Improvements And Enhancements For Multi-layer Interactive Path Planning In Virtual Reality Simulations

In recent years, product desining and industrial production process arebecoming more and more complex. In order to satisfy the rapid development ofindustrial production, virtual reality automation models are increasingly used inindustrial field. In the use of virtual reality in industrial production, such as productassembly and disassembly tasks, interactive path planning plays a crucial role.Based on LGP-ENIT lab’s research works, an interactive path planningsolution based on a multi-layer environment description has been proposed. Theidea of this solution is based on the high-level semantic information to calculate apath to move virtual object to the defined position in virtual environment, interactwith user via interactive devices and get feedbacks from user actions to correct thepath. A part of the development works have been done by Simon CAILHOL andYinshen ZHAO.This thesis mainly discusses about the enrichment and further development ofthe interactive path planning system. In order to meet the needs of industrialproduction, the system has been improved and enhanced in the following aspects:For assisting the real production, not only the position but also the orientation ofthe object should be considered. Therefore, a planning strategy based onRapidly-exploring Random Tree (RRT) algorithm for path planning in6D space (3dimentions for position and3dimentions for orientation) has been realized. Tohandle the collisions and contacts between the objects, a collision detection functionbased on the FCL collision detection open-source library has been developed. Semantic layer can be considered as the core of the interactive planning system,responsible for controlling and scheduling other layers. In order to enrich thefunctions of semantic layer and improve the architecture of the semanticinformation. The data structure of the semantic information has been defined, asemantic information management module, topological weight computationfunction and geometric planning strategy selection function based on semanticinformation have have been realized.This thesis is aiming at improving and enriching the interactive path planningsystem. It verified the correctness of the semantic information management, collision detection and6D path planning. It provided possibility for using theinteractive path planning system to assiste production.