Cooperative adaptive control of VSG rotational inertia and damping coefficient based on deep learning

The virtual synchronous machine (VSG) control is to embed the rotational inertia and damping coefficient of the synchronous generator into the inverter control, which enhances the system's anti-interference capability, but the fixed rotational inertia and damping coefficient lose the ability to adjust the transient process. To address this problem, this paper proposes a cooperative adaptive control strategy for VSG rotational inertia and damping coefficient based on the deep neural network of gated recirculation unit (GRU). Since the rotational inertia and damping coefficient are nonlinearly related to the angular frequency during the transient adjustment process, the ability of deep neural network to approximate the continuous nonlinear function is exploited, which enables these two parameters to be adjusted online in real time based on the information of system angular frequency to improve the system performance. Finally, a simulation model is built in Matlab/Simulink, and the effectiveness of the proposed strategy is verified by comparing with the fixed parameter control method.