光学陀螺随机误差特性的混合理论方差方法分析

Hybrid Theoretical Variance Analysis for Random Error Properties of Optic Gyroscope

为更有效地测试和分析光学陀螺的随机误差特性,提出采用混合理论方差(Theo H)的方法全面表征和分析陀螺输出数据。该方法建立在Allan方差和去偏的#1理论方差(Theo1)的基础上,解决了Allan方差计算的平均时间只能够达到数据持续时间长度的一半,以及长相关时间下置信度下降的问题,同时构造的新方法能够自动去除Theo1方差相对于Allan方差的偏差。对比分析了新方法对白噪声特性以及光纤陀螺实测数据中随机噪声的类型和噪声水平的辨识结果。实验结果表明,通过去除偏差和提高方差估计的置信度,新混合理论方差方法对估计各种白噪声以及长相关时间下光纤陀螺噪声特性提取更有效,精度更高,同时能够兼容Allan方差对量化噪声、角度随机游走、零偏不稳定性、速度随机游走、速率斜皮5种噪声进行无偏估计。

For the sake of testing and analyzing the random error properties and identifying the noise amplitude of optic gyroscope more effectively, a new hybrid theoretical variance （TheoH） strategy is proposed to analyze the gyroscope output, which is integrated with the Allan variance and theoretical variance #1 with auto - bias removal. The problems of that the averaging time calculated with Allan variance is half of total data length and the confidence decreases at long-term averaging time are perfectly solved. A new method for analyzing the random error properties of optic gyroscope is proposed. The proposed method is compared with the Allan and theoretical variance#1, and the noise types and noise levels of white noises and real gyroscope output are identified. The experimental results show that the TheoH can improve the confidence efficiently and has higher evaluation precision than the conventional Allan method in long-term τ-values.