened to obtain the following: Figure 4 The coding sequence shown;

[0116] Among them, the coding sequence has variables, each of which variables form a module, representing the Types of missiles Treating targets The decision variables The position in the sequence is ;

[0117] S300, an improved discrete particle swarm optimization algorithm (DPSO) is used to solve the model;

[0118] The purpose of population initialization is to construct a feasible solution as the starting point of the algorithm iteration. Therefore, when initializing, the existing constraint set in the model must be fully considered. Among them, the constraint set requires that the number of missiles of each type sent by the Red side cannot exceed the number of available missiles after deployment. Therefore, initialization needs to be performed according to the module, and the sum of the elements in each module must not exceed the total number of corresponding missiles. The constraint set requires that the number of missiles that the Red side must send must be an integer greater than zero. Therefore, the variables in each module must be natural numbers during initialization. The constraint set requires that the Red side can only send a maximum of times the strike effectiveness, but modular coding is based on missile type rather than target type. Considering that this constraint set will complicate the initialization method and reduce the solution speed, this constraint set is not restricted. At the same time, in the objective function, this constraint has been processed by the penalty function. Therefore, although there are some infeasible solutions in the initial solution, these infeasible solutions will be gradually eliminated during the continuous iteration process due to the penalty function.

[0119] In addition to considering the feasibility of the initial solution, the present invention reduces the randomness and uncertainty of the initial solution by introducing expert knowledge in the WTA field, and effectively improves the convergence speed of the algorithm by improving the quality of the initial solution; when initializing, the missile is preferentially assigned to the target with the greatest damage efficiency; For example, first rank the missile's damage effectiveness against different targets, that is, the lethality matrix No. The row elements are sorted in descending order and then replaced by random numbers The missiles are allocated to the target corresponding to each element in turn in the form of , and the number of remaining missiles needs to be updated after each allocation is completed. , the random number generation method is shown in the formula:

[0120]

[0121] The initial solution constructed in this way can not only satisfy the constraint set, but also accelerate the convergence speed. The initialization pseudo code of each module is as follows. The population initialization needs to loop this code Second-rate;

[0122] Population initialization pseudocode

[0123]

[0124] The fitness value is an important indicator that reflects the quality of the solution relative to the original objective function, and is also an important indicator for selecting individual extreme values ​​in the PSO algorithm. and global extremes Here, the fitness function is consistent with the objective function in the multi-missile coordinated target allocation model, so the fitness of any solution is:

[0125]

[0126] S400, introducing a penalty function into the objective function to gradually eliminate solutions that violate the strike effectiveness constraint, thereby ensuring the rationality of the solution;

[0127] After testing, it was found that for the WTA problem, the update method of the PSO algorithm population and the way of handling boundary conditions are key to whether it can learn the experience of excellent individuals in the population; due to the existence of the constraint set, all decision variables need to be kept as natural numbers; during the exploration process, in order to keep the updated variables still integers, the present invention approximates the updated speed to ensure that the updated decision variables are still integers;

[0128] At the same time, when particles explore, they may be constrained by the boundaries of the search space. Although theoretically this will not affect the generation of the optimal solution, in practice, useless exploration outside the boundaries will cause a huge waste of computing resources and seriously affect the convergence speed of the algorithm. Therefore, it is necessary to reasonably handle the boundary conditions to improve the performance of the algorithm. There are three common ways to handle boundary conditions: penalty function, boundary absorption, and boundary reflection, as shown in Table 1.

[0129] Table 1 Boundary processing methods of optimization algorithms

[0130]

[0131] Wherein in order to ensure that the decision variable is still a natural number, the lower bound of a single decision variable is treated by boundary absorption, that is: ; since the upper bound of a single decision variable is affected by other decision variables within the entire module, the processing of the upper bound of the decision variable is based on the entire module, and the modular boundary reflection is adopted, the reflection rule is re-random allocation, if the sum of the decision variables within the module exceeds the total number of missiles that can be used by the module, then the decision variables within the module are re-allocated according to the steps S300 .

[0132] Specific implementation scheme two: the dynamic target allocation system based on the improved DPSO algorithm of the application has a program module corresponding to the above steps, and executes the steps in the dynamic target allocation method based on the improved DPSO algorithm.

[0133] Other combinations and connection relationships of the embodiment are the same as those of specific implementation scheme one.

[0134] Specific implementation scheme three: the computer readable storage medium of the application stores a computer program, and the computer program is configured to realize the steps of the dynamic target allocation method based on the improved DPSO algorithm when called by the processor.

[0135] Other combinations and connection relationships of the embodiment are the same as those of specific implementation scheme one.

[0136] Simulation experiment

[0137] In order to test the performance of the improved DPSO algorithm of the application, the algorithm is compared with the traditional discrete particle swarm optimization algorithm (DPSO) and the hybrid mutation DPSO algorithm for single target allocation problem. Among them, algorithm one extends the standard particle swarm optimization algorithm to the discrete search space, and solves the WTA problem by approximating the integer position after each update; algorithm two forms a hybrid mutation DPSO algorithm for solving the WTA problem by introducing the concepts of mutation and crossover in genetic algorithm into the standard particle swarm algorithm. In order to test the effect of expert experience on accelerating the convergence speed of the algorithm, an improved DPSO2 algorithm is also added, that is, the improved DPSO algorithm with random initialization. The simulation case data are as follows:

[0138] Table 1 WTA simulation case data

[0139]

[0140] ​To ensure the fairness of the comparative experiment, all simulations are performed on the same computer, which is configured with Windows 10 operating system, 16 GB memory and AMD Ryzen 5 4600H processor (3.00 GHz), and all algorithms are programmed in Matlab R2021b. In the setting of simulation parameters, the variable population size and the maximum number of iterations of the four algorithms are set consistently, among which , , the inertia weight of the velocity position update particle swarm algorithm is set to , and the acceleration factor is set to .

[0141] In the comparative experiment, in order to avoid random errors caused by random numbers in the algorithm, multiple experiments are needed to increase the number of comparison samples, so based on the above case data and simulation parameter settings, 100 times of cyclic simulation are performed, and the maximum fitness value of each algorithm in each run is shown in Figure 5 . As can be seen from the scatter plot, in 100 times of cyclic simulation, the four algorithms can make the particle population find the optimal allocation scheme within 100 times, as shown in

[0142] Table 2 Optimal allocation scheme of red side missile cooperative attack

[0143]

[0144] To further analyze the performance of the above algorithms, the following three evaluation indexes are established: the average maximum fitness value of the algorithm (AMFV), the average optimization time of the algorithm (ACT) and the success rate of solving (SR), and their calculation formulas are as follows:

[0145]

[0146]

[0147]

[0148] In the above formulas, represents the total number of algorithm cycles, and represent the maximum fitness value and the optimization time in the th cycle of the algorithm, indicates the maximum fitness value corresponding to the optimal allocation scheme, and according to the above formulas, the evaluation indexes of each algorithm can be calculated as shown in the following table.

[0149] Table 3 Evaluation indexes of each algorithm

[0150]

[0151] In addition to the four algorithms described above, a mixed integer linear programming solver (Intlinprog) in MATLAB R2021b is introduced for comparative experiments. The solver uses traditional solutions such as branch and bound and simplex method. Through analysis, it is found that in terms of solution quality, the improved DPSO algorithm and the Intlinprog solver are first, followed by the hybrid DPSO algorithm and the improved DPSO2 algorithm, and the DPSO algorithm is the worst. The index of the reaction algorithm solution quality is combined with the index of the reaction algorithm optimization success rate, and it is found that the improved DPSO algorithm and the Intlinprog solver can find the optimal allocation scheme in 100 cycles with 100% success rate, and have strong stability and accuracy. The hybrid DPSO algorithm and the improved DPSO2 algorithm are slightly inferior to the above two algorithms, but they can also find solutions with small differences from the optimal allocation scheme. The DPSO algorithm has a large gap in solution quality and solution success rate compared with other algorithms, indicating that it is necessary to make specific improvements to the WTA problem in addition to extending the particle swarm algorithm to the solution of the WTA problem and expanding the exploration space to the discrete area. In terms of solution time, the last three algorithms are mainly compared, and it is found that introducing expert experience in the field in the initialization and boundary value processing of the algorithm can greatly improve the algorithm solution performance and convergence speed.

[0152] Although the present application is disclosed as above, the protection scope of the present application is not limited to this. The person skilled in the art can make various changes and modifications without departing from the spirit and scope of the present application, and these changes and modifications will fall within the protection scope of the present application.

Claims

1. A dynamic target allocation method based on an improved DPSO algorithm, characterized in that: The following steps are involved: S100. Develop a dynamic target allocation model for the mission, with the goal of completing the strike mission while conserving ammunition. Specifically, based on the OODA observe-shoot-observe-shoot cycle, in a single engagement model, assuming that the Red side has Types of missiles, denoted as The number of missiles of each type that can be used in this round is , the weight of each type of missile in this round of planning is ; Define the red team's planning matrix as: Among them, the matrix Line and The column distribution represents the number of missiles and targets to be attacked. The variable Indicates the Targets to be hit The allocated Types of missiles the number of The lethality matrix of the red team against the blue team is defined as: Among them, the variable Each missile Can damage enemy targets to be attacked ; The blue side's loss is equivalent to the sum of the red side's attack effectiveness against it. If all the red side's missiles hit the target, the blue side will get The loss model of a target to be hit is: Considering resource constraints, strike effectiveness constraints and integer constraints, the multi-missile coordinated target allocation optimization model is established as follows: Maximize st in, It is a continuous damage factor; The constraint set is used to limit the number of missiles of each type sent by the Red side to no more than the number of missiles available after deployment; the constraint set is achieved by introducing a hyperparameter that limits the maximum strike effectiveness. , limiting the maximum number of targets that the red side can send times the strike effectiveness; the constraint set is used to limit the red side's decision variables The value of is a non-negative integer; Using the external penalty function method, define the first The penalty for a constraint is: The constraint set is introduced into the objective function in the form of penalty, and the solutions that violate the constraints are eliminated during the iteration, and the new objective function is obtained as follows: in, represents the penalty factor, and ; When a solution does not meet the constraint set When there is a constraint, we know ,because ,therefore , then the corresponding objective function value decreases, and thus is eliminated in the process of continuous iteration; The multi-missile coordinated target allocation model based on penalty function is obtained as follows: Maximize st S200, through decimal coding and speed integerization method, the traditional particle swarm optimization algorithm is discretized, and through modular initialization and boundary value processing, it is used to ensure that the resource constraints and integer constraints are always satisfied when the population is updated; Specifically, in the multi-missile coordinated target allocation model established in step S100, the allocation plan of the red side's missiles is represented as a matrix with m rows and n columns. , where the i-th row represents the number of missiles assigned to each target by the i-th missile; Adopt modular coding method and perform flattening to obtain coding sequence; Among them, the coding sequence has variables, each of which variables form a module, representing the Types of missiles Treating targets The allocation scheme, decision variables The position in the sequence is ; S300, using an improved discrete particle swarm optimization algorithm to solve the model obtained in step S200, including introducing a penalty function into the objective function to gradually eliminate solutions that violate the strike effectiveness constraint; and introducing domain expert knowledge to reduce the randomness and uncertainty of the initial solution.

2. The dynamic target allocation method based on the improved DPSO algorithm according to claim 1, characterized in that: In step S300, according to the constraint set established in step S100, the updated speed is approximately rounded to an integer to ensure that the updated decision variable is an integer.

3. The dynamic target allocation method based on the improved DPSO algorithm according to claim 2, characterized in that: In order to prevent the boundary constraints of the search space, the lower bound of a single decision variable is subjected to boundary absorption. The lower bound of a single decision variable is subjected to modular boundary reflection based on the entire module, and the reflection rule is random reassignment.

4. The dynamic target allocation method based on the improved DPSO algorithm according to claim 3 is characterized in that: In the process of introducing the expertise of WTA, missiles are assigned to targets with the greatest destructive effect, including missiles The damage efficiency of different targets is ranked, that is, the lethality matrix No. The row elements are sorted in descending order and then replaced by random numbers The missiles are allocated to the target corresponding to each element in turn in the form of , and the number of remaining missiles needs to be updated after each allocation is completed. , the random number is generated as follows: The fitness function is consistent with the objective function in the multi-missile coordinated target allocation model: 。 5. A dynamic target allocation system based on an improved DPSO algorithm, characterized by: The system has a program module corresponding to the steps of the method according to any one of claims 1 to 4, and executes the steps of the dynamic target allocation method based on the improved DPSO algorithm during operation.

6. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and the computer program is configured to implement the steps of the dynamic target allocation method based on the improved DPSO algorithm according to any one of claims 1 to 4 when called by a processor.

Images