Optimizing Fault Detection in Electric Power Load Management Systems: A Reinforcement Learning Approach for Prioritized Test Case Selection

This paper introduces a novel reinforcement learning-based method for optimizing test case selection in electric power load management systems. Addressing challenges in user-end and control loop efficiency, the study employs an Actor-Critic algorithm to prioritize test cases, enhancing detection accuracy and system stability. Through a simulated environment, test cases are represented as vectors in a multidimensional space. The model, trained over 1000 episodes, demonstrates a significant improvement in selecting optimal detection paths, as evidenced by increased total rewards and decreased loss. This approach offers a promising solution for managing complex power systems, illustrating the potential of reinforcement learning in real-world applications.