Load Modeling and prediction

This project is to develop approaches to monitor and model load in power systems, utilizing data mining and machine learning techniques. The load modeling part aims to develop helpful model for simulations for transmission networks. The load forecasting part aims to predict short time load (4-hour ahead) at a submission or transformer level. Now we have tried time series approach and machine learning approaches, such as ANN, RNN, and CNN. Preliminary results reaches 95% accuracy.