摘要
在自主车辆的路径规划是否经过特定点的两种实际情况下,提出了不同的解决方案。当车辆不需要经过特定点时,引入A*算法,较传统算法将启发函数改为欧几里得函数(Euclidean Distance),并引入一个权值以降低启发函数的权重。当车辆需要经过特定点时,应用Hopfield神经网络思想优化算法,以达到理想的路径规划。仿真实验表明,改进后的算法使得路径规划寻优得到明显提高,并验证了算法的有效性。
Path planning of autonomous vehicles through a specific point of whether the actual situation of two, which proposed the respective solution ： a certain point without the need for global planning, conditions of heuristic, compared with traditional A ＊ algorithm, is equaled with euclidean distance, and add a weight which reducing the importance of heuristic relatively. In the other opposite achieve the desired path path planning, and prov planning. The ed the validity planning , the hopfiled neural network is adopted in order to simulations show that the improved algorithm enhances the performance of of the algorithm.
出处
《科学技术与工程》
2011年第34期8499-8503,共5页
Science Technology and Engineering
关键词
自主车辆
路径规划
最优路径
A＊算法
神经网络
autonomous vehiclepath planning optimal paths A ＊ algorithm neural network
