Identification of the vibration responses of a rub-impact rotor under uncertain excitations

A pseudo-periodic surrogate algorithm was introduced to analyze the vibration responses of a rub-impact rotor system under uncertain excitations, and the faulty signals were identified and classified. The dynamic model of the rub-impact rotor under the bounded noise excitation was established, where the nonlinear oil-film force and the rub-impact force were also taken into account. Then, the surrogate data of the responses of the system were generated via the pseudo-periodic surrogate algorithm, and the correlation dimensions of the original responses were compared with those of their corresponding artificial data. Meanwhile, the Poincaré cross-section portrait, the leading Lyapunov exponent and the bifurcation diagrams were employed to validate the predictions. The results show that, the pseudo-periodic surrogate algorithm can effectively identify the periodic-dominant and chaotic-dominant response signals, and can be further applied to the signal identification and fault diagnosis of rotor systems under uncertain excitations. © 2019, Editorial Office of Journal of Vibration and Shock. All right reserved.