Application of Particle Swarm Optimization in the Job-Shop Scheduling Problem

Wei-jun XIA,Zhi-ming WU,Wei ZHANG,Gen-ke YANG
DOI: https://doi.org/10.3321/j.issn:1006-2467.2005.03.011
2005-01-01
Abstract:The job-shop scheduling problem (JSP) is an NP-hard problem. A new approximation algorithm was proposed for the problem of finding the minimum makespan in JSP environment. The new algorithm is based on the principle of particle swarm optimization (PSO). PSO combines local search and global search, possessing high search efficiency. The simulated annealing (SA) employs certain probability to avoid becoming trapped in a local optimum and the search process can be controlled by the cooling schedule. By reasonably combining these two different search algorithms, a general, fast and easily implemented hybrid optimization algorithm, namly HPSO, was developed. The comparison with other results in benchmark JSP problems indicates that the PSO-based algorithm is a viable and effective approach for the JSP problem.
What problem does this paper attempt to address?