Abstract: With the grid connection of large-scale distributed photovoltaics, the reliable operation of the distribution network is facing severe challenges. It is urgent to build a stronger distribution network to improve the reliability of the power supply. In view of this, this study proposes a stochastic programming model considering reliability constraints for distribution networks It accounts for load shedding and power transfer via the tie line after the fault, and explicitly models five types of reliability indicators as constraints suitable for planning problems, that is, user interruption frequency, user interruption duration, system average interruption time, and system average interruption frequency index. Additionally, considering the uncertainty caused by the distributed photovoltaics generation, this model adopts a scenario-based stochastic programming method to enhance the robustness of the planning results. The results indicate that the proposed model can significantly reduce the planning cost while meeting the reliability requirements of the distribution network system.