区域建筑群冷热负荷预测方法研究

发布时间：2018-03-09 23:05

本文选题：建筑分类　切入点：标准建筑　出处：《湖南大学》2014年硕士论文　论文类型：学位论文

【摘要】：建筑节能是我国可持续发展战略的重要组成部分。随着我国城镇化的快速发展，以小区形式进行规划建设并建有区域供冷供热系统的建筑群越来越多，关注建筑区域的能耗已是建筑节能的重要内容之一。在区域供冷供热系统设计过程中，需要在缺乏详细建筑信息的规划阶段，对区域建筑群总冷热负荷进行较为准确地预测，用以指导能源规划、方案设计和产品开发等。现有的建筑负荷预测方法主要是针对单体建筑，而区域建筑负荷预测方法研究相对较少，预测精确度不高。本文提出一种以计算机模拟与统计回归相结合的方法，建立预测模型，用以预测区域建筑群冷热负荷。本文首先分析了已有区域建筑负荷预测方法及其局限性。然后，分类分析了建筑冷热负荷影响因素，总结出各类影响因素的衡量指标；对区域建筑进行分类，确定了不同类型建筑的标准建筑模型类型；利用影响因素衡量指标，建立各类标准建筑模型并提出了模型简化原则。通过调研选取适当因素水平，利用正交试验确定各类标准建筑模型样本量。进而，采用DesignBuilder能耗模拟软件对标准建筑进行负荷模拟，，获得了各类标准建筑的动态负荷。以获得的各类标准建筑动态负荷当作为先验信息，以调查的其他区域负荷统计值作为样本信息，建立Bayesian回归模型，求解得到后验信息，以此作为负荷预测因子。以某地区一示范区为例，利用通常的简单面积叠加模型和Bayesian回归模型计算夏季典型日该区域逐时冷负荷，再对两种模型预测结果与实测值进行对比。利用逐时相对误差、均方根相对误差和最大误差比三个指标评价了二种模型预测精度。二种模型实例预测对比结果表明，Bayesian回归预测模型预测结果的三类误差指标都小于简单面积叠加模型预测结果，表明Bayesian回归预测模型的有效性和具有更好的预测准确度。可以认为，在区域建筑群用能规划阶段，为了获得更好的预测精确度，采用计算机模拟与统计学相结合的方法比通常的简单面积扩展的方法更好。
[Abstract]:Building energy saving is an important part of the sustainable development strategy of our country. With the rapid development of urbanization in our country, more and more buildings are planned and built in the form of residential area and have regional cooling and heating system. Paying attention to energy consumption of building area is one of the important contents of building energy saving. In the design process of district cooling and heating system, it is necessary to forecast the total cooling and heat load of regional building group accurately in the planning stage of lack of detailed building information. It is used to guide energy planning, project design and product development. The existing building load forecasting methods are mainly for individual buildings, while the regional building load forecasting methods are relatively few. This paper presents a method of combining computer simulation with statistical regression to establish a prediction model to predict the cooling and heat load of regional buildings. This paper first analyzes the existing regional building load forecasting methods and their limitations. Then, classifies and analyzes the factors affecting the building cooling and heat load, summarizes the measurement indicators of various factors, and classifies the regional buildings. The types of standard building models of different types of buildings are determined, and various kinds of standard building models are established and simplified principles are put forward by using the measurement index of influencing factors. The appropriate factor level is selected through investigation and research. The orthogonal test is used to determine the sample size of all kinds of standard building models. Furthermore, the load simulation of standard building is carried out by using DesignBuilder energy consumption simulation software, and the dynamic load of all kinds of standard building is obtained. Taking all kinds of standard building dynamic loads as prior information and other regional load statistics as sample information, the Bayesian regression model is established, and the posteriori information is obtained. Taking a demonstration area as an example, a simple area superposition model and Bayesian regression model are used to calculate the hourly cooling load in a typical summer day. Then the prediction results of the two models are compared with the measured values, and the prediction accuracy of the two models is evaluated by using three indexes: time-by-hour relative error, RMS relative error and maximum error ratio. The comparison of two models shows that the three kinds of error indexes of the prediction results of Bayesian regression model are all smaller than those of the simple area superposition model. The results show that the Bayesian regression prediction model is effective and has better prediction accuracy. It can be concluded that in order to obtain better prediction accuracy in the energy use planning stage of regional buildings, The combination of computer simulation and statistics is better than the usual simple area expansion method.
【学位授予单位】：湖南大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TU831

【参考文献】