Solving the Flexible Job Shop Scheduling Problem Using an Effective Jaya Algorithm

Abstract

Flexible job shop scheduling problem (FJSSP) is considered to be NP-hard, in which the allocation of operations to machines is not predetermined. Due to the complex nature of this problem, various metaheuristics have been increasingly employed to address the problem, obtaining satisfactory solutions in a reasonable computational time. In this work, an upcoming heuristic named Jaya algorithm (JA) is employed to minimize the makespan. The JA enjoys the advantages of absence of any algorithm-specific parameter to be tuned to achieve optimal solutions and reduced computational effort. Although JA has powerful exploration capabilities, it lacks exploitation capability. To enhance this shortfall, an effective local search technique is integrated to improve the local search ability of the JA. Hence, the proposed approach possesses superior diversification and intensification search abilities. The performance of the effective Jaya algorithm is compared with other well-known reported algorithms with Kacems benchmark instances. Experimental results revealed that the proposed algorithm gave the best results for all five problem instances.