LAUSR

An original method for distributed secondary control (DSC) using averaged feedback reward pinning (AFRP) is proposed to eliminate static error caused by droop controllers in a micorgrid (MG). The method uses local and global rewards to evaluate the frequency and voltage control effects, and it further optimizes the control performance by maximizing the global feedback reward. The AFRP can adaptively adjust the pinning values step-by-step and always converges in the direction that maximizes global feedback. Only certain parts of agents are pinned to the averaged feedback reward, and then all other existing agents converge along with the pinned agents through the neighboring communication coupling. In this way, the DSC can obtain a near-optimal solution, which asymptotically eliminates frequency deviations and optimally regulates voltages in consideration of power losses in the MG. The efficiency and superiority of the AFRP-based DSC are verified by discussion and analysis of the simulation results.