Multi Stage Adaptive Hybrid Swarm Intelligence Algorithm And Its Application

Swarm intelligence optimization algorithm has obvious advantages in solving nonlinear and complex optimization problems.It is widely used in various fields of engineering practice.In this paper,we focus on the deficiency that ant colony optimization(ACO)algorithm is easy to premature convergence when optimizing discrete space,low precision of the result when optimizing continuous space and bat algorithm(BA)is easy to get local optimal solution,researching the improvement of ACO algorithm in the discrete and continuous space,the improvement of bat algorithm(BA),the hybrid of the improved bat algorithm and the improved discrete ant colony algorithm and the application of these algorithms.First of all,aiming at the defection that the traditional ACO algorithm in the discrete space optimization,the pheromone updating mechanism is single and is easy to premature convergence,a multi-stage adaptive pheromone ant colony optimization(MAPACO)algorithm which combined with the actual social activities is proposed.When the proposed algorithm appears to be stagnant for a long time,the chaotic operator is introduced to help the algorithm to jump out of the premature convergence,and to give full play to the advantages of ACO algorithm.In addition,aiming at the traveling salesman problem,the proposed algorithm is compared with other improved ACO algorithms,test results verify the stronger searching power of the proposed algorithm.Moreover,for solving the problem of ACO algorithm single searching form,single sampling mechanism and low adaptive degree,we propose an adaptive parallel ant colony algorithm(APACA)by improving the ant colony optimization for continuous domains(ACOR)from three aspects: the information sharing mechanism,the search sampling mechanism and the adaptive adjustment of the weighting factor.The experimental results show that APACA performs faster convergence speed and higher precision searching optimization.Next,we propose multi-form force bat algorithm(MFBA)that searching process is divided two stages based on periodical searching strategy drawing lessons from acting force rules in physics for dealing with the deficiency that the BA is not full use of interacting among bats in searching process.In addition,test results prove that the proposed MFBA has better optimization accuracy than bat algorithm,two-phase particle swarm optimization algorithm and so on.Finally,in order to combine the advantages of different swarm intelligence algorithms,we propose multi-stage adaptive hybrid swarm intelligence algorithm combining MFBA and MAPACO,which is applied to reliability optimization of series parallel multi-state system,reliability optimization of hydraulic system and PID parameter tuning of electro-hydraulic system,verifing the ability of the proposed hybrid algorithm to solve the actual optimization problem.