On the influence of surrounding load demand to improve primary substation STLF

Short-term load forecasting (STLF) is a major column in every day´s life of power networks nowadays. Attached or bordering primary substation may share common features such as temperature, humidity, wind direction and force, etc. and, therefore, they may present similar changes in their consumption profile. Following this idea, we address here the hypothesis of whether data on surrounding primary substations may enrich and improve the forecast on a given primary substation. Therefore, we have replicated two well-known cutting-edge forecasting methods and have validated the hypothesis empirically applying said methods with and without surrounding substation´s data to 3 public load datasets following the leave-one-out cross-validation procedure. The results show that, indeed, using the data of the connecting ones helps ameliorating the forecast of a given primary substation.