您的位置:  首页> 科学研究> 科研成果> 论文>

闪电的分形特征研究及其在自动识别中的应用

刊物全称

物理学报

论文作者

火元莲,张广庶,吕世华,袁萍

发表年度

2013

刊物类别

SCI

卷

62

期

59201

影响因子

1.027

通讯作者

Huo, YL

请求全文

huoyuanlian@163.com

关键词

lightning signal, fractal dimension, support vector machine, automatic recognition

摘要

By analyzing the fractal feature of 120 cloud-to-ground lightning signals and 77 intracloud lightning signals obtained by the fast antenna system in Qinghai area during the summer of 2009, the results show fractal dimension of cloud-to-ground lightning signal is obviously different from that of intracloud lightning signal. Then 5 characteristic values of fractal dimension are used to recognize the discharge types of lightning signal via support vector machine, and the recognition rate is higher than 95%. The construction of cloud-to-ground lightning time series signal fractal dimension trajectory map shows that the fractal dimension minimum value corresponds to the return stroke of the original time series signal, which can be used to quickly and accurately detect the return stroke of lightning signal, and the detection rate can reach 100%.The fractal dimension is a discriminatively physical property which can be used for intelligently analyzing and automatically processing the lightning signal.

相关文章