虚拟角色的一种自主运动路径规划方法

Method of Autonomous Path Planning for Virtual Character

虚拟角色的路径规划是动漫游戏的一个重要课题,如何建立高效的路径规划方法仍是一个热点话题。提出了一种基于感知记忆的路径规划方法,该算法包括全局路径规划、局部路径规划和记忆模块。全局路径规划是根据已知的路径建立的,提出了拥堵系数和容忍度的公式;而局部路径规划是根据局部感知建立的,虚拟角色能够依据局部路径规划探索一个未知环境,全局路径信息记录在记忆模块中。构造了一个包含虚拟角色的三维迷宫,实验结果表明,根据前面探索信息建立的全局路径规划是有效的。

Path planning of a virtual character is a very important subject for cartoon games, and how to build the high-effect methods of path planning is still a hot topic. A perception-memory based path planning was proposed, and the algorithm is composed of global path planning, local path planning and memory module. The global path planning was set up based on the known paths, the formulas for congestion coefficient and tolerance were proposed, and the local path planning was built by local perception. A virtual character could explore an unknown environment by local path planning, and the information of the global path was recorded in memory module. A 3D maze with a virtual character was established. The experiment shows that the global path planning by previous exploration information is effective.