Selective disassembly sequence planning under uncertainty using trapezoidal fuzzy numbers: A novel hybrid metaheuristic algorithm
Xuesong Zhang,Anping Fu,Changshu Zhan,Duc Truong Pham,Qiang Zhao,Tiangang Qiang,Mohammed Aljuaid,Chenxi Fu
DOI: https://doi.org/10.1016/j.engappai.2023.107459
IF: 8
2023-11-25
Engineering Applications of Artificial Intelligence
Abstract:Nowadays, recycling end-of-life (EoL) products has emerged as a vital approach to address resource scarcity. Within the recycling process, disassembly plays a pivotal role and has garnered substantial attention from researchers. Disassembly sequence planning (DSP) is a crucial method to enhance disassembly efficiency. Among the various DSP models, selective disassembly sequence planning (SDSP) has gained prominence as a means to save time and reduce costs. It empowers operators to locate specific components or materials in real time, thereby boosting efficiency and minimising resource wastage. However, a notable research gap exists in the domain of SDSP, particularly in uncertain environments. To bridge this gap and render SDSP solutions more practical for real-world disassembly operations, this study adopts trapezoidal fuzzy numbers to represent uncertain information within the disassembly process and formulates a comprehensive SDSP model. In response to the intricate challenges posed by this problem, we propose a hybrid approach termed nondominated sorting genetic algorithm-II with simulated large neighborhood search (NSGA–II–SLNS). This innovative algorithm leverages the strengths of the nondominated sorting genetic algorithm-II (NSGA-II), simulated annealing algorithm (SA), and large neighborhood search (LNS). Additionally, we introduce several novel search operators into NSGA–II–SLNS, including a crossover and mutation strategy based on chaotic mapping, as well as a local search operator founded on the SA criterion and LNS. To assess the effectiveness of the proposed algorithm and model, extensive numerical case studies are conducted in this research. The outcomes contribute to the advancement of rapid, nearly optimal SDSP strategies in the face of uncertainty and ambiguity in problem settings.
automation & control systems,computer science, artificial intelligence,engineering, electrical & electronic, multidisciplinary