建筑能耗分析用逐日随机气象模型

Daily stochastic meteorological model for building energy consumption analysis

针对建筑能耗分析所需气象数据的特点,应用时间序列分析方法对北京市十年六项气象参数(温度日均值、温度日较差、水蒸气分压力日均值、水蒸气分压力日较差、太阳总辐射和太阳直接辐射)建立了六维疏系数混合回归随机气象模型。为使用于建模的数据满足平稳性要求,在建模前对原始气象数据进行了平稳性变换。建模时用近似BIC准则挑选回归变元和确定模型阶数。模型建立后首先通过了白噪声检验,可用于预报和模拟。将用此模型得出的模拟值与实测气象数据进行的分析对比表明,模拟数据不仅很好地反映了气象随机过程的数量大小和分布特性,而且很好地反映了各气象参数自身和相互间的相关关系。因此,模型可实际用于建筑能耗分析、空调系统负荷计算、运行管理及动态控制。

A time sequential analysis method is adopted to build a stochastic meteorological model for building energy consumption analysis.Meteorological data accumulated over10years for Beijing are used to build a6-dimensional sparse-coefficient mixed-regressive model.The six parameters involved in the model are daily average temperature,daily temperature variation,daily average vapor partial pressure,daily vapor partial pressure variation,daily total solar radiation and daily direct solar radiation.Stationary transformation was performed before the primary meteorological data were used to build the model.During the model building,the BIC criterion was used to select regressive parameters and determine the order of the model.The model passes the white noise test,and it can be used for prediction and simulation of meteorological processes.It turns out that not only the value and distribution,but also the correlation and auto correlation of the6parameters of the meteorological stochastic processes given by this model's simulation results agree well with those of the observed data.Thus,the model is applicable to building energy consumption analysis,load calculation,operation management and dynamic control of air conditioning systems.