Scholarlink

<p>Unlike traditional assembly scenarios, sequence planning in human–robot collaborative assembly must balance efficiency with effective human–robot collaboration and dynamic decision-making capabilities. This paper proposes an assembly sequence planning method that integrates knowledge graphs, Pareto dominance relationships, and particle swarm optimization. The method optimizes assembly information aggregation and improves human–robot parallelism. First, based on key factors influencing human–robot collaboration, a dynamic allocation model for assembly strategies is proposed using fuzzy comprehensive evaluation. Next, a knowledge graph of human–robot collaborative assembly processes is constructed. Through topological sorting knowledge inference, a feasible set of assembly sequences satisfying constraints is derived. Then, using this feasible sequence set as initial particles, a novel assembly sequence planning algorithm (PD-PSO) is implemented by integrating the Pareto dominance relationship with the particle swarm optimization (PSO) algorithm. Finally, using a spur gear reducer and a planetary gear reducer as case studies, the superiority of this method in human–robot collaborative assembly scenarios is validated by comparing it with other classical heuristic algorithms across three aspects: convergence speed, solution efficiency, and sequence quality.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Human–Robot Collaborative Assembly Planning via Knowledge Graphs and PD-PSO

  • 期刊论文

  • 发表日期：

    2026年7月22日

查看全文
  • Xue Bo,
  • Tao Chen,
  • Guangyue Wang,
  • Xinying Zhang,
  • Jianzhou Ying

摘要

Unlike traditional assembly scenarios, sequence planning in human–robot collaborative assembly must balance efficiency with effective human–robot collaboration and dynamic decision-making capabilities. This paper proposes an assembly sequence planning method that integrates knowledge graphs, Pareto dominance relationships, and particle swarm optimization. The method optimizes assembly information aggregation and improves human–robot parallelism. First, based on key factors influencing human–robot collaboration, a dynamic allocation model for assembly strategies is proposed using fuzzy comprehensive evaluation. Next, a knowledge graph of human–robot collaborative assembly processes is constructed. Through topological sorting knowledge inference, a feasible set of assembly sequences satisfying constraints is derived. Then, using this feasible sequence set as initial particles, a novel assembly sequence planning algorithm (PD-PSO) is implemented by integrating the Pareto dominance relationship with the particle swarm optimization (PSO) algorithm. Finally, using a spur gear reducer and a planetary gear reducer as case studies, the superiority of this method in human–robot collaborative assembly scenarios is validated by comparing it with other classical heuristic algorithms across three aspects: convergence speed, solution efficiency, and sequence quality.