The vibration monitoring system on the belt using neural network and interface with Labview and android apps

Abstract

Detection of belt conditions as a connector on the AC motor can be done by knowing how characteristics vibration signals. All vibration signals are processed by using Fast Fourier Transform (FFT) to the frequency domain. The vibration signal spectrums are collected and processed by the Neural Network method so that conditions of the belt can be grouped. The cause of the actual condition can be monitored on Labview and Android apps. The proposed method is based on analyzing frequency-domain features with the Neural Network. On this paper, it uses two hidden layers on the Neural Network. The Result of the experiment based on three grouped of the condition can get accurate by 94.03% to 96.52%.