I would like to conduct a load forecasting and generation forecasting for a power system at TSO-DSO bus bars interfaces, I am wondering what is the best Python library to do so?
Dear Mohammed, you may use statistical methods or learning-based methods for this purpose. It depends on your choice. There are various libraries in this matter such as TensorFlow, Keras, Scikit-learn. Also, you might need other modules in addition to these modules such as NumPy, Pandas, Seaborn, etc. for preprocessing and visualization of data. The type of variables that you are going to predict is also important regarding their features.
I'm agree with Thomas Wolgast. For forecasting long short-term memory (LSTM) model of load and generation by Deep Learning, I recommend to use the Tensorflow/Keras libraries; also, I think PyTorch and Pyqlearning libraries are appropriate for Reinforcement Learning.