小波域模极大值的激光信号滤波方法研究

针对用传统滤波方法滤除激光陀螺随机噪声性能低的缺点,提出了一种基于波域的模极大值滤波方法。该算法利用小波变换模极大值滤波方法对激光陀螺零漂数据进行处理,获取模极大值点,通过交替投影算法重构信号,并采用Allan方差法对波效果进行定量分析。通过实验验证了该方法滤波效果优于基于时间序列模型的卡尔曼滤波方法,能有效减小随机误差,提高测量精度。

Study on modular maximum denoising for signal of laser gyro in wavelet domain

According to the poor performance in denoising laser gyro random noise using traditional filtering methods,a modular maximum denoising method based on wavelet domain is presented,which utilizes the modular maximum denoising method based on wavelet transform to get modular maximum points by denoising the zero drift signal of laser gyro,then reconstructs signal using the alternating projection algorithm,and uses the Allan variance method to analyze the filtering result quantitatively.The experiment results show that the filtering algorithm can obtain the better filtering effect than Kalman filter based on the time series model,decrease the random error and improve the measurement precision effectively.