Traditional centralized optimal scheduling faces issues such as complex control, low reliability, and information privacy. Compared to centralized control, distributed control has advantages such as greater flexibility and reliability. Therefore, this paper proposes a distributed optimal scheduling strategy for distribution networks with high-proportion photovoltaic (PV) integration. Firstly, based on the modularity index and cluster self-consumption capacity index, the genetic algorithm is used to partition the distribution network clusters. Then, aiming at minimizing network loss costs and curtailment costs, a distributed optimal scheduling model for the distribution network is established and solved using the alternating direction method of multipliers (ADMM). Finally, using the IEEE 33-node distribution system as an example for simulation verification, the proposed strategy can effectively enhance the PV consumption capacity of the distribution network, reduce system network losses, and improve the power quality of the system.