High-Temperature Weather Conditions-Based Risk Situation Awareness for Distribution Networks

High-temperature weather increases the failure rates of distribution lines and distribution transformers, thereby affecting the safe operation of urban distribution networks. To address this issue, a hybrid risk situation prediction framework combining model-driven and data-driven approaches is established based on analyses from both distribution lines and distribution transformers. First, from the model-driven perspective, the region is partitioned into grid units, and surface temperature is estimated based on ambient air temperature. The failure mechanisms of distribution equipment under high-temperature conditions are then analyzed. By integrating transformer winding temperature and line current-carrying capacity, a fault model for distribution network equipment is constructed. Second, the failure probabilities obtained by the Weibull proportional hazards model are introduced into the CNN–BiGRU–Attention network as prior risk features. Meanwhile, multiple environmental and geographic feature variables are collected from a data-driven perspective, which effectively integrates physical mechanism analysis and data-driven risk temporal prediction to complete risk situation forecasting. Finally, a case study is conducted by applying the IEEE 33-bus system to a specific region, verifying the prediction accuracy of the proposed method. The results demonstrate that the CNN–BiGRU–Attention model is capable of effectively performing risk situation prediction and can provide valuable support for operators in preventing potential fault risks.