Combination Systerm And Filter Algorithm Design For MEMS Gyroscopes

Recently small satellite is becoming a hot spot of the current international space technology research, because of its series of advantages, such as flexible applications, low cost, high degree integrated functions, short development cycle etc. MEMS gyroscope has a series of advantages, such as small volume, low energy consumption, light weight, small inertia, large bandwidth and ease of integration. Therefore it is expected to be widely used in orbit and attitude control system of for micro satellite.This topic is derived from a research institute of China. It aimed at designing an early form of orbit and attitude control system for micro satellite.In order to realize collecting, analysis, filter and storage MEMS gyro’s output data, a set of MEMS gyroscope combination system is builded, which includes data acquisition system and PC receiving interface. The data acquisition system mainly realizes to acquisite and filter MEMS gyroscope’s output data. The function of PC is to realize receiving, displaying, decoding and storing the data. The former using highly integrated So C, Smart Fusion2 series processor to implement acquiring and filtering high-precision MEMS gyroscope CRG20’s output data. The latter mainly realizes receiving, decoding, display and storage function.In order to facilitate the system’s debugging, transplantation, application and reduce the area of the circuit board, modular design method is adopted. The backplane provides power control, RS-422, CAN, SPI communication interface etc. Other function modules includes core board module, micro gyroscope module, ADC module and wireless module, each functional module can be pluged on the backplane. Finally the system achieved acquisition, filtering, real-time display and storage functions.Kalman filter algorithm is designed based on Autoregressive(AR).model, to deal with MEMS gyro’s application bottleneck problems such as low precision, big random noise. The noise of CRG20’s output signal is analysed by Allan variance, based on the result, AR mathematical model of CRG20 random drift error is established. To acquire the true value of gyroscope’s angular velocity, the state of true value is added in the filter’s state vector, then the state equation of Kalman filter is obtained; CRG20’s output data as the measuring signal, then the measurement equation of Kalman filter is established. The variance of system’s process noise and measurement noise Q, R respectively select gyroscope’s coefficient of RRW and ARW. After that, we simulate the filter by Matlab and analyze the effect of Q, R, AR modle parameters for the filter.Finally, we download the c language program of filter to SF2, to realize the real-time filtering.The simulation results show that,the Parameter of Kalman filter is selected properly,and the filter could effectively suppress high frequency noise and improve the smoothness of the CRG20’s output signal. The system could track to the signal of satellite attitude, if the speediness request is not too high.