CBR-based Load Estimation for Distribution Networks

Load estimation is very important for management and control of complex distribution networks. A novel method based on case-based-reasoning (CBR) is proposed for distribution network nodal load estimation. Principle of the method is analyzed, a hybrid learning algorithm is presented, and its application is discussed. The CBR-based load estimation method can build nodes and connections for a fuzzy neural network dynamically by a rapid and incremental learning procedure and can withstand the effect of bad data effectively through network self-organizing. The method is a key component of an integrated load and state estimation framework. The proposed method is tested on a 33-node system whose nodal load data come from a practical system, and test results show that it can provide high quality nodal load estimates