基于改进自适应混合遗传算法的抽水蓄能电站厂内运行优化
|
摘要
抽水蓄能电站以其独具的蓄能填谷作用和快速、灵活的启停特性,可作为系统中承担调峰、调频、调相、调压、旋转备用、事故备用和黑启动的重要技术手段,受到了日益广泛的重视。研究抽水蓄能电站的运行优化方式对提高整个电力系统的供电质量、运行可靠性与经济性具有重大意义。本文以研究抽水蓄能电站厂内运行优化为出发点,对抽水蓄能电站厂内运行优化的遗传算法数学模型、优化方法以及模型的计算机求解方面进行了研究和改进。根据遗传算法的基本原理和操作步骤,阐述了此数学模型在抽水蓄能电站厂内运行优化的具体实现过程。
本文以抽水蓄能电站厂内运行优化为出发点,结合电站运行中的实际情况和需要考虑的关键问题,对以往厂内运行优化的遗传算法数学模型进行改进,建立考虑抽水蓄能电站机组承担随机负荷、无功负荷、避开汽蚀振动等限制运行区间和考虑随时调整优化运行方式的抽水蓄能电站厂内运行优化的遗传算法数学模型。
求解最优负荷分配数学模型时,针对水电机组安全运行区是一系列离散子区间构成的特点,在初始种群生成之时,增加了对机组汽蚀、振动区的约束,使得产生的每一个初始解都在可行域内,减少了初始种群的搜索空间和时间,提高了运算效率。针对最优机组组合模型的特点,设计最小开停机和输出功率等约束,指导初始种群的生成,节省了随机方法产生可行解浪费的时间。制定抽水蓄能电站最优运行的日计划是一个双重决策的过程,在计算中,根据一个最优策略的任意子策略都是最优的理论,将第二重决策的迭代求解改为直接调入计算出的机组空间最优负荷分配表来查表求解,减少了计算的复杂性。
为了验证本文提出的模型及优化方法的可行性及有效性,通过对十三陵抽水蓄能电站抽水工况运算,结果表明：此模型及优化算法比实际运行节水11.3×104m3,提高效益1.43%,可多发电23.4×104kW·h。表明了改进自适应混合遗传算法模型及优化方法在抽水蓄能电站电站厂内经济运行的有效性。
Pumped storage power station with its unique role in energy storage and filling valleys and fast, flexible start and stop feature'can be a system to bear peaking, frequency modulation, phase modulation, voltage regulation, spinning reserve, black start accident and the importance of alternative technologies means to be an increasingly wide attention. Study of pumped-storage power station to improve the economic operation of the power system of power quality, reliability and economical significance. Based on the research pumped-storage plant of economic operation as the starting point, the pumped-storage plant of economic operation of the genetic algorithm, and the optimized mathematical model of solving method of computer models are studied and improved. According to the basic principle of genetic algorithms and operating procedures, expounds the mathematical model in the pumped-storage plant of economic operation of the concrete implementation process.
Based on the pumped-storage plant operation optimization, combining the actual situation and power operation requires consideration of the key problems for ever, the internal operation optimization of genetic algorithm, the improved model established considering the pumped-storage unit undertake random load, reactive load, avoid cavitation vibration restricted operation interval and consider the optimized operation mode adjust the pumped-storage plant operation optimization of genetic algorithm, the mathematical model.
Solving mathematical model of optimal load distribution, hydro turbine safety operation area is composed of a series of discrete intervals, in the initial population generation, increase of cavitation, vibration and the "S" shape characteristic, the constraint of every initial solution is feasible in the initial population, reduce the search space and time, improve the computational efficiency. According to the model of optimal combination of units, and the minimum output power design made such constraints, guiding the initial population, save the stochastic method feasible solutions to waste time. Formulate pumped-storage power station, the optimal operation of the plan is a double decision-making process, in computing, according to an optimal strategy of anyons strategy is optimal theory, the second iteration decision to direct the crew into compute the optimal load sharing space to solve, reduce the check the complexity of computation.
In order to verify the model proposed in this paper and optimization method of the feasibility and availability, through to the Ming tombs pumped-storage pumping condition, results showed:This model and the optimization algorithm could save water 11.3×104m3 than Traditional GA operation, improve efficiency about 1.43%,and could generate electricity, to more 23.4×104kW·h.Indicates that the improved adaptive hybrid genetic algorithm model and the optimization method pumped-storage hydropower plant in economical operation efficiency.
本文以抽水蓄能电站厂内运行优化为出发点,结合电站运行中的实际情况和需要考虑的关键问题,对以往厂内运行优化的遗传算法数学模型进行改进,建立考虑抽水蓄能电站机组承担随机负荷、无功负荷、避开汽蚀振动等限制运行区间和考虑随时调整优化运行方式的抽水蓄能电站厂内运行优化的遗传算法数学模型。
求解最优负荷分配数学模型时,针对水电机组安全运行区是一系列离散子区间构成的特点,在初始种群生成之时,增加了对机组汽蚀、振动区的约束,使得产生的每一个初始解都在可行域内,减少了初始种群的搜索空间和时间,提高了运算效率。针对最优机组组合模型的特点,设计最小开停机和输出功率等约束,指导初始种群的生成,节省了随机方法产生可行解浪费的时间。制定抽水蓄能电站最优运行的日计划是一个双重决策的过程,在计算中,根据一个最优策略的任意子策略都是最优的理论,将第二重决策的迭代求解改为直接调入计算出的机组空间最优负荷分配表来查表求解,减少了计算的复杂性。
为了验证本文提出的模型及优化方法的可行性及有效性,通过对十三陵抽水蓄能电站抽水工况运算,结果表明：此模型及优化算法比实际运行节水11.3×104m3,提高效益1.43%,可多发电23.4×104kW·h。表明了改进自适应混合遗传算法模型及优化方法在抽水蓄能电站电站厂内经济运行的有效性。
Pumped storage power station with its unique role in energy storage and filling valleys and fast, flexible start and stop feature'can be a system to bear peaking, frequency modulation, phase modulation, voltage regulation, spinning reserve, black start accident and the importance of alternative technologies means to be an increasingly wide attention. Study of pumped-storage power station to improve the economic operation of the power system of power quality, reliability and economical significance. Based on the research pumped-storage plant of economic operation as the starting point, the pumped-storage plant of economic operation of the genetic algorithm, and the optimized mathematical model of solving method of computer models are studied and improved. According to the basic principle of genetic algorithms and operating procedures, expounds the mathematical model in the pumped-storage plant of economic operation of the concrete implementation process.
Based on the pumped-storage plant operation optimization, combining the actual situation and power operation requires consideration of the key problems for ever, the internal operation optimization of genetic algorithm, the improved model established considering the pumped-storage unit undertake random load, reactive load, avoid cavitation vibration restricted operation interval and consider the optimized operation mode adjust the pumped-storage plant operation optimization of genetic algorithm, the mathematical model.
Solving mathematical model of optimal load distribution, hydro turbine safety operation area is composed of a series of discrete intervals, in the initial population generation, increase of cavitation, vibration and the "S" shape characteristic, the constraint of every initial solution is feasible in the initial population, reduce the search space and time, improve the computational efficiency. According to the model of optimal combination of units, and the minimum output power design made such constraints, guiding the initial population, save the stochastic method feasible solutions to waste time. Formulate pumped-storage power station, the optimal operation of the plan is a double decision-making process, in computing, according to an optimal strategy of anyons strategy is optimal theory, the second iteration decision to direct the crew into compute the optimal load sharing space to solve, reduce the check the complexity of computation.
In order to verify the model proposed in this paper and optimization method of the feasibility and availability, through to the Ming tombs pumped-storage pumping condition, results showed:This model and the optimization algorithm could save water 11.3×104m3 than Traditional GA operation, improve efficiency about 1.43%,and could generate electricity, to more 23.4×104kW·h.Indicates that the improved adaptive hybrid genetic algorithm model and the optimization method pumped-storage hydropower plant in economical operation efficiency.
引文
[1]邱彬如,刘连希.抽水蓄能电站工程技术.北京：中国电力出版社,2008
[2]张克诚.抽水蓄能电站水能设计.北京：中国水利水电出版社,2007,45-53
[3]陆佑楣,潘家铮.抽水蓄能电站.北京：水利电力出版社,1992
[4]张勇传.水电站经济运行原理.第2版.北京：中国水利水电出版社,1998
[5]马跃先,马希金,阎振真.小型水电站优化运行与管理.郑州：黄河水利出版社,2000.9
[6]华中工学院主编.水电站经济运行.北京：水利电力出版社,1984
[7]权先璋.动态规划原理在水电站厂内经济运行中的应用.水电能源科学,1983,1(1)：95-104
[8]张雪源,王蔚然,沈学东.关于水电厂经济运行的研究.东北电力技术,1999,No.1：36-38
[9]姚齐国,张士军,蒋传文,权先璋.动态规划法在水电厂内经济运行中的应用.水电能源科学,1999,17(1)：46-49
[10]周明,孙树栋编著.遗传算法原理及应用.北京：国防工业出版社,2000
[11]王黎,马光文.基于遗传算法的水电站厂内经济运行新算法.中国电机工程学报,第18卷第1期,1998：64-67
[12]于尔铿等.抽水蓄能电站群短期优化调度.电网技术,1996,51(5)：63-70
[13]卢显锋.最优化方法应用基础.上海：同济大学出版社,2003
[14]何坚勇.最优化方法.北京：清华大学出版社,2007
[15]王仁权等.附件梯级水电站群短期优化调度模型及其计算方法.云南水力发电,2002,18(1)：52-53,102
[16]詹森,巴恩斯.网络流规划.北京：科学出版社,1988,4
[17]赵子臣,夏清.水电站群补偿调节计算的非线性网络流法.北京：清华大学学报(自然科学版),1994,34(9)：86.94
[18]王海忠,于尔铿.抽水蓄能电站群短期优化调度.电网技术,1996,20(2)：35-38,46
[19]李长河.人工智能及其应用.北京：机械工业出版社,2006
[20]蔡自兴,姚莉.人工智能及其在决策系统中的应用.北京：国防科技大学出版社,2006
[21]王黎,马光文.基于遗传算法的水电站厂内经济运行新算法.中国电机工程学报,第18卷第1期,1998：64-67
[22]袁晓辉,张双全,王金文,张勇传.拟梯度遗传算法在水电厂内经济运行中的应用研究.电网技术,2000,24(12)：66-69
[23]程芳,高龙华,陈守伦.抽水蓄能电站优化调度技术研究综述.水利水电科技进展,2005,24(5)：64-67
[24]言茂松等.抽水蓄能电站随机模拟及运行优化.控制与决策,1992,135(9)：25-28
[25]王海忠,于尔铿.“协调方程式法”用于抽水蓄能电站群优化调度的研究.电网技术,1996,20(5)：24-25
[26]周燕屏,李承军,李洪斌.用动态网络规划法制定水电厂开停机计划.水电自动化与大坝监测,2003,27(3)：20-22
[27]A.阿尔塞等.伊泰普水电站水轮发电机级的优化调度.水利水电快报,2002,23(20)：1-4
[28]组合最优化-算法和复杂性.刘振宏,蔡茂诚译.北京：清华大学出版社,1988
[29]Holland J H. Adaptation in Nature and Artificial Systems. MIT Press,1992
[30]De Jong KA. An Analysis of the Behavior of a Class of Genetic Adaptive Systems. Ph.D Dissertation, University of Michigan, No.76-9381,1975
[31].Goldberg D E. Genetic Algorithms in Search, Optimization and Machine Learning. Addison-Wesley,1989
[32]文福拴,韩祯祥.模拟进化优化方法在电力系统中的应用综述(上).电力系统自动化,1996,20(1)：59-63
[33]文福拴,韩祯祥.模拟进化方法在电力系统中的应用综述(中).电力系统自动化,1996,20(2)：60-64同第1章[16]
[34]文福拴,韩祯祥.模拟进化方法在电力系统中的应用综述(下).电力系统自动化,1996,20(3)：72-75
[35]张潮海,赖红妮.遗传算法在电力系统中的应用.电力情报,1995,3：1-4
[36]刘云峰,曹春蕾.一维大地电磁测深的遗传算法反演.浙江大学学报(自然科学版),1997,31(3)：300-305
[37]杨旭东,张彤.遗传算法应用与系统在线识别研究.哈尔滨工业大学学报,2000,32(1)：102-105
[38]Mathias K E and Whitley L D. Transforming the Search Space with Gray Coding. In:Proc. Of the 1st IEEE Int'1. Conf. On Evolutionary Computation(ICEC' 94), 519-524
[39]Schraudolph NN and Belew RK. Dynamic Parameter Encoding for Genetic Algorithms. Machine Learning,1992,9(1):73-86
[40]Michalewicz Z. Gentic algorithms+Data structure=evolution program. Springer-Verlag, Berlin.3rd, Revised and Extended Edition,1996
[41]张晓馈,方浩,戴冠中.遗传算法的编码机制研究.信息与控制,1997,26(2)：134-139
[42]Goldberg DE. Genetic Algorithms in Search, Optimization and Machine Learning. Addison-Wesley Publishing Company, INC, New York.1989
[43]朱鳌鑫.遗传算法的适应度函数研究.系统工程与电子技术,1998(11)：58-62
[44]赵明旺.非线性动态系统参数估计的混合计算智能算法.应用基础与工程科学学报,1997.5(1)：97-102
[45]Vassilios P. Varying fitness functions in genetic algorithm constrained optimization. The Cutting Stock and Unit Commitment Problem. IEEE Trans Syst Man, Cybern, 1998.28 (5):629-639
[46]恽为民.基于遗传的机器人运动规划.上海：上海交通大学,1995
[47]Schaffer JD. A Study of Control Parameters Affecting Online Performance of Genetic Algorithms for Function Optimization. In:Proc. Of the 3rd Int'1. Conf. Genetic Algorithms, Morgan Kaufmann, Los Altos,1989:51-60
[48]AGOSTON EE, ROGER TH, ZBIGHIE WM. Parameter control in evolutionary algorithms. IEEE Transactions on Evolutionary Computation,1999,3(2):12-14
[49]周明,孙树栋.遗传算法原理及应用.北京：国防工业出版社,1999.6
[50]梅亚东,左园忠,朱教新一类含有0-1变量的厂内经济运行模型及解法.水电能源科学,1999,17(3)：27-30
[51]邓晓娟,权先璋,张勇传.三峡梯级水电站厂内经济运行.东北电力技术,2002,No.4：1-4
[52]路志宏,魏守平,罗元胜.基于开关控制策略的厂内经济运行模型及应用.水电自动化与大坝监测,2003,27(1)：11-13
[53]苑希民,李鸿雁,刘树坤,崔广涛.神经网络和遗传算法在水科学领域的应用.北京：中国水利水电出版社,2002.8
[54]玄光男,程润伟(著).汪定伟,唐加福,黄敏(译).遗传算法与工程设计.北京：科学出版社,2000
[55]王培萍.抽水蓄能电站与火电站联合电力系统优化运行的研究.北京工业大学硕士学位论文.2002.12
[56]肖黎,刘礼华,袁文阳.紧水滩4#机组振动试验研究[J].水电能源科学,2002,20(4)：73-75.
[57]柏勇,李德芳,孙成玲.沙河抽水蓄能电站2号机上导瓦烧损事故分析[J].水力发电,2004(5)：54-55.
[58]邱家俊.电机的机电耦联与磁固耦合非线性振动研究[J].中国电机工程学报,2002,22(5)：109-115.
[59]陶星明,吴新润.岩滩电厂楼板强振问题分析[J].大电机技术,2004(1)：35-40.
[60]姚大坤,李至昭,曲大庄.混流式水轮发机自激振动分析[J].大电机技术,1998(5)：43-47.
[61]颜庆津.数值分析.北京：北京航空航天大学出版社,1992.7
[62]贾嵘、蔡振华、罗兴锜;改进自适应遗传算法及其在水电站最优报价中的应用.水力发电学报。2007.26(1)：11-15.
[63]王培萍.抽水蓄能电站与火电站联合系统优化运行的研究.北京工业大学硕士学位论文.2002.
[2]张克诚.抽水蓄能电站水能设计.北京：中国水利水电出版社,2007,45-53
[3]陆佑楣,潘家铮.抽水蓄能电站.北京：水利电力出版社,1992
[4]张勇传.水电站经济运行原理.第2版.北京：中国水利水电出版社,1998
[5]马跃先,马希金,阎振真.小型水电站优化运行与管理.郑州：黄河水利出版社,2000.9
[6]华中工学院主编.水电站经济运行.北京：水利电力出版社,1984
[7]权先璋.动态规划原理在水电站厂内经济运行中的应用.水电能源科学,1983,1(1)：95-104
[8]张雪源,王蔚然,沈学东.关于水电厂经济运行的研究.东北电力技术,1999,No.1：36-38
[9]姚齐国,张士军,蒋传文,权先璋.动态规划法在水电厂内经济运行中的应用.水电能源科学,1999,17(1)：46-49
[10]周明,孙树栋编著.遗传算法原理及应用.北京：国防工业出版社,2000
[11]王黎,马光文.基于遗传算法的水电站厂内经济运行新算法.中国电机工程学报,第18卷第1期,1998：64-67
[12]于尔铿等.抽水蓄能电站群短期优化调度.电网技术,1996,51(5)：63-70
[13]卢显锋.最优化方法应用基础.上海：同济大学出版社,2003
[14]何坚勇.最优化方法.北京：清华大学出版社,2007
[15]王仁权等.附件梯级水电站群短期优化调度模型及其计算方法.云南水力发电,2002,18(1)：52-53,102
[16]詹森,巴恩斯.网络流规划.北京：科学出版社,1988,4
[17]赵子臣,夏清.水电站群补偿调节计算的非线性网络流法.北京：清华大学学报(自然科学版),1994,34(9)：86.94
[18]王海忠,于尔铿.抽水蓄能电站群短期优化调度.电网技术,1996,20(2)：35-38,46
[19]李长河.人工智能及其应用.北京：机械工业出版社,2006
[20]蔡自兴,姚莉.人工智能及其在决策系统中的应用.北京：国防科技大学出版社,2006
[21]王黎,马光文.基于遗传算法的水电站厂内经济运行新算法.中国电机工程学报,第18卷第1期,1998：64-67
[22]袁晓辉,张双全,王金文,张勇传.拟梯度遗传算法在水电厂内经济运行中的应用研究.电网技术,2000,24(12)：66-69
[23]程芳,高龙华,陈守伦.抽水蓄能电站优化调度技术研究综述.水利水电科技进展,2005,24(5)：64-67
[24]言茂松等.抽水蓄能电站随机模拟及运行优化.控制与决策,1992,135(9)：25-28
[25]王海忠,于尔铿.“协调方程式法”用于抽水蓄能电站群优化调度的研究.电网技术,1996,20(5)：24-25
[26]周燕屏,李承军,李洪斌.用动态网络规划法制定水电厂开停机计划.水电自动化与大坝监测,2003,27(3)：20-22
[27]A.阿尔塞等.伊泰普水电站水轮发电机级的优化调度.水利水电快报,2002,23(20)：1-4
[28]组合最优化-算法和复杂性.刘振宏,蔡茂诚译.北京：清华大学出版社,1988
[29]Holland J H. Adaptation in Nature and Artificial Systems. MIT Press,1992
[30]De Jong KA. An Analysis of the Behavior of a Class of Genetic Adaptive Systems. Ph.D Dissertation, University of Michigan, No.76-9381,1975
[31].Goldberg D E. Genetic Algorithms in Search, Optimization and Machine Learning. Addison-Wesley,1989
[32]文福拴,韩祯祥.模拟进化优化方法在电力系统中的应用综述(上).电力系统自动化,1996,20(1)：59-63
[33]文福拴,韩祯祥.模拟进化方法在电力系统中的应用综述(中).电力系统自动化,1996,20(2)：60-64同第1章[16]
[34]文福拴,韩祯祥.模拟进化方法在电力系统中的应用综述(下).电力系统自动化,1996,20(3)：72-75
[35]张潮海,赖红妮.遗传算法在电力系统中的应用.电力情报,1995,3：1-4
[36]刘云峰,曹春蕾.一维大地电磁测深的遗传算法反演.浙江大学学报(自然科学版),1997,31(3)：300-305
[37]杨旭东,张彤.遗传算法应用与系统在线识别研究.哈尔滨工业大学学报,2000,32(1)：102-105
[38]Mathias K E and Whitley L D. Transforming the Search Space with Gray Coding. In:Proc. Of the 1st IEEE Int'1. Conf. On Evolutionary Computation(ICEC' 94), 519-524
[39]Schraudolph NN and Belew RK. Dynamic Parameter Encoding for Genetic Algorithms. Machine Learning,1992,9(1):73-86
[40]Michalewicz Z. Gentic algorithms+Data structure=evolution program. Springer-Verlag, Berlin.3rd, Revised and Extended Edition,1996
[41]张晓馈,方浩,戴冠中.遗传算法的编码机制研究.信息与控制,1997,26(2)：134-139
[42]Goldberg DE. Genetic Algorithms in Search, Optimization and Machine Learning. Addison-Wesley Publishing Company, INC, New York.1989
[43]朱鳌鑫.遗传算法的适应度函数研究.系统工程与电子技术,1998(11)：58-62
[44]赵明旺.非线性动态系统参数估计的混合计算智能算法.应用基础与工程科学学报,1997.5(1)：97-102
[45]Vassilios P. Varying fitness functions in genetic algorithm constrained optimization. The Cutting Stock and Unit Commitment Problem. IEEE Trans Syst Man, Cybern, 1998.28 (5):629-639
[46]恽为民.基于遗传的机器人运动规划.上海：上海交通大学,1995
[47]Schaffer JD. A Study of Control Parameters Affecting Online Performance of Genetic Algorithms for Function Optimization. In:Proc. Of the 3rd Int'1. Conf. Genetic Algorithms, Morgan Kaufmann, Los Altos,1989:51-60
[48]AGOSTON EE, ROGER TH, ZBIGHIE WM. Parameter control in evolutionary algorithms. IEEE Transactions on Evolutionary Computation,1999,3(2):12-14
[49]周明,孙树栋.遗传算法原理及应用.北京：国防工业出版社,1999.6
[50]梅亚东,左园忠,朱教新一类含有0-1变量的厂内经济运行模型及解法.水电能源科学,1999,17(3)：27-30
[51]邓晓娟,权先璋,张勇传.三峡梯级水电站厂内经济运行.东北电力技术,2002,No.4：1-4
[52]路志宏,魏守平,罗元胜.基于开关控制策略的厂内经济运行模型及应用.水电自动化与大坝监测,2003,27(1)：11-13
[53]苑希民,李鸿雁,刘树坤,崔广涛.神经网络和遗传算法在水科学领域的应用.北京：中国水利水电出版社,2002.8
[54]玄光男,程润伟(著).汪定伟,唐加福,黄敏(译).遗传算法与工程设计.北京：科学出版社,2000
[55]王培萍.抽水蓄能电站与火电站联合电力系统优化运行的研究.北京工业大学硕士学位论文.2002.12
[56]肖黎,刘礼华,袁文阳.紧水滩4#机组振动试验研究[J].水电能源科学,2002,20(4)：73-75.
[57]柏勇,李德芳,孙成玲.沙河抽水蓄能电站2号机上导瓦烧损事故分析[J].水力发电,2004(5)：54-55.
[58]邱家俊.电机的机电耦联与磁固耦合非线性振动研究[J].中国电机工程学报,2002,22(5)：109-115.
[59]陶星明,吴新润.岩滩电厂楼板强振问题分析[J].大电机技术,2004(1)：35-40.
[60]姚大坤,李至昭,曲大庄.混流式水轮发机自激振动分析[J].大电机技术,1998(5)：43-47.
[61]颜庆津.数值分析.北京：北京航空航天大学出版社,1992.7
[62]贾嵘、蔡振华、罗兴锜;改进自适应遗传算法及其在水电站最优报价中的应用.水力发电学报。2007.26(1)：11-15.
[63]王培萍.抽水蓄能电站与火电站联合系统优化运行的研究.北京工业大学硕士学位论文.2002.