Improved Particle Swarm Algorithm for Production Scheduling Problems

摘要

Job shop scheduling problem (JSSP) is very common in manufacturing and industrial engineering. It is difficult to solve for its high complexity and strong constraints. In this paper, an improved PSO (IPSO) algorithm is proposed to solve job shop scheduling problem. The proposed approach employs an opposition-based learning concept to find better candidate solutions. To deal with discrete variables in JSSP, the operations among candidate solutions are modified. Simulation studies on ten well-known JSSP benchmark instances show that the proposed approach IPSO obtains promising results and outperforms other three algorithms.