Bearing fault diagnosis using time encoded signal processing and recognition

Many novel condition monitoring techniques have been invented in recent yearsand the challenge lies in coming up with a highly reliable and cost efficient monitoring system which should be capable of tracking down and give an early indication of machinery's faults. The focus of this paper is to develop advanced approaches based on advanced intelligent computations to diagnosis and prognosis bearing condition. TESPAR (Time Encoded Signal Processing and Recognition) is an effective and direct way for describing complex waveforms in digital terms. It is the generic terms set to a collection of novel signal analysis, recognition and classification approaches that can be applied to describe and classify various ranges of complicated band limited signals. The results show that vibration signal waveforms of bearing faults can be digitized and analyzed in terms of its epochs' duration and shape which are the main parameters of the TESPAR technique that provides an accurate separation between different bearing faults with different degree of severity.