摘要
针对微机械(MEMS)陀螺仪的随机误差导致其输出信噪比低的问题,该文提出一种基于ARMA模型的随机误差卡尔曼补偿方法。该方法通过对MEMS陀螺仪输出的原始信号进行均值化处理和趋势项提取后,分别采用轮次检验法和Jarque Bera检验法对随机序列平稳性和正态性进行检验,建立时间序列ARMA模型并采用卡尔曼状态方程对陀螺随机漂移进行补偿。6次卡尔曼滤波对比实验结果表明,MEMS陀螺仪信号幅度随着滤波次数增加而减小,滤波后的均值由2.454 4E-4减小到-4.830 0E-5,标准差由1.654 7减小到0.003 6,随机漂移被有效抑制。
Aiming at the problem that the random error of MEMS gyroscope can cause low noise-signal ratio,a random error Kalman compensation method based on ARMA model was proposed.Through the mean processing of the output raw signal of MEMS gyroscope and the trend item extraction,the round test and Jarque Bera test were adopted respectively to test the sequence stability and normality of the random sequence,then the time series ARMA model was established and the Kalman state equation was used to compensate the gyro random drift.The results of the six Kalman filter contrast experiments showed that the signal amplitude of the MEMS gyroscope decreased with the increase of the number of filtering times,the mean value of the filter was reduced from -2.454 4 E-4 to-4.830 0 E-5,the variance was reduced from 1.654 7 to 0.003 6 accordingly,and the random drift was effectively suppressed.
出处
《测绘科学》
CSCD
北大核心
2017年第12期52-56,82,共6页
Science of Surveying and Mapping
基金
国家自然科学基金项目(41304032)
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"培养经费资助项目(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资助项目(K201401)
国家级大学生创新训练项目(201710147000353
201710147000051)
