Research And Application On Multi-Objective Evolutionary Algorithm Based On Arena Principle And Niche

No matter in the field of natural sciences or engineering,multi-objective optimization is a very important subject and problem of research.To this day,the study on multi-objectiv evolutionary algorithm entered a boom period in the international community,and all kinds of the new dominated evolutionary algorithm appeared. Genetic algorithm is a general optimization algorithm by the genetic and evolutionary process and the formation of a creature, it is used to solve complex problems by improving the advantages.Now, the multi-objective evolutionary algorithms have been widely used in natural sciences and engineering.This article introduced the study of multi-objective optimization at present, fundamental principles and genetic algorithm in the theory and techniques, etc. By the research of the algorithm running efficiency and the diversity, this paper improve the multi-objective evolutionary algorithm. And applied to the Vehicle Routing Problem with Time Windows and the problem of Flow shop scheduling,and achieve the ideal effect.In this paper, the following tasks are covered:1. The development of multi-objective evolutionary algorithm and research status is briefly introduced.2. The basic principles of genetic algorithm and implementation technology of algorithm.3. The mathematical model of multi-objective optimization prpblem and typical multi-objective evolutionary algorithms are introduced.4. A multi-objective mathematical model of Vehicle Routing Problem with Time Windows VRPTW is built, a clustering-based method and a Niched Pareto Hybrid Genetic Algorithm is proposed based. Hybrid parallel select approach,cluster reducing the non-dominated solutions(NDS), arena's principleare are used in this algorithm and, by these methods, the difficulty of solving non-convex problems and easily premature are advoided,and the convergence speed of Genetic Algorithm is improved. The experimentation demonstrated that hybrid algorithm is quite effective method for VRPTW.5. In order to solve the problem of Flow shop scheduling effectively, a new method based on hybrid multi-objective genetic algorithm was proposed. In this method, a Niche Genetic Algorithm,a policy of double elite and a Pareto local search strategy was used. It can automatically adjust crossover probability and mutation probability according to fitness. The results show that this algorithm has good convergence speed and effective optimization.