Multi-handed Assemblv Path Planning And Repair Simulation In Low Clearance Virtual Assembly And Its Application

The multi-handed assembly truss-frame structure is assembled in narrow space and requires high assembly accuracy. It is difficult for workers to plan assembly path and repair parts properly. Currently, the truss-frame structure is assembled by repeating trial assembly, repairing and re-assembling, which prolongs the assembly proccess and decreases assembly efficiency. To tackle the problems mentioned above, this article studies virtual assembly based on low clearance model for the solution of assembly path and repair scheme. The part model which combines b-rep model, point cloud, voxel array and tolerance information is constructed. The relative position method is adopted to locate non-ideal model. An improved PRM method is applied to plan assembly path with narrow passages. The variation of hybrid model is presented to express the effects of model after repairing.The first chapter summarizes the related research, reviews the key technologies including hybrid modeling for virtual assembly, path planning of low clearance model, assembly location solving considering manufacturing variations, accuracy computing and repairming simulation. The existing problems of these technologies are discussed, and the structure of the article are introduced.The second chapter proposes a hybrid model combining b-rep model, scanned data, voxel array model and tolerance information. The b-rep data is constructed by the automated parameter modeling system. The scanned data is acquired from pre-assembly parts by Handy-Scanner to express the part variation away from the design model. The voxel array is a volume model of poly-mesh data to represent the model repairing.The third chapter investigates the assembly location of non-ideal model and assembly path planning with narrow passages in virtual environment. We translate the assembly requirements to objective functions and optimize them using NSGA-II algorithm to obtain the assembly location. This chapter puts forward an improved PRM algorithm to plan the assembly motion path with narrow passages. In order to recognize the narrow passages in global area, a hybrid sampling strategy is proposed. In order to improve the efficiency of the PRM algorithm, we only dectect collisions in the area of the matching features.The fourth chapter studies the repairing methods with the consideration of interference in assembly process and accuracy request after assembly. We present a method in transferable obstacle C-space and free C-space to plan the path for parts which can not reach the target position. The accumulated interference along pre-planned assembly path is the repair volume which has to be removed. The repairing simulation process is expressed by blanking voxel and re-meshing poly-mesh faces.The fifth chapter introduces the development of the prototype system based on CATIA/CAA platform. This prototype system consists of hybrid model construction module, assembly path planning module and model repairing simulation module. We describe the overall design, user interface and application in truss-frame assembly.The sixth chapter summarizes the achievements of this article, and discusses the future studies.