中文摘要：

为了有效提取航空发动机转子碰摩故障信号,提高碰摩故障诊断的准确性,采用广义S变换方法,把碰摩信号变换到相空间中,在相空间检测和提取故障特征,并与连续小波变换的结果进行了比较.变换分析结果表明：在广义S变换域可以更清楚地反映出信号的时频局域化特性和奇异点位置,易于检测出碰摩点的位置,广义S变换有较好的抗干扰能力,在变换域易于信号与噪声的分离.

英文摘要：

In order to effectively pick out the impact-rub fault signal of aeroengine rotor, improve fault diagnosis veracity, the impact-rub signal was transformed to phase space by using generalized S-transform, the features of fault signal were detected in phase space, and the results obtained by using generalized S-transform（GST） and continuous wavelet transform（CWT） were compared. Comparison result shows that the time-frequency localization characteristics and singular points of impact-rub signal are more distinct in GST domain than in CWT domain, the impact-rub points are easier to be detected, GST has strong capability in noise reduction, and the separation of signal and noise is also available in GST domain.