Fault Diagnosis of Rotating Machinery based on the Minutiae Algorithm

Abstract

Rotary machinery plays an important role in industry. Combined faults can be observed in rotating machinery, making fault classification difficult. In this paper, the Minutiae algorithm is used to classify the faults from the frequency domain of a particular fault. This paper provides a fault classification technique based on image processing for fault analysis of rotating machinery, recognizing function extraction automatically. Minutiae algorithm, a rising method within the discipline of image processing for characteristic extraction, is utilized in this paper to classify specific faults from the converted recurrence plot. The results reveal the effectiveness of the proposed method, providing a rather powerful tool for fault diagnosis of rotating machinery. The proposed model achieved an accuracy of 100% for combined faults, 98.33% for loosened faults, and 95% for unbalanced faults proving its applicability.