不确定条件下半湿润偏旱灌区多目标水资源优化配置模型和应用研究

Sub-humid arid irrigation district is an important base of grain production in China, however, the agricultural water use and grain production of the sub- humid arid irrigation district have slightly declined. This will affect the income level of farmers, the sustainable development of agriculture, and the quality of water nearby. The previous uncertain optimization models for water resources allocation could not avoid the risk of market economy in farm products and hardly consider the fair distribution of farmer’s income, which will hinder the sustainable development of regional agriculture. This project will describe the uncertain relationship between economical prices and food supply-demand using Copula function, measure the equality of allocation using Gini coefficient, and establish a multi-objective optimization model for water resources allocation which is capable of avoiding marketing risk, achieving equitable distribution, and guaranteeing the quality of water environment. The Pareto search algorithm coupled with multi-objective particle swarm optimization algorithm will be proposed to solve the model through programming, with Monte Carlo simulation technique and VaR risk model are integrated as well. The optimal agricultural water allocation schemes will be obtained, which will improve the structural reform of agricultural supply. The proposed multi-objective optimization model and the solving method will provide guiding significance for regional water resources optimal allocation in other regions.

半湿润偏旱灌区作为我国的重要粮食生产地区，随着农业供给侧结构性改革的推进，其农业用水量和粮食生产量略有下降，也进一步影响到农民的经济收入水平、农业可持续发展、周边水体质量。以往的优化配置模型无法规避农产品的市场经济风险，也几乎不考虑农民的经济收入公平分配问题，不利于实现区域的农业可持续发展。本项目拟结合Copula函数描述经济价格与粮食供需总量之间的不确定关系，用Gini系数衡量分配的公平程度，并建立可规避市场风险、实现公平分配、保障水域环境质量的多目标水资源优化配置模型。结合Pareto解集搜索算法和多目标粒子群优化算法，编制模型求解程序。运用MC模拟法，度量VaR风险。研究结果将得到最优农业水资源配置方案，保障农业供给侧结构性改革顺利推进。研究所得的多目标水资源优化配置模型和求解方法，对其它区域性资源的优化配置也具有重要的借鉴意义。

在农业供给侧结构性改革的前提下，综合考虑了水资源时空变化及经济参数呈尖峰厚尾分布等不确定特点与社会经济发展过程中多种不确定因素。由于农业水资源优化系统中存在着诸多不确定性参数，例如降水、地下水可用量、农产品收购价格、种植面积等，建立不确定性优化模型关键就在于正确分析数学模型中的多种不确定性参数，并且有效地描述这些不确定性参数的取值范围及分布特性。以往的优化模型不足以反映经济参数非正态变化对农业灌溉优化系统的影响。项目结合了Copula方法，建立了多重不确定性半湿润偏旱灌区灌溉水资源优化模型，并据此开发了农业灌溉系统的管理软件IOMSD。软件包括数据输入模块、模型求解模块和结果演示模块，操作简单。采用粒子群优化方法对优化模型进行求解，并利用GIS对优化结果进行显示。将开发的IOMSD应用于灌区渠道水优化配置问题。结果表明所应用灌区37 ~ 54天系统利用率特别低，仅干渠B5独立灌溉。从第1天到第36天，主干渠的流量需要反复调整，从而导致主通道流量的显著变化，主河道流量峰值偏大。因此，可以对优化结果进行改进。采用层次聚类分析方法，简化了干渠的操作，缩短了整个灌溉周期。改进后的优化方案更加简洁，干渠利用率更高(提高8.3% ~ 25%)。主渠道不对称系数降至1.375，灌溉系统更加稳定高效。

内容获取失败，请点击重试