摘要
采用一种局部连接 Hopfield神经网络 ( HNN)来构建骨架图 ,进行移动机器人路径规划。该HNN势场没有非期望的局部吸引点 ,所构建骨架图和环境连通性相一致 ,保证了路径规划的完整性。
Skeleton is useful in many engineering applications such as path planning of mobile robot. We aim to make skeleton's application more efficient by constructing it with HNN environment. In this paper we explain in much detail how to use HNN environment's field valley to construct skeleton. Because the HNN environment has no unexpected local attractive points and because the connectivity of the skeleton constructed by us is consistent with the HNN environment, the path planning is complete and efficient. Then we use this skeleton to plan path of mobile robot very quickly. Simulation results show preliminarily that, even for a problem as complex as that of path planning of mobile robot, our method can satisfy the real-time requirements of robot path planning and it is effective and robust when adapted to complex environment.
出处
《西北工业大学学报》
EI
CAS
CSCD
北大核心
2005年第1期40-44,共5页
Journal of Northwestern Polytechnical University
