Multimodal Analysis of Induction Motor Signals for Power Quality Abnormality Detection Using Wavelet-RBF Approach

Power Quality Disturbance (PQD) is the major challenges faced by the industry and power stations, etc. So the PQD detection and perfect classification of power quality abnormalities is a significant task. The PQD is increased due to single-phase half bridge inverters, full wave bridge rectifiers, unbalanced power system equipments, increased usage of electronics equipments, light controls, data processing equipments, nonlinear loads, and industrial converters. PQD such as rise and reduction in voltage, transients, THD, Notch, DC offset, short time Interrupt and flickering are the most repeated type of disturbances that arise in a power conditioning system. The precious apparatus connected to the power system gets affected due to abnormality arising for very long or short intervals. Hence, it is appropriate to detect the presence of disturbance, and classify the signal from the vibration of the motor. Here the motor vibrations are detected by using MEMS vibration sensor. The proposed setup output has been simulated using “MATLAB” toolbox for various conditions. Two hundred sampled data set is used for training the Radial Basic Function Neural Network (RBFNN) to detect the anomalous conditions and classification of PQD.