Study On Detection And Noise Reduction Technology Of MHD Angle Rata Sensor's Angle Vibration Signal

With the continuous development of space technology,the demand of spacecraft's accuracy become more. and more stringent. Micro-angle vibration exists in spacecraft which contains a lot of high-frequency components up to 1KHz due to disturbance, so we must make use of measurement and control platform to detect and compensate for micro-angle vibration to ensure the quality of observation. MHD angular rata sensor as one kind of broadband, high precision, low cost,miniaturization and long life inertial sensing element, making it the best choice for measuring satellite micro-angular vibration. However, MHD angular rata sensor is faced with many problems, like the extraction and detection of weak signal, random drift of sensor and so on. Therefore, it is necessary to study the detection and noise reduction technology of its angular vibration signal.Firstly, the structure and working principle of MHD angular rata sensor is introduced. The weak signal of the sensor is amplified by the amplifier circuit,two-stage amplifier circuit will amplify the sensor's output signal from hundreds of nanovolt or several millivolt to the volt level. And the noise of the primary amplifier circuit is analyzed, so the feasibility of the circuit is proved.Secondly, the noise source of MHD angular rata sensor header and its output signal preprocessing circuit is analyzed, and the static output of MHD angular velocity sensor is analyzed by Allan variance. The noise is mainly angular random walk and bias instability, that is white noise and 1 / f noise.Then the compensation method of random drift error of MHD angular rata sensor is studied. Based on the time series modeling of the static output data of the sensor, Kalman filter algorithm is used to compensate the random drift. The time series modeling object must be zero mean, smooth, normal time series, so before modeling we should remove the wild value, zero mean, remove the trend items and other operations, after testing to meet the smooth and normal requirements. The model is identified as ARMA (2,1), and the classical Kalman is used to filter the random drift data of the sensor. After the compensation is completed, its performance is evaluated by means of variance, spectrum and Allan variance analysis. The results show that the Kalman filter constructed on the basis of time series modeling can effectively reduce the random drift error of MHD angular rata sensor.Finally, an improved Sage-Husa adaptive Kalman algorithm is proposed. The simulation results show that the algorithm can effectively improve the signal-to-noise ratio of the sensor's data. In addition, I have designed the embedded DSP signal extraction and processing system, and the designed algorithm is applied to process the signal of MHD angular rata sensor.