论文部分内容阅读
在重要的岩土工程中,确定大样本岩土参数的概率分布对岩土工程稳定性和可靠性分析有着极其重要的意义。为此,基于积分均方误差最小计算得到的最优窗宽,提出了推断大样本岩土参数概率密度函数的正态信息扩散法。该方法基于信息扩散原理,从试验样本和信息论的角度出发,充分利用样本提供的数据信息,而不是先假定成经典的概率分布曲线拟合检验,数学意义和物理意义更加充分和严密。以推断压缩指数Cc的概率密度函数为例,分析了窗宽对信息扩散估计结果的影响规律,说明了该方法在大样本岩土参数概率密度函数推断方面的合理性。用该方法推断了大样本岩土抗剪强度参数的概率密度函数,通过K-S检验,验证了该方法的正确性和实用性。
In important geotechnical engineering, it is very important to determine the probability distribution of geotechnical parameters in large samples for the stability and reliability analysis of geotechnical engineering. Therefore, based on the optimal window width calculated by the mean square error of the integral, a normal information diffusion method is proposed to estimate the probability density function of geotechnical parameters of large samples. Based on the principle of information diffusion, this method makes full use of the data and information provided by the sample from the perspectives of experimental sample and information theory, instead of assuming the classical probability distribution curve fitting test, the mathematical meaning and physical meaning are more fully and strictly. Taking the probability density function of Cc as an example, the influence of window width on the result of information diffusion estimation is analyzed, and the rationality of this method in extrapolating the probability density function of geotechnical parameters is illustrated. Using this method, the probability density function of rock-soil shear strength parameters of large samples is inferred. The correctness and practicability of this method are verified by K-S test.