上海市大型公共建筑能耗的贝叶斯统计分析  

Bayesian statistical analysis on energy for consumption of large-scale public buildings in shanghai

作　　者：徐鹏涛[1] 刘吉彩[1] 郑鹭[1] 岳荣先[1] 

机构地区：[1]上海师范大学数理学院,上海200234

出　　处：《上海师范大学学报（自然科学版）》2017年第2期169-177,共9页Journal of Shanghai Normal University(Natural Sciences)

基　　金：上海市科学技术委员会科研计划项目(14DZ201902)

摘　　要：在建筑能耗的计量过程中,积累了大量的实时能耗数据.这些数据的特点是数量大、噪声大,存在缺失和测量误差等.如何分析和应用如此海量数据,是一个极具挑战性的问题.以2015年上海市大型建筑的电耗数据为研究对象,通过建立多层贝叶斯模型,对各类型大型建筑的月平均单耗、年平均单耗进行估计.该结果将可以帮助政府监管部门对建筑节能工作进行有效评价.In the process of measuring the power consumed in buildings, massive quantity of real-time energy consumption data have been accumulated. Salient features of these data include large samples, noise accumulations and the presence of measurement errors,etc. Thus,how to analyze and apply these massive data becomes a very challengeable problem. In this paper,based on the dataset which include the consumption of large-scale public buildings in Shanghai for 2015 ,we establish a hierarchical Bayesian model to estimate the average monthly consumption and the average annual consumption of large public-scale buildings in 2015. The results will help government regulators to conduct effective evaluation on energy saving for buildings.

关 键 词：大型公共建筑 多层贝叶斯模型 平均单耗估计 MCMC抽样 

分 类 号：O212.8[理学—概率论与数理统计]