An effective metaheuristic with adaptive delay active decoding for the no-wait job-shop scheduling problem

Abstract

The no-wait job-shop scheduling problem (NWJSP) exists in several real-life manufacturing industries. It requires no waiting time between adjacent operations of the same job because of some special process requirements. The decoding strategy is a fundamental component of metaheuristics. Most previous studies ignored the delayed start paradox of NWJSP in the decoding stage, which may directly cut out potentially good solutions. Therefore, this study designs an active decoding strategy based on adaptive delay to expand the valuable solution space. Based on this, an improved genetic algorithm (IGA) is proposed for NWJSP. To enhance the local search ability and robustness of IGA, an adaptive neighbourhood search and a reinitialization strategy, respectively, are designed. In computational experiments based on a well-known benchmark (including 44 open instances), the proposed IGA achieves the new best solutions for 24 instances, demonstrating a strong advantage in solving NWJSP. Finally, IGA effectively solves a real-life painting shop case.