AC optimal power flow (AC OPF) is a challenging task in power system operations due to the nonlinearity and complexity of power systems. In this paper, a new strategy for solving the AC OPF problem is proposed using decomposition and reinforcement learning (RL)-based cutting plane methods. The proposed strategy simplifies the original AC OPF problem by decomposing it into a DC OPF sub-problem and AC power flow (AC PF) calculation sub-problem, while also incorporating linear inequality constraints, or cuts, to limit the feasible region of the DC OPF sub-problem. Subsequently, the power generation profiles obtained from the DC OPF sub-problem are employed for solving AC PF calculation sub-problem. An RL agent is leveraged to estimate the optimal cuts and references for the voltage magnitudes at generator buses. An action selection method is also proposed to reduce the complexity of the agent training and the magnitudes of the variations in the power generations and bus voltages during a time period. Comprehensive case studies are conducted to compare the proposed strategy with the conventional strategies under various test conditions. The results demonstrate that the proposed strategy outperforms the conventional strategies in terms of feasibility and computational efficiency.

This paper proposes a new strategy for optimal grid frequency regulation (FR) in an interconnected power system where regional ac grids and an offshore wind farm are linked via a multi-terminal high voltage direct-current (MTDC) network. In the proposed strategy, decentralized H∞ controllers are developed to coordinate the operations of ac synchronous generators and hybrid MTDC converters, thus achieving optimal power sharing of interconnected ac grids and minimizing frequency deviations in each grid. To develop the controllers, robust optimization problems are formulated and solved using a dynamic model of the hybrid MTDC-linked grids with model parameter uncertainty and decentralized control inputs and outputs. The model orders of the resulting controllers are then reduced using a balanced truncation algorithm to eliminate unobservable and uncontrollable state variables while preserving their dominant response characteristics. Sensitivity and eigenvalue analyses are conducted focusing on the effects of grid measurements, parameter uncertainty levels, and communication time delays. Comparative case studies are also carried out to verify that the proposed strategy improves the effectiveness, stability, and robustness of real-time FR in MTDC-linked grids under various conditions characterized mainly by load demands, communications systems, and weighting functions.