Design Simulation and Assessment of Morphological Features and Entropy Deconvolution Based Efficient Diagnosis of Bearing Faults

Abstract

Rotating machines are well known as important machinery that require precision and efficient output in power plants, manufacturing industries and the automotive industry. The rotating equipment of the coils is the most frequently used mechanical parts and is the key trigger for machine failures. These disturbances can result in expensive delays, manufacturing errors, and even humans. A responsive and robust monitoring system is needed to detect failures early on and warn of potential failures in order to reduce downtimes of the computer. Such a system can minimise maintenance costs, avoid catastrophic failures and increase the efficiency of machines. A thorough understanding of the nature of the bearing is essential to establish an efficient diagnostic and prognostic method. In the last year, vibratory rotating system control was primarily examined from a signal processing view. However, the consequences of failure with vibration actions were not taken into account. The first effective move in the implementation of the health monitoring assistance is therefore to create a safe bearing for the baseline 's behaviour. Further, although a series of rotating machines work under various conditions of speed and loading, very few researchers have suggested rigorous techniques for failure diagnosis and prognosis. This paper proposes a new algorithm for this SE selection on the basis of the kurtosis. Rotating machinery defects are found as vibration signal impulses, but most of them are submerged in noise. A new technique with morphological operators is proposed in this research to effectively eliminate this noise and to detect the impulses. The thesis work is split into two main sources. We tried in the first part to improve bearing fault diagnosis with MEDs integrating spectral kurtogram and autoregressive techniques. The second aspect involves morphological analysis as the same fault signal bears the identification of fault bearing frequencies.