Disassembly Sequence Optimization for Profit and Energy Consumption Using Petri Nets and Particle Swarm Optimization

Bouazza, Syrine; Hassine, Hichem; Barkallah, Maher; Amari, Said; Haddar, Mohamed

doi:10.1007/978-3-030-84958-0_29

Cited by 1 publication

(1 citation statement)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…ZMeta-heuristics are gradually applied for solving disassembly scheduling problems due to their excellent performance in balancing the computation cost and solution quality [28][29][30] , including particle swarm optimization (PSO) [31][32][33][34][35][36] , artificial bee colony (ABC) [37][38][39][40] , genetic algorithm (GA) [41][42][43][44][45][46] , and variable neighborhood search (VNS) [40,[47][48][49][50] . Kalayci et al propose a PSO based on variable neighborhood mutation operators to solve DLBPs [31] .…”

Section: Introductionmentioning

confidence: 99%

Integrating meta-heuristics and a Sarsa algorithm for disassembly scheduling problems with cycle time and hazard coefficients

Li,

Gao,

Ren

et al. 2024

Green Manuf Open

View full text Add to dashboard Cite

End-of-life products recycling can reduce the waste of resources, and disassembly line scheduling planning can effectively improve the recycling efficiency and reduce the pollution of the environment. This work addresses a bi-objective disassembly line scheduling problem with considering time interference between tasks. The weighted sum of the cycle time and hazard coefficients is optimized. First, a mathematical model of the disassembly line scheduling problem is established under the constraints of priority and time interference relationships. Second, four meta-heuristics are improved to solve the concerned problems, including particle swarm optimization, artificial bee colony, genetic algorithm and variable neighborhood search. Ten objective-oriented local search operations are designed for improving meta-heuristics’ performance. A reinforcement learning algorithm, Sarsa, is employed to guide task assignment among workstations and local search selection during iterations, respectively. Finally, experiments are carried out for 10 instances with different scales. The effectiveness of the improving strategies is verified; the meta-heuristics combined with Sarsa based task assignment and local search strategies has better robustness and stability than the classical ones. Comparisons and discussions show that the particle swarm optimization with improved strategies outperforms other algorithms.

show abstract