Short-Term Load Forecasting of a Distribution Transformer using Self-Organizing Fuzzy Neural Networks

The distribution transformer Load forecasting is very essential in the control of future smart grids and an economical interfacing of Distributed Resources (DRs) to distribution networks. A distribution transformer connects DRs to the main grid. Exact distribution transformer load forecasting makes an economical DRs scheduling possible. In this regard, this paper firstly introduces a new Self-Organizing Fuzzy Neural Network (SOFNN). Then, it applies SOFNN to perform a five-minute load forecasting for a real-life distribution transformer in Lorestan Electric Power Distribution Company (LEPDC). Simulation results for active and reactive powers show that the proposed SOFNN outperforms ANFIS.