A Novel Method for Online Nodal Load Estimation of Middle Voltage Distribution Networks

A novel method is proposed for solving online nodal load estimation of middle voltage distribution networks systematically. Firstly the solution strategies are introduced, and then the algorithm is presented. The proposed method consists of three phases that are crude load estimation, load forecasting and robust load estimation. Three load allocation methods are offered in crude load estimation phase which can provide crude load values. A case-based fuzzy-neural network is utilized in load forecasting phase for supplying nodal load forecasting values. Robust load estimation has been robustified synthetically in both structure space and measurement space, which can effectively withstand the influence of gross errors and many small errors. The three phases of load estimation cooperate closely and form a closed-cycle information flow. The proposed method can run online and can provide reliable and consistent load data set for control and management of distribution networks