Research on Rub-Impact Fault Quantification of Rotor System Based on Effective Singular Value Noise Reduction and Minimum Mutual Entropy Principle

The improved effective singular value noise reduction algorithm is able to reduce the noise of the signal while retaining as many effective frequencies of the signal as possible. Meanwhile, the difference between two signals is calculated by the minimum mutual entropy principle. Therefore, the improved effective singular value noise reduction algorithm and the minimum mutual entropy principle are combined to quantify the rotor system rub-impact faults. Firstly, the obtained signals are performed noise reduction using the improved algorithm. Secondly, the difference between signals in the healthy state and the signal in the fault state of the rotor system is calculated by the minimum mutual entropy principle. Finally, the faults of the rotor system are quantified based on the difference between signals.