Enhancing the transmission capacity of long-distance power transmission channels using neural networks

With the rapid development of new energy projects, the operational characteristics of long-distance, chain-like relay transmission channels have become increasingly complex, introducing significant uncertainty into the large-scale, long-distance transmission of new energy. The article first proposes an electrical partition-based method for the automatic discovery of critical restricted sections to improve adaptability to the online operational modes of the power grid. Subsequently, a neural network model is constructed to describe the relationship between channel factors and the amount of blocked transmission capacity in the transmission channel, and a dataset is established for training purposes. By analyzing and calculating key transmission sections, adjustment rules for key sections are proposed to optimize transmission efficiency. Finally, a case study using actual operational data from a power grid in 2024 verifies the effectiveness of the method in judging key factors and enhancing the capacity of power transmission corridors. Compared to traditional sensitivity algorithms, the method presented in this paper is more feasible and can provide data support for the data analysis tasks of dispatchers.