确定岩体力学参数先验分布的随机加权Bayes方法

Random Weighting Bayes Method to Determine the Prior-information Distribution of Rockmass Mechanical Parameters

岩体力学参数在进行Bayes法统计时必须利用先验分布,但经常出现先验信息少而不能确定先验分布的情况.为了解决小样本条件下先验分布确定的问题,采用随机加权重采样技术,产生岩体力学参数再生样本来模拟先验信息的总体分布,从而获得岩体力学参数先验信息统计分布的均值与方差,将其与当前样本分布信息代入贝叶斯公式,从而实现了对岩体力学参数后验分布的确定.仿真算例证明,这种方法在进行岩体力学参数估计时比经典参数估计方法有更高的精确性.

The prior distribution must be used in the statistical process of mechanical parameters of rock mass by Bayes method,but there often appears the situation that the prior information is frequently too little to determine the prior distribution.In order to determine the prior distribution under small samples,the random weighting sampling technique was used to simulate the population distribution of the prior information based on the regeneration sample of mechanical parameters of rock mass,thus the average value and the variance of the prior information statistical distribution of mechanical parameters of rock mass was obtained.Substitute the obtained average value and variance and the current sample distribution information to the Bayes formula,then there comes the determination of the posterior distribution of mechanical parameters of rock mass.The simulation example proved that this method had a higher accuracy compared to the classical parameter estimation method when estimating the mechanical parameters of rock mass.