分析水轮机发电机轴心轨迹是水轮机机组诊断振动故障的重要方法。由于轴心轨迹中包含了大量的噪声信号，将小波分析的理论应用于轴心轨迹的提纯，在传统小波硬、软阈值去噪方法的基础上提出了一种改进的软阈值去噪方法；并且引入谐波小波分析方法。通过计算机模拟试验，达到了比较理想的去噪及轨迹提纯结果，并对上述4种方法进行了分析和比较。

Analysis on the shaft orbit of hydraulic generator is important for diagnosing fault vibration of generator sets. To identify the shaft orbit among noisy signals, an improved soft threshold denoising method is proposed on the basis of traditional soft-threshold wavelet de-noising and hard-threshold wavelet de-noising. Computer simulation results show that this new method could achieve satisfactory results for de-noising and orbit identification. Apart from the above three methods, harmonic wavelet analysis is also introduced to identify shaft orbits. Comparison and analysis of the results reveal that the harmonic wavelet analysis is more efficient and convenient with the highest signal to noise ratio.