Abstract: The mismatch between PV output and power load may increase the risk of voltage overrun and reverse overload in the distribution network area. Based on this, this study developed a data-driven anti-backflow control technology for distribution network stations under high-permeability distributed photovoltaic access, realized the real-time prediction of power load in the distribution network based on the TiDE algorithm, and proposed four anti-backflow control strategies for distribution network stations based on historical load data and predicted load data. In addition, this study also defines the risk factor based on the kernel density function as a reliable indicator to evaluate the counterflow risk of the distribution network area. Finally, based on the simulation test of a rooftop photovoltaic power station project in Huairou, Beijing, the effectiveness of the algorithm is verified, which can reduce the risk of backflow in the distribution network area and increase the utilization rate of distributed photovoltaic, which has reference significance for the safe and stable operation of the distribution network under the access of distributed photovoltaic with high permeability.