DDPG-Based Backstepping Controller for DC Solid State Transformer in DC Microgrid With Constant Power Loads

Abstract

The DC solid-state transformer (DCSST) is a key component of the DC microgrid, and stability issues due to operation with constant power loads (CPLs) have become increasingly prominent. To stabilize the system of DCSST with CPLs, in this paper, a novel intelligent backstepping controller based on a nonlinear disturbance observer (NDO) is proposed, and the control coefficients of the backstepping control are automatically adjusted by an auxiliary deep deterministic policy gradient (DDPG) controller. More specifically, the NDO estimates the large-signal nonlinear disturbance of the load power variation, and a backstepping controller is designed to improve the stability of the system under the estimated large-signal disturbance. Then, an input voltage equalizing compensator is added to ensure the identical input voltage to each sub-module of DCSST. Furthermore, a DDPG sub-controller is introduced into the backstepping control to intelligently adjust the control coefficients to improve the operating efficiency of the system. The results of the simulation and experiment show that the system is more robust due to the learning ability of the proposed controller, and the proposed controller can improve the stability of the system under the large-signal disturbance and intelligently ensure fast dynamic response with accurate tracking over different operating environments.