中文摘要：

精密磁悬浮陀螺全站仪定向系统的转子电流观测数据是随观测时间和环境的变化而变化的非线性数据序列，研究表明仪器定向系统中受强噪声干扰的非线性时序观测数据应用小波分解和重构可有效的剔除误差，分离出用于定向计算的有效的转子电流数据。该方法解决了传统的数据处理方法对仪器定向系统产生的非平稳、非线性观测数据序列滤波的局限性，有效地改善了仪器的定向结果，提高了定向角的精度。

英文摘要：

The observed rotor current data of maglev gyro total station directional system can be described as a kind of nonlinear data series relating to time and environment. Practical results show that, by using wavelet decomposition and reconstruction, this method can efficiently extract directional rotor current characteristics from the observed data series which were disturbed by strong noises. Moreover, comparing with traditional processing techniques, it has excellent features for nonstationary and nonlinear rotor series data filtering in directional system. It can effectively improve the instrument orientation results, and also improve the precision of directional angle.