基于卡尔曼滤波的机械振动WSNs时间抖动累积误差抑制方法  

A suppression method for the cumulative time jitter error of mechanical vibration WSNs based on Kalman filter

作　　者：胡建川 汤宝平[1] 黄艺 舒帅 HU Jianchuan;TANG Baoping;HUANG Yi;SHU Shuai(State Key Laboratory of Mechanical Transmission,Chongqing University,Chongqing 400030,China)

机构地区：[1]重庆大学机械传动国家重点实验室,重庆400030

出　　处：《中国测试》2022年第1期32-39,共8页China Measurement & Test

基　　金：国家自然科学基金资助项目(51675067);重庆市自然科学基金资助项目(cstc2019jcyj-zdxmX0026)。

摘　　要：针对机械振动无线传感器网络连续高频采样的时间抖动累积误差大的问题,提出一种基于卡尔曼滤波的时间抖动累积误差抑制方法。首先设计基于IEEE 802.11标准的机械振动采集节点,并分析其时间抖动累积误差产生的原因;然后通过父子链路轮询时钟同步,采用卡尔曼滤波估计节点间时钟的频率偏移;最后各节点根据实际采样频率与时钟频率的关系修正节点的采样间隔,减小节点的采样时间抖动累积误差。实验结果表明:连续采样90 s的平均同步累积误差为0.35μs,最大同步累积误差为1.58μs,时间抖动累积误差得到有效抑制。Aiming at the problem of the large accumulation error of the time jitter in continuous high-frequency sampling of mechanical vibration wireless sensor networks,a suppression method for the cumulative time jitter error based on Kalman filter was proposed.Firstly,a mechanical vibration acquisition node on the basis of the IEEE 802.11 standard is designed,and the reason of the time jitter accumulation error is analyzed.Then,the frequency deviation of clocks among nodes is estimated by Kalman filter,which synchronizes the polling clock of the parent-child link.Finally,the sampling interval of each node is corrected according to the relationship between the actual sampling frequency and the clock frequency,which is able to reduce the cumulative error of the sampling time jitter.The experimental results show that the time jitter accumulation error is effectively suppressed,since the average synchronous accumulation error of 90 s is 0.35μs,the maximum synchronous accumulation error is 1.58μs.

关 键 词：机械振动监测 无线传感器网络 时间抖动 同步采集 卡尔曼滤波 

分 类 号：TH17[机械工程—机械制造及自动化] TP393.1[自动化与计算机技术—计算机应用技术]