부산대학교 도서관

Transmission lines are exposed to the atmospheric environment for a long time and are susceptible to failures due to lightning, typhoons, heavy rains and other catastrophic weather. Wind disasters are the more hazardous meteorological disasters in Henan Province, and the spatial and temporal distribution of their disasters is of great significance to wind disaster defense. The statistical data of high wind disaster failure in each substation in Henan Province in the past four years are used. The probability of transmission line failure under strong convective windy weather conditions is studied, and conclusions based on statistical analysis are presented. By statistically analyzing the number of overhead transmission line faults due to wind disasters in each city of Henan Province in each month of the year, most of the tripping and grounding faults of transmission lines are concentrated in April, May, June and July, and the distribution network wind disaster faults are most concentrated in May. Principal component analysis is used to analyze the original data, extract principal components, and then use cross-validation to extract training samples and prediction samples for BP neural network prediction, establish an early warning model, and predict the risk of gale disasters. The results show that the prediction accuracy can reach more than 84%, which can provide a computational model basis for the research on the risk warning of gale disasters.