【作者】 马庆；

【导师】 李歧强；

【作者基本信息】 山东大学 ， 系统工程， 2012， 博士

【摘要】 持续增长的空调能量消耗和电力需求不断挑战着空调用户和电力供应商。办公建筑空调系统存在能效低下及能量浪费严重的现象,因此依据不同的电价机制,对空调系统能量消耗进行优化,实现空调用户和电力供应商的双赢,是本文研究的出发点。能量利用效率是转换效率、传输效率和终端使用效率相乘得到的,终端使用效率的提高,可以减少整个能源转换和传输链的消耗。而目前常采用的空调系统能量分析方法,主要注重空调系统的传输转换链这一环节。因此,寻求一种从使用效率开始,接着逆向分析传输转换效率的空调系统能量使用分析方法,是目前亟待解决的问题。针对该问题,不同电价机制下的室内冷负荷的需求优化是提高空调系统使用效率的有效方法。而基于空调系统能量输入比的概念,对风机盘管空调和变风量空调的传输转换链进行了优化,是提高空调系统传输转换效率的主要手段。本文对不同电价机制下的办公建筑空调系统的优化方法进行了深入的研究,主要的研究工作概括如下：1、针对在不同电价机制下,空调系统基线冷负荷难以预测的问题,提出一种基于模糊C-均值聚类预处理的神经网络的空调系统冷负荷基线预测方法。采用模糊C-均值聚类的方法对预测日各时刻的室外气象数据进行了聚类；在此基础上,采用BP神经网络对大量的训练样本进行了网络训练,得到了各时刻的负荷预测模型；实验结果表明该预测模型整体预测精度都高于线性模型,对于小时负荷变化较大的时刻,更是显示了该模型的优越性。2、针对在固定电价下,空调系统室内温度设定值常采用固定设定值的问题,提出一种根据室外逐时温度动态设定室内温度值的方法。通过获得的动态室内温度设定值,可以得到空调系统的最小冷负荷需求。首先,以一间典型的办公房间为例,通过构建一个带有热物质的简化模型,建立了空调冷负荷与室内温度设定值和室外逐时温度的热平衡方程；然后,对舒适度指标公式进行了简化,建立了舒适度指标与室内温度设定值和热物质内表面温度的关系模型；最后,基于上述模型,建立了以空调系统冷负荷需求最小为目标,舒适度指标为约束,室内温度设定值为决策变量的优化模型。实验结果表明建筑热物质能够延迟冷负荷峰值的产生时间,具有削峰填谷的作用。动态温度设定值相比于26℃固定温度设定值,不仅空调冷负荷需求减少了7%,而且还能改善PMV的指标。优化后的固定电价室内冷负荷与基线冷负荷相比,削减了11.2%。3、针对在分时电价下,采用预冷控制策略时,室内冷负荷需求难以预测的问题,提出一种以空调系统电费最省为目标的室内温度值优化设定方法。通过优化后获得的室内温度设定值,可以得到空调系统的冷负荷需求。以一间典型的办公房间为例,通过构建一个带有热物质的简化模型,建立了工作期间和预冷期间的室内温度设定值与空调冷负荷的热平衡方程；通过分析建筑热物质模型中各参数对冷负荷的影响,建立了以空调系统电费最省为目标,舒适度指标为约束,室内温度设定值和预冷时间为决策变量的优化模型。实验结果表明在两阶段分时电价下,利用建筑热物质的蓄热功能,提前两小时采用预冷控制策略,相对于没有采用预冷策略的空调系统,在保证室内舒适度的前提下,可以节省9%的空调电费。优化后的分时电价室内冷负荷与基线冷负荷相比,削减了25.5%。由此可见,采用分时电价机制可以起到移峰填谷的作用。4、针对风机盘管空调水泵控制常采用固定最不利支路位置和固定压差的问题,提出一种最不利支路位置和压差可变的优化方法。首先,从水力失调度的定义出发,以固定电价下得到的室内冷负荷需求为前提,提出了一种根据水力失调度来确定最不利支路位置的判定方法；然后,根据判定的最不利支路位置,对水系统进行了系统流量调节；最后,在固定电价下,以水泵能量输入比最小为目标函数,对各管路阻力特性系数和水泵扬程进行了优化控制；实验结果表明该优化方法相比于末端压差控制方法而言,水泵能耗不仅降低了20%左右,而且还能改善水力失调的现象。5、针对变风量空调系统,从空调系统传输转换效率的角度出发,以第二、三章得到的空调冷负荷需求为基础,提出一种基于系统能量输入比的空调系统优化模型。首先根据空调系统各设备的特性,建立了各设备功率的性能曲线模型,并根据各设备的性能曲线模型,建立了空调系统能耗的性能曲线模型；然后根据空调系统的能量守恒约束、各设备的性能约束和各设备之间的耦合约束建立了空调系统的约束模型；最后对模型参数进行了简化,以空调系统能量输入比最小为目标,建立了空调系统各设备输出设定值优化模型；实验结果表明,该方法与以系统能耗比最小为目标的优化方法相比,固定电价下的系统功率提高了6.89%,分时电价下的系统功率提高了7.28%。因此,该方法具有更好的节能效果。本文最后在总结全文工作的基础上,指出了不同电价机制下的办公建筑空调系统优化方法有待进一步研究的问题和方向。更多还原

【Abstract】 The air conditioning user and electric energy provider are challenged continuously by the sustainable growth of the buiding’s energy consumption and power demand. The low energy efficiency and serious energy waste is a universal phenomenon for the air conditioning system of office buildings, therefore, the main research purpose of this thesis is to realize the win-win results for the air conditioning user and electric energy provider by optimizing the energy consumption of air conditioning system under different price mechanism. The energy utilization efficiency is obtained by the multiplication of conversion efficiency, transmission efficiency and use efficiency of terminal energy. The improvement for the use efficiency of terminal energy can decrease greatly energy consumption in energy conversion and transmission chain. However, the actual energy analysis of air conditioning system is often focused on transimission chain.Therefore, it is essential to find a correct analysis approach which can begin from the analysis of terminal demand for air conditionding, then backward analysis for the transmission chain. To studing the cool load demand under different price mechanism is an effective method for improving use efficiency of air conditioningsystem. Based on system enegy input ratio, optimizing the transmission chain of fan coil and VAV system is an important means for improving the transmission efficiency of air conditioning system.The main research of this theis is that the air conditioning syetem of office buildings is optimized under different price mechanism. The corresponding details are shown as follows:1. Basing on the fact that the baseline cool load for air conditioning system is hard to predict effectively under different price mechanism, a kind of BP neural networks forecasting model based on FCM optimization preprocesses is proposed. The hourly outdoor meteorological data of daily demand response predicting is clustered by adopting the FCM method. The hourly load forecasting model is obstained by training network for a large of training samples based on BP neural network. The prediction results show that the prediction precision of the model is higher than that of linearity model, especially at the great variation moment of hour load.2. A dynamic method based on optimal indoor temperature set point is proposed to solve problem of fixed indoor temperature set point under fixed power price. The minimal cool load demand is obtained by optimizing indoor temperature set point. Firstly, taking a simplified office as an example, an office building model with thermal mass is constructed. According to this model, the heat balance equation about cool load with indoor temperature set point and outdoor hourly tempareture is built. Then, the relation model about PMV with indoor temperature and thermal mass inner surface temperature is built by simplified PMV. Finally, taking the least cool load demand as an objective function, PMV as constraint, the optimal model on indoor temperature set point is built. The experimental results indicate that the thermal mass can delay the creation time of cool load peak value and have the effect of peak load shifting. Compared with the26℃temperature set point, the dynamic temperature can reduce the cool load demand by7percents as well as improve the PMV index. Compared with the baseline cool load, the cool load reduces by11.2percents under fixed price.3. An optimal control method for indoor temperature set point based on minimal electricity charge is proposed to solve the problem of being hard to forecast cool load when precooling control strategy is used under TOU price. The terminal energy of the air conditioning system is obtained by optimizing indoor temperature set point. Taking a simplified office as an example, an office building model with thermal mass is constructed. Basing on this model, the heat balance equation about indoor temperature set point and cool load is built during work and precooling. It is analyzed that the parameters of thermal mass influence on cool load. Taking the least electricity charge as an objective function, PMV as constraint, and indoor temperature set point and precooling time as decision variable, the optimal model is built. Comparing with no adoption of precooling stratege, the experimental results indicate that using the precooling control strategy ahead of two hours based on two-stage TOU power price can reduce the electricity charge by9percents, as well as ensure the unchanged PMV index. Compared with the baseline cool load, the cool load reduces by by25.5percents under TOU price. Therefore, it can realize peak load shifting for adoping the TOU price.4. An optimal method of dynamic position and changed pressure difference for the fan coil pump is proposed to solve problem of fixing most disadvantaged position and pressure difference.Firstly, starting from the difinition of the water maladjustment degree, a new judgment method was obtained to ensure the most disadvantaged branch position with the water maladjustment degree. Then, according to the judgment method, the system flow is regulated for the water system based on the cool load demand under fixed power price. Finally, taking the least energy input ratio of pump as an objective function, the resistance characteristic coefficient and pump lift is optimized under fixed power price. The experimental results indicate that compared with end pressure control method, the optimal method can reduce the pump consumption by20percents as well as improve the hydraulic aberration of the coupling pipe network.5. Thinking of transmission chain of VAV of office buildings the air conditioning optimal model based on system energy input ratio is proposed to meet cool load demand under different price mechanism. Firstly, according to each equipment characteristic, the performance curve model of each equipment power is built. Based on these models, the performance curve model of system energy consumption is built too.Then, basing on energy conservation constraint, each equipment performance constraint and coupling constraint, the constraint equation is built. Finally, simplifying the model parameter, taking the least energy input ratio as an objective function, the optimal set point model of various equipments is built. The experimental results indicate that compared with the least energy consumption optimal method, the ststem power is improved by6.89percents under fixed power price, by7.28percents under TOU power price. Therefore, this method has better energy saving effect.Finally, we point out a few existing problems to be solved after summarizing the works done in this dissertation.更多还原