A novel optimization algorithm for assemblable region boundary points search in high-dimensional posture space of parts assembly

Surface deviations caused by manufacturing errors have an essential influence on the assemblability of mechanical parts. The concept of assemblable region in the posture space has been introduced as an effective tool to express the posture feasible adjustment range of non-ideal parts in the assembly process. It can reflect the complexity of posture control through ergonomic principles and thus indicates the difficulty and feasibility of assembly. The assemblability considering surface deviations can be evaluated based on the geometric and topological (G&T) properties of the assemblable region. However, one necessary precondition for realizing the analysis of G&T properties is to obtain the boundary of assemblable region in high-dimensional posture space while guaranteeing the computational efficiency. In this paper, a novel optimization algorithm is proposed for the boundary calculation of the assemblable region. The high-dimensional irregular boundary calculation problem is firstly transformed into a boundary points search problem. Then, a new heuristic boundary points diffusion search (BDS) algorithm is developed. Two main parts are designed within the proposed algorithm, including the diffusion search rules as well as the collaborative search process. By optimizing the location and quantity of points in searching group, the algorithm can efficiently achieve the boundary points search goal and realize the boundary calculation of assemblable region. Finally, two typical cases are carried out to verify the feasibility of the proposed algorithm. The comparison with an improved genetic algorithm in application also highlights the efficiency and stability of the proposed algorithm.