强激光与粒子束, 2014, 26 (7): 073208, 网络出版: 2014-06-30
基于广义极值的混沌电场峰值分布研究
Analysis of chaotic electric maximum field by generalized extreme value
混沌电场 峰值分布 广义极值分布 重采样估计 参数估计 置信区间 chaotic electric field maximum field distribution generalized extreme value distribution bootstrap estimating parameters estimating confidence intervals
摘要
通过分析峰值分布及广义极值分布理论关于混沌电场峰值分布的特点,获得一种混沌电场峰值分布的计算方法,结合Bootstrap重采样技术,提出了混沌电场峰值分布参数估计方法,该方法吸取了重采样技术估计的优点,对分布参数置信区间的估计更为准确、实用。以蒙特卡罗仿真和某机械搅拌混响室内电场数据为实例,结果表明,与传统的正态近似法、百分位法、BC-t等方法相比,其计算结果与实际结果更为吻合,证明了该方法具有更高的准确度。
Abstract
The traditional parameters estimating methods of maximum field distribution in chaotic electric field are the approximate normal method, percentile method and BC-t method, which all lead to kind of error when used in small samples and non-normal distribution. We make a careful analysis of the chaotic maximum field distribution and generalized extreme value distribution theory, and then a hybrid algorithm about the chaotic maximum field distribution is presented which based on Bootstrap resample technique, and it is found that this method is more practical and veracious through Monte Carlo simulation and reverberation chamber testing.
庄信武, 余志勇, 刘光斌, 滕向如. 基于广义极值的混沌电场峰值分布研究[J]. 强激光与粒子束, 2014, 26(7): 073208. Zhuang Xinwu, Yu Zhiyong, Liu Guangbin, Teng Xiangru. Analysis of chaotic electric maximum field by generalized extreme value[J]. High Power Laser and Particle Beams, 2014, 26(7): 073208.