Research On Separation Of Twin-Spool Aero-Engine Vibration Signals

Vibration signal is the important parameters of condition monitoring and fault diagnosis of aero-engine. In the actual acquisition of mixed signals are from the different vibration signals aroused by the vibration excitation source, but also includes the measurement noise and other interference signals. The use of blind source separation of handling the vibration signals according to the different excitation source, which is the condition monitoring and fault diagnosis of the key steps to improve engine durability and reliability, as well as to ensure flight safety.A blind source separation algorithm was expounded in this paper. BSS is a promising technique for separation of twin-spool aero-engine vibration signals that allows the recovery of unknown sources from observed signals mixed by an unknown propagation medium. Using BSS in mixtures of vibration signal would enable the specific signature of each rotating system to be used for diagnosis, thus isolating them from the interferences of the environment. BSS can be used as a promising pre-processing tool for non-destructive monitoring of rotating machine. BSS was adopted to separate imitate signals, to judge whether the BSS outputs are consistent with the source signals, in order to validate the correctness and feasibility of arithmetic. In combination with FFT separation, which was adopted to separate imitate vibration signals come from rotor experimental platform in the lab, the blind source separation of the measured vibration signals that come from a high-pressure rotor and a low-pressure rotor of a twin-spool engine. It presents blind source separation techniques, which are proved by the separation to be feasible and effective signal processing methods for twin-spool engines' fault diagnosis and vibration monitoring.