徐明威 张禹 李延斌



摘 要:六维力传感器测量数据时输出信号不可避免地被混合噪声干扰导致降噪性能不佳,同时针对测量噪声/系统噪声模型不准确使得卡尔曼滤波辨识误差大的问题,文章采用了基于Sage-Husa的自适应卡尔曼滤波算法。将六维力传感器采集的数据分别用卡尔曼滤波器/自适应卡尔曼滤波器进行处理,分析了两种算法对传感器测量数据降噪的性能。实验结果表明,基于Sage-Husa的自适应卡尔曼滤波器对测量数据曲线的拟合度与平滑性均优于卡尔曼滤波器,其能够更有效地对随机突变噪声进行降噪处理。
关键词:六维力传感器;系统噪声;测量噪声;自适应卡尔曼滤波
中图分类号:TP241.2 文献标识码:A文章编号:2096-4706(2021)23-0033-04
A Signal Noise Processing Technology of Six-Axis Force Sensor
XU Mingwei, ZHANG Yu, LI Yanbin
(Shenyang University of Technology, Shenyang 110870, China)
Abstract: When the six-axis force sensor measures data, the output signal is inevitably interfered by mixed noise, resulting in poor noise reduction performance. At the same time, aiming at the problem of large identification error of Kalman Filter due to the inaccurate measurement noise / System Noise model, this paper uses Adaptive Kalman Filter algorithm based on Sage-Husa. The data collected by the six-axis force sensor is processed by Kalman Filter/Adaptive Kalman Filter respectively, and the performance of the two algorithms for noise reduction on sensor measurement data is analyzed. Experimental results show that the Adaptive Kalman Filter based on Sage-Husa has better fitting degree and smoothness to the measured data curve than the Kalman Filter, and it can more effectively reduce the noise of random sudden noise.
Keywords: six-axis force sensor; System Noise; measurement noise; Adaptive Kalman Filter
0 引 言
六維力传感器是一种能够实时地检测出空间六个方向力/力矩信息的力传感器,通常被安装在工业机械臂的末端,用来协助工业机械臂进行力/位置控制、孔轴装配、柔顺控制、表面抛光等一些复杂的任务,在工业机械臂的末端安装六维力传感器后,可以实现工业机械臂与外界环境接触时所产生的实时接触力,可以保证人与机械臂交互的安全性并且能以低成本的方式对传统的机械臂进行优化升级[1,2]。
然而,在工程实际应用中六维力传感器由于其敏感元件、设计结构以及放大、采集电路电路带有热噪声并伴随电磁干扰等缘故,其输出信号会不可避免的被混合噪声信号干扰,使得测量的精度变低甚至输出信号不能使用[1]。……