Broadband Angular Velocity Signal Fusion Of MHD And MEMS Sensors Based On Kalman Filtering

With the rapid development of national aerospace technology,the requirements of pointing accuracy and stability performance of high-precision inertial line-of-sight(LOS)stabilization system are more and more stringent.Due to the limitation of angular velocity measurement bandwidth,the attitude measurement accuracy of the existing inertial stabilization reference unit cannot meet the requirements.The MHD angular velocity sensor has the advantages of wide bandwidth and low noise level,which can effectively obtain the micro angular vibration information in the bandwidth of 1-1000 Hz.In order to achieve the measurement of the angular velocity within the DC-1000 Hz throughout bandwidth,and considering the load applied to the inertial stabilization platform,the MHD angular velocity sensor and the MEMS gyroscope are used for combined measurement.The Kalman filter is suitable for high-precision realtime fusion in the field of inertial navigation.However,in the case of uncertain sensor models and various types of interference,the stability of the Kalman filter algorithm is lacking.Therefore,it is necessary to study and improve the fusion method based on the Kalman filter,and use the angular velocity signals of the MHD angular velocity sensor and the MEMS gyroscope to achieve full-band measurement.The main work of this paper is as follows:(1)The working principle and characteristics of the MHD angular velocity sensor and the MEMS gyroscope are analyzed,and the selection of the two sensors in this article is made clear.Several common sensor fusion methods of all-passband angular velocity are summarized,and their advantages and disadvantages are analyzed.(2)In order to improve the stability of the Kalman filter algorithm,the working process of fusion algorithm based on the classical Kalman filter for the MHD angular velocity sensor and the MEMS gyroscope is analyzed firstly.Then an adaptive Kalman filter algorithm based on frequency domain enhancement is proposed.This is an enhanced adaptive fusion algorithm combined with measurement covariance auxiliary signal frequency determination.The fusion principle is that the low frequency band outputs the signal using the MEMS gyroscope,the medium frequency band fuses the signals of two sensors,and the high frequency band outputs the signal using the MHD angular velocity sensor.(3)In order to verify the effectiveness of the enhanced Kalman filter fusion algorithm,the frequency domain models of the MHD angular velocity sensor and the MEMS gyroscope were established,and their noise characteristics are analyzed by Allan variance statistics.Then,the reliability of the algorithm is verified in the sweep frequency comparison experiment,and the multi-harmonic and step response experiment.The experimental results show that although affected by the frequency domain model error of the two sensors,the maximum amplitude fluctuation of the fused signals in the whole frequency band is only 0.0045 d B.And the fusion signal-tonoise ratio is improved.(4)After the verification of the proposed algorithm in the off-line experiment,DSP TMS320F28335 is used to realize the angular velocity signal fusion,and the realtime test is carried out on the vibration turntable.Affected by the frequency domain model of the sensor,the calculation error and the data processing ability of DSP TMS320F28335,the amplitude of the fused signal fluctuates greatly in the whole frequency band.The maximum amplitude fluctuation of the fusion signals is 0.195 d B,and the noise is too large.Therefore,it is necessary to further optimize the real-time implementation method.