Research And Application Of Path Planning Technology In Virtual Environment

Path planning is a basic intellectual behavior of human, and also is a kind of valuable research problems in computer science, which is widely used in Artificial Intellect, Interacting Virtual Reality and many other systems. In virtual environment, path planning technology provide for virtual human the capacity to perceive and identify the environment, make a decision, and find the best path, which offers a great support to control the motion of virtual human through macro orders by operator. Path planning before generally considered distance as its reference, and made the shortest path the goal of planning. As the demand gets heavier, a lot of standards are raised, for example, the least energy required or time consumed. The question is, these sorts of path planning focus only on one purpose, which seldom take terrain influence, planning individual difference or types of barriers into consideration. Besides, how to deliver the environmental information of virtual scene affects the path searching efficiency hugely, which is also an important problem to tackle in path planning research.On the basis of the studies of former path planning technology, the thesis analyses the interacting frame between virtual human and virtual environment and corresponding procedure of path planning for virtual human firstly. Then the planning is divided by several sub-modules, which are solved one after another. The representation of environment employs conjoint method of grid and quad-tree structure. A new resolution is proposed regarding both individual information and classified obstacles for an indoor scene. To be more specific, a table containing classified obstacles is built. Path planning in outdoor environment refers to terrain impact. What is done is to simplify the ground attribution information, designing affection factors for heuristic algorithm application, which indeed modify the assessing function of traditional A* algorithm. A comparison of efficiency and effect is taken for A* and ACA through experiment, and A* is picked for further realization of indoor and outdoor path planning, in which an improvement has been made based on quad-tree structure, adopting a combining tactic of global planning and local planning. In the end, the indoor path planning approach is applied in virtual safety operation simulation system, which presents a vivid intelligent path finding of virtual human. Application indicates that the recommending technique is worthy and meaningful in use.