Remanufacturing is an important method to realize resource saving and circular economy. Disassembly is a vital step of remanufacturing and it is always finished by manual labor which is low efficiency. Although robotic disassembly covers the shortages of manual disassembly, it is still difficult for the robots to independently finish the disassembly process without the operator's assistance. The human-robot collaborative disassembly is being paid increasing attention and the human-robot collaborative disassembly line balancing is to assign suitable disassembly tasks to each human-robot collaborative disassembly workstation in a balanced manner. In human-robot collaborative disassembly, the operator's safety should be considered when the robot collaboratively executes the disassembly tasks with the operator. In this paper, disassembly information model of human-robot collaborative disassembly considering the safe strategy is studied. After that, this paper solves the human-robot collaborative disassembly line balancing problem using the improved discrete bees algorithm. To verify the proposed method, case study based on a bearing coupler is carried out. The results show the proposed improved discrete bees algorithm can generate the optimal solutions of human-robot collaborative disassembly line balancing problem considering the safe strategy and finds better solutions compared with the other optimization algorithms in terms of running time, average generational distance and Pareto fronts.

再制造是实现资源节约和循环经济的重要途径。拆卸是再制造的关键步骤，而且总是由人工完成，效率低下。虽然机器人拆解弥补了人工拆解的不足，但机器人在没有操作人员协助的情况下仍难以独立完成拆解过程。人机协同拆解越来越受到重视，人机协同拆解线平衡就是将合适的拆解任务均衡地分配给每个人机协同拆解工位。在人机协同拆卸中，机器人与操作者协同执行拆卸任务时，应考虑操作者的安全。在本文中，研究了考虑安全策略的人机协同拆装信息模型。之后，本文利用改进的离散蜜蜂算法解决了人机协同拆装线平衡问题。为了验证所提出的方法，进行了基于轴承耦合器的案例研究。结果表明，所提出的改进离散蜜蜂算法能够在考虑安全策略的情况下生成人机协同拆装线平衡问题的最优解，并在运行时间、平均世代距离和帕累托前沿方面与其他优化算法相比找到更好的解。本文采用改进的离散蜜蜂算法解决人机协同拆装线平衡问题。为了验证所提出的方法，进行了基于轴承耦合器的案例研究。结果表明，所提出的改进离散蜜蜂算法能够在考虑安全策略的情况下生成人机协同拆装线平衡问题的最优解，并在运行时间、平均世代距离和帕累托前沿方面与其他优化算法相比找到更好的解。本文采用改进的离散蜜蜂算法解决人机协同拆装线平衡问题。为了验证所提出的方法，进行了基于轴承耦合器的案例研究。结果表明，所提出的改进离散蜜蜂算法能够在考虑安全策略的情况下生成人机协同拆装线平衡问题的最优解，并在运行时间、平均世代距离和帕累托前沿方面与其他优化算法相比找到更好的解。