基于神经网络解耦控制器的热网控制研究
|
摘要
随着城市建设飞速发展，节能、降耗及环保意识的增强，集中供热成为我国北方城市供热的大势所趋。它是节约燃料、改善环境及实现城市生活现代化的一项重要技术措施和政策。如何有效地控制和管理整个热力系统，提高热力系统的经济效益和社会效益，成为各供热企业急需解决的重要课题。由此导致集中供热网的运行调节和控制策略等相关问题应运而生。
本文以实验室热网装置的控制为基础，旨在对集中供热网的运行调节和控制方法进行研究，主要进行了以下几方面的工作：
对国内外集中供热系统的运行调节和控制技术进行了概述，并从按需供热和均匀供热的角度进行了进一步阐述。然后针对我国集中供热网的特点和存在的问题采用了适合国情的基于温度的流量控制方法。
系统地讨论了集中供热系统流量调节的神经网络解耦控制策略，并针对集中供热网这种具有大滞后、大惯性、非线性、强耦合特性的系统，设计了基于非线性多变量的神经网络解耦控制器，并随后深入探讨了该控制器的特性和具体的设计思想。
详细介绍了实验室热网系统硬件装置的组成、实时热网监控系统软件平台的设计开发、神经网络解耦控制器的具体实现。结果表明，该控制器能够合理的分配各热力子站一次网侧的流量，使各用户二次网供回水平均温度趋于一致，进而实现全网内热量的合理分配，提高供热网的供热质量，达到了集中供热的整体目标即稳定供热和均匀供热。
Along with cities' rapid development, people's recognition of energy saving, consume dropping and environment protecting, the district heating supply has become the main heating mode in north China. It has been used as an important technique management and policy in saving fuel, improving environment and realizing city life's modernization. So how to control and manage the whole heating system effectively, improve the heating system's economy and society benefit have become the most important problem that needs to resolve for most heating supply corporations. All of these lead to the birth of correlation questions like running regulation and control policy.
On the basis of Laboratory-scale heating system control, the problems of running regulation and control policy have been discussed. The application of Neural Networks technology to District Heating System to adjust the flux is also discussed. It mainly includes:
The running regulation and control policy are discussed in the first part of this paper; need-based heating method and even distribution heating method are also discussed deeply. Then by analyzing the characteristics and possible problems of District Heating System in China, the flow control strategy based on temperature is presented, which adapts to the situation of our country.
The method of multivariable decoupling control of Neural Networks is discussed in the second part of this paper. In accordance with the District Heating System having large delay, nonlinear and strong coupling characteristic, this paper presents a nonlinear constrained optimization controller based on multivariable decoupling controller of Neural Networks, and then the characteristic and design idea of the controller are also discussed thoroughly.
It introduces the composing of Laboratory-scale heating system, the real time monitoring software platform and the implementation of multivariable decoupling controller of Neural Networks in detail. It shows that this controller can distribute the flow of the primary pipe network in every substation, make the average temperature of the supplying water and return water tend to be equal. Furthermore, it can make well-proportioned heat distribution in the whole heating network, which improves heating quality and achieves stabilization heating and even distribution heating. Meanwhile, a comfortable living environment can be created.
本文以实验室热网装置的控制为基础，旨在对集中供热网的运行调节和控制方法进行研究，主要进行了以下几方面的工作：
对国内外集中供热系统的运行调节和控制技术进行了概述，并从按需供热和均匀供热的角度进行了进一步阐述。然后针对我国集中供热网的特点和存在的问题采用了适合国情的基于温度的流量控制方法。
系统地讨论了集中供热系统流量调节的神经网络解耦控制策略，并针对集中供热网这种具有大滞后、大惯性、非线性、强耦合特性的系统，设计了基于非线性多变量的神经网络解耦控制器，并随后深入探讨了该控制器的特性和具体的设计思想。
详细介绍了实验室热网系统硬件装置的组成、实时热网监控系统软件平台的设计开发、神经网络解耦控制器的具体实现。结果表明，该控制器能够合理的分配各热力子站一次网侧的流量，使各用户二次网供回水平均温度趋于一致，进而实现全网内热量的合理分配，提高供热网的供热质量，达到了集中供热的整体目标即稳定供热和均匀供热。
Along with cities' rapid development, people's recognition of energy saving, consume dropping and environment protecting, the district heating supply has become the main heating mode in north China. It has been used as an important technique management and policy in saving fuel, improving environment and realizing city life's modernization. So how to control and manage the whole heating system effectively, improve the heating system's economy and society benefit have become the most important problem that needs to resolve for most heating supply corporations. All of these lead to the birth of correlation questions like running regulation and control policy.
On the basis of Laboratory-scale heating system control, the problems of running regulation and control policy have been discussed. The application of Neural Networks technology to District Heating System to adjust the flux is also discussed. It mainly includes:
The running regulation and control policy are discussed in the first part of this paper; need-based heating method and even distribution heating method are also discussed deeply. Then by analyzing the characteristics and possible problems of District Heating System in China, the flow control strategy based on temperature is presented, which adapts to the situation of our country.
The method of multivariable decoupling control of Neural Networks is discussed in the second part of this paper. In accordance with the District Heating System having large delay, nonlinear and strong coupling characteristic, this paper presents a nonlinear constrained optimization controller based on multivariable decoupling controller of Neural Networks, and then the characteristic and design idea of the controller are also discussed thoroughly.
It introduces the composing of Laboratory-scale heating system, the real time monitoring software platform and the implementation of multivariable decoupling controller of Neural Networks in detail. It shows that this controller can distribute the flow of the primary pipe network in every substation, make the average temperature of the supplying water and return water tend to be equal. Furthermore, it can make well-proportioned heat distribution in the whole heating network, which improves heating quality and achieves stabilization heating and even distribution heating. Meanwhile, a comfortable living environment can be created.
引文
[1] 李天智，孙海波．城市集中供热系统控制策略．河北省科学院学报，2000，(4)：205-207．
[2] 黄文，管昌生．城市集中供热研究现状及发展趋势．国外建材科技，2004，25(5)：78-80．
[3] 邹平华．借鉴俄罗斯经验积极发展我国集中供热事业．暖通空调，2000，30(4)：33-37．
[4] Bendt Bendtsen．丹麦区域供热：百年经验．区域供热，2005，1(4)：27-37．
[5] Joanne M. Holford. Gary R. Hunt. Fundamental Atrium Design for Natrual Ventilation[J]. Building and Environment, 2003,38(3):409-426.
[6] Molodezhnikova L.I.,Konishev A.E.,Yusupova Yu.B.. Problem Solving Search for Energy Saving in Heat Networks. Proceedings of the 7th International Scientific and Practical Conference of Students, Tomsk, 2001:327-332.
[7] Jacimovic Branislav, Zivkovic Branislav, Genic Srbislav etal. Supply Water Temperature Regulation Problems in District Heating Network with Both Direct and Indirect Connection. Energy and Buildings, 1998,28(3):317-322.
[8] Fredrik Wernstedt, Paul Davidsson. An Agent-Based Approach to Monitoring and Control of District Heating Systems. Applications of Artificial Intelligence and Expert Systems, Australia: Cairns, 2002:402-411.
[9] Bojic M.,TrifunovicN.,Gustafsson S.I.. Mixed 0-1 Sequential Linear Programming Optimization of Heat Distribution in a District-heating System. Energy and Buildings, 2000,32(3):309-317.
[10] 曾享麒，蔡启林．欧洲集中供热的发展．区域供热，2002，(1)：1-8．
[11] 刘贺明．中国城市集中供热发展与改革情况．区域供热，2003，1(3)：4-8．
[12] 刘浩，王孝红．多级热网分布式微机监控系统通讯网络．煤气与动力，2000，(5)：193-195．
[13] 高捷，樊红斌．热电厂热网计算机管理的现状和发展．区域供热，2001，(2)：6-8．
[14] 李先瑞，郎四维．我国建筑供热采暖的现状及问题分析．区域供热，200l，(1)：11-22．
[15] 唐卫．热力站自动监控系统基本思路与控制模式分析．区域供热，2001，1(5)：4-13．
[16] 江亿．管网可调性和稳定性的定量分析．暖通空调，1997，27(3)：1-7．
[17] 柴天佑．多变量自适应解耦控制及应用．北京：科学出版社，2001．
[18] 刘文琦，杨建华，林艳．集中供热网智能控制方法研究．大连理工大学学报，2004，44(3)：464-468．
[19] 王耕田，王昕琦．城市集中供热网智能协调控制系统．测控技术，2000，19(10)：24-26．
[20] 贺平，孙刚．供热工程．北京：中国建筑工业出版社，1993．
[21] 贾文姣，李素芬．热电厂集中供热网最佳供回水温度及运行调节的研究．节能，2002，1(4)：9-12．
[22] 石兆玉．供热系统运行调节与控制．北京：清华大学出版社，1994．
[23] (美)M，阿坦斯等著．宗孔德译．系统、网络与计算．北京：人民教育出版社，1979．
[24] 陶永华．新型PID控制及其应用．北京：机械工业出版社，1998．
[25] 李新刚．集中供热热力站微机监控系统的设计．电力学报，2000，1(4)：308-311．
[26] 李国勇．智能控制及其MATLAB实现．北京：电子工业出版社，2005．
[27] 易继楷，候媛彬．智能控制技术．北京：北京工业大学出版社，2001．
[28] 董壮进，廖荣平．供热管网初调节的探讨．煤气与动力，2000，(7)：270-272．
[29] 高隽．人工神经网络原理及仿真实例．北京：机械工业出版社，2003．
[30] Wessels L.F.A.,Barnard E.. Avoiding False Local Minima by Proper Initialization of Connections. Neural Networks, IEEE Transactions. 1992,3(6):899-905.
[31] Huailin Shu, Xiucai Guo, Hua Shu. PID Neural Networks in Multivariable Systems. Intelligent Control, 2002. Proceedings of the 2002 IEEE International Symposium on, Vancouver, 2002.
[32] Leszek Kieltyka, Robert Kuceba, Adam Sokolowski. Application of Neural Network Topologies in the Intelligent Heat Use Prediction System. Artificial Intelligence and Soft Computing-ICAISC 2004:7th International Conference, Poland:Zakopane, 2004.
[33] 薛昊洋，刘红军．基于BP神经网络的多变量PID解耦控制．仪器仪表标准化与计量，2005，6(1)：7-9．
[34] 杨建华，张春杰，刘文琦．快速开发测控系统软件的相关技术(方法)研究．制造业自动化，1999，21(5)：49-52．
[35] 丁展，刘海英．网络通信编程案例精选．人民邮电出版社，2004．
[36] 宋朝坤，李伟明．VC++数据库系统开发案例精选．人民邮电出版社，2006．
[37] 师玉宝．自适应神经网络PID控制器．青海大学学报(自然科学版)，2005，23(2)：67-70．
[38] 张磊，胡春．BP改进算法及其在乙二醇精制软测量中的应用．自动化仪表，2005，26(6)：31-34。
[39] 曲婧华．基于自适应神经网络的电路系统故障诊断研究．空军工程大学学报(自然科学版)，2006．7(4)：82-84．
[40] Guiyou Chen, Jinlong Huang, Tongjing Sun. Distributed Supervisor System of Heat Supply Network Based on GSM Platform. Proceedings of the 6th World Congress on Intelligent Control and Automation, IEEE. 2006:6666-6669.
[41] 刘国繁，王迎旭．新型智能数字PID控制器及其应用．电机与控制学报，2006，10(4)：415-419．
[42] 李明，林永君．自适应神经元非模型多变量系统解耦控制．计算机仿真，2003，20(3)：68-71．
[43] 孟哲，刘应宗．我国集中供热的现状与对策．华东交通大学学报，2004，21(3)：66-69．
[44] 宋玉梅，张立忠等．集中供热网监控系统设计中应考虑的若干问题．区域供热，2006，(2)：10-13．
[2] 黄文，管昌生．城市集中供热研究现状及发展趋势．国外建材科技，2004，25(5)：78-80．
[3] 邹平华．借鉴俄罗斯经验积极发展我国集中供热事业．暖通空调，2000，30(4)：33-37．
[4] Bendt Bendtsen．丹麦区域供热：百年经验．区域供热，2005，1(4)：27-37．
[5] Joanne M. Holford. Gary R. Hunt. Fundamental Atrium Design for Natrual Ventilation[J]. Building and Environment, 2003,38(3):409-426.
[6] Molodezhnikova L.I.,Konishev A.E.,Yusupova Yu.B.. Problem Solving Search for Energy Saving in Heat Networks. Proceedings of the 7th International Scientific and Practical Conference of Students, Tomsk, 2001:327-332.
[7] Jacimovic Branislav, Zivkovic Branislav, Genic Srbislav etal. Supply Water Temperature Regulation Problems in District Heating Network with Both Direct and Indirect Connection. Energy and Buildings, 1998,28(3):317-322.
[8] Fredrik Wernstedt, Paul Davidsson. An Agent-Based Approach to Monitoring and Control of District Heating Systems. Applications of Artificial Intelligence and Expert Systems, Australia: Cairns, 2002:402-411.
[9] Bojic M.,TrifunovicN.,Gustafsson S.I.. Mixed 0-1 Sequential Linear Programming Optimization of Heat Distribution in a District-heating System. Energy and Buildings, 2000,32(3):309-317.
[10] 曾享麒，蔡启林．欧洲集中供热的发展．区域供热，2002，(1)：1-8．
[11] 刘贺明．中国城市集中供热发展与改革情况．区域供热，2003，1(3)：4-8．
[12] 刘浩，王孝红．多级热网分布式微机监控系统通讯网络．煤气与动力，2000，(5)：193-195．
[13] 高捷，樊红斌．热电厂热网计算机管理的现状和发展．区域供热，2001，(2)：6-8．
[14] 李先瑞，郎四维．我国建筑供热采暖的现状及问题分析．区域供热，200l，(1)：11-22．
[15] 唐卫．热力站自动监控系统基本思路与控制模式分析．区域供热，2001，1(5)：4-13．
[16] 江亿．管网可调性和稳定性的定量分析．暖通空调，1997，27(3)：1-7．
[17] 柴天佑．多变量自适应解耦控制及应用．北京：科学出版社，2001．
[18] 刘文琦，杨建华，林艳．集中供热网智能控制方法研究．大连理工大学学报，2004，44(3)：464-468．
[19] 王耕田，王昕琦．城市集中供热网智能协调控制系统．测控技术，2000，19(10)：24-26．
[20] 贺平，孙刚．供热工程．北京：中国建筑工业出版社，1993．
[21] 贾文姣，李素芬．热电厂集中供热网最佳供回水温度及运行调节的研究．节能，2002，1(4)：9-12．
[22] 石兆玉．供热系统运行调节与控制．北京：清华大学出版社，1994．
[23] (美)M，阿坦斯等著．宗孔德译．系统、网络与计算．北京：人民教育出版社，1979．
[24] 陶永华．新型PID控制及其应用．北京：机械工业出版社，1998．
[25] 李新刚．集中供热热力站微机监控系统的设计．电力学报，2000，1(4)：308-311．
[26] 李国勇．智能控制及其MATLAB实现．北京：电子工业出版社，2005．
[27] 易继楷，候媛彬．智能控制技术．北京：北京工业大学出版社，2001．
[28] 董壮进，廖荣平．供热管网初调节的探讨．煤气与动力，2000，(7)：270-272．
[29] 高隽．人工神经网络原理及仿真实例．北京：机械工业出版社，2003．
[30] Wessels L.F.A.,Barnard E.. Avoiding False Local Minima by Proper Initialization of Connections. Neural Networks, IEEE Transactions. 1992,3(6):899-905.
[31] Huailin Shu, Xiucai Guo, Hua Shu. PID Neural Networks in Multivariable Systems. Intelligent Control, 2002. Proceedings of the 2002 IEEE International Symposium on, Vancouver, 2002.
[32] Leszek Kieltyka, Robert Kuceba, Adam Sokolowski. Application of Neural Network Topologies in the Intelligent Heat Use Prediction System. Artificial Intelligence and Soft Computing-ICAISC 2004:7th International Conference, Poland:Zakopane, 2004.
[33] 薛昊洋，刘红军．基于BP神经网络的多变量PID解耦控制．仪器仪表标准化与计量，2005，6(1)：7-9．
[34] 杨建华，张春杰，刘文琦．快速开发测控系统软件的相关技术(方法)研究．制造业自动化，1999，21(5)：49-52．
[35] 丁展，刘海英．网络通信编程案例精选．人民邮电出版社，2004．
[36] 宋朝坤，李伟明．VC++数据库系统开发案例精选．人民邮电出版社，2006．
[37] 师玉宝．自适应神经网络PID控制器．青海大学学报(自然科学版)，2005，23(2)：67-70．
[38] 张磊，胡春．BP改进算法及其在乙二醇精制软测量中的应用．自动化仪表，2005，26(6)：31-34。
[39] 曲婧华．基于自适应神经网络的电路系统故障诊断研究．空军工程大学学报(自然科学版)，2006．7(4)：82-84．
[40] Guiyou Chen, Jinlong Huang, Tongjing Sun. Distributed Supervisor System of Heat Supply Network Based on GSM Platform. Proceedings of the 6th World Congress on Intelligent Control and Automation, IEEE. 2006:6666-6669.
[41] 刘国繁，王迎旭．新型智能数字PID控制器及其应用．电机与控制学报，2006，10(4)：415-419．
[42] 李明，林永君．自适应神经元非模型多变量系统解耦控制．计算机仿真，2003，20(3)：68-71．
[43] 孟哲，刘应宗．我国集中供热的现状与对策．华东交通大学学报，2004，21(3)：66-69．
[44] 宋玉梅，张立忠等．集中供热网监控系统设计中应考虑的若干问题．区域供热，2006，(2)：10-13．