CMX Student/Postdoc Seminar

Today's energy systems are increasing in unpredictability and in distribution shifts, in large part due to the increasing adoption of renewable energy. This talk presents two perspectives on the problem of learning and control of sustainable energy systems when faced with this problem of uncertainty in the systems. First, I will discuss the voltage control problem, which generally requires accurate information about the electricity grid's topology in order to guarantee network stability. However, accurate topology identification is challenging for existing methods, especially as the grid is subject to increasingly frequent reconfiguration due to the adoption of renewable energy. By combining a nested convex body chasing algorithm with a robust predictive controller, we are able to achieve provably finite-time convergence to safe voltage limits in the online setting where there is uncertainty in both the network topology as well as load and generation variations. Our approach can also incorporate existing partial knowledge of the network to improve voltage control performance. We demonstrate the effectiveness of our approach in a case study on a Southern California Edison 56-bus distribution system. Second, I examine the challenge of controlling sustainable energy systems from the perspective of reinforcement learning (RL). The lack of standardized benchmarks for RL for sustainable energy systems has made it difficult to both track progress on specific domains and identify bottlenecks for researchers to focus their effort. In this work, I present SustainGym, a suite of five environments designed to test the performance of RL algorithms on realistic sustainable energy system tasks, ranging from electric vehicle charging to carbon-aware data center job scheduling. The environments test RL algorithms under realistic distribution shifts as well as in multi-agent settings. We show that standard off-the-shelf RL algorithms leave significant room for improving performance and highlight the theoretical and practical challenges ahead for introducing RL to real-world sustainability tasks.