期刊
ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE
卷 105, 期 -, 页码 -出版社
PERGAMON-ELSEVIER SCIENCE LTD
DOI: 10.1016/j.engappai.2021.104411
关键词
Flexible assembly system; Scheduling; Blocking; Deadlock; Particle swarm optimization; Petri net
类别
资金
- Fundamental Research Funds for the Central Universities, CHD [300102320302]
- National Natural Science Foundation of China [62103062, 61573278]
- Natural Science Basic Research Program of Shaanxi [2021JQ-288]
This paper addresses the scheduling problem of flexible assembly systems without intermediate buffers using Petri nets and a hybrid particle swarm optimization algorithm. Novel strategies such as particle improvement policy and local search method are proposed to enhance the algorithm's performance. Experimental results demonstrate the superiority of the proposed algorithm in finding better solutions and maintaining performance stability compared to other scheduling algorithms.
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.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据