Enhanced Hydrozoan Algorithm with Local Searches for Job-Shop Scheduling Problems

Abstract

Job-Shop Scheduling Problem (JSSP) is one of the most difficult NP-Hard problems to solve. The goal of JSSP is to find a schedule of operations that minimizes the maximum completion time (or makespan) of n jobs with m machines. This work developed and evaluated an enhanced hydrozoan algorithm with local search strategies for efficiently solving JSSPs. The three incorporated local search strategies were one of the strengths of this algorithm, exploiting the local solution space thoroughly. Diverse exploration of global space was also enhanced by incorporation of a strong-parent crossover operator and three mutation operators. The evaluation was based on 18 widely used JSSP benchmark instances. The developed algorithm performed well, as intended; it provided the best solutions to all 18 instances that matched exactly with the best-known solutions reported in the literature. Our future work will be on attempting to apply the developed algorithm to flexible job shop problem, a more complex problem.