连续压实指标最大熵统计分布模型研究

针对现行连续压实指标概率分布求解方法精度不足,引入最大熵原理,得出基于最大熵分布的连续压实指标概率分布的统一模型。利用蒙特卡罗法生成随机数组对最大熵统计分布模型进行仿真试验,仿真结果精度及不确定度评定表明,该模型具有较高的可靠性。采用现场检测数据,对该模型和传统方法得到的概率分布函数进行了对比分析,结果表明最大熵分布模型具有更高精度。该模型计算结果精确,且概念清晰,计算简便,为连续压实指标的统计分布研究提供了新思路。

Study on Maximum Entropy Model for Statistical Distribution of Continuous Compaction Index

For the insufficient accuracy in resolution of probability distribution of current continuous compaction index, the maximum entropy principle was introduced to get a unified model for such probability distribution based on maximum entropy distribution. By use of Monte Carlo method, random numbers were generated to carry out simulation test on this maximum entropy model, and through the evaluation of accuracy and uncertainty of the results from the simulation, this model was proved to have relatively high reliability. With the on-site measurements, the probability distribution function from this model and the traditional method were compared, and the result showed that the maximum entropy model of distribution has better accuracy. This model is of more accurate calculation result, clear concept and simple computation, which offers new thoughts for study on statistical distribution of continuous compaction index.