基于智能算法与GIS的灌溉水资源多目标优化配置

发布时间：2018-01-15 05:15

本文关键词：基于智能算法与GIS的灌溉水资源多目标优化配置　出处：《中国科学院研究生院(东北地理与农业生态研究所)》2016年博士论文　论文类型：学位论文

【摘要】：随着社会、经济的发展以及人口数量的急剧增长,人类对水资源的需求量越来越大。在水资源日益短缺的情况下,科学合理地规划与管理灌溉水资源对提高水分利用率和保障粮食安全具有十分重要的意义。研究人员对此开展了大量卓有成效的工作,但多专注于特定的方面或尺度,深入性很强,系统性、整体性较弱。本文从管理的综合性与技术的集成性出发,以智能算法在各类模型求解中的应用为主线回顾和总结了前人的研究成果,提出了用于灌溉水资源时空优化配置的多层次多尺度框架,阐明了框架内各组成部分的基础理论,并选定盈科试验灌区作为研究对象,主要取得了以下几个方面的成果:(1)灌溉水资源优化配置框架的提出本文提出了多尺度多层次的灌区灌溉水资源优化配置框架,整个框架涵盖灌区尺度到田间尺度,中间以渠系尺度作为连接,从综合管理与技术集成的角度出发来考虑如何实现灌溉水资源同时在时间与空间上达到最优化配置;可以根据具体实际情况来建立各类单目标或多目标模型,并利用现代智能算法对这些模型进行求解来制定最优配水方案,既可以从空间尺度上总体控制灌溉用水总量,也可以从时间尺度上减少灌溉水量的损失;(2)灌区尺度灌溉水量优化配置首先根据研究区实际情况和具体需求构建多目标灌溉水量优化配置模型,并利用由遥感与GIS技术获取的数据来对模型进行实例化处理,采用多目标智能算法来求解此模型。在进行灌溉水量最优配置时考虑了两种情况,第一种情况是上级渠系来水量是由传统人工计算所得,此时解算结果表明,与将这些水量按传统比例关系分配至不同作物的人工配水方案相比较,利用多目标优化配置模型可节约水量23.51%;第二种情况考虑来水量是由现代科学的方法,即基于作物需水规律及田间土壤水分等状况来计算灌溉所需水量。在此情景下,较之于传统的配水方案,水量可节约37.20%,此时,虽然优化目标之一的灌溉增产总效益与第一种方案基本持平,但优化目标之二的管理部门水费收入却减少了17.79%。因此,基于帕累托最优理论,取折中解,应基于第一种方案来制定最优灌溉计划;(3)渠系尺度最优轮灌组划分得到将灌溉水资源量分配至研究区子区域不同种类作物的最优化方案之后,研究如何将这些灌溉水资源量通过各级渠系输送至研究区需水最末端,即在渠系尺度上对灌溉水资源进行优化分配,采用“组间续灌,组内轮灌”的工作方式,选择总配水时间最短与轮灌组之间引水持续时间差异值最小同时作为优化目标,来构建多目标渠系优化配水模型。研究结果表明通过渠系优化配水模型,不论是基于多目标粒子群算法,还是多目标蚁群算法,其在求解多目标渠系优化问题时所寻找到的最优轮灌组合,并依此来制定的灌溉计划均优于传统的人工方式。较之于传统人工方式所规定的轮灌周期,优化后的轮灌周期可节约时间约32.44%,且在研究中将管理部门此前未曾考虑的各轮灌组之间引水持续差异值最小这一问题也作为一个需要优化的目标;(4)基于GIS的灌溉水资源管理系统在灌区尺度和渠系尺度的灌溉水资源得到优化配置的基础上,为提升灌区管理与决策水平,更加直观、高效地对灌溉水资源进行优化配置,基于GIS技术,整合灌溉水资源多目标优化配置模型以及多目标智能算法,来研制开发基于GIS的灌溉水资源管理系统。既包括地理信息系统基本功能的实现,也反应了行业应用的特点,具有针对灌溉水资源管理的专业功能;将系统应用于灌区日常运营管理中,可以提升使用者的管理水平与科学决策能力,提高灌溉水资源利用率。基于系统性策略来研究灌溉水资源优化配置问题,既可以从时间尺度上总体控制灌溉用水总量,也可以从空间尺度上减少灌溉水量的损失,为提升灌区灌溉水资源管理水平,提高水分生产率以保障粮食安全提供科学依据与技术支撑。
[Abstract]:With the social and economic development and rapid population growth, human demand for water resources is more and more big. In the growing shortage of water resources, scientific and rational planning of water resources management and irrigation to improve water use efficiency and food security is of great significance to carry out a large number of researchers. Very fruitful work, but more focused on specific aspects or dimensions, depth is very strong, systematic, holistic weak. The integration of comprehensive technology and management of the intelligent algorithm is applied in all kinds of solution of the model as the main line to review and summarize the previous research results, for multi-level scale framework of irrigation water resources optimization disposition is proposed, clarifies the basic theory of the framework, and selected's irrigation area as the research object, mainly made the following Achievements: (1) irrigation water resources optimal allocation framework proposed by this paper puts forward the optimization of irrigation water resources allocation framework of multi scale and multi levels, the whole framework covering the irrigation to the field scale, as the middle connected by canal scale, from the perspective of comprehensive management and technology integration in view of how to achieve the irrigation water resources at the same time to achieve optimal allocation in time and space; to build all kinds of single target or multi-target model according to the actual situation, and to solve these models in order to develop optimal water allocation scheme using modern intelligent algorithm, which can control the overall irrigation water volume from the spatial scale, can also reduce the amount of irrigation water from the time scale the loss; (2) configuration according to the actual situation of the study area and the specific needs of constructing the multi-objective optimization model of optimal irrigation water irrigation irrigation, And using the obtained by remote sensing and GIS technology for data processing of model instantiation, the multi-objective intelligent algorithm to solve this model. The two cases considered in the optimal allocation of irrigation water, the first is the superior canal water is obtained from the traditional manual calculation, the calculation results show that with the the water by artificial water distribution scheme traditional proportion allocated to different crops compared to the optimal allocation model of water saving by 23.51% multi objectives; second cases considered to water is determined by modern scientific methods, namely crop water requirement and soil water status based on irrigation water required for this. Under the situation, compared with the traditional water allocation scheme, water can be saved by 37.20%, at this time, although the optimization goal of irrigation benefit and the first scheme is basically the same, but the optimization goal two tube Department of water revenue has decreased 17.79%. therefore, based on Pareto optimality theory, take compromise solution, should be the first to formulate the scheme based on the optimal irrigation plan; (3) the optimal irrigation canal scale group divided by the irrigation allocation of water resources to study area sub area crops of different kinds of optimization scheme, study how to the irrigation water resources through all levels of canal system is transported to the study area to the end of the water, water resources in irrigation canal system scale optimization allocation, the group continued irrigation, irrigation group "work, choose the shortest time and the total water irrigation diversion duration differences between groups the minimum value at the same time as the optimization goal, to build a multi-objective optimal allocation of canal water model. The results of the study show that the optimal allocation of canal water model, whether it is a multi-objective particle swarm algorithm based on multi-objective ant colony algorithm, or, in o The optimal solution of multiobjective optimization problem of canal wheel for irrigation combination, and in order to make the artificial irrigation schemes are better than the traditional artificial way. By compared with the traditional irrigation cycle, optimized irrigation cycle time can be saved about 32.44%, and in the study of the management department has not previously considered the continuous irrigation diversion group differences between the minimum value of this problem as an optimization target; (4) based on irrigation water resources management system of GIS based optimal allocation of irrigation water resources in the irrigation canal system and scale, to improve the irrigation management and decision-making level, more intuitive, are the optimal allocation of irrigation water resources efficiently, based on the GIS technology, the integration of irrigation water resources multi-objective optimization model and multi-objective intelligent algorithm developed based on GIS irrigation water resources management system. Both the packet The realization of the basic functions of the geographic information system, also reflects the characteristics of industry application, is for the professional function of irrigation water resources management; the system is applied to the daily operation of irrigation management, can improve the management level of users and scientific decision-making ability, improve the utilization rate of irrigation water resources. The system of strategy based on the research of irrigation water resources optimal allocation problem, not only can the overall control of irrigation water from the total time scale, but also from the spatial scale to reduce irrigation water loss, to enhance the level of irrigation water management in irrigation area, improve water productivity in order to ensure food safety and provide scientific basis and technical support.

【学位授予单位】：中国科学院研究生院(东北地理与农业生态研究所)
【学位级别】：博士
【学位授予年份】：2016
【分类号】：S274

【参考文献】