论文部分内容阅读
在地源热泵热物性测试中,经常会遇到测试样本数据太少的问题。根据浙江某项目地源热泵热物性测试数据,提出一种针对极小样本试验虚拟增广样本的方法。基于正态信息扩散原理,从虚拟增广样本和信息论的角度出发,充分利用虚拟样本,而不是先假设其服从某一经典概率分布,来求解样本信息扩散函数。利用K-S检验法,验证所求正态扩散函数的正确性,从而证明所得正态信息概率密度函数更加符合岩土热物性参数的真实分布,为推断整个项目岩土热物性参数的概率分布提供一定的参考。
In geothermal heat pump thermal physical testing, often encounter the problem of too few test sample data. According to the thermophysical properties test data of a ground source heat pump in a project in Zhejiang Province, a method for testing virtual augmented samples for a very small sample is proposed. Based on the principle of normal information diffusion, from the point of view of virtual augmented sample and information theory, the virtual sample is fully used instead of assuming that it obeys a classical probability distribution to solve the sample information diffusion function. KS test is used to verify the correctness of the normal diffusion function, which proves that the probability density function of the normal information obtained is more in line with the true geophysical parameter of geomaterials, and provides a certain degree of probability distribution for inferring geophysical parameter of geomaterials The reference.