论文部分内容阅读
鉴于最大信息熵原理推导水文频率分布参数估计的最大熵法(POME)通过泛函分析及Lagrange乘子法推演过程相当繁琐复杂的问题,探讨了Mellin变换在指数Γ分布、广义Pareto分布和二参数对数正态分布等参数估计POME中的应用,采用Mellin变换代替Lagrange乘子法推导了其统计分布参数估计的POME。结果表明,相比Lagrange乘子法,Mellin变换推导过程更为快捷、简单、直接,可作为推导统计分布参数估计POME的一种有效方法。
In view of the principle of maximum information entropy, the maximum entropy method (POME) of parameters estimation of hydrological frequency distribution is deduced by functional analysis and the Lagrange multiplier method. Because of the complicated and complicated problems of the exponential Γ distribution, the generalized Pareto distribution and the two parameters Logarithm normal distribution and other parameters to estimate the application of POME. The Mellin transform instead of Lagrange multiplier method is used to derive the POME of its statistical distribution parameter estimation. The results show that compared with Lagrange multiplier method, the Mellin transformation process is faster, simpler and direct, and it can be used as an effective method to estimate POME from statistical distribution parameters.