【作者】 夏国恩；

【导师】 陈建； 杨国才；

【作者基本信息】 西南农业大学 ， 农业机械化工程， 2004， 硕士

【摘要】 决策支持系统(DSS，Decision Support System)自诞生之日起，其理论、方法和应用得到了迅速的发展。目前己广泛应用到企业、军事、经济、环境、医学、能源、交通和公共安全等领域。一直以来，DSS的应用都是DSS的一个重要研究课题。雨水集蓄工程作为农业领域的一个重要组成部分，其建设和发展情况对农业的发展具有举足轻重的作用。由于我国丘陵山地雨水集蓄工程起点低、基础较差，较普遍地存在管理不善以及效率低等问题。于是建设高效的雨水集蓄工程逐渐成为丘陵山地节水灌溉的重要内容。另外，因为自然、经济、社会诸多因素的变化导致水源及需水量等发生变化，这使得丘陵山地雨水集蓄工程利用的决策较为复杂，用传统的计算方法难以甚至无法满足要求，不能给决策者比较、选择的机会。因此，丘陵山地雨水集蓄智能决策支持系统(HRHIDSS，Hilly Rainwater Harvest Intelligent Decision Support System)的研究，将有助于解决雨水集蓄工程中决策困难的问题。 本论文通过分析丘陵山地雨水集蓄的现状以及计算机在雨水集蓄中的应用情况，指出建立借助计算机在处理大量信息的基础上为决策者提供决策支持的丘陵山地雨水集蓄智能决策支持系统(HRHIDSS)，是解决丘陵山地雨水集蓄决策困难问题的一条较好途径。根据丘陵山地雨水集蓄的实际情况，丘陵山地雨水集蓄智能决策支持系统从功能上包括人机交互界面、系统设置、系统说明、数据管理、模型管理、知识支持、决策管理、信息查询及帮助等9个功能子系统。本研究由于增加了知识库以及相应的推理系统，大大地增强了DSS对于决策者的支持。就其系统结构而言，采用了四库三功能的系统结构，但是由于把方法库溶合于模型库之中，则其结构改变成三库三功能的系统结构，即数据库、模型库和知识库。论文对各个组成部分的设计与实现作了阐述，并着重论述了模型库、知识库的设计与实现。其中就模型库而言，因为雨水集蓄活动涉及许多变量，各变量之间存在复杂的关系，而且不同的决策者对雨水集蓄问题具有不同的决策风格，丘陵山地雨水集蓄智能决策支持系统要支持这些不同风格和复杂的变量关系，必须建立模型，通过模型的建立将结构不良的决策问题转化为结构化的问题，给予决策者分析问题、比较各种方案的能力。可以这样认为，丘陵山地雨水集蓄智能决策支持系统是由“模型驱动”的。系统一共建立了五个模型，包括作物需水量模型、丘陵山地需水量模型、全年可集雨水量模型、水量供需平衡模型、丘陵山地雨水集蓄评价模型。另外由于引入了人工智能技术，因此建立了知识库，其目的是用来存放决策专家的决策经验和决策知识，西南农业大学硕士学位论文中文摘要以及某一领域专家提供的进行问题求解的经验和知识，从而扩大与决策者共有的领域，以便更好地进行沟通，真正达到决策支持所提出的目标。丘陵山地雨水集蓄智能决策支持系统的知识来源于领域专家对丘陵山地雨水集蓄知识的总结和概括，以及当地或同类地区雨水集蓄的试验资料等。在丘陵山地雨水集蓄智能决策支持系统中，将雨水集蓄的领域知识用数据型知识和规则型知识两种类型来表示和组织。丘陵山区雨水集蓄包含的知识较为广泛，由于许多是属于建设性，描述性的知识，而且这些知识又往往与逻辑性、过程性、运算性知识相互混合，交织在一起运用。即有描述性知识，又有逻辑性、运算性知识。为此，系统采用描述框架和“规则架+规则体”的规则组库组成的综合知识体结构的知识表示方法.在评价雨水集蓄工程设计各个方案方面，论文引入了层次分析法和定量分析法对雨水集蓄工程设计结果进行评价。通过对雨水集蓄.「程常用的技术经济指标进行分析后，提出采用工程方案的先进性和合理性以及工程造价两个指标来对雨水集蓄工程设计方案进行综合评价。最后，通过对系统进行分析、设计、编码调试和软件测试等工作后，将HRHIDSS应用于实际，取得了较好的效果。更多还原

【Abstract】 The theory, method and application of Decision Support System(DSS) have been witnessed booming development since DSS emerged. At present, DSS has widely been applied to the fields of enterprise, military affairs, economic environment, medicine, energy sources, traffic and public security, etc. The application of DSS is an important research task at all times. The development of collecting rain engineering, which is an important part of the agricultural field, plays an important role in the development of agriculture. There are some characters on China’s collecting rain of hilly regions .such as low jumping-off point ,weak foundation, lack of management and low efficiency, so building the highly efficient HRH engineering is gradually becoming an important matter of hilly water saving irrigation engineering. In addition, because of the changes of water sources and water demand caused by the changes of the various factors including nature, economy ,society etc, the decision of HRH engineering design becomes more complicate ,and the need cannot be meet by traditional calculation methods for the decision makers while there are few chances of comparison and choice. Therefore, the research of HRHIDSS will be beneficial to resolve the difficult problems of rainwater harvest engineering decision.This study, which analyses the present state of hilly collecting rain and the application of computer for it, points out that building Hilly Rainwater Harvest Intelligent Decision Support System(HRHIDSS),which is based on vast information processed by computer, is a good way to resolve the difficult problems of HRH. The system functions include the interface, establishing system, system statement, data management, model management, knowledge support, decision management, searching information and help etc. Because this study increases knowledge base and reasoning system, the support for decision makers is greatly strengthened by the DSS. So the system construction of four bases and three functions is used. However, the model base has included the method base so that it’s construction becomes into three bases and three functions, including data base,model base and knowledge base. The thesis expatiates the design and realization of its each component and emphasizes the design and realization of model base and knowledge. As for model base, because collecting rain activity involves many variables in which there are the complex relationships ,and different decision makers have different decision characters for the problems of HRH, it must build models to support these different characters and complex relationships for HRHIDSS and give decision makers the power of analyzing comparing various methods. It is obvious that the operation of HRHIDSS is supported by the models. This system totally establishes five models ,including the need water quantity of the crop model, the hilly water demand model, the rainwater harvest quantity of the whole year model, the equilibrium of water support and demand analytic model and HRH evaluating model. In addition, because of introducing the artificial intelligent technique .knowledge base must be established in order to store decision and special field experts’ experience and knowledge .It will enlarge the common fields of decisions to communicate the thoughts and truly achieve the targets of decision support. HRHIDSS’s knowledge comes from the generalization of special field experts for HRH and the rainwater harvest trial data of local or the same region etc. In the HRHIDSS, the knowledge of rainwater harvest is articulated and organized by data knowledge and rule knowledge. The knowledge of HRH is extensive and much of them belongs to constructive and describing knowledge. In the same time, these integrates logic, process, calculation knowledge. So the system uses the knowledge method which is the synthetic knowledge construction of describing frame and the rule base of rule frame and rule body. We adopt the Analytic Hierarchy Process(AHP) and Quantitative Analytic Method to evaluate t更多还原