Unmanned ship obstacle clearance task allocation method based on improved genetic algorithm

Abstract

The invention relates to the technical field of unmanned ship autonomous control, in particular to an unmanned ship obstacle clearance task allocation method based on an improved genetic algorithm. According to the invention, through comprehensive consideration of factors such as obstacle target position distribution, the navigation speed of each boat, the number of carried blasting objects and the like, a task target is reasonably determined, task constraints are accurately analyzed, a multi-boat cooperative task allocation mathematical model fitting an actual use scene is constructed, an obstacle removal target sequence is reasonably allocated for each unmanned boat, and the unmanned boat cooperative task allocation efficiency is improved. And smooth execution of the obstacle clearing task is ensured. According to the invention, the improved genetic algorithm is adopted to solve the multi-boat collaborative obstacle clearance task allocation problem, and the comprehensiveness of individual information is ensured by adding a breeding selection strategy; through local mutation operation, the local search capability of the algorithm is enhanced; and the task execution time of each boat is balanced while the shortest obstacle clearance total route of the unmanned boat is ensured, collaborative allocation of targets of each boat is realized, and the overall task execution efficiency is greatly improved.

Description

Technical Field

[0001] The invention relates to the technical field of autonomous control of unmanned boats, and in particular to an unmanned boat obstacle removal task allocation method based on an improved genetic algorithm. Background Art

[0002] At present, a lot of research has been conducted on the task allocation problem at home and abroad. The main purpose of task allocation is to ensure that the execution of tasks consumes less energy, has a short distance, and takes a short time. However, due to the constraints of actual usage scenarios, the scale and complexity of tasks increase the difficulty of solving the task allocation problem. At present, the main methods include swarm intelligence algorithm, contract network algorithm and evolutionary algorithm. Swarm intelligence algorithm is an algorithm that simulates the behavior of biological populations in nature. Typical swarm algorithms include ant colony algorithm, particle swarm algorithm and bee colony algorithm. The contract network algorithm originates from the contract mechanism in the business process. By setting the roles of tenderers, bidders and successful bidders to imitate task negotiation, the resources are reasonably analyzed and evaluated, so as to achieve task allocation. Evolutionary algorithms include genetic algorithms, genetic programming, evolutionary programming, etc., all of which imitate the mechanism of biological evolution and inheritance in nature and are used to solve large-scale complex engineering and technical problems. Among them, genetic algorithm is a typical evolutionary algorithm. This method is a global random search and optimization method based on natural selection. It follows the basic Darwin evolution principle of "survival of the fittest" and conducts organized but random information exchange during the evolution process. As the population evolves, the genes of excellent individuals are gradually retained and reorganized, and new and better individuals are continuously produced. This process is similar to biological evolution, where excellent characteristics are continuously inherited, and repeated evolution produces optimal or suboptimal solutions for the population.

[0003] Genetic algorithm was first proposed by J. Holland in 1975. It is a method of optimizing by simulating the genetic mechanism in biological evolution in nature. This method first initializes the population and gives a set of initial solutions to the problem. Then, genetic operations such as individual selection, crossover and mutation are performed to generate new solutions that meet the fitness. The above process is repeated until the optimal solution that meets the goal is generated. In the process of "population selection" of genetic algorithm, the important reference factor is the fitness of the individual. The algorithm selects the evolutionists who will be retained to the next generation based on the individual fitness. Individuals with high fitness are more likely to be selected. These selected individuals realize the transformation and recombination of individual genes through a series of genetic operations such as crossover and mutation. Since heredity ensures the continuity and stability of genes in the process of biological evolution, and variability creates the diversity of genes, genetic algorithms not only enrich the population, but also optimize the quality of individuals, so that excellent genes can be preserved. This is reflected in the algorithm, which continuously optimizes the solution and gradually tends to the optimal solution. As a highly adaptable evolutionary algorithm, genetic algorithm is almost not restricted by the scene and problem type. There are no complex derivation operations or special rules. It directly performs random search in the solution space of the problem. At the same time, genetic algorithms also have problems such as difficult encoding methods and weak local search capabilities. The application research and improvement of genetic algorithms are currently hot issues in the field of optimization.

[0004] When performing obstacle clearance tasks, unmanned boats in the prior art travel above obstacles and use delivery equipment to drop explosives to clear the obstacles. However, when multiple unmanned boats perform obstacle clearance tasks together, the prior art lacks an efficient task allocation method, which can easily cause some unmanned boats to be idle, affecting the efficiency of obstacle clearance task execution. Summary of the invention

[0005] The technical problem to be solved by the present invention is: to propose an unmanned boat obstacle clearance task allocation method based on an improved genetic algorithm, which can comprehensively consider factors such as the obstacle target location distribution, the navigation speed of each boat and the number of explosives carried, and reasonably allocate obstacle clearance target sequences for each unmanned boat, so as to ensure that the workload of each boat is balanced under the premise of smooth execution of the task, thereby improving the overall execution efficiency of the task.

[0006] In order to solve the above technical problems, the technical solution adopted by the present invention is:

[0007] A method for allocating obstacle removal tasks for unmanned boats based on an improved genetic algorithm comprises the following steps:

[0008] S1, building a multi-boat collaborative obstacle clearance task allocation model;

[0009] S2, determine the initial position of the unmanned boat cluster, the location distribution of obstacles and targets, the speed of the unmanned boats and the number of explosives they carry;

[0010] S3, using the double chromosome encoding method to encode the correspondence between the unmanned boat and the target task sequence, and initialize the population;

[0011] S4, evaluate the fitness of the population through the fitness function, calculate the fitness value of the individual, and determine whether the evolution termination condition is met. If so, output the best individual and end the allocation process. If not, continue to execute step S5;

[0012] S5, select individuals to be retained to the next generation from the population, and calculate the number of offspring produced by the individuals, so that individuals with high fitness can generate multiple offspring, while individuals with poor fitness are retained to produce a small number of offspring;

[0013] S6, cross-match the individuals according to the crossover probability, exchange the gene fragments of the two parent individuals, and perform corresponding mutation operations on the individuals according to the mutation probability to generate new individuals; when the evolution termination condition is met, output the best individual, that is, the best allocation sequence for the unmanned boat obstacle clearance task, otherwise re-execute step S4.

[0014] Furthermore, in step S1, the construction of a multi-boat collaborative obstacle removal task allocation model specifically includes the following steps:

[0015] S11, assuming that the unmanned boat cluster performing the obstacle removal task is U = {U1, U2, ..., U n}, unmanned boat U i , the starting coordinates of i∈[1,n] are The amount of explosives carried is The navigation speed of the unmanned boat in the obstacle removal mission is recorded as The obstacle to be cleared is M = {M1, M2, ..., M m}, obstacle M j , the coordinates of j∈[1,m] are Obstacle M j With obstacle M k , the Euclidean distance between k∈[1,m] is:

[0016]

[0017] Unmanned Boat i The assigned obstacle target is denoted as Contains the set M goals;

[0018] S12, determine the multi-boat collaborative obstacle removal task allocation objective function:

[0019]

[0020] in, For unmanned boat U iFrom the initial position to the obstacle M j The distance and is the indicative function of whether the unmanned boat is associated with the obstacle, λ1 and λ2 are the distance cost and time cost coefficients respectively;

[0021] S13, determine the constraints:

[0022] 1) The number of explosives carried by the unmanned boat must be greater than the number of obstacles assigned.

[0023]

[0024] 2) The unmanned boat needs to destroy all targets, and each target can only be assigned to one unmanned boat, satisfying:

[0025]

[0026] 3) Each unmanned boat is assigned at least one target, which satisfies:

[0027]

[0028] Furthermore, in step S3, the dual chromosome encoding method is used, in which one chromosome represents the unmanned boat sequence U and the other chromosome represents the target sequence M; the gene position on the chromosome represents the unmanned boat or target number, and the alleles of the two chromosomes represent the mapping relationship between the unmanned boat and the target.

[0029] The initialization population is specifically:

[0030] For m obstacle targets, they are numbered from 1 to m, and then a sequence of chromosome genes from 1 to m is randomly generated. At the same time, any number from 1 to n is generated on the allele of another chromosome, and the visit sequence of the unmanned boat to the obstacle targets is obtained.

[0031] Furthermore, in step S4, the fitness function is set to:

[0032]

[0033] Furthermore, in step S5, the number of offspring generated by individual i in the population is:

[0034]

[0035] Among them pop max is the maximum number of offspring that an individual can generate, pop min is the minimum number of offspring that an individual can generate, f i is the fitness of individual i, minf is the best fitness of the individual, and maxf is the worst fitness of the individual.

[0036] Furthermore, in step S6, the mutation operation specifically includes:

[0037] 1) Exchange mutation

[0038] Randomly swap the positions of two targets in the target sequence, that is,

[0039]

[0040] 2) Reversal of mutation

[0041] Randomly reverse a fragment in the target sequence, that is,

[0042]

[0043] 3) Insertion mutation

[0044] Randomly insert a target in the target sequence into another position, that is,

[0045]

[0046] 4) Local variation

[0047] For the target point M that is close to j and M k , if it is assigned to different unmanned boats U in the above genetic operation p and U q , then the target points are randomly exchanged between the above unmanned boats, and the fitness value is calculated. If the fitness value decreases, the target sequence is updated; otherwise, no change is made.

[0048] Furthermore, the evolution termination condition includes: the target individual no longer changes or the maximum number of iterations is reached.

[0049] Compared with the prior art, the present invention has the following main advantages:

[0050] 1. The present invention proposes a method for allocating obstacle clearance tasks for unmanned boats based on an improved genetic algorithm. By comprehensively considering factors such as the location distribution of obstacle targets, the navigation speed of each boat, and the number of explosives carried, the task objectives are reasonably determined, the task constraints are accurately analyzed, and a mathematical model for multi-boat collaborative task allocation that fits the actual usage scenario is constructed. The obstacle clearance target sequence is reasonably allocated to each unmanned boat to ensure the smooth execution of the obstacle clearance task.

[0051] 2. The present invention adopts an improved genetic algorithm to solve the problem of multi-boat collaborative obstacle removal task allocation. By adding a breeding selection strategy, the comprehensiveness of individual information is guaranteed; the local search capability of the algorithm is enhanced through local mutation operations; while ensuring the shortest total route for unmanned boat obstacle removal, the time for each boat to perform the task is balanced, achieving collaborative allocation of the targets of each boat, and greatly improving the overall efficiency of task execution. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 A flowchart of an improved genetic algorithm according to an embodiment of the present invention;

[0053] Figure 2 Schematic diagram of a double chromosome encoding method in an embodiment of the present invention;

[0054] Figure 3 Schematic diagram of local variation in an embodiment of the present invention. DETAILED DESCRIPTION

[0055] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention. In addition, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0056] It should be pointed out that, according to the needs of implementation, the various steps / components described in this application can be split into more steps / components, and two or more steps / components or partial operations of steps / components can be combined into new steps / components to achieve the purpose of the present invention.

[0057] Embodiment 1: This embodiment provides a method for allocating unmanned boat obstacle removal tasks based on an improved genetic algorithm, which mainly includes the following steps:

[0058] The first step is to build a multi-boat collaborative obstacle removal task allocation model:

[0059] minJ=λ1d U +λ2σ t

[0060]

[0061] Among them, U={U1,U2,...,U n} is a cluster of unmanned boats, For unmanned boat U i ,i∈[1,n] is the navigation speed in the obstacle removal task, is the number of explosives carried by the unmanned boat, M={M1,M2,...,M m} is the obstacle target sequence, For unmanned boat U i From the initial position to the obstacle M j The distance The obstacle M j and obstacle M k , the distance between k and j∈[1,m], and is the indicative function of whether the unmanned boat is associated with the obstacle, λ1 and λ2 are the distance cost and time cost coefficients respectively.

[0062] The second step is to set relevant parameters and the initial position of the unmanned boat cluster. Obstacle target location distribution Speed ​​of unmanned boat and the number of explosives

[0063] The third step is to use the double chromosome coding method to encode the correspondence between the unmanned boat and the target sequence, initialize the population, number the m obstacle targets from 1 to m, and then randomly generate a sequence of chromosome genes from 1 to m. At the same time, generate any number from 1 to n on the allele of the other chromosome, and obtain the visit sequence of the unmanned boat to the obstacle target.

[0064] The fourth step is to evaluate the fitness of the population and calculate the fitness value f of the individual i , determine whether the evolution termination conditions are met. If so, output the best individual and end the process; otherwise, perform genetic operations on the population and go to the fifth step.

[0065] The fifth step is to select individuals from the population to be retained for the next generation through the following formula: Calculate the number of offspring produced by individuals so that individuals with high fitness can generate multiple offspring to ensure the continuity of good genes. At the same time, in order to avoid losing useful information from individuals with poor fitness, retain individuals with poor fitness to produce a small number of offspring.

[0066] Step 6: According to the crossover probability P c Cross-match the individuals and exchange the gene fragments of the two parent individuals; according to the mutation probability P m Perform reversal mutation, swap mutation, insertion mutation, local mutation and other operations on individuals. Through the above genetic operations, new individuals are generated and the population evolves.

[0067] Embodiment 2: This embodiment provides a method for allocating unmanned boat obstacle removal tasks based on an improved genetic algorithm, including:

[0068] (1) Basic assumptions

[0069] Assume that the unmanned boat cluster performing the obstacle removal task is U = {U1, U2, ..., U n}, unmanned boat U i , the starting coordinates of i∈[1,n] are The amount of explosives carried is The navigation speed of the unmanned boat in the obstacle removal mission is recorded as The obstacle to be cleared is M = {M1, M2, ..., M m}, obstacle M j , the coordinates of j∈[1,m] are Obstacle M j With obstacle M k , the Euclidean distance between k∈[1,m] is:

[0070]

[0071] Unmanned Boat i The assigned obstacle target is denoted as Contains the set M A goal.

[0072] (2) Objective function

[0073] When multiple boats work together to perform obstacle removal tasks, it is necessary to achieve the shortest overall obstacle removal route and balance the time spent on obstacle removal by each boat. The total route for obstacle removal by multiple boats is:

[0074]

[0075] The first half is the sum of the distances from the initial position of each boat to its first target, and the second half is the sum of the distances of all targets. and Is the indicative function:

[0076]

[0077] Unmanned Boat i Obstacle removal collection The time resources consumed are

[0078]

[0079] The average time taken by each boat to perform the obstacle removal task is

[0080]

[0081] In order to ensure that the time spent on each boat to perform the task is relatively balanced, that is, the deviation between the time spent on obstacle removal by each boat and the average time spent is small, the standard deviation is used to indicate the deviation of the time spent by each boat from the average value.

[0082]

[0083] If the standard deviation is large, the time taken by most boats is significantly different from the average time taken; if the standard deviation is small, it means that the time taken by each boat is closer to the average time taken.

[0084] In summary, in order to ensure that the total route is the shortest and the time consumption of each boat is relatively balanced when multiple boats cooperate to perform obstacle removal tasks, the multi-objective function is expressed as

[0085]

[0086] Among them, λ1 and λ2 are the distance cost and time cost coefficients respectively.

[0087] (3) Constraints

[0088] 1) The number of explosives carried by the unmanned boat must be greater than the number of obstacles assigned.

[0089]

[0090] 2) The unmanned boat needs to destroy all targets, and each target can only be assigned to one unmanned boat, satisfying:

[0091]

[0092] 3) Each unmanned boat is assigned at least one target, which satisfies:

[0093]

[0094] like Figures 1 to 3 As shown, the present application uses an improved genetic algorithm to solve the above-mentioned multi-boat collaborative obstacle removal task allocation model, regards each allocation scheme as a chromosome, and several chromosomes form a population. Through genetic operations such as crossover and mutation, chromosomes with strong adaptability are generated. The population continues to evolve and individuals with good fitness are retained. In order to ensure the diversity of population evolution and improve the efficiency of population evolution, an improved genetic algorithm with a local optimization operator is proposed. The improved genetic algorithm includes the following steps:

[0095] S1, genetic code

[0096] The dual chromosome encoding method is used to represent the unmanned boat task allocation scheme, that is, one chromosome represents the unmanned boat sequence U, and the other chromosome represents the target sequence M. The gene position on the chromosome represents the unmanned boat or target number, and the alleles of the two chromosomes represent the mapping relationship between the unmanned boat and the target.

[0097] S2, population initialization

[0098] Initialize the obstacle target sequence and unmanned boat sequence of each individual in the population, which are generally generated by random generation, that is, generating chromosome genes M1 to Mm Randomly arranged individuals, the other chromosome randomly generates numbers from U1 to U on the alleles n genes.

[0099] S3, calculate the fitness function

[0100] The fitness function is an important indicator to measure the quality of individuals and an important factor affecting the convergence speed of the algorithm. It is usually set as the objective function of the mathematical model. In the multi-boat collaborative obstacle removal task allocation model, on the one hand, it is necessary to ensure that the total route of all unmanned boats performing obstacle removal tasks is the shortest, and on the other hand, the time spent by each boat in performing obstacle removal tasks is relatively balanced. Therefore, the fitness function is set as:

[0101]

[0102] S4, select operation

[0103] Selection refers to selecting certain individuals from the current population to the next generation population. In the evolutionary process, individuals with high fitness are usually more likely to be selected. In order to ensure genetic diversity, offspring populations are generated through reproduction strategies, that is, individuals with high fitness generate multiple offspring to ensure the continuity of good genes. At the same time, in order to avoid losing useful information from individuals with poor fitness, individuals with poor fitness are retained to produce a small number of offspring, thereby more comprehensively representing population information.

[0104] The number of offspring produced by individual i in the population is:

[0105]

[0106] Among them pop max is the maximum number of offspring that an individual can generate, pop min is the minimum number of offspring that an individual can generate, f i is the fitness of individual i, minf is the best fitness of the individual, and maxf is the worst fitness of the individual.

[0107] S5, genetic operator operation

[0108] 1) Cross

[0109] Crossover refers to the exchange of some genes between two individuals in a mating exchange method to form a new individual, which can improve the diversity of chromosomes and expand the search space of solutions. The crossover operation is performed on chromosomes by partial matching crossover. In the evolution process, according to the crossover probability P c Perform a crossover operation. First, randomly generate two breakpoints and exchange the genes between the breakpoints of the two individuals. For example, the target sequences of individuals I and J are:

[0110] I:M1-M l -Mk -M j

[0111] J:M2-M m -M s -M t

[0112] The breakpoints generated are 1 and 3. The two middle gene positions are selected for exchange, and I' and J' are:

[0113] I′:M1-M m -M s -M j

[0114] J′:M2-M l -M k -M t

[0115] Since each target must be eliminated by an unmanned boat and only needs to be eliminated once, the target sequence of an unmanned boat cannot have the same number or a target with a missing number at the same time. For the phenomenon of gene duplication and / or deletion, conflict detection is required. The corresponding gene is modified according to the mapping relationship of the exchanged gene fragments and the original chromosome, and the two parent individuals are tested in turn until the detection is completed and the transformation is completed, then two new individuals are generated.

[0116] 2) Mutation:

[0117] Mutation refers to the operation of generating probabilistic gene mutations on individual genes, allowing the genes on chromosomes to recombine and form new individuals. The following mutation operations are performed according to the mutation probability:

[0118] i. Exchange variation

[0119] Randomly swap the positions of two targets in the target sequence, that is,

[0120]

[0121] ii. Reversal of mutation

[0122] Randomly reverse a fragment in the target sequence, that is,

[0123]

[0124] iii. Insertion mutation

[0125] Randomly insert a target in the target sequence into another position, that is,

[0126]

[0127] iv. Local variation

[0128] In order to enhance the local search capability of the algorithm, an individual local mutation strategy is proposed. j and M k , if it is assigned to different unmanned boats U in the above genetic operation p and U q , then the target points are randomly exchanged between the above unmanned boats, and the fitness value is calculated. If the fitness value decreases, the target sequence is updated; otherwise, no change is made.

[0129] S6, evolution termination judgment

[0130] The fitness function of the multi-boat collaborative obstacle removal task allocation model is that the total route of the unmanned boat is the shortest and the time for each boat to perform the task is balanced. Therefore, when the target no longer changes during the process, the termination condition of the genetic algorithm is met; at the same time, the maximum number of iterations of the algorithm is set. When the maximum number of iterations is reached, the termination condition is also met, and the operation of the genetic algorithm is terminated, and the best individual is output, that is, the best allocation sequence of the unmanned boat obstacle removal task. Otherwise, the individual fitness in the new population is recalculated.

[0131] Furthermore, all parts involved in this application that are not described in detail are the same as the prior art or are implemented using the prior art.

[0132] In summary:

[0133] 1. The present invention proposes a method for allocating obstacle clearance tasks for unmanned boats based on an improved genetic algorithm. By comprehensively considering factors such as the location distribution of obstacle targets, the navigation speed of each boat, and the number of explosives carried, the task objectives are reasonably determined, the task constraints are accurately analyzed, and a mathematical model for multi-boat collaborative task allocation that fits the actual usage scenario is constructed. The obstacle clearance target sequence is reasonably allocated to each unmanned boat to ensure the smooth execution of the obstacle clearance task.

[0134] 2. The present invention adopts an improved genetic algorithm to solve the problem of multi-boat collaborative obstacle removal task allocation. By adding a breeding selection strategy, the comprehensiveness of individual information is guaranteed; the local search capability of the algorithm is enhanced through local mutation operations; while ensuring the shortest total route for unmanned boat obstacle removal, the time for each boat to perform the task is balanced, achieving collaborative allocation of the targets of each boat, and greatly improving the overall efficiency of task execution.

[0135] It will be easily understood by those skilled in the art that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for allocating obstacle removal tasks for unmanned boats based on an improved genetic algorithm, characterized in that: The steps include: S1, building a multi-boat collaborative obstacle clearance task allocation model; S2, determine the initial position of the unmanned boat cluster, the location distribution of obstacles and targets, the speed of the unmanned boats and the number of explosives they carry; S3, using the double chromosome encoding method to encode the correspondence between the unmanned boat and the target task sequence, and initialize the population; S4, evaluate the fitness of the population through the fitness function, calculate the fitness value of the individual, and determine whether the evolution termination condition is met. If so, output the best individual and end the allocation process. If not, continue to execute step S5; S5, select individuals to be retained to the next generation from the population, and calculate the number of offspring produced by the individuals, so that individuals with high fitness can generate multiple offspring, while individuals with poor fitness are retained to produce a small number of offspring; S6, cross-match the individuals according to the crossover probability, exchange the gene fragments of the two parent individuals, and perform corresponding mutation operations on the individuals according to the mutation probability to generate new individuals; when the evolution termination condition is met, output the best individual, that is, the best allocation sequence for the unmanned boat obstacle clearance task, otherwise re-execute step S4.

2. The unmanned boat obstacle removal task allocation method based on improved genetic algorithm according to claim 1 is characterized in that In step S1, the multi-boat collaborative obstacle removal task allocation model is constructed, which specifically includes the following steps: S11, assuming that the unmanned boat cluster performing the obstacle removal task is U = {U1, U2, ..., U n }, unmanned boat U i , the starting coordinates of i∈[1,n] are The amount of explosives carried is The navigation speed of the unmanned boat in the obstacle removal mission is recorded as The obstacle to be cleared is M = {M1, M2, ..., M m }, obstacle M j , the coordinates of j∈[1,m] are Obstacle M j With obstacle M k , the Euclidean distance between k∈[1,m] is: Unmanned Boat i The assigned obstacle target is denoted as Contains the set M goals; S12, determine the multi-boat collaborative obstacle removal task allocation objective function: in, For unmanned boat U i From the initial position to the obstacle M j The distance and is the indicative function of whether the unmanned boat is associated with the obstacle, λ1 and λ2 are the distance cost and time cost coefficients respectively; S13, determine the constraints: 1) The number of explosives carried by the unmanned boat must be greater than the number of obstacles assigned. 2) The unmanned boat needs to destroy all targets, and each target can only be assigned to one unmanned boat, satisfying: 3) Each unmanned boat is assigned at least one target, which satisfies:

3. The unmanned boat obstacle removal task allocation method based on improved genetic algorithm according to claim 1 is characterized in that In step S3, the double chromosome encoding method is used, in which one chromosome represents the unmanned boat sequence U and the other chromosome represents the target sequence M; the gene position on the chromosome represents the unmanned boat or target number, and the alleles of the two chromosomes represent the mapping relationship between the unmanned boat and the target.

4. The unmanned boat obstacle removal task allocation method based on improved genetic algorithm according to claim 3 is characterized in that: The initialization population is specifically: For m obstacle targets, they are numbered from 1 to m, and then a sequence of chromosome genes from 1 to m is randomly generated. At the same time, any number from 1 to n is generated on the allele of another chromosome, and the visit sequence of the unmanned boat to the obstacle targets is obtained.

5. The unmanned boat obstacle removal task allocation method based on improved genetic algorithm according to claim 1 is characterized in that In step S4, the fitness function is set to:

6. The unmanned boat obstacle removal task allocation method based on improved genetic algorithm according to claim 5 is characterized in that In step S5, the number of offspring generated by individual i in the population is: Among them pop max is the maximum number of offspring that an individual can generate, pop min is the minimum number of offspring that an individual can generate, f i is the fitness of individual i, minf is the best fitness of the individual, and maxf is the worst fitness of the individual.

7. The unmanned boat obstacle removal task allocation method based on improved genetic algorithm according to claim 1 is characterized in that In step S6, the mutation operation specifically includes: 1) Exchange mutation Randomly swap the positions of two targets in the target sequence, that is, 2) Reversal of mutation Randomly reverse a fragment in the target sequence, that is, 3) Insertion mutation Randomly insert a target in the target sequence into another position, that is, 4) Local variation For the target point M that is close to j and M k , if it is assigned to different unmanned boats U in the above genetic operation p and U q , then the target points are randomly exchanged between the above unmanned boats, and the fitness value is calculated. If the fitness value decreases, the target sequence is updated; otherwise, no change is made.

8. The unmanned boat obstacle removal task allocation method based on improved genetic algorithm according to claim 7 is characterized in that: The evolution termination conditions include: the target individual no longer changes or the maximum number of iterations is reached.

Images