Disassembly lines are utilized frequently to disassemble the end-of-life products completely or partially to retain the valuable components for remanufacturing or recycling. This research introduces and solves the profit-oriented U-shaped partial disassembly line balancing problem (PUPDLBP) for the first time. A 0–1 integer linear programming model is formulated to tackle the PUPDLBP with AND/OR precedence, which is capable of solving the small-size instances optimally. As the considered problem is NP-hard, a novel discrete cuckoo search (DCS) algorithm is implemented and improved to solve the considered problem. The proposed DCS employs a two-phase decoding procedure to handle the precedence constraint, and new population update and new method to select and replace the abandoned individuals to achieve the proper balance between exploitation and exploration. Case studies demonstrate that the U-shaped line might obtain the larger total profit than a straight line. The comparative study shows that the improvements enhance the performance of DCS by a significant margin. The proposed algorithm outperforms CPLEX solver when solving large-sized instances and produce competing performance in comparison with 11 other algorithms.

拆卸线经常用于完全或部分拆卸报废产品，以保留有价值的部件以进行再制造或回收。本研究首次引入并解决了以利润为导向的U形局部拆解线平衡问题（PUPDLBP）。制定了一个 0-1 整数线性规划模型来处理具有 AND/OR 优先级的 PUPDLBP，该模型能够最优地解决小规模实例。由于所考虑的问题是 NP-hard，因此实现并改进了一种新颖的离散布谷鸟搜索 (DCS) 算法来解决所考虑的问题。所提出的 DCS 采用两阶段解码程序来处理优先约束，以及新的种群更新和新方法来选择和替换被遗弃的个体，以实现开发和探索之间的适当平衡。案例研究表明，U 形线可能比直线获得更大的总利润。比较研究表明，这些改进显着提高了 DCS 的性能。与其他 11 种算法相比，所提出的算法在求解大型实例时优于 CPLEX 求解器，并产生具有竞争力的性能。