Improvement Of PSO Algorithm And Its Application In Multi-UAV Cooperative Task Allocation

The modern battlefield environment is becoming more and more complex,and unmanned aerial vehicle can’t meet the war requirements with individual operations.Multi unmanned aerial vehicle joint operations have become a hot research topic.Multi unmanned aerial vehicle cooperative task allocation is the key problem of multi unmanned aerial vehicle joint operation.Its complexity is mainly reflected in the complexity of targets and constraints and many uncertainties.The traditional algorithm is difficult to meet the requirement of solution.Modern swarm intelligence algorithm has the advantages of zero requirement for objective function,simple implementation and parallel search,and has obvious advantages in solving such problems.In this paper,particle swarm optimization algorithm is taken as the research focus.Analyze the learning mode and key elements of particle swarm optimization algorithm,design an efficient improved algorithm based on the deficiencies of the algorithm,and applied to the problem of multi unmanned aerial vehicle collaborative task assignment.The main research contents of this paper include the following aspects.Firstly,for the problem that the basic particle swarm algorithm had less convergence precision and slower convergence,the improved particle swarm optimization algorithm based on hierarchical classification is proposed.First of all,inspired by the idea of multi-subgroup,the population is divided into three different classes according to the number of iterations and the value of fitness.On the base of the particle characteristics of different classes,local,standard and global learning models are adopted to reflect the influence of individual differences and cognition on the performance of the algorithm,so as to improve the optimization performance of the algorithm.Then,layered inertia weight adjustment strategy,acceleration and deceleration factors are introduced to balance the global and local search ability of the algorithm to improve the performance of the algorithm.Finally,the feasibility and effectiveness of the improved strategy are verified by simulation experiments.Secondly,in order to further improve the performance of the algorithm and get rid of the influence of the intermediate level standard learning model on the convergence speed and convergence accuracy of the algorithm,an improved particle swarm optimization algorithm based on adaptive cuckoo algorithm and hierarchical classification is proposed.First of all,the search step of cuckoo algorithm is adaptively adjusted according to the principle of approaching optimality to realize the closed-loop feedback control of the environment,and the discovery probability is dynamically adjusted by sinusoidal decreasing strategy.Then,the standard learning model is improved by fusing cuckoo algorithm and swarm optimization algorithm based on hierarchical classification through the mechanism of dynamic greedy selection.Finally,the feasibility and effectiveness of the fusion strategy are verified by simulation experiments.Thirdly,the improved algorithm is applied to the optimization of multi unmanned aerial vehicle cooperative task allocation.First of all,a multi unmanned aerial vehicle cooperative task allocation model is established to transform the multi-objective optimization problem into a single-objective optimization problem.Then,the mapping from the problem of multi unmanned aerial vehicle collaborative task assignment to the solution space is realized by real-coded method and the solution space restriction strategy,and the overload bidding strategy is used to solve the actual overload constraint problem.Finally,simulation experiments and analysis are carried out to verify the feasibility and effectiveness of the algorithm in solving the problem of multi unmanned aerial vehicle collaborative task assignment.The thesis includes 24 figures,15 tables and 90 literautre references.