摘要
火电机组热力系统在运行过程中具有非线性、时变、多变量和多回路耦合以及响应速度不同等各种动态过程交织在一起的特点,其建模方法一直是富于挑战性的研究课题之一。火电机组热力系统建模方法是解决机组优化运行、控制、性能评估、故障诊断、仿真等关键问题的基础。针对现有建模方法在实际应用中存在的精度不高,不能准确反映过程特点的问题,本文深入研究了一种新的建模理论——反向建模,以期提高火电机组热力系统的建模效率和建模精度。
本文首次提出了火电机组反向建模方法的基本定义、一般建模模式、要求、意义和作用,形成了较完整的理论和方法。深入研究反向建模过程中的实时数据验证、特征变量提取和神经网络算法、偏最小二乘算法、最小二乘支持向量机算法和遗传算法等具体反向建模算法以及反向建模的可行性和模型的准确性验证。
确保建模数据的正确性是反向建模的基础,本文研究实时运行数据的验证方法。根据火电厂不同测量数据布置的测点数目不同,把数据分两类验证。对于单、双测点数据,研究相邻数据变化率的方法和曲线拟合残差的方法。为了克服这两种方法对火电机组某些特殊运行工况下产生的突变数据的误判,提出用另外与它有密切关联关系的测点的变化量作为参考进行二次判断。对于多测点参数的数据,提出用格拉布斯准则进行实时验证。通过对实时运行数据的验证表明,所用方法可以有效剔除和校正实时数据中的显著误差。
特征变量是反向建模中的重要参数,对建模精度有很大影响,因此研究特征变量的提取方法是提高反向建模精度的关键。本文通过对多种特征变量提取方法的研究,提出在反向建模过程中用灰关联分析法提取特征变量,运用少量输入变量达到改善模型精度和提高模型速度的目的,并在主蒸汽流量的反向建模中得到了验证。
在确立了反向建模理论和方法后,进行反向建模的可行性研究和模型的准确性验证。对火电厂运行数据进行的反向建模分析表明,其运行数据之间的关系主要表现为两种形式:一种是自变量之间有良好的线性相关性,且自变量和因变量之间也有良好的线性相关性;另一种是自变量之间有良好的线性相关性,但自变量和因变量之间的相关性很弱。对第一种情况本文提出以偏最小二乘算法建模;对第二种情况,提出以神经网络算法和最小二乘支持向量机算法建模,并用遗传算法优化最小二乘支持向量机的参数。本文选取火电机组中具有代表性的参数为例,运用反向建模的方法分别建立了火电厂主蒸汽流量、汽轮机排汽焓、飞灰含碳量、亚临界和超(超)临界机组的高温受热面管壁温度、中间点温度等的数学模型,结果表明所建模型可以解决目前火电机组在线性能计算和关键参数的冗余分析中遇到的难题。所建模型的精度验证了反向建模方法的可行性,为火电机组建模提供了新途径。
Due to the characteristics of nonlinear, time-varying, multi-variable, multi-loop coupling and different responding speed, the thermal system modeling method of power unit is always one of the challenging projects. The thermal system modeling method is a basic project in the field of the system operation optimization, control, performance evaluation, fault diagnosis, stimulation and etc. To solve the problems of the thermal system model in the application, a new modeling method called reversed modeling method was developed in the paper which aims at overcoming the lower efficiency and lower accuracy of the thermal system model.
The basic definition, general modeling mode, requirements, significance and the applications of inversed modeling method were presented in the paper first time. A more integrated method and theory were preliminarily built. Real-time data validation, feature variables extraction, modeling algorithm including neural network, partial least squares algorithm, support vector machines and genetic algorithm, reversed modeling feasibility and model accuracy validation were studied in detail in the paper.
Real-time data validation methods were studied in the paper because to ensure the creditability of modeling data is the basis for reversed modeling method. Different measurment parameters have different measuing point numbers, so data were divided into two categories to test. For single and double measuring point data, two real-time data validation methods are studied in the paper. One is based on the rate of adjacent data change and the other is based on curve fitting residual. To avoid the misjudgment of data validation on some special conditions with above two methods, the change rate of a measuring point data associated with the testing data could be taken as the reference data so as to improve the accuracy of model. For multi-point measurement data, Grubbs was recommended for real-time data validation. The results show that the methods can removed and regulate gross error effectively.
Feature variables extraction is a very important issue for reversed modeling method, because feature variables have a great impact on the modeling accuracy. According to the exploration on the feature variables extraction methods, gray relational analysis was recommended for feature variables extraction in the paper. Useing a small amount of input variables can improve the model accuracy and speed. The example of main steam flow model demonstrates the effectiveness and availability of the variety selection method suggested in the paper for independent variables.
After reversed modeling theory and method are established, reversed modeling accuracy and feasibility are studied in the paper. There are two type of relationship between power plant operation data. The first one is that the relationship between the independent variables and relationship between independent variables and the dependent variable are linear correlation; another one is that the relationships between the independent variables are linear correlation, but the relationship between the independent variables and the dependent variable are less correlation. For the first case, partial least squares algorithm was adopted for modeling. For the second case, neural network and support vector machines algorithm were developed for modeling in the paper. In the paper, the reversed modeling method was applied to build the mathematic model of main stream flow, steam turbine exhaust enthalpy, unburned carbon in fly ash, the high-temperature surface metal temperature and boiler intermediate point temperature of supercritical unit. The research results show that the models built in the paper are useful for solving the problem of the on-line performance computing and the redundancy analysis of key parameters. The modeling examples indicated that reversed modeling method was feasible with high accuracy. Reversed modeling method is a new way for thermal power unit modeling.
引文
[1]张玉铎.系统辨识与建模[M].水利电力出版社.1995:3-27
[2]倪维斗,徐向东,李政,等.热动力系统建模与控制的若干问题[M].科学出版社.1996:1-30
[3]吕崇德,范永胜,蔡瑞忠.我国电站仿真技术进展与建模理论研究[J].中国工程科学.1999,1(1):99-103
[4]邓良才,王广军,陈红.锅炉汽温对象的在线模糊辨识[J].中国电机工程学报.2006,26(18):111-115
[5]赵亮,雎刚,吕剑虹.一种改进的遗传多目标优化算法及其应用[J].中国电机工程学报.2008,28(2):96-102
[6]张小桃,倪维斗,李政,等.基于主元分析与现场数据的过热汽温动态建模研究[J].中国电机工程学报.2005,25(5):131-135
[7]栾秀春,李士勇.基于局部神经网络模型的过热汽温多模型预测控制的研究[J].中国电机工程学报.2004,24(8):190-195
[8]荣雅君,窦春霞,袁石文.过热汽温模糊神经网络预测控制器的设计[J].中国电机工程学报.2003,23(1):177-180
[9]陈立甲,伞冶,王子才,等.锅炉过热器系统机理与神经网络组合建模方法[J].中国电机工程学报.2001,21(1):73-76
[10]卓旭升,周怀春,杨超,等.燃煤汽包炉发电机组的动态建模及其运行数据验证[J].中国电机工程学报.2008,28(17):26-31
[11]田亮,刘鑫屏,赵征,等.一种新的热量信号构造方法及实验研究[J].动力工程.2006,26(4):499-502
[12]李必成,盛赛斌.一种基于数据挖掘的入炉燃料发热量在线智能软诊断方法研究[J].热能动力工程.2007,22(1):25-28
[13]刘吉臻,刘焕章,常太华,等.部分烟气信息下的锅炉煤质分析模型[J].中国电机工程学报.2007,27(14):1-5
[14]韩忠旭,周传心,李丹,等.燃煤发热量软测量技术及其在超临界机组控制系统中的应用[J].中国电机工程学报.2008,28(35):90-95
[15]张怡强.功率(电流)法对磨煤机煤位的控制与应用[J].华东电力.2001, 29(7):48-49
[16]水利电力部华北电力设计院技术处.国外大型磨煤机的应用[M].水利电力出版社.1984:65-89
[17]Kolacz J. Measurement system of the mill charge in grinding ball mill circuits[J]. Minerals Engineering.1997,10(12):1329-1338
[18]朱明,吕震中.基于振动频谱分析的球磨机存煤量测控与优化[D].东南大学博士学位论文.2004:5-39
[19]Kang Enshun, Guo Yugang, Du Yuyuan, et al. Vibration signal processing and analysis in ball mill[C]. Proceedings of the 6th World Congress on Intelligent Control and Automation. Dalian,2006:6690-6693
[20]苏志刚,王培红,于向军,等.中储式制粉系统出力在线监测软测量建模[J].中国电机工程学报.2007,27(29):90-95
[21]王东风,宋之平.基于神经元网络的制粉系统球磨机负荷软测量[J].中国电机工程学报.2001,21(12):97-100
[22]卢勇,徐向东.烟气含氧量软测量新方法研究[J].热能与动力工程.2002,17(6):614-618
[23]韩璞,王东风,翟永杰.基于神经网络的火电厂烟气含氧量软测量[J].信息与控制.2001,30(2):189-192
[24]赵征,曾德良,田亮,等.基于数据融合的氧量软测量研究[J].中国电机工程学报,2005,25(7):7-12
[25]王培红,陈强,董益华,等.数据挖掘及其在电厂SIS中的应用[J].电力系统自动化.2004,28(8):76-79
[26]李蔚,刘长东,盛德仁,等.国内火电厂运行优化系统的现状和发展方向[J].电站系统工程.2004,20(1):59-61
[27]杨勇平,杨昆.火电机组节能潜力诊断理论与应用[J].中国电机工程学报.1998,18(2):131-134
[28]任浩仁,李蔚,盛德仁,等.火电机组变工况下运行指标应达值的分析[J].中国电机工程学报.1999,19(9):50-56
[29]李建强,刘吉臻,张栾英,等.牛成林.基于数据挖掘的电站运行优化应用研究[J].中国电机工程学报.2006,26(20):118-123
[30]张承乾,张国强.系统辨识与参数估计[M].机械工业出版社.1986:1-40
[31]杨戈,吕剑虹,刘志远.基于RBF神经网络的热工过程在线自适应建模算法研 究[J].中国电机工程学报.2004,24(1):191-195
[32]张艳秦,徐向东.用于热力系统建模的基于粗糙集的模糊神经网络[J].清华大学学报(自然科学版).2004,44(8):1084-1086
[33]王志强,徐向东.基于遗传算法的自适应模糊控制器在热力系统中的应用[J].电站系统工程.2003,19(4):42-44
[34]崔大龙,黄鹏,李政.核电汽轮机热力系统主导因素建模方法与故障诊断新思路[J].热力透平.2003,32(3):137-142
[35]Ali Niazi, Sasan Sharifi, Effat Amjadi. Least-squares support vector machines for simultaneous voltammetric determination of lead and tin:A comparison between LS-SVM and PLS in voltammetric data[J]. Electroanalytical Chemistry. 2008,623(1):86-92
[36]Noslen Hernandez, Rudolf Kiralj, Marcia M.C. Ferreira, Isneri Talavera. Critical comparative analysis, validation and interpretation of SVM and PLS regression models in a QSAR study on HIV-1 protease inhibitors[J]. Chemometrics and Intelligent Laboratory Systems.2009,98(1):65-77
[37]Ahmet Baylar, Davut Hanbay, Murat Batan. Application of least square support vector machines in the prediction of aeration performance of plunging overfall jets from weirs[J]. Expert Systems with Applications.2009,36(4):8368-8374
[38]Davut Hanbay, Ahmet Baylar, Murat Batan. Prediction of aeration efficiency on stepped cascades by using least square support vector machines[J]. Expert Systems with Applications.2009,36(3):4248-4252
[39]Nouna Kettaneh, Anders Berglund, Svante Wold. PCA and PLS with very large data sets[J]. Computational Statistics & Data Analysis.2005,48(1):69-85
[40]Jarno Kohonen, Satu-Pia Reinikainen, Kari Aaljoki, Agnar Hoskuldsson. Non-linear PLS approach in score surface[J]. Chemometrics and Intelligent Laboratory Systems.2009,97(2):159-163
[41]U. Thissen, M. Pepers, B. Ustun, W. J. Melssen, L. M. C. Buydens. Comparing support vector machines to PLS for spectral regression applications[J]. Chemometrics and Intelligent Laboratory Systems.2004,73(2):169-179
[42]Philip R.C. Nelson, John F. MacGregor, Paul A. Taylor. The impact of missing measurements on PCA and PLS prediction and monitoring applications[J]. Chemometrics and Intelligent Laboratory Systems.2006,80(1):1-12
[43]Xindi Cai, Nian Zhang, Ganesh K. Venayagamoorthy, Donald C. Wunsch Ⅱ. Time series prediction with recurrent neural networks trained by a hybrid PSO-EA algorithmv[J]. Neurocomputing.2007,70(13-15):2342-2353
[44]Dong Hwa Kim. GA-PSO based vector control of indirect three phase induction motor[J]. Applied Soft Computing.2007,7(2):601-611
[45]C. Praveen, R. Duvigneau, Low cost PSO using metamodels and inexact pre-evaluation:Application to aerodynamic shape design[J]. Computer Methods in Applied Mechanics and Engineering.2009,198(9-12):1087-1096
[46]Ali Allahverdi, Fawaz S. Al-Anzi.A PSO and a Tabu search heuristics for the assembly scheduling problem of the two-stage distributed database application[J]. Computers & Operations Research.2006,33(4):1056-1080
[47]K.M. EL-Naggar, M.R. AlRashidi, A.K. Al-Othman. Estimating the input-output parameters of thermal power plants using PSO[J]. Energy Conversion and Management.2009,50(7):1767-1772
[48]Yoshiyuki Yamashita, Supervised learning for the analysis of process operational data[J]. Computers and Chemical Engineering.2000,24:471-474
[49]张恒喜,郭基联,朱家元,等.小样本多元数据分析方法及应用[M].西北工业大学出版社.2002:12-29
[50]Burges C J C. A tutorial on support vector machines for patten recognition[J]. Data Mining and Knowledge Discovery.1998,2 (2):1-43
[51]张言俊,张科.系统辨识理论及应用[M].国防工业出版社.2003:1-8
[52]钱学森.系统科学、思维科学与人体科学[J].自然杂志.1981,(1):3-7.
[53]钱学森,于景元,戴汝为.一个科学新领域-开放的复杂巨系统及其分析方法[J].自然杂志.1990,13(1):8-12
[54]许东,吴铮.基于MATLAB6.X的系统分析与设计—神经网络[M].西安电子科技大学出版社.2002:32-42
[55]闻新,周露,王丹力,等.MATLAB神经网络应用设计[M].科学出版社.2001:169-232
[56]胡守仁.神经网络技术应用[M].国防科学技术出版社.1993:1-34
[57]边肇祺,张学工.模式识别[M].清华大学出版社.2000:1-7
[58]王惠文.偏最小二乘回归方法及其应用[M].国防工业出版社,1999:13-37
[59]Cortes C, Vapnik V N. Support vector machine[J]. Machine Learning.1995, (20): 273-297
[60]Kok Seng Chua. Efficient computations for large least squares support vector machine classifiers[J]. Pattern recognition.2003, (24):75-80
[61]Suykens J A K, Vandewalle J. Least squares support vector machine classifiers[J]. Neural Processing Letters(S 1370-4621).1999,9(3):293-300
[62]Suykens J A K, Lukas L, Van Dooren P, De Moor B, Vandewalle J. Least square support vector machine classifiers:a large scale algorithm[C]. Proceedings of the European Conference on Circuit Theory and Design. Stresa, Italy,1999, (9): 839-842
[63]宋胜利.智能控制技术概论[M].国防工业出版社,2008:182-233.
[64]李敏强,寇纪凇,林丹,等.遗传算法的基本理论与应用[M].科学出版社,2002:28-60.
[65]张文修,梁怡.遗传算法的数学基础[M].西安交通大学出版社,2003:10-53.
[66]陈丽丽,李蔚,盛德仁,等.火电厂实时监控系统测量数据预测的研究进展[J].电站系统工程2005,21(2):1-4
[67]陈坚红,盛德仁,李蔚,等.汽轮发电机组优化运行专家系统中实时数据的在线预测-验证模型[J].电站系统工程.2001,17(6):375-378
[68]司风琪,徐治皋.基于自联想神经网络和测量数据自校正检验方法[J].中国电机工程学报.2002,22(6):153-155
[69]王楠,律方成.在线监测数据预处理方法的研究[J].高压电器.2003,39(3):57-59,63
[70]董学育.基于人工神经网络的故障诊断方法在电站中的应用研究[D].东南大学博士学位论文.2000:1-20
[71]J.W.Hines, R.E.Uhrig, D.J. Wrest. Use of Autoassociative Neural Network for Signal Validation[J]. Journal of Intelligent and Robotic Systems.1997,143
[72]陈华友.组合预测方法及其应用研究[D].中国科学技术大学博士学位论文.2002:1-7
[73]董学育,胡华进,徐治皋.电站性能分析采样数据的可靠性检验方法[J].动力工程.1998,18(2):16-19,74
[74]E. Eryurek and B. R. Upadhyaya. Sensor validation for power plants using adaptive backpropagation neural network[J]. IEEE Transactions on Nuclear Science.1990,37:1040-1047
[75]Ruy L. Milidiu, Ricardo J. Machado, Raul P.Renteria. Time-series forecasting through vaveltets transformation and a mixture of expert models[J]. Neuro computing.1999,28:145-156
[76]何悦盛,叶春,杨波等.人工神经网络在汽轮机热力参数在线仿真及故障识别中的应用[J].电网技术.2002,26(5):35-38
[77]高建强,赵君友,范晓颖.2种改进的数据处理方法及其在SIS的应用[J].电力系统自动化.2009,33(1):90-92
[78]Enio Kumpinsky. Detection of stealthy outliers in thermal-kinetics modeling [J]. Thermochimica Acta.2008,478(1-2):6-12
[79]王培红,陈强,李琳,等.厂级监控系统中数据校正算法模型研究[J].东南大学学报(自然科学版).2005,35(1):11-15
[80]陈亚菲,高翔等.孤立点检测算法及其在数据流挖掘中的可用性[J].计算机科学.2007,34(10):200-203
[81]李金海.误差理论与测量不确定度评定[M].中国计量出版社.2003:1-9
[82]王春林,周昊,周樟华,等.基于支持向量机的大型电厂锅炉飞灰含碳量建模[J].中国电机工程学报.2005,25(20):72-76
[83]李智,蔡九菊,郭宏.基于神经网络的电站锅炉飞灰含碳量软测量系统[J].节能技术.2004,22(126):6-7
[84]周昊,朱洪波,曾庭华,等.基于人工神经网络的大型电厂锅炉飞灰含碳量建模[J].中国电机工程学报.2002,22(6):96-100
[85]陈强,王培红,李琳,等.电站锅炉飞灰含碳量在线软测量模型算法[J].电力系统自动化.2005,29(2):45-49
[86]方湘涛,叶念渝.基于BP神经网络的电厂锅炉飞灰含碳量预测[J].华中科技大学学报(自然科学版).2003,31(12):75-77
[87]李建强.基于数据挖掘的电站运行优化理论研究与应用[D].华北电力大学博士学位论文.2006:7-15
[88]章元,朱尔一,李静,等.模拟退火算法与遗传算法结合用于变量筛选[J].分析化学.1999,27(10):1131-1135
[89]吴晓华,陈德钊.化学计量学非线性偏最小二乘算法进展评述[J].分析化学评述与进展.2004,(32):534-540
[90]丁世飞,靳奉祥,史忠植.基于PLS的信息特征压缩算法[J].计算机辅助设计与图形学学报.2005,17(2):368-371
[91]Barros A S, Rutledge D N. Genetic algorithms applied to the selection of principal components[J]. Chemometrics and intelligent laboratory systems.1998, 40(1):65-82
[92]李勇,邵诚.一种新的灰关联分析算法在软测量中的应用[J].自动化学报.2006,32(2):311-317
[93]张雪平,殷国富.基于层次灰色关联的产品绿色度评价研究[J].中国电机工程学报.2005,25(17):78-82
[94]付国仙.灰关联-反向建模方法在球磨机料位软测量中的应用[D].华北电力大学硕士论文.2010:49-54
[95]刘素海,余岳溪.关于煤耗在线计算的一些难点分析[J].广东电力.2009,22(7):46-49
[96]冯伟忠.大机组主蒸汽流量测量刍议[J].华东电力.2000,(12):14-17
[97]何军民,李明.湘潭电厂300 MW机组主蒸汽流量计算的修改[J].湖南电力.2005,25(5):17-19
[98]孙永平,朱梅芳,王敏.北仑电厂2号机组DAS系统主蒸汽流量计算模型的修改[J].浙江电力.2000,(3):14-16
[99]张春发,赵宁,王惠杰.一种汽轮机组排汽干度的在线软测量方法[J].中国电机工程学报.2008,28(26):1-6
[100]闫顺林,王俊有,李太兴,等.汽轮机低压缸排汽焓在线计算新模型的研究及应用[J].华东电力.2007,35(4):84-86
[101]闫顺林,徐鸿,李永华,等.汽轮机排汽焓动态在线计算模型的研究[J].动力工程.2008,28(2):181-185
[102]吴葳,洪炳熔.自由浮游空间机器人捕捉目标的运动规划研究[C].中国第五届机器人学术会议论文集.哈尔滨,1997:75-80
[103]李顺良,王俊有.汽轮机排汽焓在线计算方法的研究及应用[J].东北电力技术.2007,(6):13-16
[104]高俊如,丁光彬,孟鑫,等.利用层次径向基神经网络的汽轮机排汽焓计算[J].动力工程.2005,25(4):466-469
[105]蔡杰进,马晓茜.支持向量机在电站汽轮机排汽焓在线预测中的应用[J].电力系统自动化.2006,30(18):77-81
[106]张利平,王铁生.基于免疫原理的RBF神经网络模型在汽轮机排汽焓计算中的应用[J].汽轮机技术.2008,50(5):347-349
[107]荣海娜,张葛祥,金炜东.系统辨识中支持向量机核函数及其参数的研究[J].系统仿真学报.2006,18(11):3204-3226
[108]阎维平,李钧,李加护, 等.锅炉飞灰采样装置结露堵灰的原因分析及其对策[J].动力工程.2008,28(3):345-349
[109]Vijiaya V. Kantubhukta, Mohamed Abdelrahman. A Feasibility Study on Using Neural Networks in Performance Analysis of Coal-fired Power Plant[C]. System theory, Proceedings of the thirty-sixth southeastern symposiums,2004,541-545
[110]柳京玉,郑世津.锅炉高温受热面炉外管壁温度检测与分析[J].江苏电机工程.2004,23(2):9-11
[111]王孟浩,王衡,郑民牛.过热器再热器炉外壁温测点结构和报警温度设定值的探讨[J].热力发电.2006,(07):29-33
[112]吴迪.大型电站锅炉汽温、壁温监测存在的问题及建议[J].广东电力.2005,18(10):63-65
[113]匡江红,杨庆娣.电站锅炉过热器壁温在线监测系统的开发研制[J].电力科学与工程.2004,(2):66-68
[114]陈鸿伟,陈聪,高建强,等.锅炉金属壁温在线监测系统模型的开发与实现[J].中国电机工程学报.2006,26(21):125-129
[115]顾山.超临界机组锅炉过热器和再热器炉内壁温在线监测[J].热力发电.2008,37(4):47-49
[116]高杰.电站锅炉过热器壁温的改进计算方法[J].黑龙江水专学报.2008,35(2):63-65
[117]杨红权,陶丽,丁士发,等.电站锅炉炉内壁温测量技术的研究[J].动力工程.2009,29(8):717-721
[118]王孟浩,王衡,陈朝柱,等.高参数大容量电站锅炉过热器再热器的超温问题和炉内壁温在线监测[J].锅炉技术.2009,40(1):5-11
[119]樊泉桂.超临界锅炉的汽温特性及控制[J].动力工程.2007,27(1):34-37.