Mechanical assembly belongs to the last stage of a complete mechanical product manufacturing, which usually involves many manual operations. Virtual assembly can be used to simulate a real product’s assembly process, and to assess the feasibility of the assembly scenario of a product during simulation processes. Virtual human operation simulation is an important part of virtual assembly simulation. In order to improve the fidelity and increase the accuracy of posture simulation for virtual human assembly operation, a multi-objective motion posture optimization scheme for virtual human operation is proposed in the paper. Since the real human body is a complex movement system, the virtual human is modeled as a simplified multi-rigid-body model to decrease complexity. According to ergonomics knowledge and requirements, three elements including joint angle, joint moment and operation field of vision are selected as the criteria to evaluate the virtual human’s motion. These elements are normalized and used for setting the optimization objective of human body assembly operation assessment. Optimal target functions with variables constraints used for the posture optimization problem are constructed in mathematical expression. As there are many rigid-body joint variables, it is difficult to solve the optimization model directly. The optimization model is decomposed according to different joint chains and the operating characteristics of the human body. A multi-objective NSGA-II algorithm is introduced to solve the optimization model, which finally generated a complete and continuous solution of the virtual human assembly operation motion. A case study of an engine connecting rod cap assembly is performed to demonstrate the effectiveness of the proposed optimization method.

This content is only available via PDF.
You do not currently have access to this content.