传感技术学报2017，Vol.30Issue(1)：115-119,5.DOI:10.3969/j.issn.1004-1699.2017.01.021
阈值去噪与RBF神经网络在MEMS陀螺仪误差补偿中的应用∗
Application of Threshold Denoising and RBF Neural Network in the Error Compensating of MEMS Gyro
摘要
Abstract
The large random errors in MEMS gyros have resulted in low signal noise ratio in devices outputs,which affects the application scope of gyros in turn. To solve this problem,this paper proposed a modeling forecasting method towards MEMS gyro drift using non￣stationary time series, this method is based on wavelet threshold de￣noising combined gradient radial basis(RBF)neural network. Principal random errors of MEMS gyros is analyzed using Allan variance,then the wavelet threshold de￣noise is used to separate white noise and drift error in error model of MEMS gyros,drift data finally modeled with RBF neural network. Experiments were carried to validate the error compensate method proposed in this paper, the experimental results show the effectiveness of the method, there is great significance for improving the precision of inertial navigation system based on MEMS gyroscope.关键词
MEMS/阈值去噪/RBF神经网络/误差补偿Key words
MEMS/threshold denoising/RBF neural network/error compensation分类
信息技术与安全科学引用本文复制引用
孙伟,段顺利,文剑,丁伟..阈值去噪与RBF神经网络在MEMS陀螺仪误差补偿中的应用∗[J].传感技术学报,2017,30(1):115-119,5.基金项目
国家自然科学基金项目(41304032)；高等学校博士学科点专项科研基金(新教师类)(20132121120005)；第8批中国博士后科学基金特别项目(2015T80265)；第58批中国博士后科学基金面上项目(2015M581360)；辽宁省高等学校杰出青年学者成长计划项目(LJQ2015044)；辽宁省自然科学基金项目(2015020078)；辽宁省“百千万人才工程”培养经费项目(辽百千万立项[2015]76号)；对地观测技术国家测绘地理信息局重点实验室开放基金项目(K201401)；地球空间环境与大地测量教育部重点实验室开放基金项目(14－01－05)；航空遥感技术国家测绘地理信息局重点实验室经费课题项目(2015B11)；精密工程与工业测量国家测绘地理信息局重点实验室开放基金项目(PF2015－13)；海岛(礁)测绘技术国家测绘地理信息局重点实验室项目(2014B05)；江西省数字国土重点实验室开放研究基金项目(DLLJ201501) （41304032）
