Disruption Particle Swarm Optimization Algorithm Based On Exponential Decay Weight

Due to the lack of communication between particles,the population loses diversity during iteration,causing the algorithm to converge prematurely and fall into local extremum.Because of these problems,we proposed a disruption particle swarm optimization algorithm based on exponential decay weight(EDW-DPSO).Firstly,the strategy adopted is to semi-uniformly initialize the populations,So that the particles are distributed throughout the space to overall uniformity,localized random manner.Ensure the randomness of the initial population and avoid the local aggregation of the population due to random initialization.Secondly,the split operator is introduced to perform dynamic optimization.The initial stage mainly explores particle splitting,increases the diversity of feasible solutions,expands the search space of the population,and avoids local optimization.When the iteration reaches a certain number of times,current position information replaces the split.Finally,using the characteristics of the exponential function,the exponential decay strategy is applied to the inertia weight,so that the algorithm advances in a large step in the early search stage,expanding the search area and ensuring the global search ability of the algorithm.In the later development stage,the update speed of the particles is slowed down by a smaller step size,so that the local extremum problem is avoided,and the convergence stability of the algorithm is improved.The algorithm is compared with the basic PSO algorithm and the variant of the PSO algorithm.Numerical experiments are carried out on five typical test functions.The experimental results show that the algorithm has a large search space in the early stage,and the population diversity increases.Later,it emphasizes local development to improve convergence precision and optimization ability,balances the search and development ability of the algorithm,It can also accelerate particles jumping out of the local extremum and approximate globle optimum.