基于K-Means与SVM结合的栅格分区路径规划方法

Path planning method based on K-Means and SVM combination grid partition

智能清洁机器人全局路径规划中,利用栅格法对清洁机器人的工作环境进行建模。分别介绍了K-Means聚类算法和支持向量机(SVM)算法,使用K-Means聚类算法与支持向量机(SVM)相结合的方法,以不同的约束条件进行聚类,在含有复杂障碍物的栅格地图时能有效减少分区,利用蚁群算法对分区后的栅格地图路径规划仿真,有效地提高了蚁群算法在栅格地图路径规划中的整体效率。

In the global path planning of intelligent cleaning robot,the grid method is used to model the working environment of the robot. This paper introduced K-means clustering algorithm and Support Vector Machine（ SVM） algorithm,using a combination of K-means clustering algorithm and SVM method for clustering with different constraint conditions. In the gird map containing complex obstacles,raster map can effectively reduce partition. Using ant colony algorithm to simulate the partitioned grid map path planning. It can effectively improve the ant colony algorithm in path planning of raster map of overall efficiency.