Application of cascade correlation neural network in modelling of overcurrent relay characteristics

Modelling of Overcurrent (OC) relays with inverse time relay characteristics is a vital job for coordination of these relays. There are many publications in which the OC relay characteristics have been modelled. In this paper a new model based on cascade correlation neural network is proposed. The cascade correlation neural network is used to calculate operating times of OC relays for various Time Dial Settings (TDS) or Time Multiplier Settings (TMS). This method can cover nonlinearity of the characteristic and its accuracy is much higher than the polynomial and the other neural networks models such as perceptron and backpropagation neural networks models. The method is tested on three types of OC relays and the results obtained shows, the accuracy of the new method is higher and therefore it is more useful than the others. The model is validated by comparing the results obtained from the new method with nonlinear analytical, perceptron and backpropagation neural networks models.