基于自适应EKF的移动机器人轨迹跟踪控制器

摘要

针对轮式移动机器人在实际的轨迹跟踪任务中存在传感器测量误差且测量噪声统计信息获取困难的情况,文章首先基于自适应扩展卡尔曼滤波算法(EKF)融合惯性传感器、地磁传感器与编码器的数据.在不需要传感器测量噪声协方差先验知识的情况下获取位姿数据的同时抑制了传感器测量噪声不稳定变化对位姿数据精度的影响;其次在机器人运动学模型的基础上,构造了具有全局稳定的双环轨迹跟踪控制器.仿真与实验表明,数据滤波与全局稳定轨迹跟踪控制器结合的控制器效果良好.%According to the fact that the wheeled mobile robots are influenced by sensor measurement noise in practice. First,in this paper,base on adaptive estimation Extended Kalman Filter(EKF),an integrated naviga-tion system for wheeled mobile robot with inertial sensor,geomagnetic sensor and encoder was designed. We get the attitude and position with no past experiences of sensors' noise. Thus,this method reduces the influence by variability of sensors′noise. Then,based on the robot dynamic model,a global asymptotical stable tracking con-troller was constructed,The simulation results of this paper showed that the tracking controller,which combined both the data filtering and variable structure tracking controller has better efficiency.