The Research On Multi-objective Optimization Algorithm Based On Hybrid Swarm Intelligence

Swarm intelligence optimization algorithm is an emerging technology that solves the complex optimization problem by simulating the behavior of biological groups in nature.It has a considerable performance in terms of performance improvements and practical applications in recent years.However,as optimization problems become more and more complicated,traditional single optimization methods have been difficult to find a satisfactory solution within a limited time while the optimization strategies that combine multiple swarm intelligence algorithms have become the development orientation of the field,and since most of the real world optimization problems involve the simultaneous optimization of two or more conflicting objectives,the research on multi-objective optimization problems also has important theoretical and practical significance.In addition,in scientific research and engineering applications,most optimization problems are subject to different types of constraints.Therefore,the application of optimization algorithms to solve constrained optimization problems has also received more attention,.For unconstrained optimization problems,this paper proposes a hybrid multi-objective optimization algorithm.Firstly,the Monarch Butterfly Optimization(MBO)is modified to give a multi-objective framework.Then,inspired by the idea of individual extremum in Particle Swarm Optimization(PSO)and differential strategies in Differential Evolution,a new mutation operator is used in the migration operator that combines the information of individual extremum and neighbors to guide the search of solutions;Finally,the solution is further updated in combination with the fast non-dominated sorting and the arithmetic crossover operator.For constrained optimization problems,this paper proposes a hybrid multi-objective swarm intelligence optimization algorithm.Firstly,Deb's constraint handling strategy is used to modify the definition of the dominance relation between two solutions,and the infeasible solution is mutated after each update.Then,the butterfly adjusting rate in MBO is associated with the number of iterations.Finally,a hybrid updating rule is proposed combined with particle swarm optimization algorithm.That is,the population is divided into three parts,each of which is updated by migration operator,butterfly adjusting operator and PSO updating rules respectively.A series of benchmark functions are used to verify the performance of the proposed algorithms,and the experimental results demonstrate the effectiveness of our algorithms compared with other important methods.