水电站厂房结构振动分析与动态识别
|
摘要
随着水电站建设的快速发展,机组容量及尺寸急剧增大,转速相应提高,机组振动及其诱发的水电站厂房振动问题日益突出,成为目前研究的热点和难点。水电机组在实际运行过程中所承受的各类动载荷是进行振动分析和动态设计的基础数据。然而,由于水电机组规模较大,动载荷时空分布复杂等特点,直接测量其动态力有许多困难,所以利用动载荷识别技术来分析识别机组的动载荷具有重要的工程实用价值。经过多年的发展,神经网络算法渐趋成熟,以及因其具有一些传统方法没有的优势,越来越多地被应用于工程结构的识别预测中。要想真正把握水电站厂房的振动特性,并解决其振动问题,最有效的途径就是通过现场实测,再反馈分析厂房结构的振动特性,并通过有限元数值计算分析对其进行振动复核。本文主要研究了神经网络技术在水电机组动载荷识别中的应用,主要研究内容和成果归纳如下：
(1)利用改进的BP神经网络对不同类型、不同加载方式的动荷载进行数值算例的识别计算,在取得良好效果的情况下,对水电站机墩结构及机组轴系统的动载荷进行了识别,取得了较好的识别结果。
(2)基于水电站机组及厂房振动联合测试数据,分析机组及厂房的振动规律,并对厂房结构动力响应进行了有限元数值反馈计算。根据国内外相关标准对机组及厂房的振动水平进行了评价。
(3)根据实测数据的分析论证,证实了机组振动和厂房结构振动之间存在着显著的耦联作用和相关关系,据此提出了预测厂房结构振动加速度的神经网络预测方法,实例计算表明预测效果较好。
With the rapid development of hydropower construction in our country, the capacity and dimension of hydroelectric generating unit increase constantly, and the rotational speed increases correspondingly. The vibrating problem of water turbine generator set emerging, which becomes an important problem to be studied. The dynamic load on hydroelectric generating units when operating is the basic data of dynamic design and vibration analysis. However, because the generator set is so large-sized and the load distribution is very complicated, it is difficult to measure dynamic load directly. So, it is significant that using load identification technique to identify dynamic load. Compared with other traditional methods, Neural networks has an unparalleled advantage and applied to more and more recognition of engineering structures after years of development. To master the vibration characteristic of hydropower station structure well and resolve the vibration problem, the best method is the field test. Then start a feedback analysis to the experimental data. At the same time, the finite elements method is used to calculate the dynamic response of the power house. In this paper, the author tries to identity dynamic load of generator set by neural network. This paper's primary studies and results such as:
(1) Applying improved BP network into a numeric example to identify different dynamic load, based on the good effect, the dynamic load on generator pier and hydro turbine shaft system has been identified and achieved good result.
(2) The vibration rule was analyzed according to the co-vibration test data of generating unit and hydropower house structures. The finite elements method is used to calculate the dynamic response of the power house. The vibration of the generator set and power house is evaluated according to the corresponding rules.
(3) The obviously coupling effect and correlation between the vibration of generating unit and hydropower house structures was verified. Thus, Neural network model is established so as to predict vibration acceleration of hydropower house structures. The validity of the predicted result is verified.
(1)利用改进的BP神经网络对不同类型、不同加载方式的动荷载进行数值算例的识别计算,在取得良好效果的情况下,对水电站机墩结构及机组轴系统的动载荷进行了识别,取得了较好的识别结果。
(2)基于水电站机组及厂房振动联合测试数据,分析机组及厂房的振动规律,并对厂房结构动力响应进行了有限元数值反馈计算。根据国内外相关标准对机组及厂房的振动水平进行了评价。
(3)根据实测数据的分析论证,证实了机组振动和厂房结构振动之间存在着显著的耦联作用和相关关系,据此提出了预测厂房结构振动加速度的神经网络预测方法,实例计算表明预测效果较好。
With the rapid development of hydropower construction in our country, the capacity and dimension of hydroelectric generating unit increase constantly, and the rotational speed increases correspondingly. The vibrating problem of water turbine generator set emerging, which becomes an important problem to be studied. The dynamic load on hydroelectric generating units when operating is the basic data of dynamic design and vibration analysis. However, because the generator set is so large-sized and the load distribution is very complicated, it is difficult to measure dynamic load directly. So, it is significant that using load identification technique to identify dynamic load. Compared with other traditional methods, Neural networks has an unparalleled advantage and applied to more and more recognition of engineering structures after years of development. To master the vibration characteristic of hydropower station structure well and resolve the vibration problem, the best method is the field test. Then start a feedback analysis to the experimental data. At the same time, the finite elements method is used to calculate the dynamic response of the power house. In this paper, the author tries to identity dynamic load of generator set by neural network. This paper's primary studies and results such as:
(1) Applying improved BP network into a numeric example to identify different dynamic load, based on the good effect, the dynamic load on generator pier and hydro turbine shaft system has been identified and achieved good result.
(2) The vibration rule was analyzed according to the co-vibration test data of generating unit and hydropower house structures. The finite elements method is used to calculate the dynamic response of the power house. The vibration of the generator set and power house is evaluated according to the corresponding rules.
(3) The obviously coupling effect and correlation between the vibration of generating unit and hydropower house structures was verified. Thus, Neural network model is established so as to predict vibration acceleration of hydropower house structures. The validity of the predicted result is verified.
引文
[1]彭程,钱钢粮.21世纪中国水电发展前景展望[J].水力发电,2006,32(2)：6-16.
[2]中华人民共和国国家发展和改革委员会可再生能源发展“十一五”规划[R].2008-03-03.
[3]马震岳,董毓新.水电站机组及厂房振动的研究与治理[M].北京：中国水利水电出版社,2004.
[4]Viliam Biela, Hector Beltran. Draft Tube Fins. Hydraulic Machinery and Cavitation[C], Proceedings of the XIX IAHR Symposium, World Scientific, Sigapore.1998:454-461.
[5]马震岳,王溢波,董毓新,等.红石水电站机组振动及诱发厂坝振动分析[J].水力发电,2000,(9)： 52-54.
[6]黎建康.岩滩电厂机组异常振动试验分析[J].广西电力技术,1996,(3)：10-16.
[7]赵凤遥.水电站厂房结构及水力机械动力反分析[D].大连：大连理工大学,2006.
[8]BBUCHMAIER H, QUASCHNOWITZB, MOSERW. Numerical optimization of high pump-turbines. Hydraulic Machinery and Cavitation,1996:160-169.
[9]青柳茶,五味义雄[日].正常运行时抽水蓄能厂房振动的调查和分析[J],电力土木,1984.No.3.
[10]L' vov A. V. Evaluation of the reliability of the dividing abutment of the Sayano-Shushenskoe hydroelectric station during discharge of floodwaters. Hydrotechnical Construction 1995 29(10):555-561.
[11]沙瑞华.基于神经网络的水电机组动载识别研究[D].大连：大连理工大学,2005.
[12]曹树谦,张文德,萧龙翔.振动结构模态分析[M].天津天津大学出版社,2001,1-3.
[13]梁艳春.计算智能与力学反问题中的若干问题[J].力学进展,2000,30(3)：321-331.
[14]Barlett Jr F. D., Flannelly W. G., Model verification of force determination for measuring vibration loads. Journal of the American Helicopter Society,1979,19, (4):10-18.
[15]林家浩,智浩,郭杏林.平稳随机振动载荷识别逆虚拟激励法(一)[J].计算力学学报,199815(2)：127-136.
[16]智浩,郭杏林,林家浩.平稳随机振动载荷识别逆虚拟激励法(二)[J].计算力学学报,1998,15(4)：395-400.
[17]Kreitinger T J. Non-parametric Force Identification from Structural Response. Soil Dynamics and Earthquake Engineering,1992,11(5):269-277.
[18]路敦勇,吴淼.动态载荷识别的SWAT方法研究[J].振动与冲击,1999,18(4)：51-54.
[19]张运良,林皋,王永学等.一种改进的动态载荷时域识别方法[J].计算力学学报,2004,21(2)：209-215.
[20]Cook A B, Fuller C R. Artificial neural network for predicting nonlinear dynamic helicopter load [J].AIAA Journal,1994,32(5):2337-2344.
[21]张方,朱德懋.基于神经网络模型的动载荷识别[J].Journal of Vibration Engineering,1997, 10:156-162.
[22]徐宜桂,马西庚,史铁林.基于神经网络的动力学反解算法及其应用研究[J].机械工程学报,1998,34(4)：106-110.
[23]王建中,刘朝斌,曾芳.结构动力特性的人工神经网络预测[J].武汉工业大学学报,2000,8(4)：3-36.
[24]张乾飞,王建,吴中如.基于人工神经网络的大坝渗透系数分区反演分析[J].水电能源科学,2001,19(4)：4-7.
[25]E.Kramer. Determining the hydraulic lateral force of pump-turbines [J]. Water power & dam consruction.1981.1.
[26]王贵,邹经湘等.水轮机转轮水力径向力的测试与识别[J].中国电机工程学报,2001,21(1)：89-91.
[27]张益群.立式机组动态力谱分析初探.大电机技术[J],1998,(5)：7-11.
[28]孙万泉,水电站结构振动分析及动态识别[D].大连：大连理工大学,2004.
[29]沙瑞华,张运良,李冬阳.改进BP网络用于水轮发电机组动载荷识别的探讨[J].东北水利水电,2005,23(2)：8-10.
[30]李守巨,刘迎曦,宋树川等.基于RBF神经网络的水轮机振源参数识别方法[J].大连理工大学学报,2007,47(1)：6-10.
[31]韩卫华,宁佐贵,崔蔚.神经网络在结构响应预测中的应用研究[J].信息技术,2004,28(5)：51-53.
[32]苑希民,刘树坤,崔广涛.动态系统的神经网络仿真研究[J].水利水电技术,1998,29(3)：38-42.
[33]翟东武等.基于神经网络的结构振动响应预测控制[J].清华大学学报,2000,40(S1)：61-65.
[34]王建中等.结构动力特性的神经网络预测[J].武汉工业大学学报,2000,22(4)：33-36.
[35]孙作玉.基于动态递归神经网络的半主动控制结构响应预测[J].振动工程学报,2000,13(3)：443-448.
[36]黄道平等.带中间状态反馈的多变量非线性预测控制[J].华南理工大学学报,1998,26(5)：44-50.
[37]朱长春等.基于振动传递特性的振动环境试验响应预测[J].西南交通大学学报,2002,37(1)：1-4.
[38]练继建,张耀东,王海军.水电站厂房结构振动响应的神经网络预测[J].水利学报,2007,38(3)：361-364.
[39]倪飞舟,李桂权.基于LM算法的现场测量系统的传感器非线性误差校正方法[J].测控技术,2004,23(6)：70-72.
[40]李炯城,黄汉雄.神经网络中LMBP算法收敛速度改进的研究[J].计算机工程与应用,2006,16：46-49.
[41]曾凡祥,李勤英.基于LM算法的BP神经网络在大坝变形监测数据处理中的应用[J].2008,32(5)：72-74.
[42]王勖成,邵敏.有限单元法基本原理和数值方法[M].北京：清华大学出版社,2001,1-2.
[43]博弈创作室.APDL参数化有限元分析技术及其应用实例[M].北京：中国水利水电出版社,2004.
[44]徐宜桂,马西庚.基于神经网络的动力学反解法及其应用研究[J].机械工程学报..1998,34(4)：106-110.
[45]宋志强.水电站机组及厂房结构耦合振动特性研究[D].大连,大连理工大学,2009.
[46]电力行业水电站水轮发电机标准化技术委员会.水轮发电机组启动试验规程-DL/T507-2002[S].北京：中国电力出版社,2002.
[47]马震岳,董毓新.水轮发电机组动力学[M].大连：大连理工大学出版社,2003.
[48]马震岳,董毓新.水电站机组及厂房振动的研究与治理[M].北京：中国水利水电出版社,2004.
[49]杨静,马震岳,陈靖等.张河湾抽水蓄能电站厂房机墩组合刚度复核[J].水力发电,2006,32(12)：33-35.
[50]葛哲学,孙志强.神经网络理论与MATLABR2007实现[M].北京：电子工业出版社,2007.
[51]郭磊,刘德辉,李志红.智能诊断技术在水电机组振动故障诊断中的应用[J].水电能源科学,2009,27(4)：178-180.
[52]Fullekrug U. Identification of modal parameters, generalized and effective masses during base-driven tests [J]. Aerospace Science Technology,1998,2(7):469-480.
[53]周开利,康耀红.神经网络模型及MATLAB仿真程序设计[M].北京：清华大学出版社,2005.
[2]中华人民共和国国家发展和改革委员会可再生能源发展“十一五”规划[R].2008-03-03.
[3]马震岳,董毓新.水电站机组及厂房振动的研究与治理[M].北京：中国水利水电出版社,2004.
[4]Viliam Biela, Hector Beltran. Draft Tube Fins. Hydraulic Machinery and Cavitation[C], Proceedings of the XIX IAHR Symposium, World Scientific, Sigapore.1998:454-461.
[5]马震岳,王溢波,董毓新,等.红石水电站机组振动及诱发厂坝振动分析[J].水力发电,2000,(9)： 52-54.
[6]黎建康.岩滩电厂机组异常振动试验分析[J].广西电力技术,1996,(3)：10-16.
[7]赵凤遥.水电站厂房结构及水力机械动力反分析[D].大连：大连理工大学,2006.
[8]BBUCHMAIER H, QUASCHNOWITZB, MOSERW. Numerical optimization of high pump-turbines. Hydraulic Machinery and Cavitation,1996:160-169.
[9]青柳茶,五味义雄[日].正常运行时抽水蓄能厂房振动的调查和分析[J],电力土木,1984.No.3.
[10]L' vov A. V. Evaluation of the reliability of the dividing abutment of the Sayano-Shushenskoe hydroelectric station during discharge of floodwaters. Hydrotechnical Construction 1995 29(10):555-561.
[11]沙瑞华.基于神经网络的水电机组动载识别研究[D].大连：大连理工大学,2005.
[12]曹树谦,张文德,萧龙翔.振动结构模态分析[M].天津天津大学出版社,2001,1-3.
[13]梁艳春.计算智能与力学反问题中的若干问题[J].力学进展,2000,30(3)：321-331.
[14]Barlett Jr F. D., Flannelly W. G., Model verification of force determination for measuring vibration loads. Journal of the American Helicopter Society,1979,19, (4):10-18.
[15]林家浩,智浩,郭杏林.平稳随机振动载荷识别逆虚拟激励法(一)[J].计算力学学报,199815(2)：127-136.
[16]智浩,郭杏林,林家浩.平稳随机振动载荷识别逆虚拟激励法(二)[J].计算力学学报,1998,15(4)：395-400.
[17]Kreitinger T J. Non-parametric Force Identification from Structural Response. Soil Dynamics and Earthquake Engineering,1992,11(5):269-277.
[18]路敦勇,吴淼.动态载荷识别的SWAT方法研究[J].振动与冲击,1999,18(4)：51-54.
[19]张运良,林皋,王永学等.一种改进的动态载荷时域识别方法[J].计算力学学报,2004,21(2)：209-215.
[20]Cook A B, Fuller C R. Artificial neural network for predicting nonlinear dynamic helicopter load [J].AIAA Journal,1994,32(5):2337-2344.
[21]张方,朱德懋.基于神经网络模型的动载荷识别[J].Journal of Vibration Engineering,1997, 10:156-162.
[22]徐宜桂,马西庚,史铁林.基于神经网络的动力学反解算法及其应用研究[J].机械工程学报,1998,34(4)：106-110.
[23]王建中,刘朝斌,曾芳.结构动力特性的人工神经网络预测[J].武汉工业大学学报,2000,8(4)：3-36.
[24]张乾飞,王建,吴中如.基于人工神经网络的大坝渗透系数分区反演分析[J].水电能源科学,2001,19(4)：4-7.
[25]E.Kramer. Determining the hydraulic lateral force of pump-turbines [J]. Water power & dam consruction.1981.1.
[26]王贵,邹经湘等.水轮机转轮水力径向力的测试与识别[J].中国电机工程学报,2001,21(1)：89-91.
[27]张益群.立式机组动态力谱分析初探.大电机技术[J],1998,(5)：7-11.
[28]孙万泉,水电站结构振动分析及动态识别[D].大连：大连理工大学,2004.
[29]沙瑞华,张运良,李冬阳.改进BP网络用于水轮发电机组动载荷识别的探讨[J].东北水利水电,2005,23(2)：8-10.
[30]李守巨,刘迎曦,宋树川等.基于RBF神经网络的水轮机振源参数识别方法[J].大连理工大学学报,2007,47(1)：6-10.
[31]韩卫华,宁佐贵,崔蔚.神经网络在结构响应预测中的应用研究[J].信息技术,2004,28(5)：51-53.
[32]苑希民,刘树坤,崔广涛.动态系统的神经网络仿真研究[J].水利水电技术,1998,29(3)：38-42.
[33]翟东武等.基于神经网络的结构振动响应预测控制[J].清华大学学报,2000,40(S1)：61-65.
[34]王建中等.结构动力特性的神经网络预测[J].武汉工业大学学报,2000,22(4)：33-36.
[35]孙作玉.基于动态递归神经网络的半主动控制结构响应预测[J].振动工程学报,2000,13(3)：443-448.
[36]黄道平等.带中间状态反馈的多变量非线性预测控制[J].华南理工大学学报,1998,26(5)：44-50.
[37]朱长春等.基于振动传递特性的振动环境试验响应预测[J].西南交通大学学报,2002,37(1)：1-4.
[38]练继建,张耀东,王海军.水电站厂房结构振动响应的神经网络预测[J].水利学报,2007,38(3)：361-364.
[39]倪飞舟,李桂权.基于LM算法的现场测量系统的传感器非线性误差校正方法[J].测控技术,2004,23(6)：70-72.
[40]李炯城,黄汉雄.神经网络中LMBP算法收敛速度改进的研究[J].计算机工程与应用,2006,16：46-49.
[41]曾凡祥,李勤英.基于LM算法的BP神经网络在大坝变形监测数据处理中的应用[J].2008,32(5)：72-74.
[42]王勖成,邵敏.有限单元法基本原理和数值方法[M].北京：清华大学出版社,2001,1-2.
[43]博弈创作室.APDL参数化有限元分析技术及其应用实例[M].北京：中国水利水电出版社,2004.
[44]徐宜桂,马西庚.基于神经网络的动力学反解法及其应用研究[J].机械工程学报..1998,34(4)：106-110.
[45]宋志强.水电站机组及厂房结构耦合振动特性研究[D].大连,大连理工大学,2009.
[46]电力行业水电站水轮发电机标准化技术委员会.水轮发电机组启动试验规程-DL/T507-2002[S].北京：中国电力出版社,2002.
[47]马震岳,董毓新.水轮发电机组动力学[M].大连：大连理工大学出版社,2003.
[48]马震岳,董毓新.水电站机组及厂房振动的研究与治理[M].北京：中国水利水电出版社,2004.
[49]杨静,马震岳,陈靖等.张河湾抽水蓄能电站厂房机墩组合刚度复核[J].水力发电,2006,32(12)：33-35.
[50]葛哲学,孙志强.神经网络理论与MATLABR2007实现[M].北京：电子工业出版社,2007.
[51]郭磊,刘德辉,李志红.智能诊断技术在水电机组振动故障诊断中的应用[J].水电能源科学,2009,27(4)：178-180.
[52]Fullekrug U. Identification of modal parameters, generalized and effective masses during base-driven tests [J]. Aerospace Science Technology,1998,2(7):469-480.
[53]周开利,康耀红.神经网络模型及MATLAB仿真程序设计[M].北京：清华大学出版社,2005.
© 2004-2018 中国地质图书馆版权所有 京ICP备05064691号 京公网安备11010802017129号 地址：北京市海淀区学院路29号 邮编：100083 电话：办公室：(+86 10)66554848；文献借阅、咨询服务、科技查新：66554700 |