Multi-Objective Flexible Job Shop Scheduling Problem Based On Hybrid Particle Swarm Optimization

In the past half century, the job shop scheduling problem was studied by many researchers, and abundant theories were yieled. Comparing to the ordinary job shop scheduling problem, since the characteristics of its flexible route, flexible job shop scheduling problem is more in line with the actural production environment. Therefore, how to find a solution method of multi-objective FJSP is very important.This paper is divided into two parts: the double-objective and the multi-objective. The double-objective includes makespan and machine loading, while multi-objective includes makespan, machine loading, production cost and critical machine loading. The focus of our work is to analyze and resolve the flexible job shop scheduling problem with single resource. Meanwhile, there are many properties and characteristics of the system, which can be studied and analyzed by merging the shared resources system. This article mainly discusses how to use hybrid PSO for solving multi-objective FJSP.The main works are follows:In the first place, hybrid PSO is proposed. Since annealing algorithm is good at global search, which make the algorithm easily jumps out of the partical optimal solution. By integrating the advantage of particle swarm optimization algorithm and annealing algorithm, hybrid PSO is proposed. The algorithm is applied to solving the multi-objectie FJSP and the experimental results show that the algorithm has good global search ablity.Secondly, put forward a new methord of adaptive allocation for multi-objective FJSP. Sort the single objevtive value for the population on the basis of Pareto superior, then according to the needs, weighting the objective values besed on the weight coefficients. Therefore, the condition that the result is affected by the single objective can not happen.Thirdly, draw on the mutation of operation of genetic algorithm to increase the diversity of solutions space. Utilizationof elite set to hold the non-dominated solutions. Then the contrst experiment among hybrid PSO, PSO and GA, the result show that the superiority of hybrid patircle swarm algorithm.