基于卡尔曼滤波数据融合的并联机床动态定位方法

DYNAMIC POSITIONING METHOD FOR PARALLEL MACHINE BASED ON KALMAN FILTERING DATA FUSION

针对并联机床动平台上主轴动态定位问题,将惯性传感技术运用于并联机床动平台的位姿测量,提出一种基于卡尔曼滤波(Kalman filtering,KF)数据融合的并联机床动态定位新方法。该方法融合动态惯性测量数据和机床各腿(支架)的编码器信息,通过KF算法补偿动态测量误差和计算误差,从而实现并联机床动平台位姿(位置和方向)的精确定位。详细介绍KF算法中惯性系统模型和外部测量模型的建立过程,以及测量变量与状态变量间的关系矩阵的推导过程。最后,进行一个单方向运动动态测量的简化试验,对并联机床动平台的6自由度位姿测量进行仿真研究,初步证实了该位姿算法的有效性。

For the problem of dynamic positioning of a parallel machine（PM）, inertial sensing technology to the measurement of moving platform pose is applied. A Kalman filtering based inertial sensing method is proposed for the dynamic pose measurement of a parallel machine. Through Kalman filtering algorithm, the inertial measurement measured by .inertial sensors is fused with the external ＂encoder measurement to compensate the dynamic measuring error and computational error, so as to realise accurate pose （positions and orientations） positioning of a PM platform. The modelling procedure of inertial system model and external measurement model in the KF algorithm are introduceed. The deriving process of the relation matrix of measurement variables and state variables is presented. Finally, a simplified experiment for one-axis movement is implemented to validate of the method, and the emulation work is done for the 6 degree of freedom PM pose measurement. The result shows that the proposed algorithm on PM pose determination is valid and effective.