An improved PSO search method for the job shop scheduling problem

In this paper, the job shop scheduling problem (JSP) as one of the well-known hardest combinatorial optimization problems is investigated. Due to the stubborn nature of the job shop scheduling, there is no efficient solution algorithm has been found yet for solving it to optimality in polynomial time and thus provides a challenging area for metaheuristics. We present an improved Particle Swarm Optimization (PSO) algorithm to solve it where a novel particle encoding and decoding scheme is designed for representing a scheduling solution for JSP. To improve the performance of PSO further, a local search heuristic is introduced to exploit more latitude of search space to anchor the global optimum. In addition, the proposed PSO is combined with a position bound disturbance strategy to add population diversity. Computational results on two benchmark instances indicated that the proposed PSO can find the optimal solution precisely and the performance of PSO is improved greatly by incorporating the above strategies.