Introduction: New Energy HVDC Transmission System (NEHTS) in China's western Desert-Gobi-Wasteland (DGW) regions relies on Ultra-High Voltage Direct Current (UHVDC) transmission for long-distance power delivery. The high uncertainty of power sources and unique load characteristics poses significant challenges to traditional Load Supply Capability (LSC) assessment methods. Method: On the power supply side, an Adaptive Diffusion Kernel Density Estimation (ADKDE) model and a Copula-based time series joint probability distribution model were constructed to improve the local fitting accuracy of output characteristics. On the load side, an improved adaptive step repeat power flow method was proposed to achieve an accurate simulation of the load growth pattern. Based on this, combined with probability power flow calculation, the Composite Security and Stability Index (CSSI) and Load Supply Margin (LSM) were used as the core evaluation indicators. Result: This study proposes a quantitative assessment method for evaluating the LSC of NEHTS, addressing the deficiency of conventional methods in accounting for the time-sequence correlation between wind and Photovoltaic (PV) generation and load growth patterns. Discussion: The fitting effectiveness and accuracy of the KDE and ADKDE models were quantitatively compared through simulation, confirming the superiority of the ADKDE approach. Furthermore, multiple cases were conducted to perform a detailed comparative simulation analysis between the proposed LSC assessment method and conventional methods. The results verify that the proposed method enhances LSM and also provides a time-sequence analysis. Conclusion: This method significantly improves the accuracy and practicality of LSC assessment, providing a theoretical basis for the safe operation and informed decision-making of NEHTS.
Gao et al. (Mon,) studied this question.