Currently there are more than 50,000 operating wind turbines across the entire U.S. Their performances varies based on the design, weather conditions, and geographic locations. Studying the capacity for each wind turbine individually is cumbersome and unable to scale with the rapid growth in wind energy deployment. We propose a solution that group the wind turbines with close geographical locations into wind farms of similar performances. This enables us to study the high-level zone behavior by treating all wind turbines within a zone to be identical. It not only reduces the total number of data (around 7TB) we need to analysis but also takes the building of future wind turbines into consideration. During this presentation, we will introduce the current clustering techniques and discuss the method used to cluster the wind turbines, including what parameters are taken into consideration, what makes a “good” cluster, and how to batch cluster this 7TB data across the U.S. Finally, the talk will present some preliminary results of clustered zones with various metrics to evaluate these clusters. This research has widespread applicability and is particularly relevant in grouping other renewable energy systems.
Zhaohong(Roy) Jin is a fourth-year undergraduate student at UC Berkeley double majoring in Computer Science and Economics. He Joined the Professor Daniel Kammen’s Renewable & Appropriate Energy Laboratory in Spring 2017 and worked with Professor Deborah Sunter on the SWITCH USA project. His focus is on the real-world application of machine learning technologies. Renewable energy is an ideal field for him to study because of the huge amount of data generated everyday. Roy’s goal to help introduce the cutting edge machine learning researches to the clean energy field by quickly prototyping applications and build reliable computational pipelines. He is also interested in various deep learning techniques in computer vision and natural language preprocessing. Roy has previously worked as a software engineer in Moody’s Analytics and Amazon. He will be attending graduate school after his Bachelor degree.