High Wind Power Penetration System Power Flow Off-Limit Early Warning Based on GCN and Rule Guidance

With the large-scale grid connection of new energy sources, the possibility of local power flow off-limit evolving into a global security risk has significantly increased, and the risk of power flow off-limit faults in the new transmission network is also increasing. However, traditional physical models for calculating power flow off-limit face issues such as slow calculation speed, high complexity, and redundant constraints. Therefore, this paper proposes a risk warning method for power flow off-limit in power systems based on the GCN algorithm and the physical rule guidance. Firstly, the influence of new energy sources, mainly wind power, on the risk of power flow off-limit in power systems under different operating conditions is fully considered, and the wind power scenarios data is generated. Secondly, a large number of power flow off-limit fault chains are generated using an improved OPA model based on DC current flow. Then, a risk warning model for power flow off-limit based on the GCN algorithm and the physical rule guidance is constructed, realizing the advanced warning of fault chains. Finally, simulation experiments are conducted on the IEEE 39 system, and the results show that the proposed method has fast calculation speed, high accuracy, and reasonable results. It helps improve the safety and reliability of power system operations.