一种基于路口预测的移动机器人路径规划方法

A path planning of mobile robots based on the prediction of path entry point

现实中受限于传感器数量及功能的限制,机器人难免存在各种感知盲区,故此类条件下如何实现局部路径规划,安全到达目标,是一大难点问题。以路口点代表自由路径的入口,基于无色卡尔曼滤波算法跟踪并估算盲区内路口点位置及其概率分布,从而实现当自由路径进入盲区后的信息记忆与利用。建立评价函数对盲区内及感知范围内的路口点进行统一评价,进而实现了同时基于历史信息与当前感知信息搜寻最优路口点。研究结果表明:与传统局部路径规划方法相比,该方法减少了盲区带来的机器人无效徘徊、规划失败等问题,提升了此类条件下移动机器人的路径规划能力。

Because of limited number and limited capability of sensors,mobile robot exists perception blind zone inevitably,so how to realize the path planning and safe arrive target is a common and practical problem.The＂entry point＂was introduced to represent the free road entrance,the location and probability distribution of＂entry point＂in blind zone were tracked and estimated based on uncented kalman filte algorithm.Then the historical sensor information is memorized and used after the free road enter into blind zone.All＂entry points＂were evaluated within blind zone and perception zone by set up a valuate function,so that a best free road point would be found based on both the historical and the latest sensor information.The results show that compared with the traditional local path planning approaches,the trap and hover problem are reduced with the blind zone,and path planning capacity of mobile robots is improved.1tab,18 figs,15refs.