Kishan Prudhvi Guddanti

Kishan Prudhvi Guddanti, Yang Weng, Baosen Zhang

The power flow (PF) problem is a fundamental problem in power system engineering. Many popular solvers face challenges, such as convergence issues. One can try to rewrite the PF problem into a fixed point equation, which can be solved exponentially fast. But, existing methods have their own restrictions, such as the required AC network structure or bus types. To remove these restrictions, we employ the circle geometry per-bus via rectangular coordinate representation to embed our physical knowledge of operation point selection in PV curves. Each iteration of the algorithm consists of finding intersections of circles, which can be computed efficiently with high numerical accuracy. Such analysis also helps in visualizing PV curve to always select the high voltage solution. We compare the performance of our fixed point algorithm with existing state-of-the-art methods, showing that the proposed method can correctly find the solutions when other methods cannot. In addition, we empirically show that the fixed point algorithm is much more robust to bad initialization points than the existing methods.

Amarsagar Reddy Ramapuram Matavalam, Kishan Prudhvi Guddanti, Yang Weng et al.

This paper describes how domain knowledge of power system operators can be integrated into reinforcement learning (RL) frameworks to effectively learn agents that control the grid's topology to prevent thermal cascading. Typical RL-based topology controllers fail to perform well due to the large search/optimization space. Here, we propose an actor-critic-based agent to address the problem's combinatorial nature and train the agent using the RL environment developed by RTE, the French TSO. To address the challenge of the large optimization space, a curriculum-based approach with reward tuning is incorporated into the training procedure by modifying the environment using network physics for enhanced agent learning. Further, a parallel training approach on multiple scenarios is employed to avoid biasing the agent to a few scenarios and make it robust to the natural variability in grid operations. Without these modifications to the training procedure, the RL agent failed for most test scenarios, illustrating the importance of properly integrating domain knowledge of physical systems for real-world RL learning. The agent was tested by RTE for the 2019 learning to run the power network challenge and was awarded the 2nd place in accuracy and 1st place in speed. The developed code is open-sourced for public use.