Study on the precession orbit shape analysis-based linear fault qualitative identification method for rotating machinery

The vibration responses of different linear faults all possess some common features, which make fault diagnosis very difficult. Based on the multi-sensor information fusion theory, this paper presents a new qualitative identification method for the diagnosis of linear faults. The excitation-response dynamic equation is constructed and system balancing response with full consideration of system anisotropy is analyzed. Through discussion of the precession orbit shape difference and its dispersive situation, the orbit shape average difference coefficient and the corresponding dispersion term are estimated to obtain the theoretical balancing effect. Finally, the qualitative identification of linear fault can be done according to whether the calculated balancing effect meets the safe operation requirement or not. The dynamic characteristic of the system difference coefficients is verified by a simulation experiment and the case study further testifies the capability and reliability of the proposed method.