The dynamic thermal rating (DTR) of an extra-high voltage (EHV) power system is an important safety factor during dispatching power flow. The maximum line ampacity of an EHV power grid is closely related to the line temperature which can be calculated based on the real-time weather data from meteorological observation stations through the IEEE Std. 738. However, the impacts brought by varying terrains on wind speeds are not considered, which easily lead to an inaccurate estimation of line temperature and DTR. Thus, this paper used the long-term historical wind speed data to uncover which kinds of terrains might cause the risk of inaccurately estimating line temperature based on the DTR model. Then, we proposed an artificial neural network-based terrain-type wind speed correction model so that the line temperature can be estimated accurately even using weather data from climate grid. A case study illustrates that the proposed model can effectively evaluate the wind speed of a specific valley where EHV power line was deployed and further improve the accuracy of estimating its line temperature. This fact suggests that the line temperature estimated by our method can serve as a reliable reference for the power dispatching strategy.
