Research Of Multi-Objective Shop Scheduling Based On Particle Swarm Optimization

With the development of market economy, the diversification and customization of customers′demands increase the uncertainty and dynamism of enterprise in practical production. In order to satisfy the needs of customers, the enterprise begins to pay atten to the rational allocation of resources. Therefore, the study of multi-objective scheduling problems has certain theoretical value and practical significance.A new method based on uniform design-based particle swarm optimization is proposed to deal with multi-objective flexible job shop scheduling problems. This algorithm adopts linear weighting method to change multi-objective optimization problem into the single objective optimization problem, and introduces random and uniform design method to produce weight coefficient particle, which ensures the diversity and uniform distribution of Pareto set. Besides, elite reserved strategy and dynamic neighborhood operator is designed to maintain the diversity of population and improve search capabilities of particles. Finally, the corresponding computational experiments are reported. The results indicate that the proposed algorithm is an efficient approach for the flexible job shop scheduling problems.In order to solve the allocation of parallel machines in hybrid flow shop scheduling problem, a new method called particle swarm optimization based on critical path is proposed. The encoding method based on real matrix is designed. The mutation strategy based on population diversity and the elite reserved strategy are used to maintain the diversity of population.The critical path and key block partition overcome the characteristics of population becoming premature and improves the algorithm's search ability. The experimental simulation results prove the effectiveness of the proposed algorithm.According to the characteristics of scheduling problem, this paper puts forward particle swarm algorithm based on the NEH search strategy. In the algorithm, the design of dynamic particle swarm reduces the time complexity. NEH algorithm is used to be local search strategy,while mutation strategy is utilized to maintain the diversity of population. Finally,simulation experiments prove algorithm′s efficiency.