Developed Particle Swarm Optimization Algorithm For Multiple Objective Disassembly Line Balancing Problem

In recent years, the end-of-life mechanical and electronic products strictly become a big threat to the environment and the society. The National Twelfth Five-year Plan has put the environment protection program into an important subject. Researching on the disassembly line attracts attentions from the researchers over the worldwide for its great significance to the sustainable development in economy and environment protection.According to the practical disassembly problems, a disassembly line balancing problem model focusing on complete disassembly was set up in this article. Multiple objectives were considered in the model:minimize the number of the workstation, equalize the idle time of every workstation, disassembly the high demand and hazardous part as early as possible and minimize the times of the disassembly parts direction change.1. Based on the basic particle swarm optimization algorithm was used to study the previous scholars disassembly line balancing problem, varieties of improved particle swarm optimization algorithms were proposed to improve the basic particle swarm optimization algorithm in solving discrete disassembly line balance problem solving ability. Improvement measures are mainly embodied in the continuous improvement of particle swarm optimization and discrete particle swarm optimization algorithm. The continuous particle swarm optimization algorithm mainly manifested in the species neighbor particle dynamic building DNMPSO, populations of particle global learning object for the nearest distance itself topology particles; improved discrete particle swarm optimization algorithm reflected in the crossover and mutation of the hybrid genetic algorithm, the particle swarm algorithm of "global learning" and "individual learning" generalized with corresponding particle path for crossover and mutation operation; in order to enhance the diversity of population, step by step to the particle population evolution and dynamically based on the population social population particle swarm algorithm was adopted by the particle population aggregation degree of the population evolution strategy.2. The engineering practice of empowerment method and Pareto algorithm were proposed in solving the problem of multi-objective optimization preference. Pareto algorithm includes two kinds:based on niche technology and sharing based on the phenotype. The concept of external files and non-inferior solution set were introduced in two kinds of Pareto algorithm, through the degree of population solution set is dominant in the screening of multi-objective feasible solution every iteration.The two kinds of improvement operations from continuous and discrete developed Particle Swarm Optimization according to the actual disassembly line balancing problem to improve the performance of PSO algorithm in searching the disassembly line balancing problem. By using empowerment coefficient and Pareto technique, the developed algorithm can deal with the multiple objectives optimization efficiently. Thus, the performance of the Particle Swarm Optimization Algorithm working with multiple objectives disassembly line balancing problem can be improved in this article.The article was written supported by the National natural science foundation projects (51205328).