一种广义高斯分布随机变量生成算法

An algorithm for generalized Gaussian random variable creation

为方便获得任意参数的广义高斯分布随机变量,基于伽玛分布推导出广义高斯分布随机变量的生成算法,利用变换法和舍选抽样法,给出伽玛分布随机变量的生成方法.该算法计算简单,通过调整分布参数值,可产生具有任何形状参数和任何方差的广义高斯分布随机变量.仿真结果表明,该算法的抽选效率达75%,随机变量估计参数与真实参数相符.对生成的随机数的检验结果表明,随机数服从广义高斯分布.

An algorithm combining random variable method and abandon-selection sampling method was developed to obtain random variables of generalized Gaussian distribution conveniently with any shape parameter and any variance. The algorithm is simple in calculation , and random variables of generalized Gaussian distribution with any shape parameter and any variance can be generated by adjusting the parameter values. Simulation resuits shows that the sampling rate is 75 % and the estimated parameters of the random variables agrees well with the real ones. Chi-squared test on random variables generated shows that the distribution of random number follows generalized Gaussian distribution.