公共楼宇大规模空调负荷虚拟调峰关键技术研究

发布时间：2018-08-07 15:20

【摘要】：随着经济的发展,国民经济水平的提高,夏季空调负荷成为很多城市电力负荷峰谷差不断增大的主要原因之一,加深了我国夏季高峰时段的电力供需矛盾。公共楼宇大规模空调负荷具备一定的调峰能力,因此挖掘公共楼宇大规模空调负荷的调峰潜力,并采用合理的调控手段削减高峰负荷,对减缓电网的压力具有很重要的意义。为解决公共楼宇大规模空调的虚拟调峰问题,本文主要进行了了如下研究：(1)分析了公共楼宇大规模空调负荷的预测技术。准确的公共楼宇大规模空调负荷预测对于掌握公共楼宇空调负荷可调度容量从而参与电网调控具有重要意义。提出了一种考虑夏季气温积温效应的基于Elman神经网络预测模型。将有效积温、温度、湿度、五日滑动平均温度作为气象因素考虑进Elman神经网络预测模型中,预测预测歳的公共楼宇大规模空调负荷值。(2)研究了公共楼宇大规模空调负荷参与电网虚拟调峰的调度方案的设计。公共楼宇大规模中央空调参与电网的调峰,需要充分考虑电网常规机组和公共楼宇的特点,所以立足于调度方案的设计,围绕框架建立、评估等方面开展研究。研究分析了公共楼宇大规模空调负荷参与的电网的日前、日内、实时的调度的调度技术框架,和有第三方负荷代理机构参与的公共楼宇大规模空调负荷参与调度架构,具体阐述了双层调度目标及约束条件,提出了公共楼宇大规模空调负荷参与调度效果在线评估指标及计算方法,可为公共楼宇大规模空调负荷参与调度优化协调系统提出参考,最后给出了公共楼宇大规模空调负荷参与调度方案的优化调整流程。(3)研究了公共楼大规模空调负荷虚拟调峰的策略。首先分析了公共楼宇大规模中央空调系统的物理模型构成以及主要功能,然后对中央空调进行物理建模,把中央空调的功率分为五大主要耗能部分,包含制冷机组耗能,冷冻水泵耗能,风机盘管耗能,冷却塔耗能和冷却水泵的耗能,并对五大耗能部分主要影响因素进行建模分析；然后针对中央空调的主要耗能部分的分析得到中央空调的可优化部分,分析了对公共楼宇大规模中央空调进行柔性调控的策略,包括全局的温度控制、增加冷冻机水温、风机盘管的关闭等,并分析得到了公共楼宇大规模参与柔性调控的流程。(4)建立了公共楼宇大规模空调负荷虚拟调峰组合优化算例。公共楼宇大规模空调组合参与电网日前调度,不同空调用户,根据自身的实际特征,可采取的调控策略不同。首先分析了公共楼宇大规模空调柔性调控组合方案,对中央空调典型用户的空调负荷特性进行曲线拟合,然后得到典型的中央空调参与的电网调控预设方案。实例分析了针对公共楼宇大规模中央空调参与电网日前调度的双层优化组合方案,采用的CPLEX优化软件包进行组合优化,得到早高峰、腰高峰和晚高峰的调控组合方案。
[Abstract]:With the development of economy and the improvement of national economy, air conditioning load in summer becomes one of the main reasons for the increasing peak and valley difference of power load in many cities, which deepens the contradiction between power supply and demand during the peak period of summer in China. The large-scale air conditioning load of public buildings has certain peak-shaving ability, so it is of great significance to excavate the peak-shaving potential of large-scale air-conditioner load of public buildings and to reduce the peak load by adopting reasonable control measures, which is of great significance to reduce the pressure of power grid. In order to solve the problem of virtual peak-shaving of large-scale air conditioning in public buildings, this paper has mainly carried out the following research: (1) the forecasting technology of large-scale air conditioning load in public buildings is analyzed. Accurate prediction of large-scale air conditioning load in public buildings is of great significance for grasping the dispatching capacity of air conditioning load in public buildings and participating in the regulation and control of power grid. This paper presents a prediction model based on Elman neural network considering the effect of accumulated temperature in summer. The effective accumulated temperature, humidity and five-day moving average temperature are taken into account as meteorological factors in the prediction model of Elman neural network. The large scale air conditioning load of public buildings is forecasted. (2) the design of dispatching scheme for large-scale air conditioning load of public buildings to participate in the virtual peak-shaving of power grid is studied. The large scale central air conditioning of public buildings takes part in the peak-shaving of power grid, which needs to fully consider the characteristics of conventional units and public buildings. Therefore, based on the design of dispatching scheme, the research is carried out around the establishment of framework, evaluation and so on. This paper studies and analyzes the pre-day, in-day, real-time scheduling technical framework of large-scale air-conditioning load in public buildings, and the large-scale air-conditioning load scheduling framework of public buildings with third-party load agents. In this paper, the double layer dispatching target and constraint condition are expounded, and the online evaluation index and calculation method of large-scale air conditioning load participating in dispatching of public buildings are put forward, which can be used as a reference for the optimization and coordination system of large-scale air conditioning load participation in public buildings. Finally, the optimization and adjustment process of large-scale air conditioning load in public buildings is given. (3) the strategy of virtual peak-shaving for large-scale air conditioning load in public buildings is studied. Firstly, the physical model and main functions of large-scale central air conditioning system in public buildings are analyzed, and then the physical model of central air conditioning system is built. The power of central air conditioning system is divided into five main energy consumption parts, including the energy consumption of refrigeration units. Cooling pump energy consumption, fan coil energy consumption, cooling tower energy consumption and cooling pump energy consumption. Then, according to the analysis of the main energy consumption part of the central air conditioning system, the optimizable part of the central air conditioning system is obtained. The flexible control strategy for large-scale central air conditioning in public buildings is analyzed, including the overall temperature control, increasing the water temperature of the freezer. The process of large-scale participation in flexible control of public buildings is obtained. (4) A virtual peak-shaving optimization example of large-scale air conditioning load in public buildings is established. Large scale air conditioning combinations of public buildings participate in daily dispatching of power grid. According to their actual characteristics, different air conditioning users can adopt different control strategies. Firstly, the large-scale flexible control scheme of public buildings is analyzed, and the curve fitting of the load characteristics of typical central air conditioning users is carried out, and then the presupposition scheme of power grid regulation and control in which typical central air conditioners participate is obtained. An example is given to analyze the bilevel optimal combination scheme for large-scale central air conditioning in public buildings to participate in the daily dispatching of power network. The CPLEX optimization software package is used to optimize the combination of the early peak, middle peak and late peak to obtain the control combination of early peak, low peak and late peak.
【学位授予单位】：东南大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TM73

【相似文献】