Fault Detection and Diagonics in Induction Motors using Convolutional Neural Networks

This research study proposes an Induction Motor Fault Detection Systems fueled by Convolutional Neural Networks (CNN) to development support techniques in mechanical situations. Induction automobiles are fundamental to an extensive range of business programs, making their green operation essential for maintaining productiveness. However, the cars are at risk of faults that, if no longer detected early, can result in high priced downtime and multiplied maintenance fees. To cope with this challenge, we recommend a complicated Induction Motor Fault Detection System powered with the aid of Convolutional Neural Networks (CNNs) to enhance predictive upkeep strategies in industrial environments. The proposed system leverages sound alerts generated by induction motors to perceive capacity faults. Initially, audio facts is captured below each every day and defective operating conditions using a microphone or sound sensor. This raw audio data is then converted into spectrogram images by visual representations of the frequency spectrum over time by using techniques such as the Short-Time Fourier Transform (STFT).