Research On Improved Particle Swarm Optimization And Its Application In Discrete Optimization Problem

Swarm intelligence optimization algorithm is a hot research topic in the field of computer science and evolutionary computing in recent years.It is mainly used to solve continuous optimization problems and has achieved good results,which proves the feasibility and effectiveness of swarm intelligence optimization algorithm.How to apply swarm intelligence optimization algorithm to solve discrete optimization problems(namely,combinatorial optimization problem)and improve the optimization performance of such algorithms for solving complex problems is one of the key issues and directions in this field.As a kind of swarm intelligence and evolutionary algorithm,particle swarm optimization algorithm has the advantages of simple concept,convenient operation and easy combination with other algorithms,which has been widely concerned by scholars at home and abroad.At present,particle swarm optimization algorithm has achieved good performance in function optimization,vehicle path scheduling,pattern recognition and data mining.However,particle swarm optimization algorithm also has the disadvantages of premature convergence and falling in local extremum.Especially in dealing with high-dimensional complex problems,the global search ability of the algorithm is insufficient.Due to the continuity of the attribute variables and motion equations of the standard particle swarm optimization algorithm,it cannot be directly used to solve discrete optimization problems,such as 0-1 programming and path planning.In order to further improve the performance of particle swarm optimization in discrete optimization problems and expand the application scenarios of particle swarm optimization,this paper has carried out two aspects of research,on the one hand,for two types of 0-1programming problems,an adaptive stickiness particle swarm optimization algorithm based on simulated annealing mechanism is proposed on the basis of stickiness binary particle swarm optimization algorithm.The main parameters of the algorithm are adjusted by the adaptive strategy function,and the local development ability and global exploration ability of the algorithm are balanced.The divergence index and the simulated annealing mechanism are introduced to improve the ability of the algorithm to jump out of the local optimum.The improved algorithm has better solution performance.On the other hand,for the two types of path planning problems,an edge stickiness momentum discrete particle swarm optimization algorithm is proposed.From the perspective of the edge set,the concept of edge stickiness is proposed.The position vector of the particle is defined as an ordered sequence,and the velocity vector of the particle is defined as the edge stickiness set,and the motion equation is redefined.At the same time,in order to balance the exploration and development ability of the algorithm,considering the characteristics of the solution and the structure of the algorithm,the inertia part of the particle in the continuous space is designed as momentum to enhance the exploration ability of the algorithm.Inspired by the optimization mechanism of the ant colony algorithm,the concept of selection trend is introduced in the process of particle position update,and the heuristic path construction method is adopted to improve the convergence speed of the algorithm.The simulation test and statistical analysis results show that the proposed algorithm has better solution accuracy and convergence speed.