Optimal decision schemes for agricultural water quality management planning with imprecise objective

Agricultural activities are the main sources of water pollution to surface water and groundwater in rural areas. Extensive soil disturbance and application of fertilizer and manure in agriculture cause nonpoint source losses of soil and nutrients such as nitrogen and phosphorus. How to generate preferred decision schemes for agricultural activities that cause such nonpoint source water pollution is a critical issue for the decision makers. In this study, an inexact agricultural water quality management (IAWQM) model is developed and applied to a case study to generate optimal decision schemes for integrated water quality management within an agricultural system. The model is based on a hybrid fuzzy possibilistic robust programming approach, which improves upon the existing fuzzy possibilistic programming and fuzzy robust programming methods by allowing fuzzy information in the model's objective and constraints to be directly communicated into the optimization processes and resulting solutions. Optimal decision schemes for agricultural activities can be generated, including cropping area, manure/fertilizer applied, and livestock husbandry size, where implications of water quality/quantity restrictions for achieving regional economic development objectives are studied. The results of the case study indicate that useful information can be obtained through the proposed IAWQM model for providing feasible decision schemes, which reflect tradeoffs between economic and environmental considerations. The decision variables are useful for the decision makers to justify and/or adjust the decision schemes for agricultural activities through incorporation of their implicit knowledge on water quality management.     

FSWM: A hybrid fuzzy-stochastic water-management model for agricultural sustainability under uncertainty

A hybrid fuzzy-stochastic water-management (FSWM) model is developed for agricultural sustainability under uncertainty, based on advancement of a multistage fuzzy-stochastic quadratic programming (MFSQP) approach. In MFSQP, uncertainties presented in terms of fuzziness and randomness can be incorporated within a multilayer scenario tree, such that revised decisions are permitted in each time period based on the realized values of the uncertain events. Moreover, fuzzy quadratic terms are used in the objective function to minimize the variation of satisfaction degrees among the constraints; it allows an increased flexibility in controlling the system risk in the optimization process. Results of the case study indicate that useful solutions for the planning of agricultural water management have been obtained. In the FSWM model, a number of policies for agricultural water supply are conducted. The results obtained can help decision makers to identify desired water-allocation schemes for agricultural sustainability under uncertainty, particularly when limited water resources are available for multiple competing users.     

基于双层分式规划的种植结构多目标模型研究

构建了基于双层分式规划的种植结构多目标优化模型,并通过交互式模糊规划方法进行求解。该模型可以定量得到研究区域在最小灌溉水量情况下的最大种植收益,避免多目标模型求解过程中的主观性;可以兼顾并综合上层管理者和下层管理者的种植方案决策。将构建的模型及相应求解方法应用于黑河中游甘州区、临泽县、高台县的种植结构优化中,并将优化结果与实际结果进行对比,结果发现:优化的种植结构整体效益增加3.2×108元,整体用水量减少1.3×109m3,单方水种植效益比实际增加了1.94元/m3,得到的结果验证了模型及求解方法的可行性。构建的模型与得到的优化结果可为黑河中游及类似地区作物种植结构优化提供理论依据与决策支持。     

基于2型模糊集的多目标农业-生态水土资源优化配置

针对西北干旱区灌区生态环境脆弱、水资源短缺、复杂不确定性等问题,以石羊河流域红崖山灌区为例,耦合2型模糊集、模糊可信度约束规划和多目标规划等理论方法,构建了基于2型模糊集的多目标农业-生态水土资源优化配置模型。模型以灌溉水损失最小、生态植被灌溉水满意度最大、生态植被灌溉水费用最小和主要粮食作物经济效益最大为目标,对红崖山灌区10个决策单元的地表水、地下水和粮食作物种植面积进行优化配置。求解模型得到不同可信度水平和不确定性程度下的水土资源优化配置方案。结果表明：耦合2型模糊集的模型能够提供丰富的配置方案,水量对可信度水平的敏感性高于不确定性程度,作物种植结构对可信度水平不敏感。以不确定性程度参数为0.5、可信度水平为0.7时为例,生态植被均通过地表水灌溉,作物通过地表水、地下水联合灌溉,玉米的产量和经济效益均大于小麦。相比前人研究,本研究考虑生态植被灌溉需求,优化结果更加真实合理。本研究可为决策者提供较为符合灌区实际的配置方案,为西北干旱区灌区现代化建设提供科学指导。     

额济纳地区水资源利用模式优选探

水资源是额济纳地区最具有战略意义的资源，水资源短缺是制约该地区发展的至关因素，合理有效地利用水资源对额济纳实现整体规划和“2010年远景目标”具有决定性意义。水资源的优化利用在保证满足优势产业结构用水的基础上，利用多目标线性规划模型，以农牧业产业结构为主优化因子，并考虑其它产业用水情况下，给出了水资源在最枯状态、平均状态供水量及“甘蒙”分水方案实施后的配水量进行优化配置的方案集。应用层次分析法对多目标线性规划给出的方案集做出最终决策，为该地区水资源的合理利用提供决策依据。     

黑河中游区农业节水模式研究及其

黑河中游区农业节水模式研究，是以农业节水措施的组合为变量，以区域经济效益最大为目标函数，采用多目标线性规划的方法，确定各种节水组合措施的发展面积及其经济效益，供政府行政部门进行决策，指导农业生产规划和农民生产、经营。     

基于多目标模糊规划的灌区多水源优化配置

以黑龙江省和平灌区为例,构建基于多目标模糊规划的灌区多水源优化配置模型,该模型能够在提高农业灌溉用水净效益的同时有效减少农业灌溉水量,促进和平灌区多水源高效配置。采用具有非线性隶属度函数的模糊多目标规划求解模型,得到不同流量不同水源下的最优配水方案。结果表明:不同流量水平下水稻不同生育阶段均存在缺水现象,低流量下需从柳河水库引入外调水才能保证水稻的最小需水量。为保证灌区整体效益,按照引水工程、提水工程、井灌工程的先后顺序进行配水,并得到多目标配水模型在不同情景下的运行稳定情况。该模型可以高效地进行灌区多水源在作物各生育阶段的优化配置。     

基于模糊几何加权模型的灌区水资源优化配置

针对农业水资源优化配置中多目标赋权的不确定性问题,以黑龙江省绰尔屯灌区为研究区域,构建了基于模糊几何加权法的灌区水资源优化配置模型,把多目标规划问题转化为单目标非线性问题,得到了不同权重下不同生育期的灌区水资源最佳配置方案,并通过灌区综合能力评价体系进行比选.结果表明：通过与线性多目标模糊规划模型的对比,模糊几何加权模型可以根据实际情况对不同目标的权重进行调整,得出该条件下的灌区水资源最佳配置方案;当灌水量权重为0.3和经济权重为0.7时为最优方案,水稻产值为8 245万元,分蘖期、拔节期、抽穗期、乳熟期的灌水量分别为1 154.56万、779.12万、539.45万、336.57万m³.验证了模糊几何加权法在灌区水资源优化配置的可行性应用,为灌区农业水资源高效利用提供参考方案和决策支持.     

温室小型西瓜调亏灌溉综合效益评价模型

应用多层次多目标模糊理论与方法,建立调亏灌溉综合效益多层次多目标模糊评价模型;利用信息熵理论求得各层评价指标客观熵权,结合专家法的主观权重获得模型各层评价指标综合权重,并运用该模型对2007年秋季温室小型西瓜不同亏水处理的综合效益进行评价,结果表明,在西瓜果实膨大期进行水分亏缺处理可以一定程度上提高果实的品质,但重度和中度的水分亏缺会降低果实产量,在果实成熟期采用一定程度的水分亏缺对果实品质和产量影响不大,在苗期、开花坐果期、果实膨大期和果实成熟期分别采用0.75Ep、0.75Ep、1.25Ep和1.00Ep的灌溉标准可以获得较为理想的调亏灌溉综合效益.     

自压喷灌输水管的线性规划模型

以标准管径的管长为优化变量,建立自压喷灌输水管优化设计的线性规划模型,当喷灌区边界已定条件下,利用这一模型可计算出各标准管径的最优管长,并可确定输水管首端位置.     

基于模糊定权的模糊可信性约束二次规划模型

为了最大化黑河中游作物产生的经济效益和社会效益,采用模糊定权的方法将多目标效益转化为作物的综合效益权重,并将该权重应用于作物的水分生产函数中,对甘州区、临泽县和高台县的作物(小麦和玉米)进行水量优化配置.该模型为模糊可信性约束二次规划模型,模糊可信性约束中采用三角模糊数,可信性置信水平分别取0.5,0.6,0.7,0.8,0.9和1.0.以2011年的水量优化为例,优化后较优化前作物的配水量减少了3 701.25万m³,产量增加了1 291.28万kg,经济效益和社会效益分别提升了0.28亿元和0.18万人.另外,规划年2020年作物水量配置较现状年(2011年)的小麦和玉米的配水量分别减少了4.98%~5.24%和11.01%~12.07%,缺水情况得到改善.不同条件下的配水结果能够给决策者提供不同的方案.     

基于熵权的导流标准多目标决策分析

通过分析决策方案目标集．采用信息熵权与决策者的主观权重相结合的方法确定综合权重．运用多目标决策理论，建立导流标准多目标模糊优选模型。该模型系统地考虑了决策方案目标集的客观因素和决策者的主观因素。通过锦屏一级水电站施工导流标准的决策分析，说明该方法和模型是有效的．很好地解决了导流工程的施工强度、投资及风险损失的关系。     

基于改进多目标决策模型的大桥水库灌区渠系自适应规划方法

提出基于改进多目标决策模型的大桥水库灌区渠系自适应规划方法,合理规划大桥水库灌区渠系配水,提高水资源利用率,降低无效弃水以及对生态环境的影响。以配水结束后灌区各渠系缺水量、渠道输水损失量以及农作物生产带来的灰水足迹三者最小为大桥水库灌区渠系规划多目标决策函数,以渠道输水能力、水量和时间为约束条件,构建大桥水库灌区渠系规划的多目标决策模型;以栅格法构建路径规划的运行环境,通过移动几率优化、信息素挥发系数自适应调整两方面改进蚁群算法,通过改进蚁群算法寻优获取渠道缺水量、输水损失量、灰水足迹最小的水库灌区渠系规划结果。实验证明：该方法可以有效规划大桥水库灌区渠系的水资源,规划后的灌区渠系在缺水量、渠道输水损失量以及灰水足迹方面都有较好的表现,且效率高、应用性强。     

基于信息熵的农村饮水安全评价组合权重模型

农村饮水安全评价可为政府的基础设施建设和管理政策提供科学的决策依据。为合理地确定评价指标的权重,在评价指标体系基础上,把评价指标的权重结构分解为反映评价指标不同属性对饮水安全保障的影响程度的主观权重,和反映各区域评价指标样本值差异信息对饮水安全保障的影响程度的客观权重,并用基于加速遗传算法的模糊层次分析法确定这些权重,根据最小相对信息熵原理把主观权重和客观权重综合为组合权重,根据饮水安全管理的全面、协调要求建立了基于乘法合成法的农村饮水安全评价组合权重模型(CWMRE)。CWMRE的应用结果说明:用CWMRE可充分挖掘指标中的专家主观经验信息和各区域评价指标样本值的客观差异信息,评价结果合理,方法通用,在我国农村饮水安全评价中具有一定的应用价值。     

考虑不同层次利益主体的灌溉水资源优化配置

针对灌溉水资源优化配置中存在的非线性和不确定性等特点,同时考虑灌区不同层次决策主体利益,分别构建考虑上层管理者利益的区间线性分式规划(ILFP)模型以获得最大的灌溉水生产力和考虑下层农民利益的区间二次规划(IQP)模型以获得最大的产量。在此基础上,将灌区上、下层利益主体作为整体,构建线性分式二次双层规划(LFQBP)模型,以协调灌区不同层次决策主体利益,促进灌区可持续发展。将所构建模型应用于盈科灌区的粮食作物配水中。通过配水结果的比较来分析各模型的性质,结果表明ILFP模型和LFQBP模型更适用于干旱地区。所构建模型从不同角度反映灌溉配水的实际问题,其优化结果有助于灌区管理者权衡各层决策主体间的利益。     

Performance based irrigation planning under water shortage

A linear programming (LP) based optimization model and a simulation model are developed and applied in a typical diversion type irrigation system for land and water allocation during the dry season. Optimum cropping patterns for different management strategies are obtained by the LP model for different irrigation efficiencies and water availability scenarios. The simulation model yields the risk-related irrigation system performance measures (i.e. reliability, resiliency and vulnerability) for the management policies defined by the optimization model. The alternative strategies are evaluated in terms of all performance criteria (i.e. net economic benefit, equity and reliability) simultaneously through a trade-off analysis using a multi-criteria decision making method (compromise programming). For the case study of the Kankai irrigation system in Nepal, with equal preference to the objectives, a management strategy with equal share of water among the project subareas appears to be the most satisfactory alternative under water shortage conditions. The existing water allocation policy is not economically efficient. Deficit irrigation in Early paddy appears attractive under favorable hydrologic scenario, particularly if accompanied by measures to improve existing irrigation system efficiency.     

基于多目标模糊优化模型的兰州市水资源优化配置研究

构建可持续发展视角下的兰州市水资源优化配置模型。以净效益系数、水利用系数和供水成本为指标,运用熵值法修正由层次法获得的各项指标的权重,并采用多目标模糊线性优化模型确定各用水部门的综合效益系数,代入线性规划法求解规划年2020年在不同用水保证率(多年平均、50%、75%、90%)下各用水部门分配水量。分配结果(多年平均)与预测结果相比工业用水量减少了1.363 5亿m~3,农业用水量和生态用水量分别增加了1.126 0和0.237 5亿m~3,该分配结果有利于实现兰州区域经济的可持续发展与水生态的协调发展。     

基于Pareto蚁群算法和3S技术的灌区水资源空间优化配置

为了解决灌区水资源配置中的信息采集方法落后和多目标模型求解易陷入局部最优等问题,通过伪随机并行搜索和最近邻域选择等策略对Pareto蚁群算法进行改进,然后将Pareto蚁群算法与3S技术耦合,以宁夏银北灌区为仿真对象,利用改进的Pareto蚁群算法计算灌区多目标、多约束的水资源空间优化配置模型,获得最优的井灌水量、渠灌水量、排水量以及最优的经济效益和灌排费用。基于Pareto蚁群算法和３S技术的灌区水资源空间优化配置,能加快算法的收敛速度,获得全局最优解,使优化结果更接近实际,对灌区的水资源宏观调度具有一定的参考价值。     

Evolving strategies for crop planning and operation of irrigation reservoir system using multi-objective differential evolution

In this paper multi-objective differential evolution (MODE) approach is proposed for the simultaneous evolution of optimal cropping pattern and operation policies for a multi-crop irrigation reservoir system. In general, farming community wants to maximize total net benefits by irrigating high economic value crops over larger area, which may also include water-intensive crops and longer duration crops. This poses a serious problem under water-scarce conditions and often results in crop failure. Under varying hydrological conditions, the fixed cropping pattern with conventional operating rule curve policies may not yield economically good results. To provide flexible policies, a nonlinear multi-objective optimization model is formulated. To achieve robust performance by handling interdependent relationships among the decision variables of the model, the recent MODE technique is adopted to solve the multi-objective problem. The developed model is applied for ten-daily reservoir operation to a case study in India. The model results suggest that changes in the hydrologic conditions over a season have considerable impact on the cropping pattern and net benefits from the irrigation system. Towards this purpose, the proposed MODE model can be used to evolve different strategies for irrigation planning and reservoir operation policies, and to select the best possible solution appropriate to the forecasted hydrologic condition.     

基于模糊决策的自动节水喷灌控制器的设计

结合灌溉不易建立精确的数学模型和模糊控制决策不需要建立精确数学模型的特点，应用模糊逻辑设计了以AT89C51为核心的全自动模糊智能节水灌溉控制器，该控制器能根据检测到的土壤含水量和作物的需水情况进行模糊决策从而实现全自动智能灌溉。系统已经成功运行干曲阜示范区，收到了较好的经济和社会效益。