基于即时学习策略的电厂热工参数预测模型及应用研究
|
摘要
我国电力生产企业在发电效率和能源利用率等方面相对较低,其中一个重要因素就在于许多重要的过程参数(如烟气含氧量、球磨机负荷等)和经济指标难以实现在线实时检测,严重制约了火电厂过程控制和优化运行。针对电力生产过程普遍存在的非线性、强耦合、工况范围变化大等特性,本文采用基于分治策略框架下的局部建模方法进行烟气含氧量、球磨机负荷等热工参数的软测量研究,主要研究工作及创新性如下：
(1)基于即时学习策略的改进支持向量机预测建模方法
针对锅炉燃烧过程、制粉系统等具有的非线性、强耦合、工况范围广等特点,而采用基于数据驱动的全局建模方法(如神经网络、支持向量机等)又难以建立一个有效的全局软测量模型,因此,本文依据统计的局部学习理论,提出一种基于即时学习策略的改进SVM在线建模方法。
在分析历史数据样本相似性的基础上,提出一种基于距离和角度信息的相似样本选取方法,构造即时学习算法的建模邻域,以提高相似样本的选取精度。由于建模邻域中相似样本规模较小,为此采用具有小样本建模能力的支持向量机模型作为局部模型,并采用改进粒子群优化算法对支持向量机建模方法中3个关键参数进行寻优,以获得最佳的局部预测模型,从而提高模型的预测精度。由于基于即时学习策略的局部预测方法是在线进行的,基于对样本数据集的检索精度和检索效率两方面的考虑,采用加权模糊C均值聚类算法对数据样本集进行聚类,利用两步搜索策略进行当前输入样本数据相似数据集的选取,同时提出一种样本数据集的在线更新策略。
(2)局部预测模型在烟气含氧量预测中的应用研究
通过对锅炉燃烧过程机理的深入分析,确定出烟气含氧量预测模型的初始辅助变量集,并依据锅炉燃烧过程中的大量历史数据,对相关过程参数的数据样本进行3σ异常检测和归一化处理,利用灰色关联分析方法对辅助变量集进行优化筛选,获得最终的辅助变量。利用本文提出的基于即时学习策略的改进SVM预测建模方法建立烟气含氧量预测模型,由于本文方法在本质上具有在线自适应性能力,能更好的适应锅炉燃烧过程中的不同工况。仿真分析说明相对于标准BP神经网络和标准SVM预测模型,基于即时学习策略的改进SVM预测模型具有更好的预测性能,虽然算法的计算开销有所增加,但能够满足锅炉燃烧过程烟气含氧量预测的实时性要求,并及时为生产操作提供参考。
(3)局部预测模型在球磨机负荷预测中的应用研究
通过对制粉系统运行过程机理的深入分析,确定出球磨机负荷预测模型的初始辅助变量集,并依据制粉过程中的大量历史数据,对相关过程参数的数据样本进行3σ和小波分析异常检测及归一化处理,采用灰色关联分析方法对辅助变量集进行优化筛选,获得最终的辅助变量。利用本文提出的基于即时学习策略的改进SVM预测建模方法建立球磨机负荷预测模型,由于本文方法具有较好的在线自适应性能力,能更好的适应制粉系统运行过程中的不同工况。仿真分析说明相对于标准BP神经网络和标准SVM预测模型,基于即时学习策略的改进SVM球磨机负荷预测模型具有更好的预测性能,虽然算法的计算开销有所增加,但能够满足制粉系统磨机负荷预测的实时性要求,并及时为生产操作提供参考。
The power generation efficiency and energy utilization of thermal power generation are relatively low, one of the main influence factor is that lots of important process parameters (such as flue gas oxygen content, ball mill load) and economic indicators are difficult to be real-time measured online, which seriously restricted the process control and optimal operation in thermal power plant. Because the power production process has the characteristic of nonlinear, strong coupling and large operation range, based on the local modeling algorithm of divide-and-conquer principle, the soft-sensing models of flue gas oxygen content and ball mill load are studied in this paper. The main research work and innovative achievements are listed as follows:
(1) The improved support vector machine modeling method based on just-in-time learning
Because the boiler combustion process and coal pulverizing system have the characteristic of nonlinear, strong coupling and large operation range, it is difficult to establish an effective global soft-sensor model using global modeling algorithm (such as neural network, support vector machine) based on data-driven. Therefore, by analyzing local modeling theory, an improved support vector machine modeling method based on just-in-time learning is proposed in paper.
By analyzing the similarity of historical data sample, a similar sample selection method based on the distance and angle information is proposed. The interested observations in a local neighborhood of the query point are selected by the selection method, and the selected precision can be improved. Because the local neighborhood size of the query point is small, the support vector machine modeling method is selected to establish the local model, and the three key parameters in support vector machine model are optimized by using an improved particle swarm optimization algorithm, so the prediction accuracy can be effectively improved. In order to balance the retrieval accuracy and the retrieval efficiency of the data sample set, a two-step searching strategy based on weighted fuzzy C-means clustering algorithm is offered in this paper. A kind of updating strategy for data sample set is also presented.
(2) Application research of the local prediction model in the flue gas oxygen content
The initial instrumental variables of the flue gas oxygen content prediction model are determined by analyzing the process mechanism of the boiler combustion process. Then based on lots of historical data and data preprocessing (such as 3σabnormal data recognizing and data normalization), the final instrumental variables are obtained by using gray relational analysis method to optimize the initial instrumental variables and the flue gas oxygen content prediction model is established by utilizing the improved SVM modeling method based on just-in-time learning strategy which is proposed in this paper. Because the proposed modeling method has the essential ability of online self-adaptive, so the obtained model can have better adaptation to different operating conditions. Simulation results show that compared with the standard BP neural network and the standard SVM prediction model, although the computing cost is increased, the proposed prediction model has better prediction performance and can satisfy the real-time requirements for the flue gas oxygen content in the boiler combustion process.
(3) Application research of the local prediction model in the ball mill load
The initial instrumental variables of the ball mill load prediction model are determined by analyzing the process mechanism of the coal pulverizing system. Then based on lots of historical data and data preprocessing (such as 3σand wavelet analysis abnormal data recognizing, and data normalization), the final instrumental variables are obtained by using gray relational analysis method to optimize the initial instrumental variables and the ball mill load prediction model is established by utilizing the improved SVM modeling method based on just-in-time learning strategy which is proposed in this paper. Because the proposed modeling method has good online self-adaptive ability, so the obtained model can have better adaptation to different operating conditions. Simulation results show that compared with the standard BP neural network and the standard SVM prediction model, although the computing cost is increased, the proposed prediction model has better prediction performance and can satisfy the real-time requirements for the ball mill load in the coal pulverizing system.
(1)基于即时学习策略的改进支持向量机预测建模方法
针对锅炉燃烧过程、制粉系统等具有的非线性、强耦合、工况范围广等特点,而采用基于数据驱动的全局建模方法(如神经网络、支持向量机等)又难以建立一个有效的全局软测量模型,因此,本文依据统计的局部学习理论,提出一种基于即时学习策略的改进SVM在线建模方法。
在分析历史数据样本相似性的基础上,提出一种基于距离和角度信息的相似样本选取方法,构造即时学习算法的建模邻域,以提高相似样本的选取精度。由于建模邻域中相似样本规模较小,为此采用具有小样本建模能力的支持向量机模型作为局部模型,并采用改进粒子群优化算法对支持向量机建模方法中3个关键参数进行寻优,以获得最佳的局部预测模型,从而提高模型的预测精度。由于基于即时学习策略的局部预测方法是在线进行的,基于对样本数据集的检索精度和检索效率两方面的考虑,采用加权模糊C均值聚类算法对数据样本集进行聚类,利用两步搜索策略进行当前输入样本数据相似数据集的选取,同时提出一种样本数据集的在线更新策略。
(2)局部预测模型在烟气含氧量预测中的应用研究
通过对锅炉燃烧过程机理的深入分析,确定出烟气含氧量预测模型的初始辅助变量集,并依据锅炉燃烧过程中的大量历史数据,对相关过程参数的数据样本进行3σ异常检测和归一化处理,利用灰色关联分析方法对辅助变量集进行优化筛选,获得最终的辅助变量。利用本文提出的基于即时学习策略的改进SVM预测建模方法建立烟气含氧量预测模型,由于本文方法在本质上具有在线自适应性能力,能更好的适应锅炉燃烧过程中的不同工况。仿真分析说明相对于标准BP神经网络和标准SVM预测模型,基于即时学习策略的改进SVM预测模型具有更好的预测性能,虽然算法的计算开销有所增加,但能够满足锅炉燃烧过程烟气含氧量预测的实时性要求,并及时为生产操作提供参考。
(3)局部预测模型在球磨机负荷预测中的应用研究
通过对制粉系统运行过程机理的深入分析,确定出球磨机负荷预测模型的初始辅助变量集,并依据制粉过程中的大量历史数据,对相关过程参数的数据样本进行3σ和小波分析异常检测及归一化处理,采用灰色关联分析方法对辅助变量集进行优化筛选,获得最终的辅助变量。利用本文提出的基于即时学习策略的改进SVM预测建模方法建立球磨机负荷预测模型,由于本文方法具有较好的在线自适应性能力,能更好的适应制粉系统运行过程中的不同工况。仿真分析说明相对于标准BP神经网络和标准SVM预测模型,基于即时学习策略的改进SVM球磨机负荷预测模型具有更好的预测性能,虽然算法的计算开销有所增加,但能够满足制粉系统磨机负荷预测的实时性要求,并及时为生产操作提供参考。
The power generation efficiency and energy utilization of thermal power generation are relatively low, one of the main influence factor is that lots of important process parameters (such as flue gas oxygen content, ball mill load) and economic indicators are difficult to be real-time measured online, which seriously restricted the process control and optimal operation in thermal power plant. Because the power production process has the characteristic of nonlinear, strong coupling and large operation range, based on the local modeling algorithm of divide-and-conquer principle, the soft-sensing models of flue gas oxygen content and ball mill load are studied in this paper. The main research work and innovative achievements are listed as follows:
(1) The improved support vector machine modeling method based on just-in-time learning
Because the boiler combustion process and coal pulverizing system have the characteristic of nonlinear, strong coupling and large operation range, it is difficult to establish an effective global soft-sensor model using global modeling algorithm (such as neural network, support vector machine) based on data-driven. Therefore, by analyzing local modeling theory, an improved support vector machine modeling method based on just-in-time learning is proposed in paper.
By analyzing the similarity of historical data sample, a similar sample selection method based on the distance and angle information is proposed. The interested observations in a local neighborhood of the query point are selected by the selection method, and the selected precision can be improved. Because the local neighborhood size of the query point is small, the support vector machine modeling method is selected to establish the local model, and the three key parameters in support vector machine model are optimized by using an improved particle swarm optimization algorithm, so the prediction accuracy can be effectively improved. In order to balance the retrieval accuracy and the retrieval efficiency of the data sample set, a two-step searching strategy based on weighted fuzzy C-means clustering algorithm is offered in this paper. A kind of updating strategy for data sample set is also presented.
(2) Application research of the local prediction model in the flue gas oxygen content
The initial instrumental variables of the flue gas oxygen content prediction model are determined by analyzing the process mechanism of the boiler combustion process. Then based on lots of historical data and data preprocessing (such as 3σabnormal data recognizing and data normalization), the final instrumental variables are obtained by using gray relational analysis method to optimize the initial instrumental variables and the flue gas oxygen content prediction model is established by utilizing the improved SVM modeling method based on just-in-time learning strategy which is proposed in this paper. Because the proposed modeling method has the essential ability of online self-adaptive, so the obtained model can have better adaptation to different operating conditions. Simulation results show that compared with the standard BP neural network and the standard SVM prediction model, although the computing cost is increased, the proposed prediction model has better prediction performance and can satisfy the real-time requirements for the flue gas oxygen content in the boiler combustion process.
(3) Application research of the local prediction model in the ball mill load
The initial instrumental variables of the ball mill load prediction model are determined by analyzing the process mechanism of the coal pulverizing system. Then based on lots of historical data and data preprocessing (such as 3σand wavelet analysis abnormal data recognizing, and data normalization), the final instrumental variables are obtained by using gray relational analysis method to optimize the initial instrumental variables and the ball mill load prediction model is established by utilizing the improved SVM modeling method based on just-in-time learning strategy which is proposed in this paper. Because the proposed modeling method has good online self-adaptive ability, so the obtained model can have better adaptation to different operating conditions. Simulation results show that compared with the standard BP neural network and the standard SVM prediction model, although the computing cost is increased, the proposed prediction model has better prediction performance and can satisfy the real-time requirements for the ball mill load in the coal pulverizing system.
引文
[1]姚之驹.2006年中国发电设备行业回顾及发展分析.电器工业,2007,1：13-22.
[2]黄其励.我国清洁高效燃煤发电技术.华电技术,2008,30(3)：1-8.
[3]李波.火力发电机组超临界化的发展趋势.中国电机工程学会2004年学术年会论文集,北京：中国电力出版社,2004,11-13.
[4]李振中,吴少华,黄其励.超超临界燃煤发电机组的技术选择与产业化发展.中国电机工程学会2004年学术年会论文集,北京：中国电力出版社,2004,14-18.
[5]焦庆丰,朱光明,李明,等.我国火电厂能耗现状及节能潜力分析.湖南电力,2008,28(1)：20-23.
[6]程启明,郭瑞青,杜许峰,等.火电厂烟气含氧量测量的现状与发展.电站系统工程,2008,24(6)：1-4.
[7]曹静,唐贵基.钢球磨煤机负荷测量的研究.矿山机械,2007,35(10)：29-31.
[8]B. Joseph, C. B. Brosilow. Inferential control of process:part Ⅰ. Steady state analysis and design, AICHE Journal,1978,24(3):485-492.
[9]B. Joseph, C. B. Brosilow. Inferential control of process:part Ⅱ. The structure and dynamics of inferential control system, AICHE Journal,1978,24(3): 492-500.
[10]B. Joseph, C. B. Brosilow. Inferential control of process:part Ⅲ. Construction of optimal and suboptimal dynamic estimators, AICHE Journal,1978,24(3): 500-509.
[11]M. Kano, Y. Nakagawa. Data-based process monitoring, process control, and quality improvement: recent developments and applications in steel industry. Computers and Chemical Engineering,2008,32:12-24.
[12]郜传厚,渐令,陈积明,等.复杂高炉炼铁过程的数据驱动建模及预测算法.自动化学报,2009,35(6)：725-730.
[13]王春生,吴敏,曹卫华,等.铅锌烧结配料过程的智能集成建模与综合优化控制.自动化学报,2009,35(5)：605-612.
[14]杨强大,王福利,常玉清.基于阶段辨识的诺西肽发酵过程菌体浓度软测量.控制理论与应用,2009,26(9)：1026-1030.
[15]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.信息与控制,2001,30(2)：189-192.
[16]卢勇,徐向东.烟气含氧量软测量新方法研究.热能动力工程,2002,17(6)：614-617.
[17]赵征,曾德良,田亮,刘吉臻.基于数据融合的氧量软测量研究.中国电机工程学报,2005,25(7)：7-12.
[18]熊志化,杨海滨,吴云峰,邵惠鹤.基于稀疏高斯过程的热力参数软测量.中国电机工程学报,2005,25(8)：130-133.
[19]熊志化,张卫庆,赵瑜,邵惠鹤.基于混合高斯过程的多模型热力参数测量软测量.中国电机工程学报,2005,25(7)：30-33,40.
[20]郑德忠,何群.基于新型广义预测算法的含氧量软测量研究.电子学报,2006,34(12)：2255-2258.
[21]王勇,刘吉臻,刘向杰,谭文.基于最小二乘支持向量机的软测量建模及在电厂烟气含氧量测量中的应用.微计算机信息,2006,22(10S)：241-243,290.
[22]乔弘,周黎辉,潘卫华,陈丽.基于参数自适应SVR的火电厂热工参数软测量.华北电力大学学报,2008,35(2)：94-97.
[23]王宏志,陈帅,侍洪波.基于最小二乘支持向量机和PSO算法的电厂烟气含氧量软测量.热力发电,2008,37(3)：35-38.
[24]付忠广,靳涛,周丽君,戈志华,郑玲,杨勇平.复杂系统反向建模方法及偏最小二乘法建模应用研究.中国电机工程学报,2009,29(2)：25-29.
[25]赵永磊,郑德忠.一种基于直接广义预测的烟气含氧量软测量方法研究.传感技术学报.2009,22(8)：1213-1216.
[26]王东风,宋之平.基于神经网络的制粉系统球磨机负荷软测量.中国电机工程学报,2001,21(12)：97-99,104.
[27]王东风,韩璞.基于RBF神经网络的球磨机负荷软测量.仪器仪表学报,2002,23(3S)：311-312,317.
[28]吕立华.复杂工业系统基于小波网络与鲁棒估计建模方法研究：[博士学位论文].浙江杭州：浙江大学,2001.
[29]王颖洁.钢球磨煤机料位的软测量及其动态过程建模与控制.[硕士学位论文].江苏南京：东南大学,2005.
[30]苏志刚,于向军,吕震中,赵刚.灰色软测量在球磨机料位检测中的应用.热能动力工程,2006,21(6)：578-581.
[31]司刚全,曹晖,王靖程,张彦斌,贾立新.基于复合式神经网络的火电厂简式钢球磨煤机负荷软测量.热力发电,2007,36(5)：64-67.
[32]王雷,于向军,吕震中.球磨机料位软测量及其低能耗高效运行研究.能源研究与利用,2007,(5)：16-19.
[33]司刚全,曹晖,张彦斌,马西奎.基于ANFIS的火电厂磨机负荷检测的软测量模型.2007中国仪器仪表与测控技术交流大会论文集,四川成都,2007,2,1-5.
[34]司刚全,曹晖,张彦斌,贾立新.一种基于密度加权的最小二乘支持向量机稀疏化算法.西安交通大学学报,2009,43(10)：11-15.
[35]肖应旺,徐保国.改进PCA在发酵过程监测与故障诊断中的应用.控制与决策,2005,20(5)：571-574.
[36]赵忠盖,刘飞.工业过程核概率主元模型建立及监控指标.华东理工大学学报(自然科学版),2006,32(7)：864-867.
[37]颜学峰,余娟,钱锋.基于自适应偏最小二乘回归的初顶石脑油干点软测量.化工学报,2005,56(8)：1511-1515.
[38]张学莲,胡立生,曹广益.基于过程数据的动态PLS建模.系统仿真学报,2008,20(10)：2686-2688,2692.
[39]许传龙,赵延军,黄健,孔明,汤光华,王式民.基于人工神经网络的固相质量流量软测量研究.计量学报,2006,27(3)：246-249.
[40]袁仲泉,尹兵太,尚海燕,朱群雄.基于PCA-BP算法的PTA溶剂脱水塔软测量模型.计算机与应用化学,2006,23(7)：623-626.
[41]曹先庆.基于PCA-改进算法BP神经网络的软测量技术.仪器仪表学报,2006,27(6S)：2342-2343.
[42]孟科,李绍军,钱锋.基于模糊免疫算法的溶剂脱水塔软测量.计算机工程与应用,2007,43(12)：228-230,234.
[43]许勇刚,杨辉.基于’RBF网络的稀土萃取过程组分含量软测量.稀土,2007,28(5)：19-22.
[44]徐余法,俞金寿.合成醋酸乙烯的空时得率和催化剂选择性软测量.上海交通大学学报,2007,41(8)：1339-1342.
[45]闭治跃,王庆丰,唐建中.基于径向基函数神经网络的挖泥船排泥管道泥浆浓度软测量模型研究.传感技术学报,2007,20(7)：1630-1634.
[46]常青,杨捷,裴洪卿.基于RBF神经网络的丙烯腈收率软测量方法.东南大学学报(自然科学版),2006,36(S1)：194-196.
[47]李勇,邵诚.灰色软测量在介质填充率检测中的应用研究.中国矿业大学学报,2006,35(4)：549-554.
[48]张勇,王介生.基于PCA-RBF神经网络的浮选过程软测量建模.南京航空航天大学学报,2006,38(S1)：116-119.
[49]王华秋,曹长修,李梁.基于并行核径向基神经网络的软测量与控制模型. 仪器仪表学报,2006,27(10)：1218-1223.
[50]谢承利,陆继东,沈凯,曾勇.基于焚烧运行参数的垃圾热值软测量模型.燃烧科学与技术,2007,13(1)：81-85.
[51]刘载文,王正祥,王小艺,杨斌,程志强.过程神经元网络学习算法及软测量方法的研究.系统仿真学报,2007,19(7)：1456-1459.
[52]杨马英,周芳芹,李军.基于Elman神经网络的城市污水处理水质参数软测量.东南大学学报(自然科学版),2006,36(S1)：119-123.
[53]吕强,俞金寿.量子神经网络软测量模型及应用.系统仿真学报,2007,19(24)：5696-5699.
[54]杨希,钱锋,张兵.基于核函数主元分析的SVM建模方法及应用.华东理工大学学报(自然科学版),2007,33(2)：259-262,270.
[55]桑海峰,王福利,何大阔,张大鹏,何建勇.基于多支持向量机的诺西肽发酵中菌体浓度软测量.系统仿真学报,2006,18(7)：1983-1986.
[56]段莹,范全义,熊智华.基于FA-SVR的丙酮精制产品质量软测量模型.微计算机信息,2007,23(1)：1-3.
[57]徐晔,杜文莉,钱锋.基于核主元分析和最小二乘支持向量机的软测量建模.系统仿真学报,2007,19(17)：3873-3875,3918.
[58]周林成,杨慧中.基于KPCA和最小二乘支持向量机的软测量建模.计算机仿真,2008,25(10)：94-97.
[59]陈佳,颜学峰,钱锋.基于贝叶斯学习的关联向量机及其在软测量中的应用.华东理工大学学报(自然科学版),2007,33(1)：115-119.
[60]朱建鸿,丁健,杨惠中,姜永森.贝叶斯回归支持向量机的软测量建模方法.南京航空航天大学学报,2006,38(S1)：136-138.
[61]袁平,毛志忠,王福利.基于递阶T-S模糊系统的软测量建模方法.东北大学学报(自然科学版),2006,27(10)：1071-1074.
[62]刘瑞兰,苏宏业,褚健.模糊神经网络的混合学习算法及其软测量建模.系统仿真学报,2005,17(12)：2878-2881.
[63]陈国初,徐余法,俞金寿.基于粒子群模糊神经网络的丙烯腈收率软测量建模.系统仿真学报,2007,19(23)：5370-5372.
[64]彭晓波,桂卫华,李勇刚,王凌云,陈勇.基于动态T-S递归模糊神经网络的闪速熔炼过程参数软测量.仪器仪表学报,2008,29(10)：2029-2033.
[65]程健,郭一楠,钱建生.跳汰机床层松散状况软测量建模方法研究.中国矿业大学学报,2006,35(5)：591-595.
[66]王介生,张勇.基于ANFIS的软测量模型再浮选中的应用.合肥工业大学学 报(自然科学版),2006,29(11)：1365-1369.
[67]贾润达,毛志忠,常玉清.湿法冶金萃取组分含量混合建模方法的研究.仪器仪表学报,2009,30(2)：267-271.
[68]袁平,毛志忠,王福利.基于微分进化的钢水温度预报混合模型.东北大学学报(自然科学版),2009,30(8)：1083-1086.
[69]王雅琳,桂卫华,阳春华,谢永芳,宋海鹰.基于有限信息的铜吹炼动态过程智能集成建模.控制理论与应用,2009,26(8)：860-866.
[70]H Y Qi, X G Zhou, L H Liu, W K Yuan. A hybrid neural network-first principles model for fixed-bed reactor. Chemical Engineering Science,1999,54: 2521-2526.
[71]王瑞敏,曹广益,朱新坚.PEMFC混合模型的建模及其稳态仿真.系统仿真学报,2008,20(5)：1299-1302.
[72]李向阳,朱学峰,刘焕彬.间歇制浆蒸煮过程的混合建模方法研究.中国造纸学报,2001,16(2)：24-28.
[73]桑海峰,范玮琦,王福利,何大阔.基于机理知识与最小二乘支持向量机的诺西肽发酵过程混合建模方法.系统仿真学报,2008,20(2)：468-472.
[74]T Takagi, M Sugeno. Fuzzy identification of systems and its applications to modeling and control. IEEE Transactions on Systems, Man, Cybernetics,1985, 15(1):116-132.
[75]张永,邢宗义,向峥嵘,胡维礼.基于Pareto协同进化算法的TS模糊模型设计.控制与决策,2006,21(12)：1332-1337,1342.
[76]肖健梅,王锡淮.基于微粒群优化的船舶柴油机T-S模糊模型.哈尔滨工程大学学报,2007,28(5)：532-535.
[77]T A Johansen, B A Foss. A NARMAX model representation for adaptive control based on local models. Modeling, Identification and Control,1992,13(1):25-39.
[78]R Murry-Smith, T A Johansen. Multiple model approaches to modeling and control. London: Taylor and Francis.
[79]T A Johansen, B A Foss. Operating regime based process modeling and identification. Computes and Chemical Engineering,1997,21(2):159-176.
[80]A N Venkat, P Vijaysai, R D Gudi. Identification of complex nonlinear process based on fuzzy decomposition of steady state space. Journal of Process Control, 2003,13(6):473-488.
[81]C Cheng, M S Chiu. A new data-based methodology for nonlinear process modeling. Chemical Engineering Science,2004,59(13):2801-2810.
[82]孙维,王伟.基于即时学习算法非线性系统多模型自适应控制.大连理工大学学报,2002,42(5)：611-615.
[83]崔宇,薛美盛.局部LSSVM方法在飞灰含碳量软测量中的应用.仪表技术,2009,16(5)：62-64.
[84]V N Vapnik. An overview of statistic learning theory. IEEE Transactions on Neural Networks,1999,10(5):988-999.
[85]J Kennedy, R Eberhart. Particle swarm optimization. Proceedings of IEEE Ineternational Conference on Neural Networks. Piscataway, NJ: IEEE Press, 1995:1942-1948.
[86]Y Shi, R Eberhart. A modified particle swarm optimizer. Proceedings of IEEE Ineternational Conference on Evolutionary Computation. Piscataway, NJ: IEEE Press,1998:69-73.
[87]H Y Fan. A modification to particle swarm optimization algorithm. Engineering Computations,2002,19(8):970-989.
[88]赵娜,张伏生,魏平,刘学.基于改进多粒子群算法的电力系统无功优化.西安交通大学学报,2006,40(4)：463-467.
[89]R Storn, K Price. Differential evolution: a simple and efficient adaptive scheme for global optimization over continuous spaces, tr-95-012. Berkeley: International Computer Science Institute,1995.
[90]张航,王伟,郑玲,李丹丹,熊富强.一种基于密度聚类的小生境差分进化算法.计算机工程与应用,2008,44(23)：42-45.
[91]高新波,李洁,姬红兵.基于加权模糊c均值聚类与统计检验指导的多阈值图像自动分割算法.电子学报,2004,32(4)：661-664.
[92]马强.循环流化床锅炉燃烧系统建模与控制系统研究[D].[博士学位论文].河北保定：华北电力大学,2009.
[93]王颖洁.钢球磨煤机料位的软测量及其动态过程建模与控制.[硕士学位论文].江苏南京：东南大学,2005.
[94]王冬风.多变量智能控制在电厂制粉系统中的应用研究.[博士学位论文].河北保定：华北电力大学,2001.
[95]H G Stark. Continuous wavelet transform and continuous multiscale analysis. Journal of Mathematical Analysis and Applications,1992,169(1):179-196.
[96]彭小奇,宋彦坡,唐英.基于小波分析的异常样本处理.信息与控制,2005,34(6)：676-679.
[97]宋彦坡.数据挖掘技术及其在铜转炉吹炼过程优化中的应用.[硕士学位论 文].湖南长沙：中南大学,2005.
[2]黄其励.我国清洁高效燃煤发电技术.华电技术,2008,30(3)：1-8.
[3]李波.火力发电机组超临界化的发展趋势.中国电机工程学会2004年学术年会论文集,北京：中国电力出版社,2004,11-13.
[4]李振中,吴少华,黄其励.超超临界燃煤发电机组的技术选择与产业化发展.中国电机工程学会2004年学术年会论文集,北京：中国电力出版社,2004,14-18.
[5]焦庆丰,朱光明,李明,等.我国火电厂能耗现状及节能潜力分析.湖南电力,2008,28(1)：20-23.
[6]程启明,郭瑞青,杜许峰,等.火电厂烟气含氧量测量的现状与发展.电站系统工程,2008,24(6)：1-4.
[7]曹静,唐贵基.钢球磨煤机负荷测量的研究.矿山机械,2007,35(10)：29-31.
[8]B. Joseph, C. B. Brosilow. Inferential control of process:part Ⅰ. Steady state analysis and design, AICHE Journal,1978,24(3):485-492.
[9]B. Joseph, C. B. Brosilow. Inferential control of process:part Ⅱ. The structure and dynamics of inferential control system, AICHE Journal,1978,24(3): 492-500.
[10]B. Joseph, C. B. Brosilow. Inferential control of process:part Ⅲ. Construction of optimal and suboptimal dynamic estimators, AICHE Journal,1978,24(3): 500-509.
[11]M. Kano, Y. Nakagawa. Data-based process monitoring, process control, and quality improvement: recent developments and applications in steel industry. Computers and Chemical Engineering,2008,32:12-24.
[12]郜传厚,渐令,陈积明,等.复杂高炉炼铁过程的数据驱动建模及预测算法.自动化学报,2009,35(6)：725-730.
[13]王春生,吴敏,曹卫华,等.铅锌烧结配料过程的智能集成建模与综合优化控制.自动化学报,2009,35(5)：605-612.
[14]杨强大,王福利,常玉清.基于阶段辨识的诺西肽发酵过程菌体浓度软测量.控制理论与应用,2009,26(9)：1026-1030.
[15]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.信息与控制,2001,30(2)：189-192.
[16]卢勇,徐向东.烟气含氧量软测量新方法研究.热能动力工程,2002,17(6)：614-617.
[17]赵征,曾德良,田亮,刘吉臻.基于数据融合的氧量软测量研究.中国电机工程学报,2005,25(7)：7-12.
[18]熊志化,杨海滨,吴云峰,邵惠鹤.基于稀疏高斯过程的热力参数软测量.中国电机工程学报,2005,25(8)：130-133.
[19]熊志化,张卫庆,赵瑜,邵惠鹤.基于混合高斯过程的多模型热力参数测量软测量.中国电机工程学报,2005,25(7)：30-33,40.
[20]郑德忠,何群.基于新型广义预测算法的含氧量软测量研究.电子学报,2006,34(12)：2255-2258.
[21]王勇,刘吉臻,刘向杰,谭文.基于最小二乘支持向量机的软测量建模及在电厂烟气含氧量测量中的应用.微计算机信息,2006,22(10S)：241-243,290.
[22]乔弘,周黎辉,潘卫华,陈丽.基于参数自适应SVR的火电厂热工参数软测量.华北电力大学学报,2008,35(2)：94-97.
[23]王宏志,陈帅,侍洪波.基于最小二乘支持向量机和PSO算法的电厂烟气含氧量软测量.热力发电,2008,37(3)：35-38.
[24]付忠广,靳涛,周丽君,戈志华,郑玲,杨勇平.复杂系统反向建模方法及偏最小二乘法建模应用研究.中国电机工程学报,2009,29(2)：25-29.
[25]赵永磊,郑德忠.一种基于直接广义预测的烟气含氧量软测量方法研究.传感技术学报.2009,22(8)：1213-1216.
[26]王东风,宋之平.基于神经网络的制粉系统球磨机负荷软测量.中国电机工程学报,2001,21(12)：97-99,104.
[27]王东风,韩璞.基于RBF神经网络的球磨机负荷软测量.仪器仪表学报,2002,23(3S)：311-312,317.
[28]吕立华.复杂工业系统基于小波网络与鲁棒估计建模方法研究：[博士学位论文].浙江杭州：浙江大学,2001.
[29]王颖洁.钢球磨煤机料位的软测量及其动态过程建模与控制.[硕士学位论文].江苏南京：东南大学,2005.
[30]苏志刚,于向军,吕震中,赵刚.灰色软测量在球磨机料位检测中的应用.热能动力工程,2006,21(6)：578-581.
[31]司刚全,曹晖,王靖程,张彦斌,贾立新.基于复合式神经网络的火电厂简式钢球磨煤机负荷软测量.热力发电,2007,36(5)：64-67.
[32]王雷,于向军,吕震中.球磨机料位软测量及其低能耗高效运行研究.能源研究与利用,2007,(5)：16-19.
[33]司刚全,曹晖,张彦斌,马西奎.基于ANFIS的火电厂磨机负荷检测的软测量模型.2007中国仪器仪表与测控技术交流大会论文集,四川成都,2007,2,1-5.
[34]司刚全,曹晖,张彦斌,贾立新.一种基于密度加权的最小二乘支持向量机稀疏化算法.西安交通大学学报,2009,43(10)：11-15.
[35]肖应旺,徐保国.改进PCA在发酵过程监测与故障诊断中的应用.控制与决策,2005,20(5)：571-574.
[36]赵忠盖,刘飞.工业过程核概率主元模型建立及监控指标.华东理工大学学报(自然科学版),2006,32(7)：864-867.
[37]颜学峰,余娟,钱锋.基于自适应偏最小二乘回归的初顶石脑油干点软测量.化工学报,2005,56(8)：1511-1515.
[38]张学莲,胡立生,曹广益.基于过程数据的动态PLS建模.系统仿真学报,2008,20(10)：2686-2688,2692.
[39]许传龙,赵延军,黄健,孔明,汤光华,王式民.基于人工神经网络的固相质量流量软测量研究.计量学报,2006,27(3)：246-249.
[40]袁仲泉,尹兵太,尚海燕,朱群雄.基于PCA-BP算法的PTA溶剂脱水塔软测量模型.计算机与应用化学,2006,23(7)：623-626.
[41]曹先庆.基于PCA-改进算法BP神经网络的软测量技术.仪器仪表学报,2006,27(6S)：2342-2343.
[42]孟科,李绍军,钱锋.基于模糊免疫算法的溶剂脱水塔软测量.计算机工程与应用,2007,43(12)：228-230,234.
[43]许勇刚,杨辉.基于’RBF网络的稀土萃取过程组分含量软测量.稀土,2007,28(5)：19-22.
[44]徐余法,俞金寿.合成醋酸乙烯的空时得率和催化剂选择性软测量.上海交通大学学报,2007,41(8)：1339-1342.
[45]闭治跃,王庆丰,唐建中.基于径向基函数神经网络的挖泥船排泥管道泥浆浓度软测量模型研究.传感技术学报,2007,20(7)：1630-1634.
[46]常青,杨捷,裴洪卿.基于RBF神经网络的丙烯腈收率软测量方法.东南大学学报(自然科学版),2006,36(S1)：194-196.
[47]李勇,邵诚.灰色软测量在介质填充率检测中的应用研究.中国矿业大学学报,2006,35(4)：549-554.
[48]张勇,王介生.基于PCA-RBF神经网络的浮选过程软测量建模.南京航空航天大学学报,2006,38(S1)：116-119.
[49]王华秋,曹长修,李梁.基于并行核径向基神经网络的软测量与控制模型. 仪器仪表学报,2006,27(10)：1218-1223.
[50]谢承利,陆继东,沈凯,曾勇.基于焚烧运行参数的垃圾热值软测量模型.燃烧科学与技术,2007,13(1)：81-85.
[51]刘载文,王正祥,王小艺,杨斌,程志强.过程神经元网络学习算法及软测量方法的研究.系统仿真学报,2007,19(7)：1456-1459.
[52]杨马英,周芳芹,李军.基于Elman神经网络的城市污水处理水质参数软测量.东南大学学报(自然科学版),2006,36(S1)：119-123.
[53]吕强,俞金寿.量子神经网络软测量模型及应用.系统仿真学报,2007,19(24)：5696-5699.
[54]杨希,钱锋,张兵.基于核函数主元分析的SVM建模方法及应用.华东理工大学学报(自然科学版),2007,33(2)：259-262,270.
[55]桑海峰,王福利,何大阔,张大鹏,何建勇.基于多支持向量机的诺西肽发酵中菌体浓度软测量.系统仿真学报,2006,18(7)：1983-1986.
[56]段莹,范全义,熊智华.基于FA-SVR的丙酮精制产品质量软测量模型.微计算机信息,2007,23(1)：1-3.
[57]徐晔,杜文莉,钱锋.基于核主元分析和最小二乘支持向量机的软测量建模.系统仿真学报,2007,19(17)：3873-3875,3918.
[58]周林成,杨慧中.基于KPCA和最小二乘支持向量机的软测量建模.计算机仿真,2008,25(10)：94-97.
[59]陈佳,颜学峰,钱锋.基于贝叶斯学习的关联向量机及其在软测量中的应用.华东理工大学学报(自然科学版),2007,33(1)：115-119.
[60]朱建鸿,丁健,杨惠中,姜永森.贝叶斯回归支持向量机的软测量建模方法.南京航空航天大学学报,2006,38(S1)：136-138.
[61]袁平,毛志忠,王福利.基于递阶T-S模糊系统的软测量建模方法.东北大学学报(自然科学版),2006,27(10)：1071-1074.
[62]刘瑞兰,苏宏业,褚健.模糊神经网络的混合学习算法及其软测量建模.系统仿真学报,2005,17(12)：2878-2881.
[63]陈国初,徐余法,俞金寿.基于粒子群模糊神经网络的丙烯腈收率软测量建模.系统仿真学报,2007,19(23)：5370-5372.
[64]彭晓波,桂卫华,李勇刚,王凌云,陈勇.基于动态T-S递归模糊神经网络的闪速熔炼过程参数软测量.仪器仪表学报,2008,29(10)：2029-2033.
[65]程健,郭一楠,钱建生.跳汰机床层松散状况软测量建模方法研究.中国矿业大学学报,2006,35(5)：591-595.
[66]王介生,张勇.基于ANFIS的软测量模型再浮选中的应用.合肥工业大学学 报(自然科学版),2006,29(11)：1365-1369.
[67]贾润达,毛志忠,常玉清.湿法冶金萃取组分含量混合建模方法的研究.仪器仪表学报,2009,30(2)：267-271.
[68]袁平,毛志忠,王福利.基于微分进化的钢水温度预报混合模型.东北大学学报(自然科学版),2009,30(8)：1083-1086.
[69]王雅琳,桂卫华,阳春华,谢永芳,宋海鹰.基于有限信息的铜吹炼动态过程智能集成建模.控制理论与应用,2009,26(8)：860-866.
[70]H Y Qi, X G Zhou, L H Liu, W K Yuan. A hybrid neural network-first principles model for fixed-bed reactor. Chemical Engineering Science,1999,54: 2521-2526.
[71]王瑞敏,曹广益,朱新坚.PEMFC混合模型的建模及其稳态仿真.系统仿真学报,2008,20(5)：1299-1302.
[72]李向阳,朱学峰,刘焕彬.间歇制浆蒸煮过程的混合建模方法研究.中国造纸学报,2001,16(2)：24-28.
[73]桑海峰,范玮琦,王福利,何大阔.基于机理知识与最小二乘支持向量机的诺西肽发酵过程混合建模方法.系统仿真学报,2008,20(2)：468-472.
[74]T Takagi, M Sugeno. Fuzzy identification of systems and its applications to modeling and control. IEEE Transactions on Systems, Man, Cybernetics,1985, 15(1):116-132.
[75]张永,邢宗义,向峥嵘,胡维礼.基于Pareto协同进化算法的TS模糊模型设计.控制与决策,2006,21(12)：1332-1337,1342.
[76]肖健梅,王锡淮.基于微粒群优化的船舶柴油机T-S模糊模型.哈尔滨工程大学学报,2007,28(5)：532-535.
[77]T A Johansen, B A Foss. A NARMAX model representation for adaptive control based on local models. Modeling, Identification and Control,1992,13(1):25-39.
[78]R Murry-Smith, T A Johansen. Multiple model approaches to modeling and control. London: Taylor and Francis.
[79]T A Johansen, B A Foss. Operating regime based process modeling and identification. Computes and Chemical Engineering,1997,21(2):159-176.
[80]A N Venkat, P Vijaysai, R D Gudi. Identification of complex nonlinear process based on fuzzy decomposition of steady state space. Journal of Process Control, 2003,13(6):473-488.
[81]C Cheng, M S Chiu. A new data-based methodology for nonlinear process modeling. Chemical Engineering Science,2004,59(13):2801-2810.
[82]孙维,王伟.基于即时学习算法非线性系统多模型自适应控制.大连理工大学学报,2002,42(5)：611-615.
[83]崔宇,薛美盛.局部LSSVM方法在飞灰含碳量软测量中的应用.仪表技术,2009,16(5)：62-64.
[84]V N Vapnik. An overview of statistic learning theory. IEEE Transactions on Neural Networks,1999,10(5):988-999.
[85]J Kennedy, R Eberhart. Particle swarm optimization. Proceedings of IEEE Ineternational Conference on Neural Networks. Piscataway, NJ: IEEE Press, 1995:1942-1948.
[86]Y Shi, R Eberhart. A modified particle swarm optimizer. Proceedings of IEEE Ineternational Conference on Evolutionary Computation. Piscataway, NJ: IEEE Press,1998:69-73.
[87]H Y Fan. A modification to particle swarm optimization algorithm. Engineering Computations,2002,19(8):970-989.
[88]赵娜,张伏生,魏平,刘学.基于改进多粒子群算法的电力系统无功优化.西安交通大学学报,2006,40(4)：463-467.
[89]R Storn, K Price. Differential evolution: a simple and efficient adaptive scheme for global optimization over continuous spaces, tr-95-012. Berkeley: International Computer Science Institute,1995.
[90]张航,王伟,郑玲,李丹丹,熊富强.一种基于密度聚类的小生境差分进化算法.计算机工程与应用,2008,44(23)：42-45.
[91]高新波,李洁,姬红兵.基于加权模糊c均值聚类与统计检验指导的多阈值图像自动分割算法.电子学报,2004,32(4)：661-664.
[92]马强.循环流化床锅炉燃烧系统建模与控制系统研究[D].[博士学位论文].河北保定：华北电力大学,2009.
[93]王颖洁.钢球磨煤机料位的软测量及其动态过程建模与控制.[硕士学位论文].江苏南京：东南大学,2005.
[94]王冬风.多变量智能控制在电厂制粉系统中的应用研究.[博士学位论文].河北保定：华北电力大学,2001.
[95]H G Stark. Continuous wavelet transform and continuous multiscale analysis. Journal of Mathematical Analysis and Applications,1992,169(1):179-196.
[96]彭小奇,宋彦坡,唐英.基于小波分析的异常样本处理.信息与控制,2005,34(6)：676-679.
[97]宋彦坡.数据挖掘技术及其在铜转炉吹炼过程优化中的应用.[硕士学位论 文].湖南长沙：中南大学,2005.