Machine learning applications for smart grids with solar PVs

• Main Author: Han, Weichou
• Other Authors: Hung Dinh Nguyen
• Format: Final Year Project
• Language: English
• Published: Nanyang Technological University 2024
• Subjects: Engineering; Machine learning
• Online Access: https://hdl.handle.net/10356/176932
• Institution: Nanyang Technological University

With the rising installation of Solar Photovoltaics (PVs) panel in Singapore, the impacts of renewable energy (RE) sources have been becoming clearer. While Solar PVs panels help reduce carbon emissions by using sunlight for electricity generation, they introduce a new challenge, i.e., how to mitigate the intermittency due to change in Solar irradiance. This project revolves around understanding the system operation under uncertainties. The project will focus on smart grid’s operation and behaviours with high Solar PVs’ power injections. In this project, I would use an artificial intelligence (AI) technique to forecast the Solar PV and design a Peer-to-Peer (P2P) energy trading market using Double Auction mechanism. The market mechanism will be able to find a mid-point price which is a guideline for all the sellers and consumers who participate to trade. AI used to predict the amount of electricity generated by Solar PVs to help with the supply and demand of electricity. By addressing the research gap, it has the potential to merge AI with the double auction market mechanism to allow for a better address of the uncertainties for Solar PV sellers.