Using Randomized Algorithms for Digital Mock-Up in Automotive Industry

摘要

An automobile consists of a large number of component parts that must be assembled. Even if all parts precisely fit together, it is not clear whether they can be assembled or not. The process of finding a suitable assembly sequence, which can be performed in reality, is called assembly planning. We present our probabilistic motion planner Ramona developed in cooperation with Audi AG, Germany, which is used within a digital mock-up project for checking the feasibility of assembly sequences. The heart of Ramona is a probabilistic complete motion planner, together with an efficient local path planner. We describe the basic concepts of our algorithm and investigate some details of the local planner.