Data-driven Analysis Method for Influencing Factors of Voltage Over-limit

In recent years, with the large-scale integration of distributed photovoltaics into power systems, the timing matching between power sources and loads in transformer areas has been significantly affected, resulting in a series of anomalies such as reverse transmission of active power and bidirectional power flow. Especially in the scenario of high penetration of photovoltaic power in transformer areas, the problem of voltage overlimit has intensified, posing a potential threat to the safe and stable operation of the power grid. This paper proposes a data-driven method for analyzing the influence of voltage overlimit. Based on the load operation data of a certain transformer area in Shanghai and the power flow calculation, the corresponding node voltages under different influence factors are obtained. Based on the XGboost evaluation method, the contribution of each influence factor is quantified to determine the main causes of voltage overlimit in low-voltage transformer areas, providing support for transformer area governance. This improves the safety of power supply.