摘要
无人驾驶智能车的最优路径问题是路径规划的核心问题,而算法的选择是其关键。选用的是模拟仿生类蚁群算法,针对传统的蚁群算法在搜索时间和运算速度上还有待提高,我们从信息素的更新方式及局部搜索策略方面进行了改进,并且将虚拟路径这一概念应用于动态路径规划中。在考虑了多种状态参数后,我们得出结论是实际路径最短的不一定就是最优路径,还需要取决于各状态参数的取值,这样的改进满足了车载系统的一些实时性和可行性要求。
The optimal path problem of the unmanned intelligent car is the core issues of path planning,and how to select the algorithm is the key.We selected the simulation of bionic type of ant colony algorithm in this article.There is still need to improve the search time and operation speed based on the traditional ant colony algorithm.We improved from the pheromone update method and local search strategy,and applied the concept of the virtual path to the dynamic path planning.After considering a number of parameters,we concluded that the actual shortest path is not necessarily optimal shortest paths,it also need to depend on the status parameter value.Such improvement meet some of the feasibility and real-time requirements of the on-board system.
出处
《国外电子测量技术》
2012年第9期15-17,30,共4页
Foreign Electronic Measurement Technology
关键词
无人驾驶智能车
最优路径
蚁群算法
虚拟路径
unmanned intelligent cars
optimal path
ant colony algorithm
virtual path
