With the increasing penetration of distributed generation (DG), its randomness and uncertainty pose significant challenges to distribution network reconfiguration. To address the optimal configuration of flexible smart soft open points (SOP) in distribution network reconfiguration, this paper proposes a two-stage robust optimization-based distribution network reconfiguration method incorporating SOP. A robust optimization model for the distribution network is developed, considering DG uncertainty and SOP. The Big-M method and second-order cone transformation are employed to achieve convex optimization reformulation of the robust model. The column and constraint generation (CCG) algorithm is introduced to implement the two-stage transformation and solution of the model. Additionally, three voltage indicators are selected to evaluate the voltage quality of each scheme. Simulation results demonstrate that dynamic reconfiguration incorporating SOP enhances the system’s economic efficiency and stability. In particular, the dynamic reconfiguration based on the optimal configuration of SOP further improves the economic operation level and voltage stability of the distribution network.