卡尔曼滤波在割草机器人路径追踪优化中应用  

The Application of Kalman Filtering in Path Tracking Optimization for Mowing Robot

作　　者：白刚[1] Bai Gang(Xiangxi Vocational and Technical College,Jishou 416000,China)

机构地区：[1]湘西民族职业技术学院,湖南吉首416000

出　　处：《农机化研究》2021年第6期47-51,共5页Journal of Agricultural Mechanization Research

基　　金：湖南省教育厅科研项目(18C1705)。

摘　　要：为了提高割草机器人工作效率与割草效果,设计了草机器人路径追踪系统。系统采用螺旋式路径规划方式,通过中值滤波、阈值分割等图像处理手段,得到了割草边界线。同时,提取该边界线上角点,进行线性拟合,得到系统导航方程,并采用几何追踪方法和递推方式实现导航方程追踪。采用卡尔曼滤波方法,综合考虑系统预测值和检测值,以两者协方差为依据,计算最优解。测试结果表明:该方法可以有效提高追踪精度,减少迭代次数。In order to improve efficiency and quality of mowing robot, the path tracking system was achieved. The screw-type Path was used, Boundary line lawn was found by image processing such as median filtering and Threshold segmentation. Corners were found out, and Navigation equation was achieved by linear fitting. Geometric Tracking and recurrence were used to achieve navigation tracking. Kalman filtering was used to make relationship between predictive value and detection value. The optimum solution was calculated based on covariance between predictive value and detection value. Tracking effect of Kalman filtering was tested, the result showed that this method could improve tracking accuracy and reduce iterations.

关 键 词：割草机器人 路径追踪 卡尔曼滤波 图像处理 

分 类 号：S817.11[农业科学—畜牧学] TP242[农业科学—畜牧兽医]