Abstract: In view of the complexity of the actual disassembly work, a mathematical model of the multi-objective disassembly line balancing problem considering fuzzy task processing time is constructed, and a Pareto based multi-objective genetic simulated annealing algorithm is proposed. An improved Metropolis criterion of simulated annealing operation is presented to make it to be applicable to multi-objective optimization problem. The crowding distance is introduced as an evaluation mechanism to filter and preserve the elite solutions, and then the elite solutions are performed the genetic operation to guide the convergence to the optimal direction. Comparing the proposed algorithm with a single-objective artificial bee colony algorithm based on the 25-task involved disassembly case, it identifies the validity and the superiority of the proposed algorithm. Finally, the proposed algorithm is applied to a printer disassembly instance, and it acquires 8 balancing schemes and realizes the diversity of the solution results.

Key words: disassembly line balancing, fuzzy operation times, multi-objective optimization, genetic simulated annealing algorithm, Pareto set

摘要： 针对实际拆卸作业的复杂性，建立了考虑模糊作业时间的多目标拆卸线平衡问题的数学模型，提出了一种基于Pareto解集的多目标遗传模拟退火算法进行求解。改进了模拟退火操作的Metropolis准则，使其能够求解多目标优化问题。采用拥挤距离评价非劣解的优劣，保留了优秀个体，并通过精英选择策略，将非劣解作为遗传操作的个体，引导算法向最优方向收敛。基于25项拆卸任务算例，通过与现有的单目标人工蜂群算法进行对比，验证了所提算法的有效性和优越性。最后将该算法应用于某打印机拆卸线实例中，求得8种可选平衡方案，实现了求解结果的多样性。

关键词: 拆卸线平衡, 模糊作业时间, 多目标优化, 遗传模拟退火算法, Pareto解集