A feature and optimized RRT algorithm-based assembly path planning method of complex products

Abstract

Currently, manufacturing enterprises lack strict regulations and plans for complex products. The actual assembly path mainly relies on technicians’ experience, leading to an unreasonable assembly path for each part at the workstation, an unreasonable transportation path across the workstation, and the problem of mutual interference between parts. To solve these problems, this paper studies the assembly path planning technology of complex products based on the basic geometric features and assembly sequence information. The complexity of assembly path planning is reduced by converting the assembly problem into a disassembly problem. The disassembly direction and distance of each part are determined based on the geometric feature type of each mating face, and the rapid disassembly of the assembly body is achieved. The Rapidly-exploring Random Trees (RRT) algorithm is optimized by target-tendency and double random tree strategy, and the optimal assembly path of each component is quickly generated by combining the path optimization method and interference checking method. Finally, the effectiveness of the proposed method and the superiority of the path planning algorithm are verified using the cylinder head of a Marine diesel engine (MDE) as a case study.