中央空调系统送风温度控制方法研究
|
摘要
能源是影响国民经济持续发展的重要因素,中央空调系统的能量消耗一般占整个建筑耗能的60%。但目前实际情况是,空调耗能严重,能源利用效率低。造成这种现象的主要原因:一是由于空调系统一般都按最大负荷计算,采用定工作点方式运行,而中央空调系统在满负荷下实际运行的时间非常短;二是中央空调控制器一般采用定参数设计。因此,空调系统控制器参数在线修正及工作点优化,是实现降低能耗的关键因素。
针对空调送风系统具有时变性、非线性、大滞后、大惯性的特点,无法获得较精确的数学模型或模型非常粗糙。本文分析了风系统的运行工艺和控制特点,采用具有自学习能力的智能方法,提出了一种基于模糊神经网络的控制策略。
在此基础上,通过临界灵敏度法确定PID控制器的参数,在空调运行的过程中,设定送风温度,对空调送风系统进行负反馈控制回路调节,解决系统响应的快速性、稳定性和准确性三者之间的矛盾,使系统以较快的速度稳定到设定目标位,并能达到长时间稳定。根据空凋工艺原理,找出影响空调送风温度的主要变量,模糊神经网络修正器能调整自身的控制参数,在不确定的环境下进行有条件的决策,以适应对象和环境的变化,在不需要建立空调系统模型的条件下,直接根据影响空调系统送风温度的变量值,产生送风温度的补偿量,使送风温度逐渐降低。引入空调能效比作为评价函数,对系统能源利用效率进行评估。利用BP学习算法中的梯度下降法对模糊神经网络修正器的参数进行调节,使其满足控制要求。通过试验和仿真验证,经过PID负反馈控制回路调节,空调运行状况平稳,响应速度快且具有准确性。基于模糊神经网络的修正器不断对送风温度进行补偿,使其逐渐降低,满足工况负荷后保持不变,使空调一直运行在低能耗的状态。
Energy impact of the national economy is an important factor of sustainable development,energy consumption of center system commonly occupies about 50 percent of energy consumption of the whole building. But in fact, a majority of center systems inefficiently runs,and wastes a mass of energy. The main reason for this phenomenon: First, center systems are generally in accordance with the most loads calculated adopting invariable working point, and the time of center systems running under the most loads is very short; Second,central air-conditioning controller using fixed-parameter design in general.
For air supply system with time-varying,nonlinear,large Time-Delay,large inertia characteristics are unable to obtain more accurate mathematical model or models are very rough. This paper analyzes the technology of the supply air system operation and control features. The intelligent methods of self-learning ability is adopted.
On this basis,in the process of air-conditioning running, set the air temperature, adopt critical sensitivity method to determine the parameter of PID parameter self-tuning. Applying negative feedback control loop regulation to air supply system, achieve the rapid system response, stability and accuracy of the contradiction between the three, allowing the system to have a more rapid rate of stability to set a target position, and can achieve prolonged stability. Using the modifier based on fuzzy neural network does not need to set up the air-conditioning system model, from the two variable values which could represent the state of air-conditioning system working to produce a compensation volume of air supply temperature, while ensuring the normal operation of the system on the basis of the air supply temperature so that reduced. The introduction of air conditioning energy efficiency ratio as the evaluation function of the system to assess the efficiency of energy use. Using BP learning algorithm of gradient descent algorithm for fuzzy neural network's parameters as amended regulation, reduce the air conditioning running and the ideal current state of the gap between running to try to achieve the desired energy efficiency standards. During testing and simulation, in the course of the work, through negative feedback control loop PID regulator, air-conditioning run smooth, has fast response and high accuracy. The modifier based on fuzzy neural network compensate air supply temperature in order to decrease it. When the system ran under the most loads, the temperature unchanged, air conditioning has been running on low power state ,at last.
针对空调送风系统具有时变性、非线性、大滞后、大惯性的特点,无法获得较精确的数学模型或模型非常粗糙。本文分析了风系统的运行工艺和控制特点,采用具有自学习能力的智能方法,提出了一种基于模糊神经网络的控制策略。
在此基础上,通过临界灵敏度法确定PID控制器的参数,在空调运行的过程中,设定送风温度,对空调送风系统进行负反馈控制回路调节,解决系统响应的快速性、稳定性和准确性三者之间的矛盾,使系统以较快的速度稳定到设定目标位,并能达到长时间稳定。根据空凋工艺原理,找出影响空调送风温度的主要变量,模糊神经网络修正器能调整自身的控制参数,在不确定的环境下进行有条件的决策,以适应对象和环境的变化,在不需要建立空调系统模型的条件下,直接根据影响空调系统送风温度的变量值,产生送风温度的补偿量,使送风温度逐渐降低。引入空调能效比作为评价函数,对系统能源利用效率进行评估。利用BP学习算法中的梯度下降法对模糊神经网络修正器的参数进行调节,使其满足控制要求。通过试验和仿真验证,经过PID负反馈控制回路调节,空调运行状况平稳,响应速度快且具有准确性。基于模糊神经网络的修正器不断对送风温度进行补偿,使其逐渐降低,满足工况负荷后保持不变,使空调一直运行在低能耗的状态。
Energy impact of the national economy is an important factor of sustainable development,energy consumption of center system commonly occupies about 50 percent of energy consumption of the whole building. But in fact, a majority of center systems inefficiently runs,and wastes a mass of energy. The main reason for this phenomenon: First, center systems are generally in accordance with the most loads calculated adopting invariable working point, and the time of center systems running under the most loads is very short; Second,central air-conditioning controller using fixed-parameter design in general.
For air supply system with time-varying,nonlinear,large Time-Delay,large inertia characteristics are unable to obtain more accurate mathematical model or models are very rough. This paper analyzes the technology of the supply air system operation and control features. The intelligent methods of self-learning ability is adopted.
On this basis,in the process of air-conditioning running, set the air temperature, adopt critical sensitivity method to determine the parameter of PID parameter self-tuning. Applying negative feedback control loop regulation to air supply system, achieve the rapid system response, stability and accuracy of the contradiction between the three, allowing the system to have a more rapid rate of stability to set a target position, and can achieve prolonged stability. Using the modifier based on fuzzy neural network does not need to set up the air-conditioning system model, from the two variable values which could represent the state of air-conditioning system working to produce a compensation volume of air supply temperature, while ensuring the normal operation of the system on the basis of the air supply temperature so that reduced. The introduction of air conditioning energy efficiency ratio as the evaluation function of the system to assess the efficiency of energy use. Using BP learning algorithm of gradient descent algorithm for fuzzy neural network's parameters as amended regulation, reduce the air conditioning running and the ideal current state of the gap between running to try to achieve the desired energy efficiency standards. During testing and simulation, in the course of the work, through negative feedback control loop PID regulator, air-conditioning run smooth, has fast response and high accuracy. The modifier based on fuzzy neural network compensate air supply temperature in order to decrease it. When the system ran under the most loads, the temperature unchanged, air conditioning has been running on low power state ,at last.
引文
[1]王健伟.中央空调系统节能研究:(硕士学位论文).上海:上海海事大学制冷与低温工程,2006
[2]卓明胜,梁荣光,许石高.现代大楼中央空调系统节能探讨[J].制冷,2004,12(3):79-81
[3]Peach J W.Twin fans convert dual duct decks to parallel VAVsystems[J].Heating,Piping&Air Conditioning,1974,46(4):64-66
[4]K.K.ran,T.H.Lee,＆ R.Ferdous.Simultaneous online automatic tuning of cascade control for open loop stable process.ISA Transaction 39,2000:233-243
[5]WangJun,Wang Rui-xiang.New method for optimizing control variable air volume(VAV)air conditioning system.Journal of Shanghai Jiaotong University,2006:p248-252
[6]Zhuang Wu,Roderick V.N.Melnik,Finn Borup.Model-based analysis and simulation f airflow control systems of ventilation units in building environments[J].Building and Environment,2007.Volume 42.Issue 1,pp.203-217
[7]Younggy Shin,Young Soo Chang,Youngil Kim.Controller design for a real-time air andling unit[J].Control Engineering Practice 2002.Volume 10.Issue 5,pp.511-518
[8]Ming He,Wen-Jian Cai,Shao-Yuan Li.Multiple fuzzy model-based temperature predictive control for HVAC systems[J].Information Sciences 2005.Volume 169.Issue 1-2,pp.155-174
[9]Jin Wen,Theodore F.Smith.Development and validation of online models with parameter estimation for a building zone with VAV system[J].Energy and Buildings,2007.Volume39.Issue 1,pp.13-22
[10]Jin xinqiao,Du zhimin.Fault tolerant control of outdoor air and AHU supply air emperature in VAV air conditioning systems using PCA method.Applied Thermal Engineering,2006:p1226-1237
[11]Chengyi Guo,Qing Song,Cai W.A Neural Network Assisted Cascade Control System for Air Handling Unit Industrial Electronics.IEEE Transactions on 2007 Page(s):620 - 628
[12]Guo Chengyi,SongQing,CaiWenjian.Real-time control of variable air volume system based on a robust neural network assisted PI controller.IEEE International Conference on Neural Networks - Conference Proceedings 2004:p1847-1852
[13]G.J.Rios-Moreno,M.Trejo-Perea,R.Casta(?)eda-Miranda,.M.Hernández-Guzmán & G.Herrera-Ruiz.Modelling temperature in intelligent buildings by means of autoregressive models[J].Automation in Construction 2007.Volume 16.Issue 5,pp.713-722
[14]SongQ,Hu W J,zhao T N.Robust neural network controller forVariable airflow volume system[C].Control Theory and Applications,IEEE Proceedings,2003,150:112-118
[15]Zhang Hnagnang,Cai kilong.Decentralized nonlinear adaptive control of an HVAC[G].IEEE Transactions on Systems,Man 1& Cybernetics:Part C-Applications &Reviews,2002,32(4):493-498
[16]Bertil Thomas,Mohsen Soleimani-Mohseni & Per Fahlén.Feed-forward in temperature control of buildings[J].Energy & Buildings 2005.Volume 37.Issue 7,pp.755-761
[17]Wang Yagang,Shi 7higang,Cai Wenjiam.PID autotuner and its application in HVAC systems[C].Proceedings of 2001 American Control conference,2001
[18]Bi.Q Cai,W Wang,Q G et al.Advanced controller auto-tuningandits application to HVAC systems[J].IFAC Joural of Control Engineering Practice,2000,8(6):633-644
[19]Stanke D.standard.Designing dualpath,multiple-zone systems[J].ASHRAE J,2005,47(5):20-30
[20]叶大法,杨国荣,董涛.双风道变风量空调系统设计.暖通空调,2008,38(1):64-68
[21]Komareji Mohammad.Optimal model-based control in HVAC systems.Proceedings of the American Control Conference,2008:p1443-1448
[22]Liu Guopeng.Supply fan control methods for VAV systems using a fan airflow station.ASHRAE Transactions,2008:p451-457
[23]Wang Jian-yu.Model predictive control and case simulation for VAV system.Journal of System Simulation,2008:p4446-4450
[24]晋欣桥.多区域变风量空调系统及其分配控制研究[C].暖通空调,2001,31(6):1-4
[25]白建波.空调系统的鲁棒PI控制[C].东南大学动力工程系,2005
[26]马丙场,倪国宗.参数自调整模糊控制器在中央空调控制系统中的应[J].控制理论与应用,2003
[27]晋欣桥,孙金龙,杜志敏.变风量空调系统中基于预测的末端再热控制策略.上海交通大学学报,2006,(02)
[28]吴柳波,张振国.VAV空调系统中的静压控制[J].制冷空调与电力机械,2005(3):40-43
[29]W.Huang and H.N.Lam.Using genetic algorithms to optimize controller parameters for HVAC systems[J].Energy and Buildings 1997,Volume 26.Issue 3,pp.277-282
[30]施志钢.应用神经元自适应PSD控制变频空调系统的仿真研究[J].暖通空调,2006
[31]廖芳芳,肖建.基于BP神经网络PID参数自整定的研究[J].系统仿真学报,2005
[32]王军,王雁,邵惠鹤等.一种优化控制变风量空调系统的新方法.上海交通大学学报,2006,(02).
[33]任庆昌,向 虎,纪加木.变风量空调机组送风温度系统建模与控制研究.暖通空调,2007,37(12):125-128
[34]李卫锋,黄磊.自适应模糊控制在VAV末端装置中的应用.黄石理工学院学报,2008
[35]刘静纨,王汝琳,曹爽.基于LonWorks现场总线技术的变风量空调系统温度PID控制.太原理工大学学报,2008,39(1):57-59
[36]王建玉,任庆昌.变风量空调系统的分布式预测控制.计算机工程与应用,2008,44(5):225-228
[37]李祥阳,严洁.集散控制系统在中央空调温度控制应用.冶金自动化,2008,(6):603-604
[38]李金川,郑智慧.空调制冷自控系统运行与管理.北京:中国建材工业出版社,2002
[39]R.A.Hegberg.Why consider a primary-secondary hydronic pumping system?.ASHRAE Trans.1994,Vol 100(1)
[40]高养田.空调变流量水系统设计拄术发展.暖通空调,1996,3
[41]Lu Lu,Wenjian Cai,Lihua Xie et al.HVAC system optimization-in-building section.Energy and buildings,v37,nl,Jan.2005,p11-22
[42]陆亚俊.暖通空调.北京:中国建筑工业出版社,2002
[43]ASHARE,ASHARE Handbook.Refrigeration System andApplications.2001
[44]Jonas Berge.Fieldbus Enables Innovative Measurements.Advances in Instrumentation and Control,2001
[45]王华春.基于模糊神经网络的主汽温控制系统的研究:(硕士学位论文).河北:华北电力大学,2003
[46]李国勇.智能控制及其MATLAB实现.北京:电子工业出版社,2007
[47]易继锴,侯媛彬.智能控制技术.北京:北京工业大学出版社,2003
[48]房振勇,游文虎,冯汝鹏.改进BP算法在模糊神经网络中的应用.北京:北京航空航天大学学报,2007
[49]Wen Guo,Yizhen Qiao.BP neural network optimized with PSO algorithm and its application in forecasting.Proceedings of IEEE ICIA 2006 - 2006 IEEE International Conference on Information Acquisition,2006:p617-621
[50]Wang Lu.BP neural network to distinguish pregnant woman by arterial pressure pulse waveform analysis.Proceedings - 2008 4th International Conference on Intelligent Information Hiding and Multimedia Signal Processing,2008:p1463-1466
[51]Li Zuo-Yong.New method of BP network optimized based on particle swarm optimization and simulation test.Tien Tzu Hsueh Pao/Acta Electronica Sinica,v 36,n 11,November,2008,p 2224-2228
[52]向婕,吴敏.一种基于改进遗传算法的模糊神经网络控制器及其在烧结终点控制中的应用.信息与控制,2008,37(2):235-241
[53]张承慧,石庆升,程金.一种多电机同步传动模糊神经网络控制器的设计.控制与决策,2007,22(3):30-34
[54]柳朝军,廖晓钟,张宇河.动态模糊神经网络控制器在伺服系统中的应用.控制与决策,2001.6(3):347-350
[55]俞建成,张艾群,王晓辉.基于模糊神经网络水下机器人直接自适应控制.自动化学报,2007,33(8):840-846
[2]卓明胜,梁荣光,许石高.现代大楼中央空调系统节能探讨[J].制冷,2004,12(3):79-81
[3]Peach J W.Twin fans convert dual duct decks to parallel VAVsystems[J].Heating,Piping&Air Conditioning,1974,46(4):64-66
[4]K.K.ran,T.H.Lee,＆ R.Ferdous.Simultaneous online automatic tuning of cascade control for open loop stable process.ISA Transaction 39,2000:233-243
[5]WangJun,Wang Rui-xiang.New method for optimizing control variable air volume(VAV)air conditioning system.Journal of Shanghai Jiaotong University,2006:p248-252
[6]Zhuang Wu,Roderick V.N.Melnik,Finn Borup.Model-based analysis and simulation f airflow control systems of ventilation units in building environments[J].Building and Environment,2007.Volume 42.Issue 1,pp.203-217
[7]Younggy Shin,Young Soo Chang,Youngil Kim.Controller design for a real-time air andling unit[J].Control Engineering Practice 2002.Volume 10.Issue 5,pp.511-518
[8]Ming He,Wen-Jian Cai,Shao-Yuan Li.Multiple fuzzy model-based temperature predictive control for HVAC systems[J].Information Sciences 2005.Volume 169.Issue 1-2,pp.155-174
[9]Jin Wen,Theodore F.Smith.Development and validation of online models with parameter estimation for a building zone with VAV system[J].Energy and Buildings,2007.Volume39.Issue 1,pp.13-22
[10]Jin xinqiao,Du zhimin.Fault tolerant control of outdoor air and AHU supply air emperature in VAV air conditioning systems using PCA method.Applied Thermal Engineering,2006:p1226-1237
[11]Chengyi Guo,Qing Song,Cai W.A Neural Network Assisted Cascade Control System for Air Handling Unit Industrial Electronics.IEEE Transactions on 2007 Page(s):620 - 628
[12]Guo Chengyi,SongQing,CaiWenjian.Real-time control of variable air volume system based on a robust neural network assisted PI controller.IEEE International Conference on Neural Networks - Conference Proceedings 2004:p1847-1852
[13]G.J.Rios-Moreno,M.Trejo-Perea,R.Casta(?)eda-Miranda,.M.Hernández-Guzmán & G.Herrera-Ruiz.Modelling temperature in intelligent buildings by means of autoregressive models[J].Automation in Construction 2007.Volume 16.Issue 5,pp.713-722
[14]SongQ,Hu W J,zhao T N.Robust neural network controller forVariable airflow volume system[C].Control Theory and Applications,IEEE Proceedings,2003,150:112-118
[15]Zhang Hnagnang,Cai kilong.Decentralized nonlinear adaptive control of an HVAC[G].IEEE Transactions on Systems,Man 1& Cybernetics:Part C-Applications &Reviews,2002,32(4):493-498
[16]Bertil Thomas,Mohsen Soleimani-Mohseni & Per Fahlén.Feed-forward in temperature control of buildings[J].Energy & Buildings 2005.Volume 37.Issue 7,pp.755-761
[17]Wang Yagang,Shi 7higang,Cai Wenjiam.PID autotuner and its application in HVAC systems[C].Proceedings of 2001 American Control conference,2001
[18]Bi.Q Cai,W Wang,Q G et al.Advanced controller auto-tuningandits application to HVAC systems[J].IFAC Joural of Control Engineering Practice,2000,8(6):633-644
[19]Stanke D.standard.Designing dualpath,multiple-zone systems[J].ASHRAE J,2005,47(5):20-30
[20]叶大法,杨国荣,董涛.双风道变风量空调系统设计.暖通空调,2008,38(1):64-68
[21]Komareji Mohammad.Optimal model-based control in HVAC systems.Proceedings of the American Control Conference,2008:p1443-1448
[22]Liu Guopeng.Supply fan control methods for VAV systems using a fan airflow station.ASHRAE Transactions,2008:p451-457
[23]Wang Jian-yu.Model predictive control and case simulation for VAV system.Journal of System Simulation,2008:p4446-4450
[24]晋欣桥.多区域变风量空调系统及其分配控制研究[C].暖通空调,2001,31(6):1-4
[25]白建波.空调系统的鲁棒PI控制[C].东南大学动力工程系,2005
[26]马丙场,倪国宗.参数自调整模糊控制器在中央空调控制系统中的应[J].控制理论与应用,2003
[27]晋欣桥,孙金龙,杜志敏.变风量空调系统中基于预测的末端再热控制策略.上海交通大学学报,2006,(02)
[28]吴柳波,张振国.VAV空调系统中的静压控制[J].制冷空调与电力机械,2005(3):40-43
[29]W.Huang and H.N.Lam.Using genetic algorithms to optimize controller parameters for HVAC systems[J].Energy and Buildings 1997,Volume 26.Issue 3,pp.277-282
[30]施志钢.应用神经元自适应PSD控制变频空调系统的仿真研究[J].暖通空调,2006
[31]廖芳芳,肖建.基于BP神经网络PID参数自整定的研究[J].系统仿真学报,2005
[32]王军,王雁,邵惠鹤等.一种优化控制变风量空调系统的新方法.上海交通大学学报,2006,(02).
[33]任庆昌,向 虎,纪加木.变风量空调机组送风温度系统建模与控制研究.暖通空调,2007,37(12):125-128
[34]李卫锋,黄磊.自适应模糊控制在VAV末端装置中的应用.黄石理工学院学报,2008
[35]刘静纨,王汝琳,曹爽.基于LonWorks现场总线技术的变风量空调系统温度PID控制.太原理工大学学报,2008,39(1):57-59
[36]王建玉,任庆昌.变风量空调系统的分布式预测控制.计算机工程与应用,2008,44(5):225-228
[37]李祥阳,严洁.集散控制系统在中央空调温度控制应用.冶金自动化,2008,(6):603-604
[38]李金川,郑智慧.空调制冷自控系统运行与管理.北京:中国建材工业出版社,2002
[39]R.A.Hegberg.Why consider a primary-secondary hydronic pumping system?.ASHRAE Trans.1994,Vol 100(1)
[40]高养田.空调变流量水系统设计拄术发展.暖通空调,1996,3
[41]Lu Lu,Wenjian Cai,Lihua Xie et al.HVAC system optimization-in-building section.Energy and buildings,v37,nl,Jan.2005,p11-22
[42]陆亚俊.暖通空调.北京:中国建筑工业出版社,2002
[43]ASHARE,ASHARE Handbook.Refrigeration System andApplications.2001
[44]Jonas Berge.Fieldbus Enables Innovative Measurements.Advances in Instrumentation and Control,2001
[45]王华春.基于模糊神经网络的主汽温控制系统的研究:(硕士学位论文).河北:华北电力大学,2003
[46]李国勇.智能控制及其MATLAB实现.北京:电子工业出版社,2007
[47]易继锴,侯媛彬.智能控制技术.北京:北京工业大学出版社,2003
[48]房振勇,游文虎,冯汝鹏.改进BP算法在模糊神经网络中的应用.北京:北京航空航天大学学报,2007
[49]Wen Guo,Yizhen Qiao.BP neural network optimized with PSO algorithm and its application in forecasting.Proceedings of IEEE ICIA 2006 - 2006 IEEE International Conference on Information Acquisition,2006:p617-621
[50]Wang Lu.BP neural network to distinguish pregnant woman by arterial pressure pulse waveform analysis.Proceedings - 2008 4th International Conference on Intelligent Information Hiding and Multimedia Signal Processing,2008:p1463-1466
[51]Li Zuo-Yong.New method of BP network optimized based on particle swarm optimization and simulation test.Tien Tzu Hsueh Pao/Acta Electronica Sinica,v 36,n 11,November,2008,p 2224-2228
[52]向婕,吴敏.一种基于改进遗传算法的模糊神经网络控制器及其在烧结终点控制中的应用.信息与控制,2008,37(2):235-241
[53]张承慧,石庆升,程金.一种多电机同步传动模糊神经网络控制器的设计.控制与决策,2007,22(3):30-34
[54]柳朝军,廖晓钟,张宇河.动态模糊神经网络控制器在伺服系统中的应用.控制与决策,2001.6(3):347-350
[55]俞建成,张艾群,王晓辉.基于模糊神经网络水下机器人直接自适应控制.自动化学报,2007,33(8):840-846