集中供热系统动态负荷预测与节能控制策略研究

发布时间：2018-10-08 11:07

【摘要】：随着我国国民经济的快速发展,人们对生活环境的舒适性要求越来越高。集中供热事业是我国重要的基础性事业,是保证我国北方地区人们冬季生活环境舒适性的重要手段。当前我国能源于环保形势严峻,粗放式集中供热方式不满足绿色发展要求。依靠先进的技术手段与控制策略来保证集中供热系统的高效节能运行是建筑集中供热系统发展的趋势。随着计量与监控技术、网络控制技术、信息处理技术的发展,依托这些先进技术,研究建筑集中供热系统节能控制策略,提升集中供热系统运行能效,已经成为相关领域研究和关注的重要内容。本文基于陕西地区某热力公司的供热技术平台的数据,考虑负荷动态调节要求,以末端负荷预测为目的,研究具有动态调节特征的负荷预测方法,在末端动态负荷预测的基础上,提出了集中供热系统末端设备节能控制策略和换热站节能控制策略,以期提升集中供热系统运行节能效率。本文首先分析了现行供热系统变流量控制策略与热负荷预测方法中存在的问题。现行供热系统的变流量控制策略主要有“温差”和“压差”控制策略,这两种方法都是基于系统热负荷的集中效应进行的控制,不能完全满足供热系统末端用户动态调节的要求和保证所用末端用户的热舒适性。随着供热系统计量与监控技术发展,我们可以比较方便地获得所用末端用户的环境与运行参数,利用这些参数,研究合适的动态负荷方法,改善现行供热系统变流量控制策略的不足,是本文研究的核心内容。通过对影响热负荷变化的因素进行分析,发现供热系统负荷变化受多种因素影响,并具有很强的非线性和不确定性。通过供热系统常用的负荷预测算法进行简单的分析,发现目前热负荷预测算法具有算法复杂、关注集中效应、不具备动态调节特征等局限。为了目前热负荷预测算法中存在的问题。本文从负荷数据预测与曲线拟合的相似性出发,引入了移动多项式最小二乘预测模型;考虑供热系统负荷变化较为平缓而且趋势性较为明显的特点,采用改进加权移动平均算法对末端负荷进行了预测。本文用相同的、来自实际工程的热负荷数据,分别采用移动多项式最小二乘预测模型、改进加权移动平均算法预测模型与目前负荷预测研究领域的热点算法BP神经网络算法进行了预测计算,并对预测结果进行了对比分析。预测结果表明:移动多项式最小二乘算法和改进加权移动平均算法较BP神经网络算法的平均预测误差小,算法更为简单,所需数据也较少,更能适用于集中供热系统末端热负荷预测的工程应用。最后,根据对末端用户热负荷动态预测的结果,结合供热系统的网络控制与管理平台的功能,本文提出了基于末端用户热负荷动态预测的节能控制策略,可分别实现用户末端设备调节和换热站负荷调节的节能运行优化控制,力图实现整个热网的按需供热,有助于提高整个供热系统运行效率,实现节能减排的目标。
[Abstract]:With the rapid development of our national economy, people's comfort in living environment is getting higher and higher. Central heating is an important basic cause of our country, and it is an important means to ensure the comfort of people living in winter in northern China. The current energy of our country is severe in environmental protection, and the mode of centralized heating does not meet the requirement of green development. By means of advanced technical means and control strategy, the efficient energy-saving operation of central heating system is the trend of the development of central heating system. With the development of measurement and monitoring technology, network control technology and information processing technology, based on these advanced technologies, this paper studies the energy-saving control strategy of central heating system and improves the energy efficiency of central heating system. It has become an important content of research and attention in relevant fields. Based on the data of the heat supply technology platform of a thermal company in Shaanxi area, the load forecasting method with dynamic adjustment characteristics is studied in consideration of the load dynamic adjustment requirement, and the load forecasting method with the dynamic adjustment characteristic is researched, and on the basis of the end dynamic load forecasting, The energy-saving control strategy and energy-saving control strategy of heat exchange station in the end of central heating system are put forward in order to improve the efficiency of energy-saving in central heating system. In this paper, the existing problems in current heating system variable flow control strategy and thermal load forecasting method are analyzed. The variable flow control strategy of the current heating system mainly has the temperature difference and the pressure difference control strategy, both methods are based on the centralized effect of the thermal load of the system, can not completely meet the requirement of the dynamic regulation of the end user of the heating system and guarantee the thermal comfort of the end user. With the development of metering and monitoring technology of heat supply system, we can obtain the environment and operating parameters of the end user conveniently, utilize these parameters, study the suitable dynamic load method, improve the deficiency of the current variable flow control strategy of the current heating system, It is the core content of this paper. By analyzing the factors affecting the change of heat load, it is found that the load variation of heating system is influenced by many factors, and has strong nonlinearity and uncertainty. Through simple analysis of load forecasting algorithm commonly used in heat supply system, it is found that the current thermal load forecasting algorithm has the limitation of complex algorithm, focus effect, dynamic adjustment feature and so on. in order to solve the problems existing in the current thermal load prediction algorithm. In this paper, based on the similarity of load data prediction and curve fitting, the paper introduces the model of the least two-multiplication prediction of mobile polynomial, considers that the load variation of heat supply system is gentle and the trend is obvious, and adopts the improved weighted moving average algorithm to forecast the end load. In this paper, using the same data of thermal load from the actual engineering, we use the least squares prediction model of the mobile polynomial, improve the prediction model of the weighted moving average algorithm and the BP neural network algorithm of the hot spot algorithm in the current load forecasting research field, and make the prediction calculation. The prediction results are compared and analyzed. The results show that the least squares algorithm and the improved weighted moving average algorithm of the mobile polynomial are smaller than the average prediction error of the BP neural network algorithm, the algorithm is simpler, the required data is less, and more applicable to the engineering application of the heat load forecasting at the end of the central heating system. Finally, according to the result of the dynamic forecast of the end user's thermal load, combined with the function of the network control and management platform of the heat supply system, this paper proposes the energy-saving control strategy based on the dynamic forecast of the end user's thermal load. the energy-saving operation optimization control of the user end equipment regulation and the heat exchange station load regulation can be respectively realized, the demand of the whole heat supply network is realized, the operation efficiency of the whole heating system is improved, and the aim of energy saving and emission reduction is realized.
【学位授予单位】：广州大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TU995

【参考文献】