误差白化与Kalman滤波的微硅加速度传感器动态补偿

Dynamic Compensation of Micro-silicon Accelerometers Based on Error Whitening and Kalman Filtering

通过模型参考的系统辨识方法建立微硅加速度传感器的动态补偿器。由于测量噪声和补偿器对传感器的频带扩展，使得补偿器的输入／输出信号存在严重的噪声干扰。在噪声干扰下，采用均方误差为代价函数的系统辨识方法，无法得到补偿器参数的无偏估计。补偿器参数的偏差和传感器频带的扩展将会使补偿器的输出信号出现严重失真和高频噪声干扰。为解决噪声对硅加速度传感器的动态补偿的影响，研究了一种新的动态补偿方法，该方法采用误差白化为代价函数的系统辨识方法得到补偿器的参数，可消除补偿器的参数在估计中的测量噪声影响，并通过卡尔曼实时滤波减小因传感器频带扩展所产生的高频噪声干扰。

A dynamic compensator of micro-silicon accelerometer can be modeled with model reference and system identification. The compensator has severely noisy input/output by measurement noise and frequency response of compensator. Conventional techniques based on the mean squared-error criterion can at best provide a biased parameter estimate of the unknown system being modeled with noisy input/output. The output of compensator has severely distortion and noisy disturbance by biased parameter and widened frequency band. A new approach to solve the problem of compensator parameter estimation and noisy disturbance is proposed. With this approach, the parameter of the compensator is optimized by the error whitening criterion, according to the measurement data of the step response of the sensor and reference model. At the same time, Kalman filter is constructed with reference mode for eliminating the high frequency effects.