Research On Drift Instrumentation Modeling And Compensation For MEMS Gyroscope In MTi

Micro inertial measurement unit is composed of MEMS gyroscope and accelerometer. In size, weight, cost, etc., micro inertial measurement unit than traditional inertial measurement unit has obvious advantages. It has broad application prospects In attitude control, inertial navigation, military and civilian fields. However, high-precision MEMS gyroscope involving military applications, not only expensive, but the embargo on China, In the current market, most of the MEMS gyroscope is a low precision, and limits its application scope. Therefore,It has the important significance in the inertial navigation how to enhance inertial measurement unit's precision based on the existing hardware . An effective method is to improve software.Micro inertial measurement unit of the MEMS gyroscope in this article are the main research subjects. According to the measured data of the gyro drift, the law of drift and random drift were analyzed,then the mathematical model of MEMS gyroscope drift was delinked and compensated.Firstly, the working principle and hardware component of Micro Inertial Measurement Unit was introduced,then, the causes of systematic error and gyro error of the system measurement were analyzed.According to the characteristics of MEMS gyroscope and the error mathematical model, the system parameter were calibrated by use of three-axis turntable,such as zero, scale factor, the installation error.It identified the test methods of the rate gyro calibration and temperature characteristics and data processing method.MEMS gyroscope random drift more complex than systematic drift, and its modeling and compensation is also the focus of this paper.Based on the measured output data of gyroscope and the dynamic data system method of timeseries modeling theory, the model and of a MEMS gyroscope'drift were established. It revealed the components and features of the gyro random drift, and discussed the temperature on the effects of random drift model.Based on this mathematical model, the MEMS gyroscope random drift data was processed by use of Kalman filter algorithm, In practical application,the output of gyro often contains outliers.So an improved Kalman filter based on the identification of ARMA model is proposed to resolve the presence of outline.The results show that the method of software filtering compensation can enhance the MEMS gyroscope precision, and achieve good results.