Flexible Neural Network-based Compensation Method for Saturation Characteristic of Current Transformer in Power System

GE Bao-ming,Aníbal T.de Almeida,Fernando J.T.E.Ferreira

DOI: https://doi.org/10.3321/j.issn:0258-8013.2006.16.026

2006-01-01

Abstract:The distorted secondary current of current transformer(CT)due to the saturation often results in many problems in applications of protections and measurements.A compensation method of CT saturation characteristic is proposed to improve its measure performance.The proposed algorithm uses a new flexible neural network with the activation function of two parameters changeable to realize estimation of magnetizing current.The accurate primary current is calculated by adding the estimated magnetizing current to the measured secondary current in real time.The flexible neurons of the established compensator will flexibly change the shape of each unit to adapt its role in learning process,which greatly simplifies the network with fewer neurons and reduces iterative learning epochs.A 900:5A CT is applied in the simulated operation,all kinds of cases such as the different direct current component of CT primary current,different CT remanent flux level,and load characteristic,are considered when testing the performance of the compensator.The simulation results verify the high precision of the proposed compensation method regardless of the waveform of CT primary current,remanent flux level,and load characteristic.”