An Efficient Meta-Heuristic for Multi-Objective Flexible Job Shop Inverse Scheduling Problem

In reality, uncertainties may still encounter after a scheduling scheme is generated. These may make the original schedule non-optimal or even impossible. Traditional scheduling methods are not effective in dealing with these situations. In response to this phenomenon, by introducing the idea of inverse optimization into the scheduling field, a new scheduling strategy called "inverse scheduling'' has been proposed. To the best of our knowledge, this is the first study to be conducted on flexible job shop inverse scheduling problem (FJISP). In this paper, first, a comprehensive mathematical model with adjustable processing time is established. Then, a hybrid multi-objective evolutionary algorithm based on decomposition and particle swarm optimization is adopted for solving FJISP. To make the proposed algorithm solving FJISP more efficiently, some new strategies are used. A 3-D coding scheme is employed to represent the particles, multiple strategies are designed for generating a high-quality initial population, and effective discrete crossover and mutation operators are specially designed according to the FJISP's characteristics. Finally, computational experiments are carried out using extended benchmarks, and the results demonstrate the effectiveness of the proposed algorithm for solving the FJISP.