The recent advancements in the field of power system optimization and control are significantly shifting towards leveraging data-driven and machine learning techniques to address the complexities introduced by the integration of renewable energy sources and power electronics. Deep reinforcement learning (DRL) is emerging as a powerful tool for optimizing the operation of energy storage systems and inverter control, offering adaptive and real-time solutions that can handle the dynamic nature of modern power grids. Safe imitation reinforcement learning frameworks are being developed to ensure operational safety and constraint compliance, which is crucial for the reliable dispatch of energy storage systems in distribution networks. Additionally, adaptive control strategies are being proposed to stabilize grid-connected power converters using online data, enhancing the robustness and adaptability of power systems. Distributionally robust control strategies are also gaining traction, providing a framework for ensuring frequency safety under uncertain prediction errors, which is vital for maintaining system stability in the face of increasing renewable energy penetration. These developments collectively aim to enhance the efficiency, safety, and reliability of power systems in the face of evolving challenges.