节水灌溉智能预测系统构建

发布时间：2018-08-09 15:40

【摘要】：滴灌技术是一种局部灌溉方式,是目前提高农业用水利用率最为有效的节水灌溉技术之一。掌握土壤水分运移规律是正确进行滴灌系统设计以及有效进行节水灌溉的前提,而对作物灌溉需水量的准确预测是进行有效节水的重要保障,对灌区农业的可持续生产具有重要的指导和借鉴意义。本文主要从灌水方式以及灌溉预测模型两个方面对节水灌溉进行了研究。主要研究内容概括如下:1)以室内无作物生长的燥红壤土为研究对象,利用土壤湿度传感器对单点源灌溉条件下土壤水分运移进行动态监测和数据采集,主要考察了滴头流量、灌水量以及初始含水量对土壤湿润锋运移的影响以及停灌后土壤含水量的再分布情况。2)提出了一种基于键逾渗模型的灌溉区基本控制单元自适应划分方法。通过定义的连通性量化指标判断灌区灌溉情况,并利用MATALB软件进行该算法的仿真,验证了该方法的有效性,最后在草坪地进行了实际验证,结果证明了该算法的可行性及实用性。3)选取温室黄瓜作为研究对象,采用遗传算法优化的BP神经网络构建了黄瓜灌溉需水量预测模型,并利用检验样本验证了该模型的预测效果。结果表明,该模型预测精度高,预测准确性好,可将其应用于温室黄瓜需水量的预测。4)利用MATLAB的GUI界面设计了温室内黄瓜需水量预测系统,该系统具有友好的人机界面,能很好的方便用户进行温室黄瓜需水量的预测。
[Abstract]:Drip irrigation is a local irrigation method, which is one of the most effective irrigation techniques to improve the utilization of agricultural water. It is a prerequisite to correctly design drip irrigation system and to effectively carry out water-saving irrigation to master the law of soil moisture migration. The accurate prediction of crop irrigation water demand is an important guarantee for effective water-saving. It has important guidance and reference significance for sustainable agricultural production in irrigation district. In this paper, water-saving irrigation was studied from two aspects: irrigation mode and irrigation prediction model. The main research contents are summarized as follows: (1) the dry red loam growing without crops is used as the research object. The soil moisture transfer and data collection under single point source irrigation are carried out by using soil moisture sensor, and the flow rate of dripper is mainly investigated. The effects of irrigation amount and initial water content on soil moisture front migration and the redistribution of soil moisture content after irrigation closure (.2) an adaptive partition method of basic control unit of irrigation area based on key percolation model was proposed. The irrigation condition of irrigation area is judged by the defined connectivity quantitative index, and the simulation of the algorithm is carried out by using MATALB software. The validity of the method is verified. Finally, the validity of the method is verified in the lawn field. The results show that the algorithm is feasible and practical. 3) Cucumber in greenhouse is selected as the research object, and the forecast model of irrigation water demand of cucumber is constructed by BP neural network optimized by genetic algorithm. The prediction effect of the model is verified by test samples. The results show that the model has high prediction accuracy and good accuracy. It can be applied to the prediction of water demand of greenhouse cucumber. The GUI interface of MATLAB is used to design the water demand prediction system of cucumber in greenhouse. The system has friendly man-machine interface. It is convenient for users to forecast the water demand of cucumber in greenhouse.
【学位授予单位】：昆明理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：S274

【参考文献】