Three phase power quality disturbance detection and classification by SCGB Neural Network

An artificial neural network based three phase real time power quality analysis is proposed in this paper. The scaled conjugate gradient descent algorithm is used as a learning algorithm for neural network. The training and testing data required is generated from the experimental set up of three phase power quality disturbance generator. The exprimental setup is based on the changing voltage range which is controlled by solid state relays. The solid state relays are controlled by the microcontroller. The disturbances generated are sag, swell, interruption, harmonics with and without sag/swell. These disturbances are acquired with the help of NI USB 6361 data acquisition system. The initial training and testing is done in the MATLAB with the dataset collected from disturbance generator. The mathematical model of the trained network is developed and implemented in the LabVIEW for the real time testing. The real time results of the analysis are displayed in the LabVIEW graphic user interface console.