建筑供暖室内温度预测及控制研究

发布时间：2018-02-24 04:00

本文关键词： 人工神经网络室温预测专家控制模糊控制　出处：《大连海事大学》2017年硕士论文　论文类型：学位论文

【摘要】：随着世界能源与环境问题的日益加剧,节约能源和保护环境已成为了各国政府的重要工作之一。建筑能耗在所有的能源消耗中占世界总能耗的1/5左右。我国的建筑系统的能源消耗占能源消耗总量的1/3以上,在全部的建筑系统的能源消耗中其供暖能耗又占60%左右,因此建筑的供暖能耗浪费最为严重,而且其节能的潜力巨大。对于目前现有的建筑供暖系统大多缺乏先进的调控手段以及自动化水平不高,不仅降低了用户的热舒适性,还造成了能源的浪费。建筑供暖系统是一个具有大惯性、大滞后的复杂非线性系统,建筑室内温度与围护结构、室外环境及供暖设施所提供的热量等存在着一定程度的非线性关系,利用机理建模难度较大,涉及参数众多,预测结果也有较大误差。通过利用人工神经网络良好的非线性逼近能力以及不依赖模型本身的特点,选择了 BP(Back Propagation)与RBF(Radial Basis Function)人工神经网络的方法,根据采集到K时刻的建筑室温、K-T时刻的建筑室温、K时刻供回水温差、控制状态、室外温度以及光照分别对建筑K+T时刻的室温进行建模和预测。通过仿真对比研究表明,RBF神经网络相对于BP神经网络的预测更加稳定,其均方差低于BP神经网络10.5%,更适合于建筑室内温度的预测。在对室温预测的基础上,需要对建筑的室温进行合理的调节。本文结合大连某高校建筑供暖实际情况设计了专家控制器,首先对建筑供暖控制策略进行了分析,针对不同建筑的不同供暖需求,采取分时供暖的策略;针对室温控制大惯性、大滞后的特点,采取利用室温预测模型与模糊控制相结合的调节方法;同时为保证供暖的安全,设计了一系列的保护策略。然后基于无线通讯技术设计了建筑供暖专家控制平台,该平台具备远程监控、故障及紧急事故处理等功能,大大提高了建筑供暖的自动化水平,保证了供暖的舒适与安全。最后经过实际实验分析,利用专家控制能够及时对室内温度进行调控,减少了超调,实现合理用热,平均节能率可达9.23%,具有良好的经济和社会价值价值。
[Abstract]:With the increasing energy and environmental problems in the world, Energy conservation and environmental protection have become one of the important tasks of the governments of various countries. Building energy consumption accounts for about 1/5 of the total energy consumption in the world. The energy consumption of the building system in China accounts for more than 1/3 of the total energy consumption. Of the total energy consumption of the building system, the heating energy consumption accounts for about 60%, so the heating energy consumption of the building is the most serious. For most of the existing building heating systems, the lack of advanced means of regulation and control and the low level of automation, not only reduce the thermal comfort of users, The building heating system is a complex nonlinear system with large inertia, large lag, building indoor temperature and enclosure structure. There is a certain degree of nonlinear relationship between the outdoor environment and the heat supplied by heating facilities. It is difficult to model the model by using the mechanism and involves a large number of parameters. The method of BP(Back propagation and RBF(Radial Basis function is chosen by using the good nonlinear approximation ability of artificial neural network and the characteristics of independent model. According to the temperature difference between K time and K time of building room temperature and K-T time, The outdoor temperature and light are used to model and predict the room temperature of the building's K T moment. The simulation results show that the RBF neural network is more stable than the BP neural network. The mean square deviation is lower than that of BP neural network, which is more suitable for the prediction of indoor temperature. It is necessary to adjust the room temperature of buildings reasonably. In this paper, an expert controller is designed according to the actual situation of building heating in a university in Dalian. Firstly, the control strategy of building heating is analyzed, aiming at the different heating requirements of different buildings. The strategy of time-sharing heating is adopted. In view of the characteristics of large inertia and large lag in room temperature control, the regulation method which combines the predictive model of room temperature with fuzzy control is adopted, and in order to ensure the safety of heating, A series of protection strategies are designed, and then the building heating expert control platform is designed based on wireless communication technology. The platform has the functions of remote monitoring, fault and emergency handling, which greatly improves the automation level of building heating. The comfort and safety of heating are ensured. Finally, through practical experiment and analysis, the indoor temperature can be adjusted in time by using expert control, which can reduce overshoot and realize rational use of heat. The average energy saving rate can reach 9.23, which has good economic and social value.
【学位授予单位】：大连海事大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TU832;TP183

【相似文献】