基于UKF的两轮自平衡机器人姿态最优估计研究  

UKF-based Optimal Attitude Estimation of Two-wheeled Self-balanced Robots

作　　者：赵杰[1] 王晓宇[1] 秦勇[1] 蔡鹤皋[1] 

机构地区：[1]哈尔滨工业大学机器人研究所,黑龙江哈尔滨150001

出　　处：《机器人》2006年第6期605-609,共5页Robot

摘　　要：针对扩展卡尔曼滤波器(EKF)设计困难并且容易发散的问题,提出基于采样卡尔曼滤波(UKF)的方法解决滤波器设计及收敛问题,并补偿低成本的惯性传感器陀螺仪和加速度计的误差,从而得到机器人姿态的最优估计.将滤波后的模型应用到两轮自平衡机器人系统,实验结果表明UKF参数设计简单,姿态估计误差小于EKF,方差估计优于EKF,估计精度、计算量基本与EKF相当.因此,UKF能够满足两轮自平衡机器人快速机动过程中的实时姿态估计要求.For the problem that the extended Kalman filter （EKF） is difficult to design and prone to diverge, the unscented Kalman filter（UKF） algorithm based method is presented to solve the problems of the filter design and convergence. The error from the low-cost inertial gyro and accelerometer is compensated to achieve optimal attitude estimation. The filtered model is applied to the two-wheeled self-balanced robot system. Experimental results demonstrate that for the UKF, the parameter design is easier, the attitude estimation error is smaller, the covariance estimation is better than those of the EKF, while the estimation precision and the computational costs are comparable. Consequently, the UKF is suitable for the realtime attitude estimation of the two-wheeled self-balanced robot in the fast and maneuverable process.

关 键 词：采样卡尔曼滤波 采样变换 姿态估计 两轮自平衡机器人 

分 类 号：TP24[自动化与计算机技术—检测技术与自动化装置]