Student: Kaung Sett Hein
Mentor: Bennett Meyers
Project Title: Image-Based Solar Forecasting System
Project Description: Photovoltaic (PV) solar power generation is growing exponentially. PV energy boasts at least a 10-fold reduction in carbon footprint compared to other energy sources, near-zero cost to operate, and an energy payback period of less than 2 years, making it an important component of the 21st century energy mix. Unlike traditional power plants, however, PV power depends on available irradiance, which can be highly variable. The ability to forecast future irradiance (and therefore PV power output), is crucial for planning and load-balancing operations in a modern grid. Direct assessment of cloud movement through the processing of 360° fisheye images is a very accurate way to generate high-frequency, short time horizon forecasts of irradiance. Currently, the GISMo lab does not have access to our own fisheye image capture system for experimenting with short-term solar forecasting. The goal of this project would be to install an end-to-end image capture system and implement methodologies from existing literature to process the images for forecasting. The Intern will be responsible for installing a fisheye camera at the GISMo lab, including mounting, data connection, and power supply. The data will then be piped into the GISMo team’s AWS cloud storage platform and implement image processing algorithms from established literature. The student will also implement some basic image processing algorithms on the new image data stream. Skills desired: (1) Prior experience with, or desire to learn about, data acquisition system installation and real-world sensing; (2) Proficiency in any of the modern programming languages, Python preferred ; (3) Prior experience in, or desire to learn, database management and AWS; (4) Prior experience in, or desire to learn, data analysis tools and libraries like Jupyter, Numpy, Pandas, and Matplotlib.