Improvement of safety operating conditions in overhead conductors based on ampacity modeling using artificial neural networks

Abstract

Thermal ratings are usually considered for planning the operating conditions for overhead lines and are usually obtained with static parameters. These conditions can be improved using dynamic ratings based on the region weather forecasts, and this improvement can be ever higher when a local prediction is performed at the point where the line is located. In this work, a model based on artificial neural networks techniques is applied to predict the ampacity property of a transmission overhead line, in order to adjust and optimize the operation point of the grid under safety conditions. These predictions are calculated for a time horizon of 24 hours and are validated with actual conditions of a real overhead line monitored by sensors. With the conclusion that applying the selected model, the operational security of the conductor can be improved, passing from a 17.82% of overheating conditions to only a 3.91%.