A Vibration Signal Filtering Method Based on KL Divergence Genetic Algorithm – with Application to Low Speed Bearing Fault Diagnosis

Fault signal plays a crucial role in bearing fault detection and diagnosis. However, in some low speed condition, the fault signal is weak and overwhelmed by the strong noise. It is hard to extract fault signal information effectively. To handle this problem, a vibration signal filtering method based on Kullback-Liebler (KL) divergence genetic algorithm (GA) is presented. This method employs binary string of GA to match the frequency of fault signal frequency spectrum. Based on the KL divergence as GA fitness function, the signal similarity between normal condition and the KL divergence GA filtered signal is evaluated and GA employs genetic operations to obtain the optimal noise binary string. After that, negating this optimal noise binary string can obtain optimal fault signal binary string, which will then be multiplied by diagnosis vibration signal frequency spectrum, to yield a new frequency spectrum that only has fault components. Based on envelope demodulation method and bearing fault characteristic frequency, the effective and advanced performance of the proposed method is evaluated by three different rolling bearing faults in low speed bearing vibration signal.