基于改进遗传算法的供水管网多工况优化设计
|
摘要
随着城市的扩大化发展以及供水管理水平的不断提高,供水管网系统的规划设计相比之前也有了高的起点和要求,因此正急需新的规划设计方法或优化设计。
针对现有的供水管网优化模型,通常是基于最高日最高时单一工况,忽略其他工况如平均时、消防时、事故时等运行工况的约束条件,不属于全局优化的结果。在已有的研究基础上,考虑到现有的多目标函数中的水质目标、安全可靠性目标的量化很含糊和不成熟性,本文提出了在满足管网各个用水工况不同的约束条件基础上,以管网造价和泵站运行费用之和的管网年费用折算值最低为单一目标的给水管网多工况优化设计数学模型,该模型中多工况的选择包括最高日最高时、平均时、消防时、事故工况时等四个工况,且模型中引入了不同工况条件下水泵机组运行时间占全年运行时间的比例。
在选用模型求解算法时仔细分析研究了常用的求解算法,如拉格朗日函数法、界限流量法、线性规划算法、非线性规划法、动态规划法、遗传算法等。最后考虑到管网优化模型中管径这一决策变量的离散性,以及多工况优化模型的研究还比较不成熟,本文暂时在比较成熟的遗传算法的基础上做了简单的改进得到的改进的遗传算法,并将其运用到实际工程中的多工况优化设计模型中。论文使用MATLAB编制供水管网优化算法的求解程序。最后通过一个实际工程的供水管网优化设计实例,进行传统遗传算法和改进遗传算法优化结果比较,证实了改进遗传算法优化结果的优越性,同时也验证了多工况优化模型的优势。?
As the city enlargement and supply management level of development of unceasing enhancement, water supply network system planning design before also had high compared to the starting point and requirements, therefore are urgently needed new planning methods or optimization design.
Current network optimization design are usually based on a single modes of the single objective or multi-objective optimization model, ignoring other conditions such as an average, fire, such as the operation condition of accidents constraints, so that those optimization models do not belong to global optimization results. Moreover, the some target of existing multi-objective function optimization models, such as safety and reliability of the quantification target, are very ambiguous and immaturity. Based on the existing research, this paper puts forward the multiple operations optimal design model of water distribution network, which select the sum of network cost and operation cost of the pipeline pump fees for a single target conversion value the lowest water networks working single target, can meet different constraint of different water conditions. This model think about four conditions and account for pump unit running time running time scale. Four conditions are peak conditions,average conditions, fire condition , accident condition.
This paper analyzed and studied carefully used common solving algorithm, such as Lagrange function method, boundaries flow method, linear programming method, nonlinear programming method, the dynamic programming, genetic algorithms, etc. Finally considering the discreteness of pipe diameter, this paper temporarily do some improve based on mature genetic algorithm to get the improved genetic algorithm, which was applied to the practical engineering working optimization design model. By using MATLAB compiling water supply network optimization algorithm of solving program. Finally, through a practical engineering example of water distribution network optimization design, with the comparison of the optimization results between the traditional genetic algorithm and improved genetic algorithm, confirmed the improved genetic algorithm, and the optimized results verified superiority and the advantage of working optimization model.?
针对现有的供水管网优化模型,通常是基于最高日最高时单一工况,忽略其他工况如平均时、消防时、事故时等运行工况的约束条件,不属于全局优化的结果。在已有的研究基础上,考虑到现有的多目标函数中的水质目标、安全可靠性目标的量化很含糊和不成熟性,本文提出了在满足管网各个用水工况不同的约束条件基础上,以管网造价和泵站运行费用之和的管网年费用折算值最低为单一目标的给水管网多工况优化设计数学模型,该模型中多工况的选择包括最高日最高时、平均时、消防时、事故工况时等四个工况,且模型中引入了不同工况条件下水泵机组运行时间占全年运行时间的比例。
在选用模型求解算法时仔细分析研究了常用的求解算法,如拉格朗日函数法、界限流量法、线性规划算法、非线性规划法、动态规划法、遗传算法等。最后考虑到管网优化模型中管径这一决策变量的离散性,以及多工况优化模型的研究还比较不成熟,本文暂时在比较成熟的遗传算法的基础上做了简单的改进得到的改进的遗传算法,并将其运用到实际工程中的多工况优化设计模型中。论文使用MATLAB编制供水管网优化算法的求解程序。最后通过一个实际工程的供水管网优化设计实例,进行传统遗传算法和改进遗传算法优化结果比较,证实了改进遗传算法优化结果的优越性,同时也验证了多工况优化模型的优势。?
As the city enlargement and supply management level of development of unceasing enhancement, water supply network system planning design before also had high compared to the starting point and requirements, therefore are urgently needed new planning methods or optimization design.
Current network optimization design are usually based on a single modes of the single objective or multi-objective optimization model, ignoring other conditions such as an average, fire, such as the operation condition of accidents constraints, so that those optimization models do not belong to global optimization results. Moreover, the some target of existing multi-objective function optimization models, such as safety and reliability of the quantification target, are very ambiguous and immaturity. Based on the existing research, this paper puts forward the multiple operations optimal design model of water distribution network, which select the sum of network cost and operation cost of the pipeline pump fees for a single target conversion value the lowest water networks working single target, can meet different constraint of different water conditions. This model think about four conditions and account for pump unit running time running time scale. Four conditions are peak conditions,average conditions, fire condition , accident condition.
This paper analyzed and studied carefully used common solving algorithm, such as Lagrange function method, boundaries flow method, linear programming method, nonlinear programming method, the dynamic programming, genetic algorithms, etc. Finally considering the discreteness of pipe diameter, this paper temporarily do some improve based on mature genetic algorithm to get the improved genetic algorithm, which was applied to the practical engineering working optimization design model. By using MATLAB compiling water supply network optimization algorithm of solving program. Finally, through a practical engineering example of water distribution network optimization design, with the comparison of the optimization results between the traditional genetic algorithm and improved genetic algorithm, confirmed the improved genetic algorithm, and the optimized results verified superiority and the advantage of working optimization model.?
引文
[1]严熙世,刘遂庆,龙腾锐.给水排水管网系统[M].北京:中国建筑工业出版社,2000,3-3
[2]杨玉奎,唐剑晖.规划设计中的给水管网计算方法的改进[J].广州大学报(自然科学版) ,2004,3(5): 466-469
[3]董文平,李红卫,吕谋等.城市排水系统的优化设计研究[J].青岛理工大学学报,2006,27(1): 113-116
[4]周云,王兆平.给水管网优化设计[J] .兰州铁道学院学报,1995, 15(1):27-35
[5]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): 363-367
[6]李树平.进化算法在排水管道系统优化设计中的应用[J].中国给水排水,2003,19(7): 74‐76
[7]中国城镇供水协会.城市供水行业2010年技术进步发展规划及2020年远景目标[M].北京:中国建筑工业出版社,2005,14-18
[8] D. Karmeli, Y. Gadish, S.Meyers.Design of Optimal Dislribution Network[J]. Journal of Pipeline,ASCE,1968,94(PLI) : 1-10
[9] E.rally. Computerized Planning of the Least Cost Water Distribution Network[J]. WaterSewage Works,1972,15(1),121-127
[10] E. Alperovits, U. Shamir. Design of Optimal Water Distribution Systems[J].Water ResouceReasercIl,1977,13(6) : 885-900
[11] U. Shamir. Optimal Design and Opration of Water Distribution Systems[J].Water ResoucesResarch,1974,10(1) : 27-36
[12]王荣和.给水管网系统多工况优化设计及拟稳定状态水力模拟研究(工程实用软件开发) [D] .上海:同济大学,1998,13-16
[13]董深,吕谋,陆海.基于遗传算法给水管网优化模型的改进研究[J].中国给水排水,2007, 23(17) : 1-4
[14]吕谋,董深,王磊.基于遗传算法的给水管网多工况优化设计[J] .水利学报,2007,38(12) : 1507-1511
[15] T. T. Tanyimboth. Caculating Maximum Entropy Flows in Networks [J]. Operational Research Society,1993,44(4) : 383-396
[16] T. M. Klebber, F. H. Formiga, P.B.C.Chaudhry. Optimal Design of Water Distribution System by Multi-objective Evolutionary Methods EvolutionaryMulti-Criterion Optimization: Second International Conference. 2003,15(4): 3-6
[17]蒋怀德.给水管网多目标优化设计[D].上海:同济大学环境科学与工程学院.2007: 33-36
[18] D. Karmeli, Y. Gadish, S. Meyers. Design of Optimal Distribution Network[J] . Journal of Pipeline, ASCE,1968,94(PLI) : 1-10
[19] E. rally. Computerized Planning of the Least Cost Water Distribution Network[J] . Water Sewage Works,1972,121-127
[20] Bhave. Noncomputer Optimization of single-source Networks, Journal of Hydraulic Division[J]. 1979,V01.104: 799-813
[21]周恒良.模糊线性规划在供水管网优化中的应用[J].安徽理工大学学报(自然科学版),2005(2) : 21‐23
[22]白丹.树状给水管网的优化[J].水利学报,1996,(11) : 52-56
[23] Jacoby. Design of optimal hydraulic networks Journal of Hydraulic DiviSion[J]. ASCE, 1968, V01. 94: 641-661
[24]王新坤,程冬玲等.干早地区农业研究[J]. 2001No.2:61‐66
[25] T. Liang. Design of conduit system by dynamic Programming, Journal of Hydraulic, 1971, 97(HY3) : 167-187
[26]高飞.管道造价的曲线拟合及其精度[J].给水排水, 1994, 20(1) : 12-14
[27]李树平,黄廷林,刘遂庆.用麦夸尔特法推求给水管道造价公式参数[J].西安建筑科技大学学报, 1999, 32(1) : 16-19
[28] A. Ostfeld. Optimal Design and Operation of Multiquality Networks under Unsteady Conditions [J] .Joumal of Water Resources Planning and Manegement, ASCE 2005, 131(2) : 116-124
[29]张增荣.给水管网单目标多工况优化设计的研究与应用[D].同济大学环境科学与工程学院.2008:9-15
[30]储诚山.改进混合遗传算法用于给水管网优化设计的研究[D].天津大学环境科学与工程.2006:33-38
[31]许可,张鸿涛,郭强等.赤峰市平庄城区给水管网改造工程优化设计研究[J].给水排水, 2007,33(增刊): 343-346
[32]上海建设与交通委员会.室外消防用水规范[M].北京:中国计划出版社,2006:112-113
[33] Haestad Methods.WaterCAD5.0 Help[M] .2002:55-78
[34] A.Tuber Saud, W.Labadie John. 0ptimal design of water—distribution networks with GIS[J] .Journal of Water Resources Planning and Management,1996,122(4) : 301-311
[35] D. E. Glodberg, C, H. Kuo. Genetic algorithms in pipeline optimization[J]. Computing in Civil Engineering,1987,1(2) : 128-141
[36]雷英杰,张善文,李绪武等.Matlab遗传算法工具箱及应用[M] .西安:西安电子科技大学出版社.2005:44‐50
[37] J E.Baker. Reducing Bias And Inefficiency in the Selection Algorithm. Proc ICGA 2,1987.16(3) : 14-21
[38]周克民,胡云昌.遗传算法计算效率的改进[J].控制理论与应用,2002,19(5):812-813
[39]李书全,遗传算法性能分析及其应用研究[D].天津:天津大学,1998:22‐34
[40]周荣敏,雷延峰,周芳等.给水管网遗传优化设计中惩罚系数的确定[J].中国给水排水,2008,24(6),44-47
[41]周荣敏,林性粹.用基于整数编码的管径遗传算法进行环状管网优化设计[J].灌溉排水,2001,20(3) : 49-52
[42]姚宇,俞国平,胡少华.基于MATLAB的给水管网可视化模型和仿真计算[J] .苏州科技学院学报工程技术版.2006.19(2) : 54-59
[43] J E. Baker. Adaptive Selection Methods for Genetic Algorithms. Proc ICGA 1,1985: 101-111
[2]杨玉奎,唐剑晖.规划设计中的给水管网计算方法的改进[J].广州大学报(自然科学版) ,2004,3(5): 466-469
[3]董文平,李红卫,吕谋等.城市排水系统的优化设计研究[J].青岛理工大学学报,2006,27(1): 113-116
[4]周云,王兆平.给水管网优化设计[J] .兰州铁道学院学报,1995, 15(1):27-35
[5]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): 363-367
[6]李树平.进化算法在排水管道系统优化设计中的应用[J].中国给水排水,2003,19(7): 74‐76
[7]中国城镇供水协会.城市供水行业2010年技术进步发展规划及2020年远景目标[M].北京:中国建筑工业出版社,2005,14-18
[8] D. Karmeli, Y. Gadish, S.Meyers.Design of Optimal Dislribution Network[J]. Journal of Pipeline,ASCE,1968,94(PLI) : 1-10
[9] E.rally. Computerized Planning of the Least Cost Water Distribution Network[J]. WaterSewage Works,1972,15(1),121-127
[10] E. Alperovits, U. Shamir. Design of Optimal Water Distribution Systems[J].Water ResouceReasercIl,1977,13(6) : 885-900
[11] U. Shamir. Optimal Design and Opration of Water Distribution Systems[J].Water ResoucesResarch,1974,10(1) : 27-36
[12]王荣和.给水管网系统多工况优化设计及拟稳定状态水力模拟研究(工程实用软件开发) [D] .上海:同济大学,1998,13-16
[13]董深,吕谋,陆海.基于遗传算法给水管网优化模型的改进研究[J].中国给水排水,2007, 23(17) : 1-4
[14]吕谋,董深,王磊.基于遗传算法的给水管网多工况优化设计[J] .水利学报,2007,38(12) : 1507-1511
[15] T. T. Tanyimboth. Caculating Maximum Entropy Flows in Networks [J]. Operational Research Society,1993,44(4) : 383-396
[16] T. M. Klebber, F. H. Formiga, P.B.C.Chaudhry. Optimal Design of Water Distribution System by Multi-objective Evolutionary Methods EvolutionaryMulti-Criterion Optimization: Second International Conference. 2003,15(4): 3-6
[17]蒋怀德.给水管网多目标优化设计[D].上海:同济大学环境科学与工程学院.2007: 33-36
[18] D. Karmeli, Y. Gadish, S. Meyers. Design of Optimal Distribution Network[J] . Journal of Pipeline, ASCE,1968,94(PLI) : 1-10
[19] E. rally. Computerized Planning of the Least Cost Water Distribution Network[J] . Water Sewage Works,1972,121-127
[20] Bhave. Noncomputer Optimization of single-source Networks, Journal of Hydraulic Division[J]. 1979,V01.104: 799-813
[21]周恒良.模糊线性规划在供水管网优化中的应用[J].安徽理工大学学报(自然科学版),2005(2) : 21‐23
[22]白丹.树状给水管网的优化[J].水利学报,1996,(11) : 52-56
[23] Jacoby. Design of optimal hydraulic networks Journal of Hydraulic DiviSion[J]. ASCE, 1968, V01. 94: 641-661
[24]王新坤,程冬玲等.干早地区农业研究[J]. 2001No.2:61‐66
[25] T. Liang. Design of conduit system by dynamic Programming, Journal of Hydraulic, 1971, 97(HY3) : 167-187
[26]高飞.管道造价的曲线拟合及其精度[J].给水排水, 1994, 20(1) : 12-14
[27]李树平,黄廷林,刘遂庆.用麦夸尔特法推求给水管道造价公式参数[J].西安建筑科技大学学报, 1999, 32(1) : 16-19
[28] A. Ostfeld. Optimal Design and Operation of Multiquality Networks under Unsteady Conditions [J] .Joumal of Water Resources Planning and Manegement, ASCE 2005, 131(2) : 116-124
[29]张增荣.给水管网单目标多工况优化设计的研究与应用[D].同济大学环境科学与工程学院.2008:9-15
[30]储诚山.改进混合遗传算法用于给水管网优化设计的研究[D].天津大学环境科学与工程.2006:33-38
[31]许可,张鸿涛,郭强等.赤峰市平庄城区给水管网改造工程优化设计研究[J].给水排水, 2007,33(增刊): 343-346
[32]上海建设与交通委员会.室外消防用水规范[M].北京:中国计划出版社,2006:112-113
[33] Haestad Methods.WaterCAD5.0 Help[M] .2002:55-78
[34] A.Tuber Saud, W.Labadie John. 0ptimal design of water—distribution networks with GIS[J] .Journal of Water Resources Planning and Management,1996,122(4) : 301-311
[35] D. E. Glodberg, C, H. Kuo. Genetic algorithms in pipeline optimization[J]. Computing in Civil Engineering,1987,1(2) : 128-141
[36]雷英杰,张善文,李绪武等.Matlab遗传算法工具箱及应用[M] .西安:西安电子科技大学出版社.2005:44‐50
[37] J E.Baker. Reducing Bias And Inefficiency in the Selection Algorithm. Proc ICGA 2,1987.16(3) : 14-21
[38]周克民,胡云昌.遗传算法计算效率的改进[J].控制理论与应用,2002,19(5):812-813
[39]李书全,遗传算法性能分析及其应用研究[D].天津:天津大学,1998:22‐34
[40]周荣敏,雷延峰,周芳等.给水管网遗传优化设计中惩罚系数的确定[J].中国给水排水,2008,24(6),44-47
[41]周荣敏,林性粹.用基于整数编码的管径遗传算法进行环状管网优化设计[J].灌溉排水,2001,20(3) : 49-52
[42]姚宇,俞国平,胡少华.基于MATLAB的给水管网可视化模型和仿真计算[J] .苏州科技学院学报工程技术版.2006.19(2) : 54-59
[43] J E. Baker. Adaptive Selection Methods for Genetic Algorithms. Proc ICGA 1,1985: 101-111
© 2004-2018 中国地质图书馆版权所有 京ICP备05064691号 京公网安备11010802017129号 地址：北京市海淀区学院路29号 邮编：100083 电话：办公室：(+86 10)66554848；文献借阅、咨询服务、科技查新：66554700 |