基于人工势场法与入侵杂草法路径规划研究

Path Planning Research Based on Artificial Potential Field Method with Invasive Weed Method

针对移动机器人路径规划研究中,移动机器人路径规划易陷入局部极小值,缺乏全局指导性及路径规划效率不高,甚至目的地不可达的问题,这里给出相应的研究方法。通过合理布局超声波探头位置,利用改进人工势场法进行移动机器人的路径规划;对移动机器人陷入局部极小值点的问题,采用入侵杂草算法在全局内有指导性的产生最优子目的地,并根据子目的地重新分配空间内的引力势,引导移动机器人摆脱"陷阱"。Matlab仿真实验表明,本文提出的路径规划算法不仅在一般环境中,而且在相对复杂的环也能引导陷入局部极小值点的移动机准确、安全到达指定目的地。为此算法主要参数选取匹配合理时,可对路径进行优化。该算法在解决局部极小值点的问题具有较高全局指导性。

This paper focuses on problems that mobile robot path planning is easy to fall into local minimum, the path planning lacks overall guidance and its efficiency is not high, and even destination can not be reached in the mobile study of robot path planning. Some appropriate research methods are presented in this paper. We presents a new mobile robot path planning program based on improved artificial potential field through the reasonable layout of ultrasonic probe. For the problem that mobile robot falls into local minimum, we use Invasive Weed Optimization to generate the sub-destination in the global scope, reallocate the gravitational potential in the space according to sub-destination and guide mobile robots to get rid of“trap”. Matlab simulation results show that the proposed path planning algorithm can make the mobile robot which falls into local minimum point reach the destination correctly and safely both in the general environment and in the relatively complex environment. When the main matched parameters of the algorithm are selected, the route can be optimized. This algorithm to solve the problem of local minimum has higher overall guidance.