Feature extraction method based on NOFRFs and its application in faulty rotor system with slight misalignment

• Published in: Nonlinear dynamics Vol. 99; no. 2; pp. 1763 - 1777
• Main Authors: Liu, Yang, Zhao, Yulai, Li, Jintao, Lu, Huanhuan, Ma, Hui
• Format: Journal Article
• Language: English
• Published: Dordrecht Springer Netherlands 01.01.2020; Springer Nature B.V
• Subjects: Automotive Engineering; Classical Mechanics; Computer simulation; Control; Dynamical Systems; Engineering; Fault detection; Fault diagnosis; Feature extraction; Frequency response functions; Higher harmonics; Mechanical Engineering; Military helicopters; Misalignment; Nonlinear response; Review; Structural damage; Vibration; Feature extraction; Weak feature enhancement; Variable weighted; NOFRFs; Slight misalignment
• ISSN: 0924-090X; 1573-269X
• DOI: 10.1007/s11071-019-05340-8

Fault features extraction method for slight misalignment of rotor systems is researched in this paper. When a rotor system with faults is excited by harmonic inputs, system response will contain higher harmonic components. Phenomenon of these higher harmonics implies the presence of nonlinear features in a rotor system. When extracting these nonlinear features, traditional methods have certain limitations, and it is easy to ignore weak features of a rotor system. Nonlinear output frequency response functions (NOFRFs) can extract effectively nonlinear features of a rotor system from noise-containing vibration signals and are used in many fault diagnosis fields. However, as for weak damage to structures and in the early stages of a system faults, fault features reflected by high-order NOFRFs are not still obvious enough. Therefore, a new method, named variable weighted contribution rate of NOFRFs, is proposed in this paper. This method enhances the ratio of high-order output frequency response to total output response of a rotor system. Based on this method, a new index MR is proposed to detect the faults of a rotor system. In addition, lumped mass model is used in this paper to simulate misaligned rotor system with slight misalignment, and the sensitivities of the traditional methods and new index for slight misalignment fault features extraction are compared. The results indicate that the new index is more sensitive for the slight misalignment rotor system. Additionally, the rotor system with misalignment fault experiment table is built, and the effectiveness of this new index for detecting the slight misalignment fault of the rotor system is verified.