Regional Frequency Dynamics Modeling of Power Systems Incorporating Frequency Support from HVDC Links and Renewable Energy Sources

Jianguo Zhou,Ye Guo,Zichao Meng,Lun Yang,Hongbin Sun
DOI: https://doi.org/10.1109/pesgm51994.2024.10689182
2024-01-01
Abstract:The increasing penetration of Renewable Energy Sources (RESs) and High Voltage Direct Current (HVDC) challenges the system frequency stability and the frequency shows distinct spatial-temporal characteristics. A Physical-Informed Neural Network (PINN)-based Regional Frequency Security Metrics (RFSM) is proposed for power systems with frequency support from HVDC and RESs. First, the RFSM models are modeled based on the physical principle. Then, neural networks are used to determine the parameters informed with the developed physical models. The proposed model can be utilized to establish frequency constraints implemented in the scheduling model. This type of constraints can be easily integrated into the optimization model due to its linearity. Moreover, the proposed model enables the system operators to find the optimal location and frequency response to be procured to guarantee the frequency stability in every region of the power system, and make pricing schemes for the corresponding ancillary services. The case study shows that the model has better physical interpretation and higher accuracy.
What problem does this paper attempt to address?