A Bottom-up Method for Probabilistic Short-Term Load Forecasting Based on Medium Voltage Load Patterns

Load forecasting has always been an essential part of power system planning and operation. In recent decades, the competition market requirements renewable integration lead more attention to probabilistic load methods, which can reflect uncertainties through prediction intervals hence benefit decision-making activities in Moreover, with development smart grid metering techniques, companies have collected enormous data about electricity customers substations. The abundant allow us utilize medium voltage measurement achieve better accuracy high transmission substation forecasting. this paper, a bottom-up method is proposed for short-term forecasting, probability distributions day-ahead values are estimated added up form predictions. Two frameworks based on patterns from outgoing lines substations respectively, mismatches between at different levels correcting comparison predictions obtained by traditional methods demonstrates that obtains accurately give narrower intervals.