Optimal Phase Angle Synchronization Control of Nonlinear Generator Systems Via Homotopy‐Based Policy Iteration

ABSTRACT This article addresses the optimal synchronization of phase angles in multi‐generator power systems by reformulating it as an optimal synchronization control problem for multi‐agent systems. First, an augmented system based on the states of the leader and followers and a distributed performance index are established. Then, an artificially stable augmented system with a homotopy constant is designed to eliminate the requirement of the initial admissible control policy in traditional policy iteration (PI). Based on the artificially stable augmented system, an online off‐policy algorithm is proposed to solve the optimal synchronization problem, enabling zero control initialization for PI. The artificial system is restored to its original state, and an admissible control policy has been automatically found in a finite number of iterations. The effectiveness of the algorithm is verified with a multiple‐generator system example.