Dynamic voltage stability enhancement in wind integrated power systems using a deep reinforcement learning based FACTS controller

Doubly-fed induction generators (DFIGs), commonly used in a wind power plant (WPP), effectively convert wind energy into electrical power. However, a DFIG-based WPP (DWPP) has limited flexibility in generating reactive power ( Q ) due to interface converters that cause delays in fault recovery voltage. Insufficient Q support leads to dynamic voltage stability (DVS) and reliability problems. To address this, FACTS devices such as SVC and STATCOM with proportional–integral (PI) controllers are widely employed to regulate reactive power and enhance voltage stability. Conventional PI-based control often exhibits limited performance under severe grid disturbances. To this end, a deep reinforcement learning (DRL) based control is proposed to enhance the Q management capability of an SVC and a STATCOM, both in steady-state and transient conditions. The performance analysis of the suggested controller is conducted in the MATLAB/Simulink environment, considering different types of disturbances using a modified IEEE nine-bus network. The simulation results show that the suggested controller improves the Q management capability and DVS of the test network. Moreover, a comparative analysis of the DRL controller with PI and fuzzy logic-based controllers shows the superior performance of the proposed controller when used with either an SVC or a STATCOM.