遗传算法在都江堰渠首配水中的应用研究
|
摘要
资料表明,都江堰灌区岷江来水有逐年下降的趋势,同时为顺应灌区经济的发展,灌区需水量又逐年增加。为解决这一矛盾,本文针对都江堰渠首六大干渠进行了优化配水的研究,而选择合适的优化算法又是整个配水过程中的关键部分。
遗传算法是一种模仿自然界生物种群选择和进化的随机搜索算法,是一种新型的优化技术。在比较了传统优化算法和遗传算法的基础上,得出选用遗传算法进行优化配水的结论。本文研究具体内容如下:
(1) 基本遗传算法的学习和讨论。系统学习了基本遗传算法的生物学模型、基本思想、实现技术、模式理论以及遗传算法的特点等,讨论了几种传统优化算法和它们的不足,得出遗传算法进行优化的优势所在。
(2) 都江堰渠首配水模型研究。系统阐述了都江堰渠首水资源配置的基本概念、配置原则和目标以及优化配置机理。给出配水模型的功能,探讨了经济效益最大原则、损失水量最小原则、水库蓄放水优先级原则等,并给出三类目标函数表达式。由于都江堰灌区的水主要用于灌溉,因此本文研究的模型是以灌溉效益最大为目标的,依据经济效益最大原则,提出本文的目标函数和约束条件。
(3) 改进遗传算法研究。针对基本遗传算法存在早期收敛和后期收敛速度缓慢等缺点,对基本遗传算法进行改进。改进包括以下几方面:编码设计、适应值函数的选取、选择策略、遗传操作、控制参数选择以及进化终止准则等。本文采用非均匀算术交叉和非均匀变异,同时设计了与进化代数相关的交叉概率以及自适应变异概率,有效改进了遗传算法的性能,提高了运行速度。
History data shows the number of Min River is declining year by year. At the same time, the economy development of Dujiangyan needs more and more water. To solve the contradiction, this paper discusses how to distribute water resource among 6 canals. Which optimal method is to be chosen is the key element.Genetic Algorithm(GA) is a new-style optimizing technique, it imitates the selection and evolution of biology in the nature and it is a random search algorithm, which can make up the deficiencies of conventional optimization algorithms, "he concrete contents are as follows:(1) Study and discussion of SGA. Biology model, basic idea, implement techniques, schema theory and feature are systematically learned.(2) Study of water distribution model. Basic theory, principle and mechanism are introduced. Model function is given and evaluation criteria of economy, amount of water and energy are discussed. Considered the water is mainly used to irrigate in Dujiangyan Irrigated Area, this paper mainly discusses irrigated benefit. According to three principle mentioned above, object function and restriction condition are given.(3) Study of improved GA (IGA). To solve the problem that SGA has some disadvantages such as early convergence and the slow upper
convergence rate, in this paper SGA is improved on the following aspects: coding design, selection of fitness, selection strategy, genetic operation, selection of control parameters and principle of ending of evolution. Inconsistent crossover operator and mutation operator are adopted, and crossover rate correlative to evolution generation and adaptive mutation rate are design in this paper. The application indicates that this algorithm has the optimum results with improved precision and convergence speed.(4) Simulation study of water distribution. Based on forecasting water data of last ten-day of May, 2005, benefit functions of 6 canals are fitted with polynomial fitting method. Finally SGA and ISGA are applied to water distribution respectively.The outstanding point of this paper consists in SGA is pointedly improved based on concrete problem of water resource distribution. The results indicate that IGA is effective and has application value which has referenced value to other similar problems.
遗传算法是一种模仿自然界生物种群选择和进化的随机搜索算法,是一种新型的优化技术。在比较了传统优化算法和遗传算法的基础上,得出选用遗传算法进行优化配水的结论。本文研究具体内容如下:
(1) 基本遗传算法的学习和讨论。系统学习了基本遗传算法的生物学模型、基本思想、实现技术、模式理论以及遗传算法的特点等,讨论了几种传统优化算法和它们的不足,得出遗传算法进行优化的优势所在。
(2) 都江堰渠首配水模型研究。系统阐述了都江堰渠首水资源配置的基本概念、配置原则和目标以及优化配置机理。给出配水模型的功能,探讨了经济效益最大原则、损失水量最小原则、水库蓄放水优先级原则等,并给出三类目标函数表达式。由于都江堰灌区的水主要用于灌溉,因此本文研究的模型是以灌溉效益最大为目标的,依据经济效益最大原则,提出本文的目标函数和约束条件。
(3) 改进遗传算法研究。针对基本遗传算法存在早期收敛和后期收敛速度缓慢等缺点,对基本遗传算法进行改进。改进包括以下几方面:编码设计、适应值函数的选取、选择策略、遗传操作、控制参数选择以及进化终止准则等。本文采用非均匀算术交叉和非均匀变异,同时设计了与进化代数相关的交叉概率以及自适应变异概率,有效改进了遗传算法的性能,提高了运行速度。
History data shows the number of Min River is declining year by year. At the same time, the economy development of Dujiangyan needs more and more water. To solve the contradiction, this paper discusses how to distribute water resource among 6 canals. Which optimal method is to be chosen is the key element.Genetic Algorithm(GA) is a new-style optimizing technique, it imitates the selection and evolution of biology in the nature and it is a random search algorithm, which can make up the deficiencies of conventional optimization algorithms, "he concrete contents are as follows:(1) Study and discussion of SGA. Biology model, basic idea, implement techniques, schema theory and feature are systematically learned.(2) Study of water distribution model. Basic theory, principle and mechanism are introduced. Model function is given and evaluation criteria of economy, amount of water and energy are discussed. Considered the water is mainly used to irrigate in Dujiangyan Irrigated Area, this paper mainly discusses irrigated benefit. According to three principle mentioned above, object function and restriction condition are given.(3) Study of improved GA (IGA). To solve the problem that SGA has some disadvantages such as early convergence and the slow upper
convergence rate, in this paper SGA is improved on the following aspects: coding design, selection of fitness, selection strategy, genetic operation, selection of control parameters and principle of ending of evolution. Inconsistent crossover operator and mutation operator are adopted, and crossover rate correlative to evolution generation and adaptive mutation rate are design in this paper. The application indicates that this algorithm has the optimum results with improved precision and convergence speed.(4) Simulation study of water distribution. Based on forecasting water data of last ten-day of May, 2005, benefit functions of 6 canals are fitted with polynomial fitting method. Finally SGA and ISGA are applied to water distribution respectively.The outstanding point of this paper consists in SGA is pointedly improved based on concrete problem of water resource distribution. The results indicate that IGA is effective and has application value which has referenced value to other similar problems.
引文
[1] Cheng, R., M, Gen, and Y. Tsujimura, A tutorial survey of job-shop scheduling problems using genetic algorithms: Ⅰ. Representation, Computers and Industrial Engineering, vol. 30, pp. 983—997, 1996
[2] Gen, M. and R. Cheng, Genetic Algorithm and Engineering Design, Wiley, New York, 1997
[3] Michalewicz, Z., Genetic Algorithm+Data Structure=Evolution Programs, 3rd edition, Springer-Verlag, New York, 1996
[4] 马少平,朱小燕.人工智能.北京:清华大学出版社.2004
[5] 玄光男,程润伟.遗传算法与工程优化.北京:清华大学出版社.2004
[6] 彭述明.都江堰:持续发展的光辉典范.中国水利.2004,516(18):15—18
[7] 钱清泉.对都江堰水资源科学利用的几点看法.中国水利.2004,516(18):24—25
[8] 孙砚方.与时俱进人水和谐的千秋伟业.中国水利.2004,516(18):12—14
[9] 徐玉野,王全凤等.遗传算法在工程结构优化中的研究.基建优化,2002,23(6):50—52
[10] 王小平,曹立明.遗传算法——理论、应用与软件实现.西安:西安交通大学出版社.2002
[11] 王凌.智能优化算法及应用.北京:清华东西出版社.2001
[12] 周明,孙树栋.遗传算法原理及应用.北京:国防工业出版社.2001
[13] 王旭,王宏等.人工神经网络原理及应用.沈阳:东北大学出版社.2000
[14] 张文修,梁怡.遗传算法的数学基础.西安:西安交通大学出版社.2000
[15] Z.米凯利维茨.演化程序——遗传算法与数据编码的结合.北京:科学出版社.2000
[16] 王正志,薄涛.进化计算.长沙:国防科技大学出版社.2000
[17] 周明,孙树栋.遗传算法原理及应用.北京:国防工业出版社.1999
[18] 李人厚.智能控制理论和方法.西安:西安科技大学出版社.1999
[19] 汪恕诚.实现由工程水利到资源水利的转变做好面向21世纪中国水利这篇大文章.水利经济.1999,4:1—6
[20] 潘正君.演化计算.北京:清华大学出版社.1998
[21] Eshelman, L. and J. Schaffer, Real-coded genetic algorithms and intervalschemata, in Whitley, L., editor, Foundations of Genetic Algorithms, vol. 2, pp. 187—202, Morgan Kaufmann Publishers, San Francisco1993
[22] Walters, G. A. and D. K. Smith, Evolutionary design algorithm for optimal layout of tree networks, Engineering Optimization, vol. 24, pp. 261—281, 1995
[23] Rechenberg, I., Evolutionstrategie: Optimieriung technischer Systeme nach Prinzipien der biologischen Evolution, Frommann-Holzboog, Stuttgart, Germany, 1973
[24] 陈国良,王煦法等.遗传算法及应用.北京:人民邮电出版社.1996
[25] 刘勇,康立山等.非数值并行算法——遗传算法.北京:科学出版社.1995
[26] Back, T., Evolutionary Algorithms in Theory and Practice, Oxford University Press, New York, 1996
[27] Fogel, D., Evolutionary Computation: Toward a New Philosophy of Machine Intelligence, IEEE Press, Piscataway, NJ, 1995
[28] Gen, M. and Y. Tsujimura, editors. Genetic Algorithms and Intelligence Systems, Gordon and Breach, New York, (forthcoming)
[29] Goldberg, D., Genetic Algorithms in Search, Optimization and Machine Learning, Addison Wesley, Reading, MA, 1989
[30] Holland J H. Adaptation in Natural and Artificial Systems. MIT Press, 1975
[31] 张晓缋,方浩等.遗传算法的编码机制研究.信息与控制,1997,26(2):134—139
[32] 王秀峰.实数编码的遗传算法及其在逆变器馈电交流电机中的应用.自动化学报,1998,24(2):250—253
[33] 吴平勇.都江堰发展与保护的若干问题.中国水利.2004,516(18):59—61
[34] 甘泓.水资源合理配置浅析.中国水利.2000,4:20—24
[35] 崔远来.非充分灌溉优化配水技术研究综述.灌溉排水,2000,19(1):66—70
[36] 黄宝全,沈菊燕.Jensen模型在灌区“三种水”合理调度中的应用.灌溉排水,1992,11(3):25—28
[37] 崔远来,尹正杰等.作物产量与全生育期总腾发量之间关系的研究.中国农村水利水电,2000,10:24-25
[38] 杨路华,夏辉等.非充分灌溉制度制定过程中Jensen模型的求解与应用.2002,21(4):13—15
[39] 陈玉民,肖俊夫等.非充分灌溉研究进展及展望.2001,20(2):73—75
[40] 王劲峰,刘昌明等.水资源空间配置的边际效益均衡模型.中国科学(D辑).2001,31(5)
[41] 马孝义.节水灌溉新技术.北京:中国农业出版社.2000
[42] 彭新竹.遗传算法的改进策略及其应用.华东船舶工业学院学报(自然科学版).2002,3:53—58
[43] 关志华,寇纪淞.一种改进的遗传算法Scatter GA.控制与决策.2002,5:579—582,634
[2] Gen, M. and R. Cheng, Genetic Algorithm and Engineering Design, Wiley, New York, 1997
[3] Michalewicz, Z., Genetic Algorithm+Data Structure=Evolution Programs, 3rd edition, Springer-Verlag, New York, 1996
[4] 马少平,朱小燕.人工智能.北京:清华大学出版社.2004
[5] 玄光男,程润伟.遗传算法与工程优化.北京:清华大学出版社.2004
[6] 彭述明.都江堰:持续发展的光辉典范.中国水利.2004,516(18):15—18
[7] 钱清泉.对都江堰水资源科学利用的几点看法.中国水利.2004,516(18):24—25
[8] 孙砚方.与时俱进人水和谐的千秋伟业.中国水利.2004,516(18):12—14
[9] 徐玉野,王全凤等.遗传算法在工程结构优化中的研究.基建优化,2002,23(6):50—52
[10] 王小平,曹立明.遗传算法——理论、应用与软件实现.西安:西安交通大学出版社.2002
[11] 王凌.智能优化算法及应用.北京:清华东西出版社.2001
[12] 周明,孙树栋.遗传算法原理及应用.北京:国防工业出版社.2001
[13] 王旭,王宏等.人工神经网络原理及应用.沈阳:东北大学出版社.2000
[14] 张文修,梁怡.遗传算法的数学基础.西安:西安交通大学出版社.2000
[15] Z.米凯利维茨.演化程序——遗传算法与数据编码的结合.北京:科学出版社.2000
[16] 王正志,薄涛.进化计算.长沙:国防科技大学出版社.2000
[17] 周明,孙树栋.遗传算法原理及应用.北京:国防工业出版社.1999
[18] 李人厚.智能控制理论和方法.西安:西安科技大学出版社.1999
[19] 汪恕诚.实现由工程水利到资源水利的转变做好面向21世纪中国水利这篇大文章.水利经济.1999,4:1—6
[20] 潘正君.演化计算.北京:清华大学出版社.1998
[21] Eshelman, L. and J. Schaffer, Real-coded genetic algorithms and intervalschemata, in Whitley, L., editor, Foundations of Genetic Algorithms, vol. 2, pp. 187—202, Morgan Kaufmann Publishers, San Francisco1993
[22] Walters, G. A. and D. K. Smith, Evolutionary design algorithm for optimal layout of tree networks, Engineering Optimization, vol. 24, pp. 261—281, 1995
[23] Rechenberg, I., Evolutionstrategie: Optimieriung technischer Systeme nach Prinzipien der biologischen Evolution, Frommann-Holzboog, Stuttgart, Germany, 1973
[24] 陈国良,王煦法等.遗传算法及应用.北京:人民邮电出版社.1996
[25] 刘勇,康立山等.非数值并行算法——遗传算法.北京:科学出版社.1995
[26] Back, T., Evolutionary Algorithms in Theory and Practice, Oxford University Press, New York, 1996
[27] Fogel, D., Evolutionary Computation: Toward a New Philosophy of Machine Intelligence, IEEE Press, Piscataway, NJ, 1995
[28] Gen, M. and Y. Tsujimura, editors. Genetic Algorithms and Intelligence Systems, Gordon and Breach, New York, (forthcoming)
[29] Goldberg, D., Genetic Algorithms in Search, Optimization and Machine Learning, Addison Wesley, Reading, MA, 1989
[30] Holland J H. Adaptation in Natural and Artificial Systems. MIT Press, 1975
[31] 张晓缋,方浩等.遗传算法的编码机制研究.信息与控制,1997,26(2):134—139
[32] 王秀峰.实数编码的遗传算法及其在逆变器馈电交流电机中的应用.自动化学报,1998,24(2):250—253
[33] 吴平勇.都江堰发展与保护的若干问题.中国水利.2004,516(18):59—61
[34] 甘泓.水资源合理配置浅析.中国水利.2000,4:20—24
[35] 崔远来.非充分灌溉优化配水技术研究综述.灌溉排水,2000,19(1):66—70
[36] 黄宝全,沈菊燕.Jensen模型在灌区“三种水”合理调度中的应用.灌溉排水,1992,11(3):25—28
[37] 崔远来,尹正杰等.作物产量与全生育期总腾发量之间关系的研究.中国农村水利水电,2000,10:24-25
[38] 杨路华,夏辉等.非充分灌溉制度制定过程中Jensen模型的求解与应用.2002,21(4):13—15
[39] 陈玉民,肖俊夫等.非充分灌溉研究进展及展望.2001,20(2):73—75
[40] 王劲峰,刘昌明等.水资源空间配置的边际效益均衡模型.中国科学(D辑).2001,31(5)
[41] 马孝义.节水灌溉新技术.北京:中国农业出版社.2000
[42] 彭新竹.遗传算法的改进策略及其应用.华东船舶工业学院学报(自然科学版).2002,3:53—58
[43] 关志华,寇纪淞.一种改进的遗传算法Scatter GA.控制与决策.2002,5:579—582,634