As distributed energy increases, the reserve capacity in distribution networks decreases, and the complexity of distribution network structures grows, the reliability faces significant challenges. To address this issue, a reliability analysis method based on multi-source IoT dynamic data is proposed. By analyzing the probabilistic models of power supply and load consumption in distribution networks, incorporating various stochastic data such as distributed renewable energy, distribution network equipment, structure, and external environmental factors that impact network operation, multi-scenario generation and reduction algorithms for power supply and load consumption are studied. A power supply-consumption scenario matching model for distribution networks is also proposed. These scenarios provide dynamic reliability indicators. Finally, the simulation results indicate that dynamic reliability assessment based on multi-source IoT data can effectively reflect changes in actual dynamic reliability levels under different influencing factors.