城市燃气输配系统的模拟及其应用
|
摘要
在城市燃气管网的设计过程中,水力计算是最关键的环节。本论文介绍了燃气管网水力计算的数学模型和求解方法,应用牛顿法解环形管网,很好的解决了燃气输配中的水力计算问题。
同时,在本论文中阐述了国内外燃气管网优化问题的发展现状,对应用于管网优化问题中的一些常用方法进行了比较与分析,指出了各常用算法的优缺点以及适用性,归纳总结出管网优化的发展趋势,给出了管径优化的目标函数,然后应用遗传算法对燃气输配管网进行了管径优化。优化结果表明,遗传算法适合于管道优化这一工程领域。
根据国内外燃气负荷研究的进展,从多角度对燃气负荷预测进行了分类。论文中提出了需要按燃气负荷预测的应用要求,燃气负荷变化特性等具体情况选择预测方法。在matlab环境下基于遗传算法的全局搜索和BP网络局部精确搜索的特性,对城市燃气管网短期负荷进行了预测。根据城市燃气负荷变化的特性,应用人工神经网络误差反向传播模型可以反映燃气负荷随机性、周期性及其变化趋势。通过采用遗传算法优化网络初始权重的方法,将遗传算法和BP算法有机结合起来,为储气设施的设计与建设提供理论依据。在大连市锦绣小区燃气负荷预测中,以历史资料进行检验的结果证明了这种方法的有效性和可靠性。
国内外的各种管道泄漏检测方法大致可以分为硬件法和软件法两大类,本论文分析了各种检测技术的检测原理和优缺点,对各种检测方法进行了比较。将序贯概率比检验法(SPRT)应用到管道泄漏的在线故障诊断中,考虑描述其性能的两个参数——操作线和平均样本数函数,提出了将序贯概率比检测应用于管道泄漏判断的方法。实际应用表明,该方法克服了序贯概率比检测的原有缺点,用其进行管道泄漏检测提高了检测灵敏度和报警速度。
The calculation of waterpower is the key step during the design of gas network of city. Model and method of calculating waterpower of gas network are introduced in this paper. The calculation of distributing gas problem has been well solved by using Newton law.
The paper introduces simulation optimal calculation of development in pipeline design. Hydraulic calculation formula and familiar solutions were adopted and sumed up by summarization. Through compared to many techniques for pipe optimization, it offers that we must make use of the adjust measures to local conditions and synthesis appliance of the development principle, which well applies to analyses and distribution manage. This paper establishes the target function of pipe diameter optimization, and utilizes Genetic Algorithm to optimize the diameters of gas networks. The result shows that Genetic Algorithm can apply to pipeline optimization efficiently.
The present article forecasted the short load of the city gas pipe net under matlab environment which was founded on the global searching Genetic arithmetic and searching the part of BP net in detail. According to the character of the city gas load, the randomicity, periodicity and the trend of the gas charge can be reflected by the application of BP ANN. GAs and BP ANNs algorithms were joined together and supplemented mutually by optimizing the initial weight s of ANNs with GAs. It provided the academic evidence with design and construction of the storage gas. It proved that this method is efficient and credible by examining the historic data during the load forecasting of the city gas pipe net.
The research works about gas load in both domestic and abroad have been reviewed. Based on statistics and modem optimization methods, literatures of gas load forecasting and practice experiences were reviewed and induced, and categories of gas load forecasting have been made from different viewpoints. Forecasting method would be slected in accordance with application need and gas load variation characteristic.
The research achievements of pipeline leak detection at home and abroad in recent years has been reviewed. There are two main methods of leak detection, hardware method and software method. The superiority and imperfections of every method in their performance and fields are discussed. The sensitivity of leaks detection method based on dynamic quantity balance cannot satisfy the measuring requirement because of influences of errors and requirements on false alarm and detection speed. The sequential probability ratio test (SPRT), one of hypothesis test methods, is analyzed and improved. In the experimental test, it is found that this method can overcome the SPRT's shortcomings, and it can improve the detection sensitivity and detection speed.
同时,在本论文中阐述了国内外燃气管网优化问题的发展现状,对应用于管网优化问题中的一些常用方法进行了比较与分析,指出了各常用算法的优缺点以及适用性,归纳总结出管网优化的发展趋势,给出了管径优化的目标函数,然后应用遗传算法对燃气输配管网进行了管径优化。优化结果表明,遗传算法适合于管道优化这一工程领域。
根据国内外燃气负荷研究的进展,从多角度对燃气负荷预测进行了分类。论文中提出了需要按燃气负荷预测的应用要求,燃气负荷变化特性等具体情况选择预测方法。在matlab环境下基于遗传算法的全局搜索和BP网络局部精确搜索的特性,对城市燃气管网短期负荷进行了预测。根据城市燃气负荷变化的特性,应用人工神经网络误差反向传播模型可以反映燃气负荷随机性、周期性及其变化趋势。通过采用遗传算法优化网络初始权重的方法,将遗传算法和BP算法有机结合起来,为储气设施的设计与建设提供理论依据。在大连市锦绣小区燃气负荷预测中,以历史资料进行检验的结果证明了这种方法的有效性和可靠性。
国内外的各种管道泄漏检测方法大致可以分为硬件法和软件法两大类,本论文分析了各种检测技术的检测原理和优缺点,对各种检测方法进行了比较。将序贯概率比检验法(SPRT)应用到管道泄漏的在线故障诊断中,考虑描述其性能的两个参数——操作线和平均样本数函数,提出了将序贯概率比检测应用于管道泄漏判断的方法。实际应用表明,该方法克服了序贯概率比检测的原有缺点,用其进行管道泄漏检测提高了检测灵敏度和报警速度。
The calculation of waterpower is the key step during the design of gas network of city. Model and method of calculating waterpower of gas network are introduced in this paper. The calculation of distributing gas problem has been well solved by using Newton law.
The paper introduces simulation optimal calculation of development in pipeline design. Hydraulic calculation formula and familiar solutions were adopted and sumed up by summarization. Through compared to many techniques for pipe optimization, it offers that we must make use of the adjust measures to local conditions and synthesis appliance of the development principle, which well applies to analyses and distribution manage. This paper establishes the target function of pipe diameter optimization, and utilizes Genetic Algorithm to optimize the diameters of gas networks. The result shows that Genetic Algorithm can apply to pipeline optimization efficiently.
The present article forecasted the short load of the city gas pipe net under matlab environment which was founded on the global searching Genetic arithmetic and searching the part of BP net in detail. According to the character of the city gas load, the randomicity, periodicity and the trend of the gas charge can be reflected by the application of BP ANN. GAs and BP ANNs algorithms were joined together and supplemented mutually by optimizing the initial weight s of ANNs with GAs. It provided the academic evidence with design and construction of the storage gas. It proved that this method is efficient and credible by examining the historic data during the load forecasting of the city gas pipe net.
The research works about gas load in both domestic and abroad have been reviewed. Based on statistics and modem optimization methods, literatures of gas load forecasting and practice experiences were reviewed and induced, and categories of gas load forecasting have been made from different viewpoints. Forecasting method would be slected in accordance with application need and gas load variation characteristic.
The research achievements of pipeline leak detection at home and abroad in recent years has been reviewed. There are two main methods of leak detection, hardware method and software method. The superiority and imperfections of every method in their performance and fields are discussed. The sensitivity of leaks detection method based on dynamic quantity balance cannot satisfy the measuring requirement because of influences of errors and requirements on false alarm and detection speed. The sequential probability ratio test (SPRT), one of hypothesis test methods, is analyzed and improved. In the experimental test, it is found that this method can overcome the SPRT's shortcomings, and it can improve the detection sensitivity and detection speed.
引文
[1]Cross H.Analysis of flow in networks or conduits or conductors.Bulletin 286,University of Illinois Engineering Experimental Station.Urbana,IL.1936
[2]严煦世,赵洪宾.给水管网理论和计算.北京:中国建筑工业出版社,1986:191-210
[3]Karmeli.D,Gadish,Y and Meyers.S.Design of optimal water distribution networks.J.Journal of the Pipeline Division,1968,94(1):1-10
[4]Goulter.I.C.System Analysis in water ---distribution network design:from theory To practice.J.Water Resour.Planning and Mgrnt.ASCE,1992,(3):238-248
[5]倪林安,李榕,许群和.管网平差中 Hardy Cross法的分析和改进.西安冶金建筑学院学报,1992,24(1):91-96
[6]李猷嘉.燃气管道水力计算公式的分析与比较.煤气与热力,2000,20(5):323-330
[7]史登旺.煤气管网的设计计算.河北冶金,2000:45-46
[8]严煦世,赵洪宾.给水管网理论和计算.北京:中国建筑工业出版社,1986.
[9]Robert.J.Gordan.Edword.Tsaland,Kenneth O.Simpson,Robert.R Olson.Optimal Gas Pipeline Design via Dynamic Programming with Variable Strages.Pipeline Simulation Intemet Group,1986:34
[10]S.Bhaduri and R.K.Talachi.Optimization of Natural Gas Pipeline Design.ASME Petroleum Division,1988:67-76.
[11]D.E.Goldberg.Computer-aided Gas Pipeline Operation Using Genetic Algorithms and Rule Learning.Doaoral dissertation,University of Michigan,1983
[12]初飞雪.吴长春输气管道优化运行的研究现状.油气储运,2004,23(11):3-6
[13]刘扬,关晓晶.环形集输管网拓扑优化设计.石油学报,1993,13(2):71-74
[14]刘扬,关晓晶.油气集输系统优化设计研究.石油学报,1993,14(3):110-117
[15]刘扬,关晓晶.油气集输管网系统多目标优化设计.石油规划设计,1993,4(1):26-28
[16]李波,余红伟.管网布局规划技术综述.石油规划设计,2001,12(1):16-18,40
[17]朱文建.燃气输配管网系统优化设计.上海煤气,2001(2):29-34
[18]宋东显,肖芳淳,严大凡.多目标灰色局势决策在油气管路设计中的应用.石油学报,1994,15(1):112-119
[19]肖芳淳.灰色物元分析在优选油气管道设计方案中的应用.油气田地面工程.1996,15(1):1-6
[20]李长俊.长距离输气管道工程混合变量优化设计研究.管道技术与设备,2001(3):1-4,8
[21]郑凤,李长俊.天然气管网系统输配气运行优化研究.油气储运,2004,23(11):22-24
[22]段常贵,徐彦峰,晏继臣等.遗传算法在煤气管径优化中的应用研究.煤气与热力,1993,19(2):19-22.
[23]孙伟,王立峰.环状燃气管网优化计算的数学模拟.煤气与热力,2001,21(6):494-495,498
[24]刘扬.集输管网系统模糊优化设计.大庆石油学院学报,1992,16(2):32-36
[25]汪玉春.输气管优化设计新方法.天然气工业,1993,13(6):64-69
[26]汪玉春,赵胜登.天然气长输管道设计方案优选系统.天然气工业,2003,23(6):136-140
[27]新坤,程冬玲,林性粹.枚举法与动态规划法结合优化田间管网.干旱地区农业研究,2001,19(2):61-66
[28]D.E.Goldberg.Computer-aided Gas Pipeline Operation Using Genetic Algodthrns and Rule Learning.Doctoral dissertation,University of Michigan,1983
[29]Randall Marcus,McMahon Graham,Sugden Stephen.A Simulated Annealing Approach to Communication Network Design.Journal of Combinatorial Optimization.2002,6(1):55-65.
[30]李卫华.天然气管网系统优化设计研究.硕士论文,2003
[31]姚莉.城市天然气管网规划设计.硕士论文,2003
[32]康正凌.复杂天然气管网系统工艺参数研究.硕士论文,1999
[33]C.A.博布罗夫斯基.天然气管路输送.北京:石油工业出版社,1985.8
[34]谢如贤.变权重组合预测模型的建立与应用.预测,1992.4
[35]门可佩.“九五”中国能源形势分析与预测.国际石油经济,1996.1
[36]朱麟.张玉润,徐卫等.城市煤气负荷预报.煤气与热力,1998,18(2):27-28
[37]田一梅,赵元,赵新华.城市煤气负荷的预测.煤气与热力,1998,18(4):20-23
[38]钱凯.中国天然气资源.石油工业出版社,1999.10
[39]汪玉春,李建民,朱勇.天然气消费需求动态预测新方法.石油规划设计,1999,10(5):29-32
[40]赖晓平,周鸿兴,云昌钦.混合模型神经网络在短期负荷中的应用.控制理论与应用,2000,17(1):69-72
[41]谭羽非,陈家新,焦文玲等.基于人工神经网络模型预测城市煤气系统短期负荷预测.煤气与热力,2001,21(3):199-202
[42]谢开贵,李春燕,俞集辉.基于遗传算法的短期负荷组合预测模型.电网技术,2001,25(8):20-23
[43]焦文玲,廉乐明,崔建华等.城市燃气短期周期负荷预测的时序模型.天然气工业,2002,22(1):92-94
[44]严铭卿,廉乐明,焦文玲等.燃气负荷及其模型研究.煤气与热力,2003,23(5):207-210
[45]谭羽非.城市燃气管网日负荷预测的灰色神经网络模型.系统工程理论与实践,1997,35(6):679-682
[46]徐洁,丁金婷,江皓.管道泄漏检测方法综述.管道技术与设备,2004(4):14-16
[47]陈华波,涂亚庆.输油管道泄漏检测方法综述.管道技术与设备,2000,(1):38-41.
[48]钱存普,蒲家宁.运用微机判定管道泄漏范围.油气储运,2000,19(5):9-13.
[49]Xue Jun Zhang.Statistical leak detection in gas and liquid pipelines.Pipes &Pipelines International,1993,(7-8):26-29.
[50]夏炜东.气体和液体管道的统计检漏法.国外油气储运,1995,13(1):42-45.
[51]陈华波,涂亚庆.基于统计推断的管线检漏方法.自动化仪器仪表,2000(4):5-8.
[52]姜正侯.燃气工程技术手册.上海:同济大学出版社,1993
[53]田贯三,张增刚,姜亿.城镇天然气管网水力分析数学模型与计算方法.天然气工业,2002,122(3):96-98
[54]吕谋,李红卫.给水管网平差计算中各种流量法的分析和比较.青岛建筑工程学院学报,1994,15(4):41-47
[55]金志刚.燃气管网平差电算基础及发展.天津大学学报,1992(10):2-7
56]管延文,张鸣,李波.燃气管网节点方程牛顿法数学模拟.煤气与热力,2003,23(8):454-458
[57]陈国良,王煦法,庄镇泉等.遗传算法及其应用.北京:人民邮电出版社,2001
[58]王文远.用基因算法求管网经济管径,给水排水.1997,23(12):22-25
[59]飞思科技产品研发中心.Matlab6.5辅助神经网络分析与设计.北京:电子工业出版社,2004
[60]罗仲达,关治洪.人工神经网络在电力系统短期负荷预测中的应用.湖南电力,2002,22(2):10-12
[61]付道明,孙军,贺志刚.国内外管道泄漏检测技术研究进展.石油机械,2004,32(3):48-51
[62]陈春刚,王毅,杨振坤.利用假设检验理论进行管道泄漏检测.石油化工设备,2004,33(2):46-49
[63]蔡正敏,彭飞,易发新等.长输管道泄漏故障诊断方法的研.应用力学学报,2002,19(2):38-43
[2]严煦世,赵洪宾.给水管网理论和计算.北京:中国建筑工业出版社,1986:191-210
[3]Karmeli.D,Gadish,Y and Meyers.S.Design of optimal water distribution networks.J.Journal of the Pipeline Division,1968,94(1):1-10
[4]Goulter.I.C.System Analysis in water ---distribution network design:from theory To practice.J.Water Resour.Planning and Mgrnt.ASCE,1992,(3):238-248
[5]倪林安,李榕,许群和.管网平差中 Hardy Cross法的分析和改进.西安冶金建筑学院学报,1992,24(1):91-96
[6]李猷嘉.燃气管道水力计算公式的分析与比较.煤气与热力,2000,20(5):323-330
[7]史登旺.煤气管网的设计计算.河北冶金,2000:45-46
[8]严煦世,赵洪宾.给水管网理论和计算.北京:中国建筑工业出版社,1986.
[9]Robert.J.Gordan.Edword.Tsaland,Kenneth O.Simpson,Robert.R Olson.Optimal Gas Pipeline Design via Dynamic Programming with Variable Strages.Pipeline Simulation Intemet Group,1986:34
[10]S.Bhaduri and R.K.Talachi.Optimization of Natural Gas Pipeline Design.ASME Petroleum Division,1988:67-76.
[11]D.E.Goldberg.Computer-aided Gas Pipeline Operation Using Genetic Algorithms and Rule Learning.Doaoral dissertation,University of Michigan,1983
[12]初飞雪.吴长春输气管道优化运行的研究现状.油气储运,2004,23(11):3-6
[13]刘扬,关晓晶.环形集输管网拓扑优化设计.石油学报,1993,13(2):71-74
[14]刘扬,关晓晶.油气集输系统优化设计研究.石油学报,1993,14(3):110-117
[15]刘扬,关晓晶.油气集输管网系统多目标优化设计.石油规划设计,1993,4(1):26-28
[16]李波,余红伟.管网布局规划技术综述.石油规划设计,2001,12(1):16-18,40
[17]朱文建.燃气输配管网系统优化设计.上海煤气,2001(2):29-34
[18]宋东显,肖芳淳,严大凡.多目标灰色局势决策在油气管路设计中的应用.石油学报,1994,15(1):112-119
[19]肖芳淳.灰色物元分析在优选油气管道设计方案中的应用.油气田地面工程.1996,15(1):1-6
[20]李长俊.长距离输气管道工程混合变量优化设计研究.管道技术与设备,2001(3):1-4,8
[21]郑凤,李长俊.天然气管网系统输配气运行优化研究.油气储运,2004,23(11):22-24
[22]段常贵,徐彦峰,晏继臣等.遗传算法在煤气管径优化中的应用研究.煤气与热力,1993,19(2):19-22.
[23]孙伟,王立峰.环状燃气管网优化计算的数学模拟.煤气与热力,2001,21(6):494-495,498
[24]刘扬.集输管网系统模糊优化设计.大庆石油学院学报,1992,16(2):32-36
[25]汪玉春.输气管优化设计新方法.天然气工业,1993,13(6):64-69
[26]汪玉春,赵胜登.天然气长输管道设计方案优选系统.天然气工业,2003,23(6):136-140
[27]新坤,程冬玲,林性粹.枚举法与动态规划法结合优化田间管网.干旱地区农业研究,2001,19(2):61-66
[28]D.E.Goldberg.Computer-aided Gas Pipeline Operation Using Genetic Algodthrns and Rule Learning.Doctoral dissertation,University of Michigan,1983
[29]Randall Marcus,McMahon Graham,Sugden Stephen.A Simulated Annealing Approach to Communication Network Design.Journal of Combinatorial Optimization.2002,6(1):55-65.
[30]李卫华.天然气管网系统优化设计研究.硕士论文,2003
[31]姚莉.城市天然气管网规划设计.硕士论文,2003
[32]康正凌.复杂天然气管网系统工艺参数研究.硕士论文,1999
[33]C.A.博布罗夫斯基.天然气管路输送.北京:石油工业出版社,1985.8
[34]谢如贤.变权重组合预测模型的建立与应用.预测,1992.4
[35]门可佩.“九五”中国能源形势分析与预测.国际石油经济,1996.1
[36]朱麟.张玉润,徐卫等.城市煤气负荷预报.煤气与热力,1998,18(2):27-28
[37]田一梅,赵元,赵新华.城市煤气负荷的预测.煤气与热力,1998,18(4):20-23
[38]钱凯.中国天然气资源.石油工业出版社,1999.10
[39]汪玉春,李建民,朱勇.天然气消费需求动态预测新方法.石油规划设计,1999,10(5):29-32
[40]赖晓平,周鸿兴,云昌钦.混合模型神经网络在短期负荷中的应用.控制理论与应用,2000,17(1):69-72
[41]谭羽非,陈家新,焦文玲等.基于人工神经网络模型预测城市煤气系统短期负荷预测.煤气与热力,2001,21(3):199-202
[42]谢开贵,李春燕,俞集辉.基于遗传算法的短期负荷组合预测模型.电网技术,2001,25(8):20-23
[43]焦文玲,廉乐明,崔建华等.城市燃气短期周期负荷预测的时序模型.天然气工业,2002,22(1):92-94
[44]严铭卿,廉乐明,焦文玲等.燃气负荷及其模型研究.煤气与热力,2003,23(5):207-210
[45]谭羽非.城市燃气管网日负荷预测的灰色神经网络模型.系统工程理论与实践,1997,35(6):679-682
[46]徐洁,丁金婷,江皓.管道泄漏检测方法综述.管道技术与设备,2004(4):14-16
[47]陈华波,涂亚庆.输油管道泄漏检测方法综述.管道技术与设备,2000,(1):38-41.
[48]钱存普,蒲家宁.运用微机判定管道泄漏范围.油气储运,2000,19(5):9-13.
[49]Xue Jun Zhang.Statistical leak detection in gas and liquid pipelines.Pipes &Pipelines International,1993,(7-8):26-29.
[50]夏炜东.气体和液体管道的统计检漏法.国外油气储运,1995,13(1):42-45.
[51]陈华波,涂亚庆.基于统计推断的管线检漏方法.自动化仪器仪表,2000(4):5-8.
[52]姜正侯.燃气工程技术手册.上海:同济大学出版社,1993
[53]田贯三,张增刚,姜亿.城镇天然气管网水力分析数学模型与计算方法.天然气工业,2002,122(3):96-98
[54]吕谋,李红卫.给水管网平差计算中各种流量法的分析和比较.青岛建筑工程学院学报,1994,15(4):41-47
[55]金志刚.燃气管网平差电算基础及发展.天津大学学报,1992(10):2-7
56]管延文,张鸣,李波.燃气管网节点方程牛顿法数学模拟.煤气与热力,2003,23(8):454-458
[57]陈国良,王煦法,庄镇泉等.遗传算法及其应用.北京:人民邮电出版社,2001
[58]王文远.用基因算法求管网经济管径,给水排水.1997,23(12):22-25
[59]飞思科技产品研发中心.Matlab6.5辅助神经网络分析与设计.北京:电子工业出版社,2004
[60]罗仲达,关治洪.人工神经网络在电力系统短期负荷预测中的应用.湖南电力,2002,22(2):10-12
[61]付道明,孙军,贺志刚.国内外管道泄漏检测技术研究进展.石油机械,2004,32(3):48-51
[62]陈春刚,王毅,杨振坤.利用假设检验理论进行管道泄漏检测.石油化工设备,2004,33(2):46-49
[63]蔡正敏,彭飞,易发新等.长输管道泄漏故障诊断方法的研.应用力学学报,2002,19(2):38-43