An improved genetic-based particle swarm optimization for job shop scheduling problem

The particle swarm optimization (PSO) is a stochastic, population-based optimization technique that has been applied to a wide range of problems, but there is little reported in respect of application to scheduling problems due to its unsuitability for them. In this paper, PSO is redefined and modified by introducing genetic operations such as crossover and mutation to update the particles, which is called GPSO and successfully employed to solve Job shop scheduling problem (JSSP) as described one of the most general and difficult of all traditional scheduling problems. Genetic-based random particle swarm optimization (RGIPSO) is proposed to avoid convergence to a local minimum, which introduces randomizing a few individuals and replacing the local best (pbest) of the particles with the pbest of others. The scheduling methods based on GPSO and RGPSO are tested on some standard benchmarks of JSSP, which are well known to measure the quality of some optimization algorithms. The effectiveness of the proposed RGPSO is demonstrated compared with GPSO and genetic algorithm. Experimental results show RGPSO yields significant improvement in solution quality and is a potentially good approach for JSSP.