Wide-frequency oscillation suppression strategy for grid-connected photovoltaic systems based on WOA-BP optimization of VSG parameters

The integration of a high proportion of renewable energy and power electronic devices in new power systems can trigger wide-frequency oscillation issues. Grid-forming devices, which have the capability to actively form the grid and provide inertia support, have become an important method to address wide-frequency oscillation problems caused by large-scale renewable energy integration. However, improper parameter design of grid-forming devices may introduce new coupled oscillations. To address these issues, a sequence impedance model of the virtual synchronous generator (VSG) is established, and wide-frequency impedance sweeping is used to identify the system’s oscillation characteristics. The impact of virtual inertia and damping coefficients on oscillations and system stability is analyzed. By combining the power angle characteristics of synchronous generators with frequency response curves, an adaptive collaborative control strategy based on frequency deviation and frequency rate of change is proposed. An innovative optimization mechanism combining Whale Optimization Algorithm (WOA) and Backpropagation algorithm (BP) neural networks is introduced to achieve real-time dynamic collaborative optimization of virtual inertia and damping coefficients, thereby suppressing wide-frequency oscillations and improving system stability. A grid-forming photovoltaic system was built on the Real Time Laboratory (RTLAB) platform for simulation verification. The results show that the anti-interference capability of the photovoltaic VSG increased from 20% to 35%. The proposed strategy effectively suppresses wide-frequency oscillations, significantly improves grid-connected power quality, reduces harmonic distortion, and enhances the system’s adaptability to grid impedance variations.