摘要
变形分析及变形趋势预测问题,不仅在工程建设及保障人民生命财产安全方面具有重要意义,就单纯从技术理论的角度而言,它也是一个复杂的系统工程。随着变形监测技术的更新及工程实际的需要,如何引用先进的数学理论和分析方法来深入地了解变形的非线性、复杂性,就是本论文研究的重点所在。
本论文围绕变形监测数据处理的各个环节展开。研究成果和具体内容包括:
(1)系统归纳了附加基准法进行监测网平差的理论及计算过程,阐述了用该方法取代经典控制网平差的优越性,研究和完善了附加基准法中考虑起算数据带有误差时的平差理论及方法,并推导出了具体的计算公式及进行了实例验算;
(2)鉴于非线性监测网平差最好能在非线性系统中进行的需求,本论文深入研究了遗传算法的非线性全局寻优特点及不同遗传算子组合对计算的收敛速度带来的不同影响,提出了最优遗传算子组合的确定方法;此外,提出了用单纯形法取代遗传算法中的交叉操作而形成一种收敛速度更快、计算精度更高的混合遗传算法,并用测试函数及非线性控制网平差实例验证了用混合遗传算法进行非线性优化与传统遗传算法相比较具有的优越性;
(3)现代变形数据分析与传统数据处理的重要区别,就是要求建立动态、非线性预报模型。针对动态、非线性的传统人工神经网络建模方法存在的缺陷,本论文研究采用了遗传算法来改进神经网络算法,即用遗传算法来优化网络参数。实例证明:这种改进算法的学习能力很强,可以高精度地逼近训练样本,预测效果好。
(4)Kalman滤波在减弱监测数据受噪声的影响、预测系统未来状态方面具有独到优势。论文根据时间序列分析及Kalman滤波均是建立在递推形式上的动态建模方法的情况,推导出两种方法间的相互转换关系,提出可在噪声统计信息缺失的情况下利用观测数据构造时序分析的新息模型,再利用模型参数推求预测增益矩阵和滤波增益矩阵进而建立Kalman滤波模型和预测模型,从而避免了传统Kalman滤波建模中必须知道噪声统计信息、必须进行复杂的Riccati方程解算的过程。
Deformation analysis and deformation trend prediction not only has important significance inthe aspects of engineering construction and ensure of people's life and property safety, but also isa complex systematic engineering, even if purely speaking from the angle of technical theory. Incompany with the update of deformation monitoring technology and the requests of engineeringpractice, this research has focused on how to introduce advanced mathematical theories andanalytic methods to give an insight into the nonlinearity and complexity of deformation.
This paper has outspreaded around each step of deformation monitoring data processing.Following are research fmdings and specific contents.
(1) This paper has systematically summed up the theory and computational procedure ofmonitoring network adjustment using the method of appending datum, expatiated advantages ofreplacement of classical control network adjustment with this method, investigated and perfectedthe adjustment theory and methodology of the method of appending datum when consideringerrors in initial numerical data. Besides, this paper has deduced concrete computation formulasand checked out it using examples.
(2) In view of the request that nonlinear monitoring network adjustment can be used in nonlinearsystem, this paper has indepth researched the nonlinear global optimation characteristics ofgenetic algorithm and effects of different genetic operator combinations to the computationalconvergence rate, proposed a method to confirm the optimal genetic operator combination.Morever, this paper has put forward substitution of the crossover operation with simplex method,shaping a hybrid genetic algorithm which can get higher convergence rate and computationalaccuracy. Besides, using trial functions and examples of nonlinear control network adjustment,this paper has verified the advantages of hybrid genetic algorithm compare to traditional onewhen used in nonlinear optimation.
(3) The important difference between modern deformation data analysis and traditionaldeformation data processing lies in the request for the establishment of dynamic and nonlinearprediction models. Aiming at defects exist in artificial neural network which is a dynamic andnonlinear model method, this paper has studied using genetic algorithm to modify neural network algorithm, that is to say using genetic algorithm to optimize network parameters. Examples haveproved that the modified algorithm has good learning capacity, which can approach trainingsamples with high accuracy and can get good predictive validity.
(4) Kalman filter has particular advantages when used to weaken the affects of noise inmonitoring data and predict the future state of a system. According that both time serried andKalman filter are dynamic model methods which are established on recursion modal, this paperhas deduced the interconversion relationship between them, and put forward that an innovationmodel of time series analysis using observation data can be utilized when statistical informationof noise is absent. Then the prediction gain matrix and filter gain matrix can be deduced usingmodel parameters. After that, Kalman filter model and prediction model can be established.Consequently, both noise statistical information and complex Riccati formula computation whichare required in the traditional Kalman filter are no longer necessary in the new model.
引文
[1] 黄声享,尹晖,蒋征.变形监测数据处理[M].武汉:武汉大学出版社,2003.
[2] 尹晖.时空变形分析与预报的理论和方法[M].北京:测绘出版社,2002.
[3] 吴世棋,孔健,张德强.近景摄影测量在隧道变形监测中的应用[J].铁路航测,1994(4).
[4] 王静瑶,吴云.现代地壳运动与地震监测预报研究现状和发展趋势[J].地球科学进展,2000,15(1).
[5] 梅文胜,张正禄,黄全义.测量机器人在变形监测中的应用研究[J].大坝与安全,2002(5).
[6] 王晓华,胡友健,柏柳.变形监测研究现状综述[J].测绘科学,2006,31(2).
[7] 潘宝玉,王贵祥.3S技术集成及其在地质领域中的应用[J].山东地质,1998,14(4).
[8] 施闯.大规模高精度GPS网平差与分析理论及其应用[M].北京:测绘出版社,2002.
[9] 孙海燕.测量平差模型的若干讨论[J].武汉大学学报—信息科学版,2003,28(特刊).
[10] 陶本藻.自由网平差与变形分析[M].北京:测绘出版社,1998.
[11] 陈俊勇.地面参照系的现代定向理论和地球自转运动[M].北京:测绘出版社,1991.
[12] 叶叔华,黄城.天文地球动力学[M].山东:山东科技出版社,2000.
[13] 黄立人.GPS观测结果分析的参考框架及其合理性[J].测绘学报,2001(1).
[14] 陈俊勇.建设我国现代大地测量基准的思考[J].武汉大学学报信息科学版,2003,28(特刊).
[15] 刘大杰,冯延明.动态大地测量的平差基准[J].武汉大学学报信息科学版,2003,28(特刊).
[16] 刘经南,施闯,陈俊勇.’92、’93国家高精度GPS A级网整体平差结果分析与我国块体运动模型研究[J].武汉测绘科技大学学报,1998,23(4).
[17] 施一民.近代大地控制测量[M].北京:测绘出版社,2003.
[18] 王穗辉,刘大杰.附合网平差的基准与起始数据误差的影响[J].大地测量与地球动力学,2004,24(3).
[19] 刘大杰.平面控制点相对点位精度[J].测绘学报,1984,13(3).
[20] 潘国荣,王穗辉.上海地铁二号线隧道贯通定向技术及精度控制[J].同济大学学报,2000.28.
[21]卓健成,路伯祥等.原始数据误差对工测多级三角网精度的影响及其野外实验验证[J].测验学报1987,16(2,3).
[22] 吴俊昶,刘大杰.控制网测量平差[M].北京:测绘出版社,1985.
[23] 王穗辉.顾及起算数据误差的附加基准平差[J].大地测量与地球动力学,2005,25(1).
[24] W.Welsch.对变形和控制网进行监视和分析的若干技术问题[M].武汉:武汉测绘学院,1984.
[25] 陈永奇,吴子安,吴中如.变形监测分析与预报[M].北京:测绘出版社,2003,8.
[26] 邢文训,谢金星.现代化计算方法[M].清华大学出版社,2003,12.
[27] 熊兴华,钱曾波等.基于遗传优化的分段线性影像增强[J].测绘学报,2004,33(4).
[28] 郑宏.遗传算法在影像处理与分析中的应用[M].测绘出版社,2003.
[29] 刘国林.非线性最小二乘与测量平差[M].北京:测绘出版社,2002.
[30] 王新洲.非线性模型参数估计理论及应用[M].武汉:武汉大学出版社,2002.
[31] 田玉刚,王新洲,花向红.非线性最小二乘估计的遗传算法[J].测绘工程,2004,13(4).
[32] 韩茂安,顾圣士.非线性系统的理论和方法[M].北京:科学出版社,2001.
[33] 王小平,曹立明.遗传算法一理论,应用与软件实现[M].西安交通大学出版社,2002.
[34] 李敏强,寇纪淞等.遗传算法的基本理论与应用[M].科学出版社,2002.
[35] 王新洲.基于信息扩散原理的估计理论、方法及其抗差性[J].武汉测绘科技大学学报,1993(3).
[36] 王穗辉.遗传算法在非线性最小二乘平差中的应用[J].大地测量与地球动力学,2006,26(2).
[37] 冯大雍.回归分析方法[M].北京:科学出版社,1985.
[38] 陈希孺,王松桂.近代回归分析一原理、方法及应用[M].合肥:安徽教育出版社,1987.
[39] 韦博成.近代非线性回归分析[M].南京:东南大学出版社,1989.
[40] 王志国,陈立,刘玉成,张春燕.遗传算子对径流预报神经网络的影响[J].武汉大学学报.信息科学版,2005,30(11).
[41] 朱家松,龚健雅,郑皓.遗传算法在管网优化设计中的应用[J].武汉大学学报.信息科学版,2003,28(3).
[42] 方开泰,马长兴.正交与均匀设计[M].北京:科学出版社,2001.
[43] 王柱,方开泰.含有定性因素的均匀设计[J].数理统计与管理,1999,18(5).
[44] 郑肇葆.遗传算法与单纯形法组合的影像纹理分类方法[J].测绘学报,2003,32(4).
[45] 袁亚湘.非线性规划数值方法[M].上海:科学出版社,1993.
[46] 邓建军.模糊逻辑系统学习和优化方法的研究及其应用[D].同济大学博士论文,2001.
[47] 潘国荣,王穗辉.引入时变递增因子的加权等维新息模型及预测[J].同济大学学报,2004,32(6).
[48] 潘国荣,王穗辉.变形预测中顾及时变特性的指数加权组合法[J].同济大学学报,2003,31(1).
[49] 潘国荣,王穗辉.多点变形动态灰色模型辨识及预测[J].测绘学报,2002,30(5).
[50] 潘国荣,王穗辉,刘大杰.用于预测地铁施工中隧道内外变形的动态系统方法[J].同济大学学报,2001,29(3).
[51] 王穗辉,潘国荣.基于MATLAB多变量灰色模型及其在变形预测中的应用[J].土木工程学报,2005,38(5).
[52] 庄镇泉.神经网络与神经计算机[M].北京:科学出版社,1994.
[53] 李士勇.模糊控制·神经控制和智能控制[M].哈尔滨:哈尔滨工业大学出版社,1996.
[54] 钟义信,潘新安,杨义先.智能理论与技术—人工智能与神经网络[M].北京:人民邮电出版社,1992.
[55] 姜绍飞.基于神经网络的结构优化与损伤检测[M].北京:科学出版社,2002.
[56] 黄全义,张正禄.神经网络专家系统在大坝变形预测中的应用[J].测绘信息与工程,2001(3).
[57] 徐晓华,李征航,罗佳.隔河大坝GPS形变监测数据分析[J].测绘信息与工程,2001(3).
[58] 刘芳,李人厚.基于遗传算法的进化神经网络[J].系统仿真学报,2003,15(10).
[59] 张立明.人工神经网络的模型及其应用[M].上海复旦大学出版社,1993.
[60] 杨晓红,刘乐善.用遗传算法优化神经网络结构[J].计算机应用与软件,1997(3).
[61] 胡夏闽.基于神经网络的组合梁栓钉连接件抗剪强度预测[J].南京建筑工程学院学报,1997(1).
[62] 李德仁,袁修孝.误差处理与可靠性理论[M].武汉大学出版社,2002.
[63] 周江文,黄幼才,杨元喜,欧吉坤.抗差最小二乘法[M].华中理工大学出版社,1997.
[64] 欧吉坤.粗差的拟准检定法(QUAD法)[J].测绘学报,1999,28(1).
[65] 施闯,刘经南.基于相关分析的粗差理论[J].测绘学报,1999,28(1).
[66] 张方仁,张金通.测量误差的统计分布和检验[M].中国计量出版社,1991.
[67] 陈朝阳,胡乐群,万鹤群.基于遗传算法的神经网络经济预测模型的建立[J].预测,1997,1.
[68] 贺昌政,徐小湛.基于自组织数据挖掘的组合预测[J].预测,2005,24(4).
[69] 唐小我.组合预测研究的若干新成果[J].预测,1992,(2).
[70] 王穗辉,潘国荣.人工神经网络在隧道地表变形预测中的应用[J].同济大学学报,2001,20(10).
[71] 梅连友.卡尔曼滤波在滑坡监测中的应用[J].测绘工程,2004,13(3).
[72] 邓自立,郭一新.现代时间序列分析及其应用[M].北京:知识出版社,1989.
[73] 马攀,孟令奎,文鸿雁.基于小波分析的Kalman滤波动态变形模型研究[J].武汉大学学报.信息科学版,2004,29(4).
[74] 杨元喜.动态Kalman滤波模型误差的影响[J].测绘科学,2006,31(1).
[75] 吕伟才,徐绍铨.Kalman滤波在似单差方法中的应用[J].测绘通报,2006,3.
[76] 王双喜,黄文骞,许坚.基于非线性滤波的单频GPS精密单点动态定位[J].测绘通报,2006,3.
[77] 胡丛伟,刘大杰,姚连壁.带约束条件的自适应滤波服及其在GPS定位中的应用[J].测绘学报,2002,31(增刊).
[78] 刘大杰,陶本藻.实用测量数据处理方法[M].北京:测绘出版社,2000.
[79] 杨叔子,吴雅.时间序列分析的工程应用[M].武汉:华中理工大学出版社,1996.
[80] 甘仞初.动态数据的统计分析[M].北京:北京理工大学出版社,1991.
[81] 韩京清,何关钰,许可康.线性系统理论代数基础[M].沈阳:辽宁科学技术出版社,1985.
[82] 崔希璋,於宗俦,刘大杰等.广义测量平差[M].北京:测绘出版社,1992.
[83] 须田英.自动控制中的矩阵理论[M].北京:科学出版社,1979.
[84] 邓自立.自校正滤波理论及其应用[M].哈尔滨:哈尔滨工业大学出版社,2003.
[85] 邓自立.卡尔曼滤波与维纳滤波[M].哈尔滨:哈尔滨工业大学出版社,2001.
[86] 邓自立.最优滤波理论及其应用[M].哈尔滨:哈尔滨工业大学出版社,2001.
[87] 刘丁酉.矩阵分析[M].武汉:武汉大学出版社,2004.
[88] 余学祥,吕伟才.抗差卡尔曼滤波模型及其在GPS监测网中的应用[J].测绘学报,2001,30(1).
[89] 王国富,朱建军.含有粗差观测值的自适应滤波[J].测绘通报,2004,4.
[90] 许国辉,余春林.卡尔曼滤波模型的建立及其在施工变形测量中的应用[J].测绘通报, 2004, 4.
[91] A. Chrzanowski. Development trends in deformation Monitoring and Analysis The 8th FIG International Symposium on Deformation Measurements. 1996 .
[92] A. Chrzanowski and Y. Q. Chen. Deformation Monitoring, Analysis and Prediction-Report. FIG XIX International Congress, 1990 .
[93] Andreas Pfeufer. On the Analysis of Dynamic Deformation Processes. FIG XX International Congress. 1994.
[94] Roman Kadaj and Marek Plewako. A New Modeling Approach to Geodetic Deformation Analysis, The 8th FIG International Symposium on Deformation Measurements, 1996 .
[95] Hans Pelzer. Kinematische und Dynamische Fragestellungen der Ingenieur geodasie,Vermessungste -chnische Rundschau, 1993.
[96] Y. L. Cai et al. A Fuzzy Neural for Fuzzy Classification, Proceeding of SCI'94,1994.
[97] Schaffrin B. Ober die Verwendung Stochastscher Vorinformation in freine Nrtzen[J]. Symposium der FG Kommissionen 5/6 Graaz Osterreich,1984.
[98] Holland J H. Adaptation in Natural and Artificial Systems[M]. Michigan: University of Michigan Press. 1975.
[99] Goldberg, D. E. Genetic Algorithms in Search, Optimization and Machine Learning[M].MA:Addison -Wesley Publishing Company, 1989.
[100] Michalewicz Z. Genetic Algorithms + Data Structures = Evolution Programs[M]. Cambridge University Press, 1988.
[101] Bhandarkar S M et al. An Edge Detection Technique Using Genetic Algorithm-BasedOptimization[J]. Pattern Recognition, 1994,27(9).
[102] Wright A H. Genetic algorithms for Real Parameter Optimization. Foundations of Genetic algorithms, Morgan Kaufmann. San Mateo, CA 1991.
[103] Cai Z, Fan J and Yao Q. Functional-coefficient regression models for nonlinear times series. J. Amer,statist,Ass.2000,95(451)
[104] Carroll R J, Ruppert D and Stefanski L A. Nonlinear Measurement Error Models[M].New York: Chapman & Hall, 1998.
[105] Lui Guolin, Tao Huaxue. Nonlinear Measures and their Applications to Explicit Nonlinear Adjustment.A.V.N.,1999.
[106] Vose, M. D., and Liepins, G. E. Schema Disruption. In Proceedings of the Fourth International Conference on Genertic Algorithms(ICGA4), Belew, R, and Brooker,1.(eds.), San Francisco: Morgan Kaufmann Publishers, Inc.,1991.
[107] De Jong, K. A. An analysis of the behavior of a class of genetic adaptive systems[D] .University of Michigan, 1975.
[108] Frank Hoffmeister, Thomas Baeck. Algorithms and evolutionstrategies: Similarities and differences, in parallel problem solving from nature[M]. Berlin: Springer-Verlag,1991.
[109] Schwefel H P. Numerical optimization of computer models[M]. Chichester:Wiley & Sons, 1981.
[110] Michalewicz Z. Genetic algorithms+data structures=evolution programs[M]. New York: Springer-Verlag,1992.
[111] Hagan M T, Demuth H B and Beale M. Neural network design[M]. Boston, MA: PWS Publishing Co., 1995.
[112] Hopfield J J. Neural networks and physical systems with emergent collective computing abilities[J].Proc. Acad. SCI, U. S. A., 1982,79.
[113] Kohonen T. Associative memory: A system theory approach[M]. Springer. New York ,1979.
[114] Saro Lee et al. Use of an artificial neural network for analysis of the susceptibility to landslides at Boun[J]. Korea. Environmental Geology, 2003(44).
[115] LI YITIAN et al. Modeling Flow and Sediment Transport in a River System Using an Artificial Neural Network[J]. Environmental Management, 2003, 31(1).
[116] Rumelhart D E. Parallel distributes processing: Exploration in the microstructure of Cognition[M]. Cambridge, MA: MIT Press, 1986.
[117] R. Hecht-Nielsen. Theory of the Back Propagation Neural Network[J]. Proc, of IJCNN, vol 1,1989.
[118] Dolores Barrios et al. Cooperative binary~real coded genetic algorithms for generating and adapting artificial neural network[J]. Neural Computing & Applications, 2003(12).
[119] Messner J I, Sanvido V E and Kumara S R T. Struct~Net: A network for structural system selection[J]. Microcomputers in Civil Engineering , 1994, 9(2).
[120] Grubbs, F. E. Sample criteria for Testing outlying observations[J]. Ann Math Statist, 1995(21).
[121] Tietjen, G. L. and Moore, R, H, Some Grubbs-Type statistics of the detection of several outliers[J]. Technometrics, 1972(14).
[122] J. M. Bates, C. W. J. Granger. Combination of Forecasts[J]. Ops. Res.Q.1969,20.
[123] Granger CWJ, Ramanathan R. Improved method of forecasting[J]. Journal of Forecasting, 1984,3.
[124] Clemen RT, Winkler RL. Combining economic forecasts[J]. Journal of Business and Economic Statistics, 1986, 4.
[125] Diebold FX, Pauly P. Structural Change and the Combination of Forecasts[J]. Journal of Forecasting ,1987,6.
[126] Donaldson RG, Kamstra M. Forecast combining with neural networks. [J]. Journal of Forecasting ,1996,15.
[127] Cho S, Cho Y, Yoon S. Reliable roll force prediction in cold mill using multiple neuralnetworks[J]. IEEE Trans. On Neural Networks, 1997,8.
[128] Kalman R E.A New Approach to Linear Filtering and Prediction Problem[M]. Trans. ASME. J.Basic Eng.1960.
[129] Anderson B D O and Moore J B. Optimal Filtering[M]. New Jersey: Prentice-Hall, Inc., 1979.
[130] KOCH K R, YANG Y. Robust Kalman Filter for Rank Deficient Observation Models[J]. Journal of Geodesy, 1998,72(8).
[131] Chen C T. Linear System Theory and Design[M]. New York: Holt Rinehart and Winston, 1984.
[132] Kailath T., Sayed A H., and Hassibi B. Linear Estimation[M]. New Jersey:Prentice-Hall,Inc.2000.
[133] Gevers M., and Wouyers W R E. An Innovations Approach to the Discrete-Time Stochastic Realization Problem[J]. Ouarterly Journal on Automatic Control, 1978,19(6).
[134] Box G E P, Jenkins G M. Time Series Analysis, Forecasting and Control. San trancisco: Holden Day,1970.
[135] Sage A P,and Husa G W. Adaptive Filtering with Unknown Prior Seatistics[M].JACC,1969.