Open-source codes: PV Generation and Load Forecasting for Community Microgrids
A Gitlab repository contains the Python source of the machine-learning forecasting for load and generation, as well as the Alternating Direction Method of Multipliers (ADMM) distributed optimization techniques for managing networked microgrids. These examples, which use open-source libraries, demonstrate how computational techniques can be effectively applied to enhance the performance and efficiency of microgrid networks.
The links below allow open access to the software. Users will first need to create an external user account. Instructions for creating an account are provided below.
PV Generation and Load Forecasting for Adjuntas PR Community Microgrids
This code presents a framework for forecasting photovoltaic (PV) output power and consumer load in microgrid operations under conditions of disrupted data availability, employing lightweight recursive time-series models to independently forecast solar irradiance, ambient temperature, and load, ensuring reliable decision-making during extreme weather events.
Networked Microgrids Optimization
This code presents focuses on optimizing the operation of networked microgrids using centralized and distributed optimization approaches with considerations for both grid-connected and islanded modes, including power exchange limitations and slack bus coordination at the distribution substation.