Human-Robot Collaboration Multi-objective Disassembly Line Balancing Subject to Task Failure via Multi-objective Artificial Bee Colony Algorithm

Abstract

Disassembly is a crucial process of remanufacturing end-of-life products. It can be done by either humans or robots. Manual disassembly has a high cost and low disassembly efficiency, while a robot is not flexible to perform different dynamic, complicated tasks. The combination of them can improve the disassembly efficiency. Because it is hard to know the detailed information about the wear and tear of the used products, there is a risk of disassembly failure. In this work, disassembly failure risks, constraints of disassembly priority, cycle time, and actual cost are taken into full consideration based on an AND/OR graph of used products. Besides, a mathematical model is constructed with the objectives of maximizing the profit and minimizing consumption, disassembly difficulties, and the number of workstations. Then, a new multi-objective artificial bee colony algorithm is proposed, which is combined with stochastic simulation. The effectiveness and feasibility of the proposed algorithm are verified and experimental results show that the proposed algorithm outperforms both nondominated sorting genetic algorithm II and multi-objective discrete grey wolf optimization algorithm.