Hybrid Particle Swarm Optimization Algorithm for Scheduling Flexible Assembly Systems with Blocking and Deadlock Constraints
Xiaoling Li,Keyi Xing,Qingchang Lu
DOI: https://doi.org/10.1016/j.engappai.2021.104411
IF: 8
2021-01-01
Engineering Applications of Artificial Intelligence
Abstract:This paper focuses on the scheduling problem of flexible assembly systems (FASs) without intermediate buffers. The main characteristic of the problem is that as no intermediate buffer exists between consecutive machines, blocking and deadlock constraints must be considered. Petri nets are used to model the considered FASs, and a novel hybrid particle swarm optimization (HPSO) algorithm is proposed to minimize the makespan. The proposed algorithm is the combination of the discrete PSO, particle repairing algorithm, particle improvement strategy, and local search method. First, each candidate solution for the problem is encoded as a permutation with repetition of part numbers, and can be uniquely decoded into a sequence of transitions. To ensure the feasibility of solutions, a repairing algorithm is developed, in which a deadlock avoidance policy is used. Then, a particle improvement policy is proposed to improve the performance of particles. Meanwhile, a local search method is designed and incorporated into HPSO to improve its search ability. Experiments are conducted to verify the effectiveness of the particle improvement policy and local search method. Comparisons between HPSO and ten other algorithms are performed. The comparison results and analysis show that our proposed algorithm can find feasible solutions for all tested instances, and is superior to other scheduling algorithms in terms of finding better solutions and performance stability.