摘要
为保证农用履带机器人在农田环境下的稳定行驶,减少滑动现象,提出了基于UKF(Unscented Kalman Filter)和SMO(Sliding Mode Observer)的滑移参数估算方法。利用UKF算法融合机器人的运动方程和测量方程,补偿传感器的测量误差,重建机器人的姿态。建立左、右履带滑动参数和滑移角的滑模观测器,引入双曲正切函数作为滑模观测器的切换函数,减弱滑模观测器符号切换函数引起的抖振,实现对滑动参数的精确估算。仿真和实验结果表明:该方法可以提供准确和高更新率的滑动量,为农用履带机器人的精确控制提供依据。
In order to make ensure an Agricultural Tracked Robot(ATR) stable driving in farmland and reduce the slip phenomenon parameters, a novel sliding parameters estimation scheme of ATR based on UKF(Unscented Kalman Filter) and SMO(Sliding Mode Observer) was proposed. The motion equation and measurement equation of ATR were fused by UKF to compensate the measurement error from sensors and to reconstruct the precision pose parameters of ATR. Three slip parameter observers were designed for the sideslip angle, left and right track slip rate, which used the tangent function was used as the switching function to cover chattering effects and realize the slip parameters accurately estimation for the slip parameters. The numerical simulation results show that this method is able to provide reliable and high update rate slip parameters with high update rate, which is taken as the basis for accurate control of ATR.
出处
《系统仿真学报》
CAS
CSCD
北大核心
2015年第7期1577-1583,共7页
Journal of System Simulation
基金
安徽省教育厅自然基金资助项目(KJ2014A074)
关键词
农用履带机器人
无迹卡尔曼滤波
滑模观测器
滑动参数
ATR(agricultural tracked robot)
UKF(unscented kalman filter)
SMO(sliding mode observer)
Slip parameter
