A new method to evaluate the lightning flashover risk of transmission line based on data mining

摘要

Lightning flashover risk evaluation is the basis for choosing proper measures to improve the lightning performance of the transmission lines. Thus, the effectiveness of the measures adopted is directly determined by the evaluation accuracy. At present, the risks are mainly evaluated through comparing the calculated lightning flashover rate (LFOR) of every tower. However, there are so many assumptions and simplifications in the process of calculation. This way will often cause some discrepancies between the evaluated results and operating experience. A new evaluation method based on data mining will be proposed to study the characteristics of the lightning failure history. The association relationship between the lightning failure and altitude, height difference, relative height difference, ground flash density, etc. will be considered to build classification rules library to determine the tower lightning flashover risk. Comparison of the proposed evaluation method with the traditional methods shows that the results deduced through the proposed method is more agree with the operating experience.