Galloping Early Warning Technology of Transmission Lines Based on RBF Neural Network

In view of strong nonlinear characteristics of transmission line galloping, firstly, the characteristics of historical dancing are analyzed, and the distribution areas and influencing factors of galloping are also revealed, which can be provided as the theoretical support for the establishment of galloping prediction model. Based on the characteristics of the galloping influence factors, several meteorological factors are selected for the galloping forecast, including surface temperature, rainfall, Upper air temperature, wind speed. Then the galloping forecast model based on Radial Basis Function neural network is built to predict whether galloping is happening. These essential meteorological factors are taken as the input parameters, while the output of the model is the galloping risk result. The galloping samples in the past five years of 2010-2014 has been taken as the training case, and the galloping samples in the November of 2015 in Liaoning Province has been taken as the testing case. The results show that it is feasible to use this artificial forecast model to predict the galloping risk of transmission line. The galloping prediction testing result using the proposed model shows that the forecast precision rate can reach as high as 91%, which can provide great technical support for the anti-galloping work for power grid. The research results can provide support for the transmission line galloping prediction and early warning technology, so as to improve the level of intelligent operation and maintenance of power grid.