论文部分内容阅读
山地场景数据量大,路径规划算法复杂,难以在网页上精确显示,通常采用的基于等高线的势能路径规划,往往得不到最优路径,且容易被隔断在悬崖下面。为解决以上问题,提出并实现了Web3D上的mACO(mountain ACO)路径规划算法,并在Web3D上实现了基于平面网格的pgACO(planar grid ACO)路径规划算法,以及一个Web3D上的A*路径规划算法。再以典型战斗场景为案例,针对mACO算法、pgACO算法以及A*算法,就实现效果、效率、网页刷新率(FPS)做了对比实验,结果显示,三种算法均可达到实时性,但mACO算法规划的路径更加精确。最后根据规划出来的最优路径,采用leader-follower思想,在Web3D上实现了实时高效的多智能体路径规划方案。
Due to the large amount of data and complexity of the path planning algorithm in the mountainous scene, it is difficult to display accurately on the webpage. Usually the path planning based on the potential contour of the contour line often fails to get the optimal path and is easily cut off under the cliff. In order to solve the above problems, this paper proposed and implemented the mACO (mountain ACO) path planning algorithm on Web3D, and implemented a planar grid ACO (Planar Grid ACO) path planning algorithm on Web3D and an A * path on Web3D Planning algorithm. Taking the typical battle scene as a case, a comparative experiment on the effect, the efficiency and the page refresh rate (FPS) is made for the mACO algorithm, the pgACO algorithm and the A * algorithm. The results show that all three algorithms can achieve real-time performance, The path of algorithm planning is more accurate. Finally, according to the planned optimal path, leader-follower is used to implement a real-time and efficient multi-agent path planning solution on Web3D.