A Neural Network Based Electrical Loss Prediction of Bare Overhead ACSR Conductors

Due to the current resource crisis, reliable electrical loss analysis and prediction attract more attention than ever for efficient power system management. This paper successfully predicts the electrical loss of ACSR with actual daily load variations using an accurate and effective neural network model, which is validated by visual comparison and strict statistical criterion. Since electrical loss for ACSR is closely related with ac resistance, case study is performed on a typical ACSR to investigate and simulate its ac resistance properties