MEMS陀螺仪随机漂移误差滤波处理研究.pdfVIP

29 2 Vol. 29 No.2 2010 4 JOURNAL OF SHENYANG LIGONG UN IVERSITY Apri. 2010 : 1003- 1251( 2010) 02- 0082- 04 MEM S 1 2 张德明, 任 波 (, 1. ; 2. , 110159) : MEMS陀螺仪是一类新型惯性器件, 具有体积小成本低重量轻可靠性高等 优点,但是精 较低随机漂移误差比较大. 从实际工程应用角 出发,针对MEMS陀 螺仪的随机漂移误差, 首先进行实时均值滤波处理, 然后基于随机序列时序分析法的基 本原理, 建立MEMS陀螺仪随机漂移误差的 一阶AR模型,然后依据 kalman滤波算法 的马尔科夫特性, 提出了对MEMS 陀螺仪每次输出进行实时多次滤波处理的新方法. 通过对具体测量数据进行处理, MEMS陀螺仪的随机漂移误差减小到原来的百分之二 左右. : MEMS陀螺仪; 随机漂移误差; AR模型; kalman滤波 : TP212. 1 : A Research onMEM S GyroscopeRandom DriftFiltering ZHANG em ing, REN Bo ( Shenyang LigongU niversity, Shenyang 110159, China) Abstract: MEMS gyroscope is a new type of inertial sensor w ith small size, low cost, light weight, high reliability and so on. However, it is lessprecisew ith random errors. A mi ing at application in practical engineering and dealingw ithMEMS gyroscope random errors, real tmi e average filtering is first processed, and then the firstorder AR model ofMEMS gyro scope random error is established on thebasis of principles of tmi e series analysis of random sequence. Finally, based anMarkor properties of kalman filtering algorithm, a new method is proposed that each of theMEMS gyroscope output ismultiple realtmi e filtered. Through the specificmeasurement dataprocessing, MEMS gyroscope random errors are reduced to a bout wt o percent of the originalone. K ey words: MEMS gyroscope; random error; AR model; kalman filte [ 1] , , . MEMS MEMS , , ., ,