水电机组故障诊断专家系统中图形征兆的提取

The symptom extraction of axis orbit in fault diagnosis of hydropower generating units

将小波包分解及重构技术用于转子轴心轨迹提纯 ,利用离散余弦变换实现轴心轨迹的自动识别 ,结合水轮发电机组故障诊断系统 ,实现了轴心轨迹图形征兆的提取 ,论证了所述方法的合理性和可行性 .

The decomposition of the wavelet packets and its reconstruction to purify the rotor axis orbit was used and the automatic identification of rotor axis orbit based on the application of discrete cosine transform was carried out. Combined with the fault diagnosis of hydropower generating units, the graphic symptom extraction of axis orbit was done. This paper discussed the rationality of this method and its feasibility. By this method, the diagnosis speed and precision of the system were improved.