Fast-Converging Deep Reinforcement Learning for Optimal Dispatch of Large-Scale Power Systems Under Transient Security Constraints | SGEPRI Journals & Magazine | IEEE Xplore