Lightning Trip Warning Based on GA-BP Neural Network Technology

Abstract

Due to the complex operation environment of transmission lines and many factors affecting lightning trip, it is difficult to realize real-time warning of lightning trip of transmission lines. This paper presents a new real-time warning method for lightning trip based on the combination of lightning location system, transmission line data and GA-BP neural network technology. Firstly, the factors affecting line tripping are determined and input data are preprocessed. Then, a lightning trip prediction model based on GA-BP neural network is established. The total number of lightning trips can be predicted by summing up the results of all the intervals of the line to be warned. Finally, according to the classification standard of early warning, the lightning trip warning of the whole line is realized. The example shows that the correct classification rate is 80%, the accuracy rate is 86.67%, the false alarm rate and the leakage alarm rate are 23.53% and 13.33% respectively. The performance index of the model is ideal, so the model has a good prediction effect for the real-time lightning trip warning of transmission lines.