摘要
针对嵌入式角位移传感器在工作过程中受到机械结构、电气系统以及外界环境等因素影响导致的原始测量误差较大,传统单一算法难以适用的问题,提出了一种基于集合经验模态分解和极限学习机的嵌入式角位移传感器动态测量误差补偿方法。通过对转子在整周范围内的原始测量误差进行集合经验模态分解,得到了一系列本征模态函数,利用相关性系数阈值法对本征模态函数进行筛选,选出其中对测量误差影响较大的分量,并对其进行希尔伯特变换,计算相应的瞬时幅值和瞬时频率,重构误差序列,利用极限学习机对残余误差进行预测补偿,得到传感器最终测量误差。实验结果表明:嵌入式角位移传感器的原始测量误差的峰峰值由117.9″降至4.5″,大幅提高了传感器的测量精度。
For embedded angular displacement sensor during working process exists large original measurement errors caused by mechanical structure,electrical system,external environment and other factors,which is difficult to apply a single algorithm,dynamic measurement error compensation method of the embedded angular displacement sensor based on ensemble empirical mode decomposition and extreme learning machine is proposed.This method obtains the intrinsic mode functions(IMF)by ensemble empirical mode decomposition(EEMD)of the original measurement error in the whole circumference,and uses the correlation coefficient threshold method to filter the intrinsic mode function.Hilbert transform(HT)is introduced to select intrinsic mode function which is performed on the components with greater influence the measurement error,the corresponding instantaneous amplitude and instantaneous frequency are calculated to reconstruct error sequence,and the residual error is predicted and compensated by the extreme learning machine(ELM)to obtain the final measurement of the sensor error.An experimental platform is built to carry out the dynamic error compensation experiment of the embedded angular displacement sensor.The experimental results show that the peak-to-peak value of the original measurement error of the embedded angular displacement sensor is reduced from 117.9″to 4.5″,which greatly improves the measurement accuracy of the sensor.
作者
孙世政
韩宇
党晓圆
李洁
SUN Shizheng;HAN Yu;DANG Xiaoyuan;LI Jie(School of Mechatronics and Vehicle Engineering,Chongqing Jiaotong University,Chongqing 400074,China;School of Intelligent Engineering,Chongqing College of Mobile Communication,Chongqing 401520,China)
出处
《重庆理工大学学报（自然科学）》
CAS
北大核心
2022年第3期78-85,共8页
Journal of Chongqing University of Technology：Natural Science
基金
重庆市教委科学技术研究计划重点项目(KJZD-K202002401)
上海市轨道交通结构耐久与系统安全重点实验室开放基金(202004)
“成渝地区双城经济圈建设”科技创新项目(KJCX2020032)。
关键词
嵌入式角位移传感器
集合经验模态分解
希尔伯特变换
极限学习机
误差补偿
embedded angular displacement sensor
ensemble empirical mode decomposition
Hilbert transform
extreme learning machine
error compensation
