Day-ahead dynamic thermal line rating forecasting and power transmission capacity calculation based on ForecastNet

With the increasing demand for cross-regional electricity trading between interconnected power grids, accurate evaluation of transmission capacity between regional power grids becomes important. Dynamic thermal rating (DTR) can be utilized to accurately calculate and evaluate the transmission capacity between the interconnected power grids. The affecting factors of DTR are complex. These factors are time-variant and some of them are affected by seasonal and diurnal cycles. The effect degrees of different factors on DTR is also disparate. Therefore, it is beneficial to forecast DTR accurately by using these factors. In this paper, a day-ahead DTR forecasting model based on ForecastNet is proposed. The model uses the numerical weather predictions to forecast DTR. The time-variant weight coefficient is utilized to modify the weight sharing structure of feed-forward neural network, which can dynamically tracks the time-variant characteristics of meteorologic factors and their effect degrees. The real data in Northeast China are utilized to verify the performance of the DTR forecasting model. Further, the day-ahead delivery capacity between two interconnected power grids in China is calculated. The results show that the DTR obtained by the proposed method has high accuracy and can realize accurate evaluation of transmission capacity.