Particle Swarm Optimization Algorithm And Its Application On The Production Scheduling Problem Of Rubber Vulcanization Workshop

This dissertation researches on production scheduling problems of rubber vulcanization workshop, sums up the characteristics of rubber vulcanization workshop, and put forward two algorithms to compute the number of batch by partition makespan and by splitting order forms. The batch Scheduling model of multi-product single-procedure on parallel machines processes of rubber vulcanization workshop is built.An improved particle swarm optimization (IPSO) algorithm is proposed. By combining the strongpoint of local search and global search, the new algorithm reduces the possibility of trapping at the local optimum. The advanced algorithm can maintain the speedability in the early convergence phase and improve the global searching ability. The IPSO algorithm is applied to optimization,and SPV method of transforming the real number into the integer code is introduced when using the IPSO algorithm to optimizes the scheduling problem with fixed batch number.According to the characteristics of the batch Scheduling problem of rubber vulcanization workshop, the batch number is computed by partitioning makespan, and then its coding method is designed, as well as decoding methods and unlawful particle estimate algorithm. Given the problem, this dissertation proposes the similar particle swarm optimization algorithm which adopts mutation operator and crossover operator of GA, and applies it to the scheduling problem's optimization of rubber vulcanization workshop in Yellowsea Rubber Group. Compared with the results of GA, the result verifies its efficiency and optimum is better than GA.We study the characteristics of multi-object optimization, build the mathematics model based on multi-object batch scheduling problem of rubber vulcanization workshop, and analyse its complexity. Co-evolutionary multi-object optimization is systemically and homely researched. This dissertation proposes the co-evolutionary multi-object optimization algorithm based on similar particle swarm optimization algorithm, which is used to obtain the solution to multi-object scheduling problems of rubber vulcanization workshop in Yellowsea Rubber Group, the result of which is compared with the result obtained by GA and SPSO algorithm. The result shows that the new algorithm is more effective than the two latter ones.