Hybrid of metaheuristic approaches and fuzzy logic for the integrated flowshop scheduling with predictive maintenance problem under uncertainties

Maintenance interventions must be properly integrated in the production scheduling in order to prevent failure risks. In this context, we investigate the permutation flowshop scheduling problem subjected to predictive maintenance based on prognostics and health management (PHM). To solve this problem, two integrated metaheuristics are proposed with the objective of minimising the makespan: a carefully tailored genetic algorithm (GA), and a variable neighbourhood search (VNS) incorporating well designed local search procedures. Moreover, we hybridise the two metaheuristics where the GA best solution is introduced as initial solution of VNS. The proposed metaheuristics use the fuzzy logic framework to deal with the uncertainties. To gain insight in the performance of the proposed methods, several computational experiments were conducted against Taillard's benchmarks endowed with the prognostics and predictive maintenance data. The results show a clear superiority of the proposed algorithms, especially for the genetic algorithm, regarding both solution quality and computational times. [Received: 10 June 2019; Accepted: 27 October 2020]