基于监测数据的建筑耗能设备运行性能预测与分析研究

发布时间：2018-01-17 23:34

本文关键词：基于监测数据的建筑耗能设备运行性能预测与分析研究　出处：《天津大学》2014年硕士论文　论文类型：学位论文

【摘要】：当今世界能源形势日益严峻,引起了世界各国的高度重视。我国能源消耗在近些年也呈现出急剧上升的态势,建筑能耗是主要增长点,其耗电量占全国电耗的22%左右,其中空调和供暖电耗约占全部建筑能耗的50%-70%,这部分能耗主要是在设备运行过程中产生的。耗能设备按周期规律运行可分为运行阶段和空闲阶段,其中运行阶段又进一步分为启动阶段、稳定运行阶段和关闭阶段。本文通过研究分析实测能耗数据,确定以空闲阶段为研究对象,建立了设备运行空闲时间预测模型,旨在分析设备运行规律,了解设备运行性能,从而有效减少建筑物生命周期内各设备的能耗,为设备节能和故障诊断奠定理论基础。主要研究方法及研究成果有:(1)综合分析耗能设备用电实测数据与周围环境的气象数据,通过程序实现了两者的自动耦合,采集了一定时期内不同时间段的建筑物能耗设备每分钟的用电量及相应的气象数据,进而对相应时期内建筑物高能耗设备的运行周期进行了统计,获得了建筑物耗能设备运行周期统计数据,为设备运行性能预测提供了数据基础。(2)在获得上述数据的基础上,提出以空闲阶段作为研究对象,对比分析支持向量机算法(SVM)、人工神经网络算法(ANN)、朴素贝叶斯算法(NB)和最邻近算法(KNN)四种数据挖掘算法,计算结果显示前两种算法性能更优。因此基于SVM算法和ANN算法建立预测模型并计算预测误差,实现了建筑耗能设备运行性能的预测。在这个过程中,依据样本数据特征将其划分成训练集和预测集,并在挖掘前对两个集合实现非线性归一化处理,此外通过交叉验证法或实验法确定了预测模型的参数。(3)基于三维建模软件建立了建筑物几何模型并对其网格划分。利用红外温度仪监测建筑物室内墙面温度后将温度值定位于墙面,利用空间反比加权插值算法理论,采用MATLAB数据处理技术,计算了室内各剖分块体中心温度值,获得了建筑物室内墙面及内部空间各中心点的所有温度值,最终建立了室内三维温度场模型;结合所建立的温度场模型和舒适度理论分析室内热环境,宏观评价了耗能设备的运行性能。
[Abstract]:Nowadays, the energy situation in the world is becoming more and more serious, which has aroused great attention from all over the world. In recent years, the energy consumption of our country has also shown a sharp rise, and the building energy consumption is the main growth point. Its electricity consumption accounts for about 22% of the country's electricity consumption, of which air conditioning and heating power consumption accounts for about 50% -70% of the total building energy consumption. This part of energy consumption is mainly generated in the equipment operation process. According to the cycle of energy consumption equipment operation can be divided into operation phase and idle stage, in which the operation phase is further divided into start-up phase. By studying and analyzing the measured energy consumption data, this paper determines the idle stage as the research object, and establishes a prediction model of the idle time of the equipment operation, aiming at analyzing the operation rule of the equipment. Understand the performance of the equipment to effectively reduce the building life cycle of each equipment energy consumption. It lays a theoretical foundation for energy saving and fault diagnosis of equipment. The main research methods and results are: 1) Comprehensive analysis of the measured data of energy consumption equipment and the meteorological data of the surrounding environment. Through the program to realize the automatic coupling between the two, the energy consumption per minute and the corresponding meteorological data of the building energy consumption equipment in a certain period of time are collected. Furthermore, the operation cycle of the building energy consuming equipment in the corresponding period is counted, and the statistical data of the operation cycle of the building energy consumption equipment are obtained. This paper provides a data base for performance prediction of equipment. (2) on the basis of obtaining the above data, this paper presents a comparative analysis of support vector machine (SVM) algorithm in the idle stage as the research object. Ann algorithm, naive Bayesian algorithm (NB) and nearest neighbor algorithm (KNN) are four kinds of data mining algorithms. The results show that the performance of the first two algorithms is better. Therefore, based on SVM algorithm and ANN algorithm, the prediction model is established and the prediction error is calculated to achieve the performance prediction of building energy consumption equipment. It is divided into training set and prediction set according to the feature of sample data, and the nonlinear normalization of the two sets is realized before mining. In addition, the parameters of the prediction model are determined by the cross validation method or the experimental method. Based on the 3D modeling software, the geometric model of the building is established and its mesh is divided. The temperature of the building indoor wall is monitored by the infrared temperature meter, and the temperature value is located on the wall. Based on the theory of inverse proportional weighted interpolation algorithm and MATLAB data processing technique, the temperature values of the center of each subdivision block in the room are calculated. All the temperature values of the interior wall and each center point of the interior space are obtained, and finally the indoor three-dimensional temperature field model is established. Combined with the established temperature field model and comfort theory, the indoor thermal environment was analyzed, and the operating performance of the energy dissipation equipment was evaluated macroscopically.
【学位授予单位】：天津大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TU111.195

【相似文献】