A collaborative mission planning method for UAV swarms based on adaptive search teaching and learning algorithm

Abstract

A method for collaborative task planning of a swarm of drones based on an adaptive search teaching and learning algorithm belongs to the field of drone task planning. The steps are: first, extract the key factors of actual constraints, clarify the decision variables and constraints to construct a combination optimization model; construct a coding form, and transform the objective function into a fitness function suitable for the coding; second, design a high-quality initial solution generation scheme based on the coding setting method to provide a high-quality initial solution for the teaching and learning algorithm, and finally summarize each coding to obtain an initial population, determine the coding update rule to obtain an evolutionary population; finally, introduce a neighborhood search operator in the evolutionary process, adaptively carry out neighborhood search operations, and obtain a solution. The present invention takes into account the real combat process in actual combat, can reduce the risk of the vehicle being discovered by the enemy, ensure the safety of the vehicle and the smooth execution of subsequent tasks; and can achieve efficient allocation of drone resources and optimization of task planning in a complex battlefield environment.

Description

Technical Field

[0001] The present invention belongs to the field of unmanned aerial vehicle (UAV) mission planning, and relates to a UAV swarm collaborative mission planning method based on an adaptive search teaching and learning algorithm. Background Art

[0002] With the rapid development of drone technology, drone swarms are increasingly being used in military and civilian applications. In modern warfare, drone swarm collaborative operations have become a crucial tool for improving combat effectiveness. However, with the expansion of battlefields and the increasing complexity of missions, the task planning problem for drone swarms has become increasingly complex. Traditional task planning methods often struggle to achieve efficient task allocation and path planning while satisfying multiple constraints (such as drone range, ammunition loadout, and task sequence). The drone swarm task planning problem is inherently a complex multi-objective optimization problem, typically involving the rational allocation of drone resources, optimizing the task execution sequence, and minimizing task completion time. Due to the NP-hard nature of this problem, traditional optimization algorithms (such as genetic algorithms and particle swarm optimization) often suffer from high computational complexity, slow convergence, and a tendency to get stuck in local optima when handling large-scale tasks. Therefore, designing an efficient and stable task planning algorithm has become a key issue in the current research on drone swarm collaborative operations.

[0003] Most existing UAV mission planning algorithms are based on a single optimization objective (such as mission completion time or the number of UAVs used), and rarely consider the temporal relationships between tasks or the dynamic allocation of UAV resources. For example, a Chinese invention patent (CN112733421A) provides a mission planning method for manned-UAV coordinated ground operations. However, this method fails to consider the possibility of multiple mission types for the same target during modeling, making it incapable of performing the dual strike-assessment missions required in actual combat. Another example is a Chinese invention patent (CN107330588A) provides a mission planning method for coordinated reconnaissance of multi-base heterogeneous UAVs. This method considers the multi-base mission planning problem and utilizes a cuckoo algorithm for its solution. Similarly, it only considers the single-task scenario and fails to respond to actual combat requirements. Furthermore, traditional heuristic algorithms lack initial solution generation and local search capabilities, resulting in poor solution efficiency and stability in complex mission scenarios. For example, a Chinese invention patent (CN112926825A) provides a multi-UAV task allocation method based on a multi-objective quantum krill swarm mechanism. However, this method's encoding method is not suitable for multi-tasking problems. Modifying the encoding format accordingly would incur significant search overhead, failing to meet the timeliness and high-quality solutions required for actual combat. Therefore, a task planning algorithm is urgently needed that can comprehensively consider multiple constraints, possesses efficient initial solution generation capabilities, and possesses strong local search capabilities to address the complex UAV swarm collaborative combat requirements of modern warfare.

[0004] To address these issues, this paper proposes a UAV mission planning algorithm based on the Teaching-Learning-Based Optimization (TLBO) algorithm. By incorporating a heuristic initial solution generation strategy and a multimodal neighborhood search operator, this algorithm significantly enhances its global search and local optimization capabilities. This allows for efficient allocation of UAV resources and optimized task execution sequencing while satisfying multiple constraints. Summary of the Invention

[0005] To address the aforementioned technical issues, the present invention proposes a method for collaborative mission planning for drone swarms based on an adaptive search teaching-learning algorithm. This method is primarily applicable to combat mission planning for drone swarms. In the context of this invention, for each target point, a drone first performs an attack mission, followed by an evaluation mission performed by drones not intended to attack the target point. Once all missions are completed, all drones return to the delivery point for collection. This invention constructs a combinatorial optimization model tailored to the specific situation and employs a TLBO algorithm mixed with a neighborhood search operator to achieve rapid search results.

[0006] In order to achieve the above purpose, the specific technical solution adopted by the present invention is:

[0007] A method for collaborative task planning of a swarm of unmanned aerial vehicles (UAVs) based on an adaptive search teaching and learning algorithm, the method comprising the following steps:

[0008] Step 1: Extract the key factors of actual constraints, clarify the decision variables and constraints, and build a combined optimization model;

[0009] Step 2: Construct an encoding form based on the decision variables and constraints in the model. Then, based on the encoding form, transform the objective function into a fitness function suitable for the encoding, which is used to evaluate the quality of the solution corresponding to the encoding.

[0010] Step 3: Based on the encoding setting method, a high-quality initial solution generation scheme is designed to provide high-quality initial solutions for the teaching and learning algorithm. Multiple encodings are generated through this strategy, and finally the initial population is obtained by summarizing the encodings.

[0011] Step 4: Based on the initial population, determine the code update rule, that is, the evolution method of the code in the teaching and learning phases of the teaching and learning algorithm, to achieve the self-evolution of the algorithm and obtain the evolved population;

[0012] Step 5: Introduce the neighborhood search operator in the evolution process and adaptively perform neighborhood search operations to further improve the search efficiency of the algorithm, enhance the search capability of the population, and ultimately obtain a solution.

[0013] Furthermore, the step 1 is specifically as follows:

[0014] First, we extract the key factors and problem-solving requirements in the actual constraints, clarify the decision variables and constraints to build a combined optimization model, and set the task goal set as , Indicates the number of drones that can be allocated, Indicates the number of target points, represents a collection of drones, Represents the target point set, where all the locations passed by the drone are .in, Indicates the drones; Indicates the target points, and then carry out model representation.

[0015] Step 1.1: Construct decision variables and related representations according to the solution requirements;

[0016] In combinatorial optimization problems, the decision variables are usually the desired mission solutions, which need to include various information about the solutions to be solved. In this problem, the information to be solved includes the drone numbers to be used and the missions to be performed by each drone. Therefore, the decision variables are set as follows:

[0017] (1)

[0018] In formula (1), represents the decision variable of this problem, Indicates the task type. Indicates drone From the target point Departure to the destination implement Task, this formula means that each task at each target point must be completed. Accordingly, the following formula can be obtained from the decision variables:

[0019] (2)

[0020] Formula (2) is used to express the navigation route of the UAV. When indicating drone From the target point Departure to the destination , obviously when There must be , and also:

[0021] (3)

[0022] Formula (3) is used to express the state of the UAV. , drones Need to go out on a mission.

[0023] Step 1.2: Construct task completion guarantee constraints based on task requirements;

[0024] (4)

[0025] Formula (4) indicates that each task at each target point must be completed, where It indicates that there are two types of missions, namely attack missions and evaluation missions, and each target point has two types of missions that need to be performed by the drone.

[0026] Step 1.3: Construct task execution requirement constraints based on task execution requirements;

[0027] (5)

[0028] Formula (5) shows that for drones , can only execute the same target point at most a task.

[0029] Step 1.4: Construct navigation constraints based on the UAV’s navigation status;

[0030] (6)

[0031] Formula (6) means that for any UAV , starting from the delivery point and finally returning to the delivery point, where and The 0 in the subscript indicates the delivery point. Similarly, there is the following formula:

[0032] (7)

[0033] Formula (7) means that for any UAV , go from any destination point to the destination point After that, the subsequent journey must start from the destination point Departure, that is, the out-degree of each target point is equal to the in-degree.

[0034] Step 1.5: Construct task completion constraints based on task execution requirements;

[0035] (8)

[0036] Formula (8) indicates that both types of tasks must be executed for each target point to complete.

[0037] Step 1.6: Establish the task execution time constraint based on the actual flight patterns and execution capabilities of the UAV;

[0038] (9)

[0039] Formula (9) indicates that each target point has a task sequence and a minimum interval time, where is the task execution time, is the minimum interval between two tasks, and Represent the target points The attack task start time and the evaluation task start time.

[0040] Step 1.7: Construct the UAV capability constraints (maximum flight time constraint, ammunition load constraint) based on the UAV's own capability limitations.

[0041] (10)

[0042] Formula (10) represents the flight time constraint of the UAV, that is, the flight time of any UAV should be less than its maximum flight time, where Indicates drone Flight time, Indicates drone In addition, since the drone consumes a certain amount of ammunition when performing an attack mission, it is also necessary to set the drone's ammunition constraint. In this problem, it is assumed that the drone consumes one piece of ammunition each time it performs an attack mission. The formula is as follows:

[0043] (11)

[0044] Formula (11) represents the UAV ammunition loading constraint, i.e., The number of attack missions must be less than the upper limit of its ammunition load, where Indicates the maximum ammunition load of the drone.

[0045] Step 1.7: Consider the task requirements and complete the expected construction of the combinatorial optimization objective function;

[0046] (12)

[0047] The objective function of this problem is shown in formula (12), where Indicates the estimated maximum number of drones in use, Indicates drone The maximum flight time, and To adjust the weight of each item, and satisfy The optimization index of this objective function is to minimize the number of drones used and the time to complete the task. It is obtained by the following formula, the specific form is as follows:

[0048] (13)

[0049] in, Represents the straight-line distance from the delivery point to the target point. Indicates the number of target points, Indicates the drone's range, represents the target point, Represents a set of target points.

[0050] Furthermore, the step 2 is specifically as follows:

[0051] Step 2.1: Construct algorithm code based on decision variables;

[0052] In intelligent algorithms, codes are strings of numbers used to store decision variables and can be recognized by computers. Therefore, in the present invention, it is also necessary to construct corresponding teaching and learning algorithm codes around the decision variables and constraint settings in the model. For the convenience of description, the teaching and learning algorithm is collectively referred to as the TLBO algorithm below.

[0053] The encoded information includes the target and the UAV (hereinafter referred to as UAV), and the mission sequence of the same UAV is executed in the order of encoding from front to back. The encoding contains two lines of information, the first line stores the target point, and the second line stores the UAV number. The length of each line is twice the number of target points, which can be regarded as a 2-line Column matrix. In the first row of code, when a target appears for the first time, it is represented as an attack task for that point. Therefore, the second appearance of the target represents an evaluation task. The second row represents the drone that performs the task indicated by the first row of code in the same column. In summary, the decision variables of the combinatorial optimization model proposed in this invention are stored using this type of encoding. The decision variables corresponding to the encoding determine the complete execution plan for a round of tasks. The task execution plan obtained by encoding is generally called the decoded plan, and this term will be used in the following.

[0054] Step 2.2: According to the constraint setting and encoding form, transform the objective function into a fitness function;

[0055] In algorithms similar to TLBO, the objective function in the combinatorial optimization model needs to be converted into a corresponding fitness function to describe the quality of the solution corresponding to a certain encoding. Since the problem studied by the present invention has high requirements for solution time, the present invention also needs to transform the objective function of the combinatorial optimization model proposed in step 1 to obtain a fitness function that can quickly evaluate the quality of the corresponding encoding solution. On the other hand, since the generated encoding cannot fully guarantee to meet the constraints during the TLBO solution process, it is necessary to introduce a penalty term to indicate whether the decoding solution of the current encoding can be implemented. The specific form is as follows:

[0056] (14)

[0057] * (15)

[0058] (16)

[0059] in, represents the fitness function; and To adjust the weights of the two main indicators, and to meet ; represents the penalty term for the decoding solution exceeding the flight time constraint; The penalty term for the decoding solution exceeding the ammunition loading constraint; Indicates the estimated maximum number of drones in use; Indicates the maximum flight time of the drone; Denotes the drone in the decoding scheme Flight time; Represents the penalty coefficient of the penalty term of the flight time constraint; represents the penalty coefficient of the ammunition loading constraint penalty term; Indicates drone Number of attack missions executed; Indicates the maximum ammunition load of the drone.

[0060] In formula (14), the fitness function consists of three parts. The first part (the first two terms in the formula) represents the objective function given in step 1, which is to minimize the number of drones used and the time to complete the task. The second part Indicates that in the encoding decoding scheme, all drones that exceed the flight time limit will be penalized. Part 3 Indicates the penalty for exceeding the maximum ammunition load for all drones in the encoding decoding scheme.

[0061] Furthermore, the step 3 is specifically as follows:

[0062] Considering the temporal relationship between tasks at the same target point, which makes the corresponding high-quality feasible solution space narrow, in order to improve the search efficiency of the algorithm, this paper designs a strategy for generating high-quality initial solutions. The specific process is as follows:

[0063] First, the target set to be attacked is replicated to generate an offensive mission target set and an evaluation mission target set. These two generated sets are identical to the targets in the target set. The initial number of drones is calculated by dividing the number of targets by the ammunition load, rounded up, and stored in a specific set, here named the sortie set. Without loss of generality, the subsequent process uses the offensive mission target set as an example. The evaluation mission target set allocation process is the same as the following process.

[0064] Step 3.1, randomly select a target from the attack mission target set;

[0065] Step 3.2, calculate the time required for each UAV in the motivation set to perform the task of the target selected in step 3.1, including the flight time from the previous target to the selected target and the task execution time;

[0066] Step 3.3: Under the premise of satisfying formula (11) and formula (17), select the drone that takes the shortest time to execute the task. If no drone that can execute the task is found, add a drone to the set of launch vehicles and assign the task to the newly added drone. Formula (17) is as follows:

[0067] (17)

[0068] in, Indicates drone Current sailing time; It is a manually set parameter used to balance the time a drone spends performing offensive and evaluation missions; Indicates the maximum flight time of the drone; the above formula indicates that the time a drone spends performing an attack mission must not exceed a certain ratio of its maximum flight time. This will have different effects on the initial solution.

[0069] Step 3.4: Delete the target randomly selected in step 3.1 from the offensive mission target set, and convert the corresponding mission execution information into coded information and store it in the code, i.e., a column on the code introduced in step 2.1;

[0070] Step 3.5: Repeat steps 3.1 to 3.4 until the attack mission target set becomes empty.

[0071] In step 3.6, the process similar to that in step 3.1 to step 3.4 is performed on the evaluation task target set, except that the iterative process in step 3.1 to step 3.4 for the attack task target set is replaced with the evaluation task target set. In addition, when the drone with the shortest time required to perform the evaluation task is also the drone that performs the attack task of the target, the drone with the second shortest time is selected or a new drone is added. Finally, when performing the evaluation task allocation, in formula (17) The value is 1.

[0072] After determining the initial solution generation strategy, the above strategy is used to generate several initial solutions as the search starting point of TLBO, where the number of generated initial solutions is determined by the preset parameter We call several generated initial solutions a population, and = is the population size, and each solution in the population is called an individual, which will be used in the following sections. Finally, the fitness function proposed in step 2 is used to calculate the fitness of each individual and save it. Due to the setting of the fitness function, individuals with lower fitness have better execution solutions.

[0073] Furthermore, the step 4 is specifically as follows:

[0074] The evolution of individuals in a population is essentially the exchange of coded information between individuals and their independent exploration of new codes. Step 4 primarily describes how this coded information is exchanged between individuals, namely the education and learning phases of the TLBO algorithm. Before the TLBO solution begins, a preset number of iterations is set: the number of times the entire population evolves after the initial population is generated. In each iteration, each individual in the population performs the education and learning phases mentioned in Step 4, followed by the neighborhood search proposed in Step 5. After all individuals have completed these three steps, the fitness function proposed in Step 2 is used to calculate the fitness function value for each individual. The individual with the lowest function value is selected as the optimal individual in the population, or teacher.

[0075] Step 4.1: Education phase;

[0076] Since the evolutionary method for encoding in the traditional TLBO algorithm is not applicable to the integer encoding proposed in step 2, this paper proposes an inter-individual evolutionary method applicable to this problem. The specific update method is as follows:

[0077] Based on the encoding form set in step 2, assume that there are two individuals, individual and individuals First, randomly select from each target point target points, which will be selected later. target points guide the information exchange between two individuals. Without loss of generality, if individuals To the individual Learning, positioning the previous selection The goal is individual position in the code and use it to guide individuals Replace the encoding position. The specific operation is as follows: First create a Empty encoding of the same size . Go through the selected target points, let the currently traversed target point be , at this time visit individual The code is found exist The position in the encoding of (appears twice in total, one represents the attack task, assuming that at this time it Column. One represents the evaluation task, assuming that at this time it is in the encoding Column). Encoded Copy all the information in the column (including the target point and the drone number of the corresponding task) to No. Column. Similarly, Encoded Copy all the information of the column to No. The target point is now complete. Repeat the above process until the selected Target points in Copy all the information on the code to middle.

[0078] The remaining The task information of the target point has not been saved At this time, visit The encoding is traversed from the first column to the back. If the target point recorded on the first row of the current column belongs to the remaining One of the target points, the column information is copied to On the first blank column in the . The information exchange process ends when there are no more blank columns. That is The individual obtained after learning is completed.

[0079] In each round of iteration, the teacher will conduct an education process for each individual in the population, assuming is a teacher in the population, and an individual in the population is , then the individual student The process of learning from the teacher is shown in formula (18):

[0080] (18)

[0081] in, represents the individual code obtained after learning. In formula (18), Represents the individual described above To the individual learning process.

[0082] Step 4.2: Mutual learning stage;

[0083] In the learning phase, individuals learn about their own shortcomings through interactive communication, thereby improving their knowledge. This phase begins after all individuals have completed the education phase, and each individual must learn from each other. The mechanism is as follows:

[0084] Let the currently selected student be , randomly select an individual from the population To communicate and learn, Represents the individual fitness function, student individual The communication process is shown in formula (19):

[0085] (19)

[0086] When individuals Fitness greater than When , the learning result is directly retained;

[0087] When individuals Fitness is less than When, first get , and then calculate If its fitness does not develop to a lower level, the learning will be regarded as invalid learning, and the individual will be degraded to the state before learning, that is, ; If the fitness successfully develops to a lower level, the learning effect will be retained.

[0088] When each individual in the population has completed the education stage and the mutual learning stage, proceed to step 5.

[0089] Furthermore, the step 5 is specifically as follows:

[0090] Since step 4 focuses on information exchange within the population, it is also necessary to supplement TLBO's "exploration" ability in the solution space. After the process of education between teachers and students and mutual learning among students, the present invention gives individuals autonomous thinking and learning behavior on this basis, that is, individuals can expand to a certain extent based on current knowledge, further improving the efficiency of finding the optimal solution.

[0091] There are three search behaviors: (1) randomly selecting a column in the code and randomly selecting a drone from the replaceable drone pool to replace the drone number at that gene position; (2) selecting a column in the code and randomly selecting a gene position in the corresponding feasible domain for exchange; (3) selecting a column in the code and randomly selecting any valuable insertion position in the corresponding feasible domain for insertion. In each round of iteration, each individual will conduct a neighborhood search. The above three search behaviors are executed with a certain probability. After all are completed, the search results are output and the corresponding results are retained.

[0092] The beneficial effects of the present invention are:

[0093] (1) Efficient task planning capability: This paper significantly improves the efficiency and accuracy of UAV swarm task planning by combining the adaptive teaching and learning algorithm (TLBO) with a neighborhood search operator. Its innovation lies in the design of a high-quality initial solution generation strategy and a multimodal neighborhood search operator, which enables the algorithm to quickly converge to a high-quality solution. In the specific implementation, the initial solution generation strategy in step 3 avoids the inefficient search of random initial solutions, and the neighborhood search operator in step 5 enhances the algorithm's local search capability, thereby achieving optimized task allocation under complex constraints.

[0094] (2) Global optimization under multiple constraints: This method comprehensively considers multiple constraints, such as the UAV's range, ammunition loadout, and mission timing. Through the combined optimization model constructed in step 1 and the fitness function designed in step 2, the multi-objective optimization problem is transformed into a solvable form. The penalty terms in the fitness function (such as flight time constraints and ammunition constraints) ensure that the generated solution meets the actual requirements, thus achieving global optimization.

[0095] (3) Dynamic task allocation capability: This invention can handle multiple tasks (attack and assessment tasks) for the same target point, and through the constraint design of steps 1.2 and 1.3, it ensures the timing and uniqueness of task execution. The education and mutual learning phases in step 4 dynamically adjust the task allocation scheme through information exchange between individuals, further improving the flexibility and adaptability of the algorithm.

[0096] (4) Fast convergence and stability: Through the TLBO algorithm design in step 4, combined with teacher guidance and student mutual learning mechanisms, the algorithm efficiently explores the solution space, avoiding the problem of traditional heuristic algorithms easily falling into local optimality. The neighborhood search operator in step 5 further enhances the algorithm's search capability, enabling it to quickly converge to a stable solution within a limited number of iterations. The convergence speed and convergence of the present invention are superior to traditional algorithms.

[0097] (5) Wide Applicability: The technical solution of this invention is not only applicable to military combat scenarios, but can also be extended to UAV collaborative mission planning in civilian fields, such as disaster relief and logistics distribution. Its core innovations (such as coding design, constraint processing, and adaptive search mechanism) are universal and can adapt to mission requirements of different scales and complexities.

[0098] In summary, this invention takes into account the actual combat process in actual combat. By optimizing the number of drones delivered, it shortens vehicle delivery time, reduces the risk of enemy detection, and ensures the safety of the vehicles and the smooth execution of subsequent tasks. At the same time, through a high-quality initial solution generation strategy and neighborhood search operator, the algorithm's search efficiency and solution quality are significantly improved. This method can achieve efficient allocation of drone resources and optimize mission planning in complex battlefield environments, ensuring the efficient and safe execution of missions, and has important practical application value. BRIEF DESCRIPTION OF THE DRAWINGS

[0099] Figure 1 Flowchart of the invention.

[0100] Figure 2 A coding diagram.

[0101] Figure 3 Decoding diagram.

[0102] Figure 4 Schematic diagram of individual learning.

[0103] Figure 5 Generate a graph for the neighborhood search of an optional drone pool.

[0104] Figure 6 Search for exchangeable and pluggable diagrams for the neighborhood.

[0105] Figure 7 Schematic diagram of searching for valuable insertion locations for the first type of neighborhood.

[0106] Figure 8 Schematic diagram of searching for valuable insertion locations for the second type of neighborhood.

[0107] Figure 9 Execute the flow chart for the algorithm.

[0108] Figure 10 This is the Gantt chart for Example 1.

[0109] Figure 11 This is the Gantt chart for Example 2.

[0110] Figure 12 This is the Gantt chart for Example 3.

[0111] Figure 13 This is the convergence curve diagram of Example 1.

[0112] . Figure 14 This is the convergence curve diagram of Example 2.

[0113] Figure 15 This is the convergence curve diagram of Example 3. DETAILED DESCRIPTION

[0114] The present invention is further described below with reference to specific examples. To ensure the rationality of the discussion, we compared the different performances under six sets of examples. These included 16, 13, 8, 13, 10, and 15 target points, respectively. The remaining drone and algorithm-related parameters are shown in the table:

[0115] Table 1 shows the relevant parameters of the UAV

[0116]

[0117] Table 2 shows the parameters related to the underlying UAV mission planning.

[0118]

[0119] The step 1 is specifically as follows:

[0120] First, we extract the key factors and problem-solving requirements in the actual constraints, clarify the decision variables and constraints to build a combined optimization model, and set the task goal set as , Indicates the number of drones that can be allocated, Indicates the number of target points, represents a collection of drones, Represents the target point set, where all the locations passed by the drone are .in, Indicates the drones; Indicates the target points, and then carry out model representation.

[0121] Step 1.1: Construct decision variables and related representations according to the solution requirements;

[0122] In combinatorial optimization problems, the decision variables are usually the desired mission solutions, which need to include various information about the solutions to be solved. In this problem, the information to be solved includes the drone numbers to be used and the missions to be performed by each drone. Therefore, the decision variables are set as follows:

[0123] (1)

[0124] In formula (1), represents the decision variable of this problem, Indicates the task type. Indicates drone From the target point Departure to the destination implement Task, this formula means that each task at each target point must be completed. Accordingly, the following formula can be obtained from the decision variables:

[0125] (2)

[0126] Formula (2) is used to express the navigation route of the UAV. When indicating drone From the target point Departure to the destination , obviously when There must be , and also:

[0127] (3)

[0128] Formula (3) is used to express the state of the UAV. , drones Need to go out on a mission.

[0129] Step 1.2: Construct task completion guarantee constraints based on task requirements;

[0130] (4)

[0131] Formula (4) indicates that each task at each target point must be completed, where It indicates that there are two types of missions, namely attack missions and evaluation missions, and each target point has two types of missions that need to be performed by the drone.

[0132] Step 1.3: Construct task execution requirement constraints based on task execution requirements;

[0133] (5)

[0134] Formula (5) shows that for drones , can only execute the same target point at most a task.

[0135] Step 1.4: Construct navigation constraints based on the UAV’s navigation status;

[0136] (6)

[0137] Formula (6) means that for any UAV , starting from the delivery point and finally returning to the delivery point, where and The 0 in the subscript indicates the delivery point. Similarly, there is the following formula:

[0138] (7)

[0139] Formula (7) means that for any UAV , go from any destination point to the destination point After that, the subsequent journey must start from the destination point Departure, that is, the out-degree of each target point is equal to the in-degree.

[0140] Step 1.5: Construct task completion constraints based on task execution requirements;

[0141] (8)

[0142] Formula (8) indicates that both types of tasks must be executed for each target point to complete.

[0143] Step 1.6: Establish the task execution time constraint based on the actual flight patterns and execution capabilities of the UAV;

[0144] (9)

[0145] Formula (9) indicates that each target point has a task sequence and a minimum interval time, where is the task execution time, is the minimum interval between two tasks, and Represent the target points The attack task start time and the evaluation task start time.

[0146] Step 1.7: Construct the UAV capability constraints (maximum flight time constraint, ammunition load constraint) based on the UAV's own capability limitations.

[0147] (10)

[0148] Formula (10) represents the flight time constraint of the UAV, that is, the flight time of any UAV should be less than its maximum flight time, where Indicates drone Flight time, Indicates drone In addition, since the drone consumes a certain amount of ammunition when performing an attack mission, it is also necessary to set the drone's ammunition constraint. In this problem, it is assumed that the drone consumes one piece of ammunition each time it performs an attack mission. The formula is as follows:

[0149] (11)

[0150] Formula (11) represents the UAV ammunition loading constraint, i.e., The number of attack missions must be less than the upper limit of its ammunition load, where Indicates the maximum ammunition load of the drone.

[0151] Step 1.7: Consider the task requirements and complete the expected construction of the combinatorial optimization objective function;

[0152] (12)

[0153] The objective function of this problem is shown in formula (12), where Indicates the estimated maximum number of drones in use, Indicates drone The maximum flight time, and To adjust the weight of each item, and satisfy The optimization index of this objective function is to minimize the number of drones used and the time to complete the task. It is obtained by the following formula, the specific form is as follows:

[0154] (13)

[0155] in, Represents the straight-line distance from the delivery point to the target point. Indicates the number of target points, Indicates the drone's range, represents the target point, Represents a set of target points.

[0156] Step 2: Construct the algorithm code based on the decision variables and constraints in the model. Then, based on the code form, transform the objective function into a fitness function suitable for the code, which is used to evaluate the quality of the solution corresponding to the code.

[0157] Step 2.1: Construct algorithm code based on decision variables;

[0158] In intelligent algorithms, codes are strings of numbers used to store decision variables and can be recognized by computers. Therefore, in the present invention, it is also necessary to construct corresponding teaching and learning algorithm codes around the decision variables and constraint settings in the model. For the convenience of description, the teaching and learning algorithm is collectively referred to as the TLBO algorithm below.

[0159] The encoded information includes the target and the UAV (hereinafter referred to as UAV), and the mission sequence of the same UAV is executed in the order of encoding from front to back. The encoding contains two lines of information, the first line stores the target point, and the second line stores the UAV number. The length of each line is twice the number of target points, which can be regarded as a 2-line Column matrix. In the first line of code, when the target appears for the first time, it is represented as the attack task at that point. Therefore, the second appearance of the target represents an evaluation task. The second line represents the drone that performs the task, and the task it performs is the task indicated in the first line of code in the same column. In summary, the decision variables of the combined optimization model proposed in the present invention are stored through this type of coding. The decision variables corresponding to the coding determine the complete execution plan for a round of tasks. The task execution plan obtained by the coding is generally called the decoded plan, and this name will be used later. For a more intuitive display, the specific form of the coding and the decoding process are as follows. Figure 2 and Figure 3 shown. Figure 2 middle, To reach the destination Execute the task The rest of the symbols have the same meanings as described above. The figure only shows one encoding scenario. Under normal circumstances, the target points in the first row of the encoding can be arranged arbitrarily. Figure 3 The decoding process of the code is shown. That is, how to get a specific task execution plan through a code. The upper half of the figure shows the code, and the lower half shows the decoded plan. UAV refers to the drone that performs the task. Go to the destination Execute the task .

[0160] Step 2.2: According to the constraint setting and encoding form, transform the objective function into a fitness function;

[0161] In algorithms similar to TLBO, the objective function in the combinatorial optimization model needs to be converted into a corresponding fitness function to describe the quality of the solution corresponding to a certain encoding. Since the problem studied by the present invention has high requirements for solution time, the present invention also needs to transform the objective function of the combinatorial optimization model proposed in step 1 to obtain a fitness function that can quickly evaluate the quality of the corresponding encoding solution. On the other hand, since the generated encoding cannot fully guarantee to meet the constraints during the TLBO solution process, it is necessary to introduce a penalty term to indicate whether the decoding solution of the current encoding can be implemented. The specific form is as follows:

[0162] (14)

[0163] (15)

[0164] (16)

[0165] in, represents the fitness function; and To adjust the weights of the two main indicators, and to meet ; represents the penalty term for the decoding solution exceeding the flight time constraint; The penalty term for the decoding solution exceeding the ammunition loading constraint; Indicates the estimated maximum number of drones in use; Indicates the maximum flight time of the drone; Denotes the drone in the decoding scheme Flight time; Represents the penalty coefficient of the penalty term of the flight time constraint; represents the penalty coefficient of the ammunition loading constraint penalty term; Indicates drone Number of attack missions executed; Indicates the maximum ammunition load of the drone.

[0166] In formula (14), the fitness function consists of three parts. The first part (the first two terms in the formula) represents the objective function given in step 1, which is to minimize the number of drones used and the time to complete the task. The second part Indicates that in the encoding decoding scheme, all drones that exceed the flight time limit will be penalized. Part 3 Indicates the penalty for exceeding the maximum ammunition load for all drones in the encoding decoding scheme.

[0167] Step 3: Based on the encoding setting method, a high-quality initial solution generation scheme is designed to provide high-quality initial solutions for the teaching and learning algorithm. Multiple encodings are generated through this strategy, and finally the initial population is obtained by summarizing the encodings.

[0168] Considering the temporal relationship between tasks at the same target point, which makes the corresponding high-quality feasible solution space narrow, in order to improve the search efficiency of the algorithm, this paper designs a strategy for generating high-quality initial solutions. The specific process is as follows:

[0169] First, the target set to be attacked is replicated to generate an offensive mission target set and an evaluation mission target set. These two generated sets are identical to the targets in the target set. The initial number of drones is calculated by dividing the number of targets by the ammunition load, rounded up, and stored in a specific set, here named the sortie set. Without loss of generality, the subsequent process uses the offensive mission target set as an example. The evaluation mission target set allocation process is the same as the following process.

[0170] Step 3.1, randomly select a target from the attack mission target set;

[0171] Step 3.2, calculate the time required for each UAV in the motivation set to perform the task of the target selected in step 3.1, including the flight time from the previous target to the selected target and the task execution time;

[0172] Step 3.3: Under the premise of satisfying formula (11) and formula (17), select the drone that takes the shortest time to execute the task. If no drone that can execute the task is found, add a drone to the set of launch vehicles and assign the task to the newly added drone. Formula (17) is as follows:

[0173] (17)

[0174] in, Indicates drone Current sailing time; It is a manually set parameter used to balance the time a drone spends performing offensive and evaluation missions; Indicates the maximum flight time of the drone; the above formula indicates that the time a drone spends performing an attack mission must not exceed a certain ratio of its maximum flight time. This will have different effects on the initial solution.

[0175] Step 3.4: Delete the target randomly selected in step 3.1 from the offensive mission target set, and convert the corresponding mission execution information into coded information and store it in the code, i.e., a column on the code introduced in step 2.1;

[0176] Step 3.5: Repeat steps 3.1 to 3.4 until the attack mission target set becomes empty.

[0177] In step 3.6, the process similar to that in step 3.1 to step 3.4 is performed on the evaluation task target set, except that the iterative process in step 3.1 to step 3.4 for the attack task target set is replaced with the evaluation task target set. In addition, when the drone with the shortest time required to perform the evaluation task is also the drone that performs the attack task of the target, the drone with the second shortest time is selected or a new drone is added. Finally, when performing the evaluation task allocation, in formula (17) The value is 1.

[0178] After determining the initial solution generation strategy, the above strategy is used to generate several initial solutions as the search starting point of TLBO, where the number of generated initial solutions is determined by the preset parameter We call several generated initial solutions a population, and = is the population size, and each solution in the population is called an individual, which will be used in the following sections. Finally, the fitness function proposed in step 2 is used to calculate the fitness of each individual and save it. Due to the setting of the fitness function, individuals with lower fitness have better execution solutions.

[0179] Step 4: Based on the initial population, determine the encoding update rule, that is, the evolution method of individuals in the teaching and learning phases of the TLBO algorithm, to achieve self-evolution of the algorithm and obtain the evolved population;

[0180] The evolution of individuals in a population is essentially the exchange of coded information between individuals and their independent exploration of new codes. Step 4 primarily describes how this coded information is exchanged between individuals, namely the education and learning phases of the TLBO algorithm. Before the TLBO solution begins, a preset number of iterations is set: the number of times the entire population evolves after the initial population is generated. In each iteration, each individual in the population performs the education and learning phases mentioned in Step 4, followed by the neighborhood search proposed in Step 5. After all individuals have completed these three steps, the fitness function proposed in Step 2 is used to calculate the fitness function value for each individual. The individual with the lowest function value is selected as the optimal individual in the population, or teacher.

[0181] Step 4.1: Education phase;

[0182] Since the evolutionary method for encoding in the traditional TLBO algorithm is not applicable to the integer encoding proposed in step 2, this paper proposes an inter-individual evolutionary method applicable to this problem. The specific update method is as follows:

[0183] Based on the encoding form set in step 2, assume that there are two individuals, individual and individuals First, randomly select from each target point target points, which will be selected later. target points guide the information exchange between two individuals. Without loss of generality, if individuals To the individual Learning, positioning the previous selection The goal is individual position in the code and use it to guide individuals Replace the encoding position. The specific operation is as follows: First create a Empty encoding of the same size . Go through the selected target points, let the currently traversed target point be , at this time visit individual The code is found exist The position in the encoding of (appears twice in total, one represents the attack task, assuming that at this time it Column. One represents the evaluation task, assuming that at this time it is in the encoding Column). Encoded Copy all the information in the column (including the target point and the drone number of the corresponding task) to No. Column. Similarly, Encoded Copy all the information of the column to No. The target point is now complete. Repeat the above process until the selected Target points in Copy all the information on the code to middle.

[0184] The remaining The task information of the target point has not been saved In, visit The encoding is traversed from the first column to the back. If the target point recorded on the first row of the current column belongs to the remaining One of the target points, the column information is copied to On the first blank column in the . The information exchange process ends when there are no more blank columns. That is The individual obtained after learning is completed. Figure 4 A visual display

[0185] In each round of iteration, the teacher will conduct an education process for each individual in the population, assuming is a teacher in the population, and an individual in the population is , then the individual student The process of learning from the teacher is shown in formula (18):

[0186] (18)

[0187] in, represents the individual code obtained after learning. In formula (18), Represents the individual described above To the individual learning process.

[0188] Step 4.2: Mutual learning stage;

[0189] In the learning phase, individuals learn about their own shortcomings through interactive communication, thereby improving their knowledge. This phase begins after all individuals have completed the education phase, and each individual must learn from each other. The mechanism is as follows:

[0190] Let the currently selected student be , randomly select an individual from the population To communicate and learn, Represents the individual fitness function, student individual The communication process is shown in formula (19):

[0191] (19)

[0192] When individuals Fitness greater than When , the learning result is directly retained;

[0193] When individuals Fitness is less than When, first get , and then calculate If its fitness does not develop to a lower level, the learning will be regarded as invalid learning, and the individual will be degraded to the state before learning, that is, ; If the fitness successfully develops to a lower level, the learning effect will be retained.

[0194] When each individual in the population has completed the education stage and the mutual learning stage, proceed to step 5.

[0195] Step 5: Introduce the neighborhood search operator in the evolution process and adaptively perform neighborhood search operations to further improve the search efficiency of TLBO, enhance the search ability of the population, and ultimately obtain a solution;

[0196] Since step 4 focuses on information exchange within the population, it is also necessary to supplement TLBO's "exploration" ability in the solution space. After the process of education between teachers and students and mutual learning among students, the present invention gives individuals autonomous thinking and learning behavior on this basis, that is, individuals can expand to a certain extent based on current knowledge, further improving the efficiency of finding the optimal solution.

[0197] There are three search behaviors in total, namely: (1) randomly select a column in the code, randomly select a drone from the replaceable drone pool to replace the drone number on the gene position, (2) select a column in the code, randomly select a gene position in the corresponding feasible domain for exchange, (3) select a column in the code, and randomly select any valuable insertion position in the corresponding feasible domain for insertion. In each round of iteration, each individual will conduct a neighborhood search, and the above three search behaviors are executed with a certain probability. After all are completed, the search results are output and the corresponding results are retained. Some specific information of the above three search behaviors is as follows. Figure 5 、 6 , 7, and 8.

[0198] Figure 10 、 11 Figures 12 and 13 show Gantt charts for drone missions for three examples. The light yellow rectangles in the figure represent drone flight time, the blue rectangles numbered 1 represent the time it takes for a drone to execute an attack mission, the dark yellow rectangles numbered 2 represent the time it takes for a drone to execute an assessment mission, and the light gray, unnumbered rectangles represent the time it takes for a drone to wait. Through the solutions in this section, we can calculate the number of drones required for each delivery point and determine the order in which each drone's missions should be executed. As shown in the Gantt charts, the mission execution time for each aircraft and the number of attack missions executed clearly meet the drone's maximum flight time constraints and ammunition load constraints. Furthermore, the overall mission execution time for each drone is balanced, and the task allocation scheme is reasonable. Furthermore, we can determine the mission completion time for time-constrained target points, allowing us to determine the mission time window for each delivery point. Figure 13 、 14 , 15 show the convergence curves of the three examples respectively. It can be seen that the algorithm achieves rapid convergence in the early stage of iteration.

[0199] In summary, simulation results demonstrate that the proposed method comprehensively analyzes the location information of delivery points, their associated target points, and drone resource information, effectively allocating tasks to each drone. The resulting task allocation scheme simultaneously minimizes the number of drones used and maximizes drone resource utilization, while also ensuring the shortest possible task execution time. This fully demonstrates the efficiency and rationality of TLBO.

[0200] The above-described embodiments merely express the implementation methods of the present invention, but should not be understood as limiting the scope of the present invention. It should be pointed out that those skilled in the art can make several modifications and improvements without departing from the concept of the present invention, which all fall within the scope of protection of the present invention.

Claims

1. A method for collaborative mission planning of drone swarms based on an adaptive search teaching and learning algorithm, characterized in that: The UAV swarm collaborative task planning method includes the following steps: Step 1: Extract the key factors of actual constraints, clarify the decision variables and constraints, and build a combined optimization model; Step 2: Construct an encoding form based on the decision variables and constraints in the model. Then, based on the encoding form, transform the objective function into a fitness function suitable for the encoding, which is used to subsequently evaluate the quality of the solution corresponding to the encoding. Step 3: Based on the encoding setting method, a high-quality initial solution generation scheme is designed to provide high-quality initial solutions for the teaching and learning algorithm. Multiple encodings are generated through this strategy, and finally the initial population is obtained by summarizing the encodings. Step 4: Based on the initial population, determine the code update rule, that is, the evolution method of the code in the teaching and learning phases of the teaching and learning algorithm, to achieve the self-evolution of the algorithm and obtain the evolved population; Step 5: Introduce the neighborhood search operator in the evolution process and adaptively perform neighborhood search operations to further improve the search efficiency of the algorithm, enhance the search capability of the population, and ultimately obtain a solution.

2. The method for collaborative mission planning of a swarm of unmanned aerial vehicles based on an adaptive search teaching and learning algorithm according to claim 1, characterized in that: The step 1 is specifically as follows: First, we extract the key factors and problem-solving requirements in the actual constraints, clarify the decision variables and constraints to build a combined optimization model, and set the task goal set as , Indicates the number of drones that can be allocated, Indicates the number of target points, represents a collection of drones, Represents the target point set, where all the locations passed by the drone are ;in, Indicates the drones; Indicates the target points, and then carry out model representation; Step 1.1: Construct decision variables and related representations according to the solution requirements; In a combinatorial optimization problem, the decision variable is the desired mission solution, which must include all the information needed to solve the solution. In this problem, the information to be solved includes the drone number to be used and the mission to be performed by each drone. Therefore, the decision variables are set as follows: （1）， In formula (1), represents the decision variable of this problem, Indicates the task type. Indicates drone From the target point Departure to the destination implement Task, this formula means that each task at each target point must be completed. Accordingly, the following formula is obtained from the decision variables: （2）， Formula (2) is used to express the navigation route of the UAV. When indicating drone From the target point Departure to the destination , obviously when There must be , and also: （3）， Formula (3) is used to express the state of the UAV. , drones Need to go out on mission; Step 1.2: Construct task completion guarantee constraints based on task requirements; （4）， Formula (4) indicates that each task at each target point must be completed, where There are two types of missions: attack missions and assessment missions, and each target point has two types of missions that need to be performed by the drone; Step 1.3: Construct task execution requirement constraints based on task execution requirements; （5）， Formula (5) shows that for drones , can only execute the same target point at most a task; Step 1.4: Construct navigation constraints based on the UAV’s navigation status; （6）， Formula (6) means that for any UAV , starting from the delivery point and finally returning to the delivery point, where and The 0 in the subscript indicates the delivery point; similarly, there is the following formula: （7）， Formula (7) means that for any UAV , go from any destination point to the destination point After that, the subsequent journey must start from the destination point Departure, that is, the out-degree of each target point is equal to the in-degree; Step 1.5: Construct task completion constraints based on task execution requirements; （8）， Formula (8) indicates that both types of tasks must be executed when each target point completes them; Step 1.6: Establish the task execution time constraint based on the actual flight patterns and execution capabilities of the UAV; （9）， Formula (9) indicates that each target point has a task sequence and a minimum interval time, where is the task execution time, is the minimum interval between two tasks, and Represent the target points The attack task start time and the evaluation task start time; Step 1.7: Construct the UAV capability constraints based on the UAV's own capability limitations, including maximum flight time constraints and ammunition load constraints; （10）， Formula (10) represents the flight time constraint of the UAV, that is, the flight time of any UAV should be less than its maximum flight time, where Indicates drone Flight time, Indicates drone Maximum flight time; In addition, the drone's ammunition constraint needs to be set. Assume that the drone consumes one piece of ammunition each time it performs an attack mission. The formula is as follows: （11）， Formula (11) represents the UAV ammunition loading constraint, i.e., The number of attack missions must be less than the upper limit of its ammunition load, where Indicates the maximum ammunition load of the drone; Step 1.7: Consider the task requirements and complete the expected construction of the combinatorial optimization objective function; （12）， The objective function of this problem is shown in formula (12), where Indicates the estimated maximum number of drones in use, Indicates drone The maximum flight time, and To adjust the weight of each item, and satisfy ; The optimization indicators of this objective function are to minimize the number of drones used and the time to complete the task.

3. The method for collaborative mission planning of a swarm of unmanned aerial vehicles based on an adaptive search teaching and learning algorithm according to claim 2, characterized in that: In the formula (12) It is obtained by the following formula, the specific form is as follows: （13）， in, Represents the straight-line distance from the delivery point to the target point. Indicates the number of target points, Indicates the drone's range, represents the target point, Represents a set of target points.

4. The method for collaborative mission planning of a swarm of unmanned aerial vehicles based on an adaptive search teaching and learning algorithm according to claim 2, characterized in that: The step 2 is specifically as follows: Step 2.1: Construct algorithm code based on decision variables; Based on the decision variables and constraint settings, the corresponding teaching and learning algorithm coding is constructed, which is collectively referred to as the teaching and learning algorithm as the TLBO algorithm; The coded information includes targets and drones. UAVs are collectively referred to as UAVs, and the mission sequence of the same UAV is executed in the order of coding from front to back. The code contains two lines of information. The first line stores the target points, and the second line stores the drone number. The length of each line is twice the number of target points. It is regarded as a 2-line a matrix of columns; In the first line of code, the first appearance of a target is represented as an attack mission for that point; therefore, the second appearance of the target represents an evaluation mission. The second line represents the drone that performs the mission indicated by the first line of code in the same column. The decision variables of the proposed combinatorial optimization model are stored using this encoding. The decision variables corresponding to the encoding determine the complete execution plan for a round of missions. The mission execution plan obtained from the encoding is called the decoded plan. Step 2.2: According to the constraint setting and encoding form, transform the objective function into a fitness function; The objective function of the combinatorial optimization model proposed in step 1 is modified to obtain a fitness function that can quickly evaluate the quality of the corresponding encoding scheme. A penalty term is also introduced to indicate whether the decoding scheme of the current encoding can be implemented. The specific form is as follows: （14）， （15）， （16）， in, represents the fitness function; and To adjust the weights of the two main indicators, and to meet ; represents the penalty term for the decoding solution exceeding the flight time constraint; The penalty term for the decoding solution exceeding the ammunition loading constraint; Indicates the estimated maximum number of drones in use; Indicates the maximum flight time of the drone; Denotes the drone in the decoding scheme Flight time; Represents the penalty coefficient of the penalty term of the flight time constraint; represents the penalty coefficient of the ammunition loading constraint penalty term; Indicates drone Number of attack missions executed; Indicates the maximum ammunition load of the drone; In formula (14), the fitness function consists of three parts. The first two terms in the formula are: the first part represents the objective function given in step 1, which is to minimize the number of drones used and the time to complete the task; the second part Indicates that in the encoding decoding scheme, all drones that exceed the flight time limit will be punished; the third part Indicates the penalty for exceeding the maximum ammunition load for all drones in the encoding decoding scheme.

5. The method for collaborative mission planning of a swarm of unmanned aerial vehicles based on an adaptive search teaching and learning algorithm according to claim 4, characterized in that: The step 3 is specifically as follows: First, the target set to be attacked is copied to generate an offensive mission target set and an evaluation mission target set. The two generated sets are exactly the same as the targets in the target set. The number of targets divided by the ammunition load is rounded up as the initial number of drones and stored in a specific set, here named the sortie set. The offensive mission target set is used for illustration, and the allocation process of the evaluation mission target set is the same as this process. Step 3.1, randomly select a target from the attack mission target set; Step 3.2, calculate the time required for each UAV in the motivation set to perform the task of the target selected in step 3.1, including the flight time from the previous target to the selected target and the task execution time; Step 3.3: Under the premise of satisfying formula (11) and formula (17), select the UAV that takes the shortest time to execute the task. If no UAV that can execute the task is found, add a UAV to the set of launch vehicles and assign the task to the newly added UAV. Formula (17) is as follows: （17）， in, Indicates drone Current sailing time; It is a manually set parameter used to balance the time a drone spends performing offensive and evaluation missions; Indicates the maximum flight time of the drone; the above formula indicates that the time a drone spends performing an attack mission must not exceed a certain ratio of its maximum flight time. will have different effects on the initial solution; Step 3.4: Delete the target randomly selected in step 3.1 from the offensive mission target set, and convert the corresponding mission execution information into coded information and store it in the code, i.e., a column on the code introduced in step 2.1; Step 3.5: Repeat steps 3.1 to 3.4 until the attack mission target set becomes empty. In step 3.6, the process similar to that in step 3.1 to step 3.4 is performed on the evaluation task target set, except that the iterative process of step 3.1 to step 3.4 for the attack task target set is replaced by the evaluation task target set. In addition, when the drone with the shortest time required to perform the evaluation task is also the drone that performs the attack task of the target, the drone with the second shortest time is selected or a new drone is added. Finally, when performing the evaluation task allocation, in formula (17) The value is 1; After determining the initial solution generation strategy, the above strategy is used to generate several initial solutions as the search starting point of TLBO, where the number of generated initial solutions is determined by the preset parameter Determine; several generated initial solutions are called a population, and That is the population size, and each solution in the population is called an individual; finally, the fitness function proposed in step 2 is used to calculate the fitness of each individual generated and save it.

6. The method for collaborative mission planning of a swarm of unmanned aerial vehicles based on an adaptive search teaching and learning algorithm according to claim 5, characterized in that: The step 4 is specifically as follows: In step 4, the method of exchanging coded information between individuals is mainly given, namely the education stage and learning stage in the TLBO algorithm. In the stage before the TLBO solution begins, a number of iterations is preset, that is, the number of times the entire population evolves after the initial population is generated. In one round of iteration, each individual in the population will perform the education stage and learning stage mentioned in step 4, as well as the neighborhood search proposed in the subsequent step 5. After all individuals have completed the above three processes, the fitness function proposed in step 2 is used to solve the fitness function value of each individual, and the individual with the lowest function value is selected as the optimal individual of the population, that is, the teacher. Step 4.1: Education phase; Propose an inter-individual evolution method applicable to this problem; Step 4.2: Mutual learning stage; During the learning stage, individuals recognize their own gaps through interactive communication between two individuals and thereby improve their knowledge level; this stage is carried out after all individuals have completed the education stage, and each individual must learn from each other.

7. The method for collaborative mission planning of a swarm of unmanned aerial vehicles based on an adaptive search teaching and learning algorithm according to claim 6, characterized in that: In step 4.1, the specific updating method of the inter-individual evolution mode is as follows: Based on the encoding form set in step 2, assume that there are two individuals, individual and individuals ; First, randomly select from each target point target points, which will be selected later. A target point guides the information exchange between two individuals; if individuals To the individual Learning, positioning the previous selection The goal is individual position in the code and use it to guide individuals Replace the coding position; the specific operation is as follows: first create a Empty encoding of the same size ; One by one through the selected target points, let the currently traversed target point be , at this time visit individual The code is found exist The position in the encoding of Encoded Copy all the information of the column to No. Column; similarly, Encoded Copy all the information of the column to No. Column; at this point, the target point is completed Repeat the above process until the selected Target points in Copy all the information on the code to middle; The remaining The task information of the target point has not been saved At this time, visit The encoding is traversed from the first column to the back. If the target point recorded on the first row of the current column belongs to the remaining One of the target points, the column information is copied to On the first blank column in the middle; when The information exchange process ends when there are no more blank columns; That is The individual obtained after learning is completed; In each round of iteration, the teacher will conduct an education process for each individual in the population, assuming is a teacher in the population, and an individual in the population is , then the individual student The process of learning from the teacher is shown in formula (18): （18）， in, represents the individual code obtained after learning. In formula (18), Represents the individual described above To the individual learning process.

8. The method for collaborative mission planning of a swarm of unmanned aerial vehicles based on an adaptive search teaching and learning algorithm according to claim 7, characterized in that: In step 4.2, the mechanism of the mutual learning phase is as follows: Let the currently selected student be , randomly select an individual from the population To communicate and learn, Represents the individual fitness function, student individual The communication process is shown in formula (19): （19）， When individuals Fitness greater than When , the learning result is directly retained; When individuals Fitness is less than When, first get , and then calculate If its fitness does not develop to a lower level, the learning will be regarded as invalid learning, and the individual will be degraded to the state before learning, that is, ; If the fitness successfully develops to a lower level, the learning effect is retained; When each individual in the population has completed the education stage and the mutual learning stage, proceed to step 5.

9. The method for collaborative mission planning of a swarm of unmanned aerial vehicles based on an adaptive search teaching and learning algorithm according to claim 8, characterized in that: The step 5 is specifically as follows: After the process of education between teachers and students and mutual learning among students, individuals expand based on their current knowledge, which includes three search behaviors: (1) randomly select a column in the code, randomly select a drone from the replaceable drone pool to replace the drone number on the gene position, (2) select a column in the code, randomly select a gene position in the corresponding feasible domain to exchange, (3) select a column in the code, and randomly select any valuable insertion position in the corresponding feasible domain to insert; In each round of iteration, each individual will conduct a neighborhood search, perform the above search behavior, output the search results after all are completed, and retain the corresponding results.

Images