Incorporating learning effect and deterioration for solving a SDST flexible job-shop scheduling problem with a hybrid meta-heuristic approach

This paper considers flexible job-shop scheduling (FJSP) where set-up times were sequence-dependent, learning effect and deterioration in jobs are concurrently considered too. The optimisation criterion is the minimisation of the maximum completion time. Since the genetic algorithm (GA) and the variable neighbourhood search (VNS) are successfully applied to some NP-hard problems, we have been motivated to employ both of them to solve the considered FJSP. To further enhance and improve the hybrid algorithm, a well-known procedure called affinity function is used in our proposed algorithm genetic variable neighbourhood search with affinity function (GVNSWAF). In order to evaluate the performance of the proposed algorithm, some experiments are designed where the proposed method is compared against some new and successful algorithms found in the literature (i.e. GA and VNS). In view of the fact that the performances of meta-heuristic algorithms are considerably influenced by the proper setting of their parameters, we employed Taguchi’s robust design method to define the best parameter values. Related results are analysed by statistical tools. The experimental results and statistical analyses demonstrate that the proposed GVNSWAF is effective for the problem.