The Study Of Two Swarm Intelligence Algorithms For Solving Multi-objective Optimization Problems

Particle swarm optimization(PSO)and artificial bee colony al-gorithm(ABC)are two widely used swarm intelligence algorithms,especially the artificial bee colony algorithm was proposed in recent years.The multi-objective optimization problem exists in all field-s of society.It is also a hot issue in current research to solve the multi-objective optimization problem by using swarm intelligence al-gorithm.The PSO algorithm and ABC algorithm are widely used to solving multi-objective optimization problems because of their own advantages.However,these improved algorithms still have some de-fects.In order to better solve the multi-objective optimization prob-lem,this paper proposes two improved algorithms based on the basic PSO algorithm and ABC algorithm.In view of the disadvantage of the basic PSO algorithm tend-s to fall into the local optimum,scholars have proposed various of PSO algorithms based on different topological neighborhoods,and found that topological neighborhoods have a great influence on the performance of the algorithm.In order to make the particle swar-m algorithm have a better performance in solving multi-objective optimization problems,a quantum particle swarm optimization al-gorithm based on self-organizing map(SOMQPSO)is proposed in this paper.According to trajectory analysis of particles in PSO by Clerc and Kennedy[41],the position of particle in PSO is sampled by the quantum ? potential model around the mean best position.The quantum-inspired particle swarm optimization algorithm(QP-SO)[42]is adopted to solve multi-objective problems and the self-organizing maps is used to adaptively select neighbors for particles.The SOMQPSO algorithm is tested on 14 benchmark functions and compared with classical algorithms and state-of-the-art popular al-gorithms.The results show that the algorithm has significant com-petitiveness.Inspired by the literatures[53,61],based on the basic artificial bee colony algorithm,this paper proposes a new artificial bee colony algorithm(NEWMOABC)for solving multi-objective optimization problems.The algorithm introduces the elite strategy in the basic ABC algorithm,that is,in the process of updating food sources,not only bees randomly select neighbors,but also the global best position is also used as a reference for learning.In addition,the average posi-tion of all individuals in the external archive is considered too.At the end of each iteration,cauchy mutation is performed on the individuals ranked in the top 5%in the external archive to increase the diversi-ty of the solutions and make the algorithm have the ability to jump out the local optimum when solving complex optimization problems.The experimental results of NEWMOABC algorithm prove that the algorithm has certain advantages.