Study On Particle Swarm Optimization With Surprisingly Popular Strategy

Most particle swarm optimization(PSO)algorithms construct learning objectives by selecting the best fitness individuals of the population,or select the best fitness individuals among the neighbors of the particles as the local optimal learning objectives.Target selection method with fitness as its core can be regarded as democratic voting method in sociology.However,the democratic approach in sociology tends to emphasize the most popular opinions of the collective rather than necessarily select the most correct one.When most people do not have enough knowledge about a problem,democratic methods can easily mislead groups into a local optimal region.This problem will lead to the rapid loss of population diversity in the early stage of particle swarm optimization,and in the later stage of the algorithm,the population will easily fall into the sub-optimal solution.In order to deal with the problems of democracy,sociologists have proposed a surprisingly popular decision-making method.Surprisingly popular decision-making is a swarm wisdom decision-making technology originating from sociology.For a specific problem,when the majority of people in the group have unreliable opinions,the surprisingly popular strategy can maximize the extraction of a small number experts in the swarm.The precondition of unreliable opinions of most people in swarm intelligence optimization algorithm also coincides with the chaotic state in the initial stage of algorithm execution.Therefore,this paper proposes a method to realize surprisingly popular decision-making in particle swarm optimization algorithm.The particle swarm optimization(PSO)algorithm proposed in this paper is called the Surprisingly Popular Algorithm-Based Comprehensive Adaptive Topology Learning PSO(SPA-CatlePSO).The algorithm uses small-world dynamic topological connection to update the neighborhood relationship between particles in the population,simulates the knowledge dissemination mechanism in the population,and plays a role in maintaining the diversity of the population.On the basis of small-world neighborhood topology,the surprisingly popular decision is used to select the particles with the highest surprisingly popularity among the groups as the learning goal of this generation to guide the exploitation direction.To verify the effectiveness of the algorithm,we use 30 benchmark functions on CEC2014 benchmark suite to run the SPA-CatlePSO proposed in this paper,and compare the results with OLPSO,TSLPSO,ASDPSO,HCLPSO,OptBees and LShade.The experimental results show that SPA-CatlePSO algorithm is more competitive than the most advanced swarm-based intelligent algorithm.Finally,this paper introduces an optimization problem in biology,that is,the optimization model of the ordinary differential equation of the biological clock of red bread fungus.The traditional particle swarm optimization algorithm and the algorithm proposed in this paper are used to optimize this problem.It is proved that this problem can be solved by the optimization algorithm,and the algorithm proposed in this paper is superior to the traditional particle swarm optimization algorithm.