Online transient stability assessment for AC/DC power systems based on clustering BCU and hybrid neural networks

The widespread integration of power electronic devices including renewable energy sources and Line Commutated converter (LCC)-based High Voltage Direct Current (HVDC) poses significant challenges to the transient stability of AC/DC power grids. To realize the accurate and fast online transient stability assessment of AC/DC power systems, this paper firstly constructs terminal potential energy expressions through the integration of active power with respect to voltage phase angle and reactive power with respect to voltage magnitude. Subsequently, this paper proposes an online transient stability assessment method based on clustering Boundary of stability based Controlling Unstable-equilibrium-point (BCU) and hybrid neural networks to address the computational inefficiency issue of pure mechanism analytical approaches in critical energy determination. The clustering BCU method significantly reduces computational complexity when generating offline critical energy datasets by eliminating redundant calculations for different fault scenarios within the same coherent fault cluster. Besides, the convolutional neural networks-gate recurrent unit (CNN-GRU) based hybrid neural network is presented for real-time online transient stability assessment, which can quickly and accurately require the critical energy during system faults and compare it with the transient energy at fault clearing time to determine system stability. Compared with the multiple parallel CNN-GRU model, the proposed approach further provides physically interpretable indicators, including critical energy, critical clearing time, and stability margin, enabling more reliable evaluation of system stability boundaries. Finally, simulation results under different renewable penetration levels, transmission conditions, and fault scenarios demonstrate that the proposed method achieves high assessment accuracy, strong robustness, and improved computational efficiency for online transient stability assessment of renewable-integrated AC/DC hybrid power systems.