AUTHOR=Bao Song , Bao Hua , Jin Miao , Ruan Yong , Shi Yunfei , Yang Chao 

TITLE=Prediction of bundle-conductor ampacity based on transformer-LSTM

JOURNAL=Frontiers in Physics

VOLUME=Volume 13 - 2025

YEAR=2025

URL=https://www.frontiersin.org/journals/physics/articles/10.3389/fphy.2025.1603239

DOI=10.3389/fphy.2025.1603239

ISSN=2296-424X

ABSTRACT=The traditional method cannot meet the demand of new power system for dynamic regulation of transmission lines. In order to solve this defect, based on finite element simulation and neural network, an overhead bundle-conductor dynamic bundle-conductor ampacity prediction method is proposed in this paper. Considering the four bundle- JL/G1A-400/35 steel-core aluminum stranded wire, the three-dimensional electric-thermal-fluid coupling model of the conductor is established by using the synergistic optimization of transformer and long-short-term memory neural network (LSTM). The results show that the mean square error and average absolute error of the proposed model are 31.14 and 6.93, respectively. Compared with the bidirectional long and short-term memory network (BiLSTM), the mean square error and average absolute error are reduced by 74.55% and 7.35%, respectively. The maximum improvement of load capacity prediction margin is 10.04%. It can effectively tap the dynamic potential of transmission lines, and provide technical support for real-time scheduling of smart grid.