Multi-objective flexible job shop scheduling based on Pareto quantum particle swarm optimization algorithm

The single particle swarm algorithm exhibits deficiencies in optimality, diversity, and convergence speed when addressing the multi-objective optimal scheduling problem in flexible job shops. In this research, we introduce a multi-objective Pareto quantum particle swarm algorithm. To aviod the algorithm falling into the problem of local convergence, three initialization strategies are proposed to enhance population quality. Additionally, a crowding degree mechanism is employed to enhance the distribution of the Pareto front, thus improving solution diversity and quality. Experimental results on Kacem and Mk standard examples validate the efficiency of the proposed approach.

I. Introduction

Flexible job shop scheduling problem (FJSP) is closer to actual production, which has received widespread attention. FJSP can be divided into total flexible job shop scheduling(TFJSP) and partially flexible job shop scheduling(PFJSP), while TFJSP means that any processes can be processed on all machines, and the processes of PFJSP can only be done on available machines. Process path flexibility makes FJSP a more complex NP-hard problem.