【作者】 王立坤；

【导师】 刘庆华；

【作者基本信息】 东北农业大学 ， 农业水土工程， 2002， 硕士

【摘要】 本研究是中国博士后科学基金《三江平原井灌水稻节水增温灌溉管理决策支持系统》的一个子专题。研究通过系统的建模与优化，希望为三江平原农业及水资源的可持续发展提供理论依据。 三江平原是我国重要的商品粮基地，近年来井灌水稻面积不断扩大，截止1998年底，已发展水稻面积近70万hm~2，其中井灌水稻约占80％。无节制、无计划人为超采地下水，致使地下水动态平衡遭到破坏、每年出现大量的“吊泵”和“漏斗”现象，地下水资源紧缺已成为其今后发展的制约因素。 针对上述问题，本研究运用现代建模与优化技术，以三江平原富锦试区井灌水稻为研究对象，利用其田间实测及当地气象局资料，进行数据处理。根据《随机水文学》理论中的时间序列分析法，建立了适合三江平原井灌水稻需水量预报的非平稳时序随机模型；通过分析降雨随机特性，选定季节性时序随机模型，建立了适合三江平原井灌水稻降雨预报的月平均降雨模型；根据最小二乘法，计算出几种常用水分生产函数中的敏感指数及敏感系数，进而得到三江平原适宜采用的水分生产函数漠型及模型中敏感指数的变化规律；本文提出遗传动态规划法(RAGA—DP)，即采用改进的基于实数编码的加速遗传算法(Real coding based Accelerating Genetic Algorithm，简称RAGA)与动态规划法(DPSA)相结合，推求非充分灌溉条件下三江平原井灌水稻的最优灌溉制度。优化的灌溉制度可用于指导在不同的缺水程度下如何在时间上定量分配水量，以使相对产量最大，减产损失最小。通过系统建模预测与优化研究，不仅可以给出科学灌水、用水的数量指标与数学模型，而且对稳定、提高三江平原粮食生产水平，缓解水资源紧缺状况都具有重要意义。更多还原

【Abstract】 The content of the thesis is a part of "China Postdoctoral Science Fund" na COi na Pos on Irrigation Management DSS of Saving Water & Increasing Temperature for Well Irrigation Rice on Sanjiang Plain." Through building models and usin Trro ug hbui 1 d ng m the writer eager to provide some academic gist for the continuum development about agriculture and water resource.Sanjiang Plain is the most important base of commodity grain of China. The area under cultivation of well irrigation for rice has been expanding year after year of lately year. There are 0.7 million hm2 of rice area by the end of 1998. And nearly 80 percent are well irrigation. The homeostasis of groundwater resources has been destroyed because of exploiting by human without plan. More and more pumps can’t work and the appearance of funnel has been arisen. So, the lacking of groundwater resources has been become the restricted factor for its development.Aim at the above problems, the writer uses the modern modeling and optimum technique to studying the well irrigation rice in Fujin Area in Sanjiang Plain. Through handling lots of testing data, the writer builds up the model of non-stable time series model to forecasting the well irrigation rice water requirement in Sanjiang Plain. At the same time, the author analyses the random characteristic of rainfall and selects the season time series model to forecasting the rainfall. According to the least twin multiplication to calculating the sensitivity index in several water production functions. Thus, the writer obtains the fitted the value of the sensitivity index and the varied rule. At the same time, the writer puts forward a new method named RAG A (Real coding based Accelerating Genetic Algorithm) and combines RAGA with DPSA to calculating the best irrigation system under the non-sufficient irrigation of well irrigation rice in Sanjiang Plain. The best irrigation system can instructs the best distributed water quantity and the irrigated time under the condition of lacking water. Thus, we can obtain the maximum relative yield and the least loss. Thus, through building models and optimization, we not only can provide the models that can instruct to irrigating, but also have the important significance to advancing the provision yield and handle the water resource crisis.Postgraduate: Wang LikunSupervisor: Liu QinghuaMajor: Water and Soil of Agricultural Engineering更多还原