中文摘要：

混响室（reverberation chamber,RC）电场分布的经典概率分布模型在实际混响室电场拟合度良好性检测中拒绝率较大,不能准确地描述电场统计分布规律。为此,研究了混响室电场统计分布规律,建立了新的电场概率分布模型。基于电场Cartesian分量的实部和虚部均服从正态分布的假设,引入了联合Gaussian分布密度函数来描述电场Cartesian分量的实部和虚部构成的2维随机变量的概率分布模型。在此基础上利用构造边缘分布函数的方法建立了混响室电场Cartesian分量幅值和相位的概率分布模型。以实际混响室电场为对象进行拟合度良好性检测,结果表明,与Rayleigh分布、Weibull分布和均一分布模型相比,边缘分布模型具有最小的拒绝率,从而证明了边缘分布模型能够更准确地描述混响室电场分布的统计规律。

英文摘要：

Because of higher rejection rates in goodness-of-fit tests, the classical probability distribution model of electric field in reverberation chamber （RC） cannot exactly describe the statistical distribution law of electric field in actual RC. So a novel statistical model of the electric field is derived based on the study of the statistical distribution law of the electric field in RC. On the basis of the assumption that both the real and imaginary parts of each Cartesian field component within the working volume of RC follow a Gauss-normal probability density function （ PDF）, a complete bivariate joint Gaussian distribution function is induced to describe two variables contained probability distribution model, which is formed by the real and imaginary parts of each Cartesian field component. Novel theoretical probability distribution models for the magnitude and phase of electric field Cartesian component inside RC are derived by means of marginal PDFs integrated with the joint Gaussian distribution model. The marginal probability distribution models have smaller rejection rates than Rayleigh, Weibull and uniform distribution in goodness-of-fit tests in an actual RC. The results prove that marginal probability distribution models are able to describe the statistical distribution of the electric field in RC more precisely.