Vibration Fault Mode Determination For Rotor Based On Correlation Dimension And Kolmogorov Entropy

Fractal and Chaos theory developed in recent years gives a new approach to the research on nonlinear dynamical system and has been used in rotor vibration fault diagnosis of aero-engine and other rotating machine. In this thesis, methods of correlation dimension and Kolmogorov entropy for rotor vibration fault diagnosis are studied in combination. Through the study of fractal and chaos theory, the algorithm for the calculation of correlation dimension and Kolmogorov entropy was introduced and was carried out using MATLAB codes. Then three typical rotor faults which are rotor unbalanced, rotor with loosen and rotor radical friction were simulated on a rotor test rig. According to the analysis of chaos identity and the correlation dimension and Kolmogorov entropy of rotor faults, a new method show that integrated correlation dimension with Kolmogorov entropy was proposed. The results show that the different faults have different correlation dimension and Kolmogorov entropy because of their different kinetics mechanisms. The correlation dimension and Kolmogorov entropy is the biggest for rotor radical friction fault due to it's higher value of Chao phenomena, the fault of rotor loosen is in the next place, the last is the rotor unbalanced fault. It proves that the method is more sensitive to radical friction fault and loosen fault than unbalanced fault. In addition to, the reseach indicate that thus correlation dimension and Kolmogorov entropy can be used to identify the faults symptoms and provide quantificational faults mode determination. The results of this thesis were valuable for faults diagnosis of rotors such as areo engine rotoes.