Industry 5.0 characterized by human-centricity and sustainability has sparked a new wave of the industrial revolution. Human-cyber-physical system is the cornerstone of human-centric manufacturing while remanufacturing has gained considerable attention in fostering sustainability. Disassembly planning is a vital but challenging task in remanufacturing. In order to accommodate the frequent changes resulting from mass personalization, a human-robot collaboration cell is valued for its ability to achieve flexible disassembly. However, recent studies treated disassembly sequence and task allocation as independent aspects. And tasks are allocated by predefined metrics and thresholds without considering the ergonomics of operators. This paper proposes a human-cyber-physical framework for the human-robot collaboration cell by illustrating human-in-the-loop and human-on-the-loop paradigms. A multi-objective sequential disassembly planning model is presented by simultaneously considering disassembly sequence and task allocation. The task complexity and operator ergonomics are evaluated based on cloud models to handle the fuzziness and randomness in human cognition. Moreover, an improved multi-objective hybrid grey wolf optimization algorithm is proposed to obtain Pareto optimal disassembly plans. A case study on the disassembly planning of a control box is presented to illustrate the feasibility and practicability of the proposed approach. The results underscore the superiority of the proposed approach in handling rigorous constraints inherent in the sequential disassembly planning model, which acquires more convergent, diverse, and uniform disassembly plans along the Pareto front.

以人为本、可持续发展的工业5.0引发了新一轮工业革命。人-网络-物理系统是以人为中心的制造的基石，而再制造在促进可持续性方面受到了广泛关注。拆卸计划是再制造中一项至关重要但具有挑战性的任务。为了适应大规模个性化带来的频繁变化，人机协作单元因其实现灵活拆卸的能力而受到重视。然而，最近的研究将拆卸顺序和任务分配视为独立的方面。任务是按预定义的指标和阈值分配的，而不考虑操作员的人体工程学。本文通过说明人在环和人在环范式，提出了人机协作单元的人-网络-物理框架。提出了同时考虑拆卸顺序和任务分配的多目标顺序拆卸规划模型。基于云模型评估任务复杂性和操作工效，以处理人类认知的模糊性和随机性。此外，提出了一种改进的多目标混合灰狼优化算法来获得Pareto最优拆卸计划。以控制箱拆卸规划为例，说明该方法的可行性和实用性。结果强调了所提出的方法在处理顺序拆卸规划模型中固有的严格约束方面的优越性，该模型沿着帕累托前沿获得了更加收敛、多样化和统一的拆卸计划。