节点文献

砀山酥梨营养诊断与矫治专家系统研究

Study on Expert System of Nutrition Diagnosis and Correction for Dangshansu Pear

【作者】 刘后胜

【导师】 李绍稳; 杭小树;

【作者基本信息】 安徽农业大学 , 果树学, 2003, 硕士

【摘要】 砀山酥梨原产于安徽省砀山县,是全国名特水果,已有2500多年的栽培历史。砀山酥梨栽种区地处黄河故道,其土壤结构不好,有机质含量低,果农为了过分追求水果的产量,大量施用化肥及农药。在消灭害虫提高产量增收效益的同时,也对当地的生态系统造成了不同程度的破坏。如土壤的污染,结构的破坏,并导致树体营养元素的失调,营养元素的缺乏,特别是砀山酥梨的品质也随之下降了。随着中国加入WTO,我国的水果生产将面临严重挑战。为了使我国水果在国际上占有一席之地,提高水果品质将是最根本的问题。 本文针对砀山酥梨栽培措施不当,引起砀山酥梨树体营养失调,表现树体外部缺素症状,造成砀山酥梨品质下降的现象,根据果树矿质营养学的原理,设计了不同诊断方法,并进行系统集成,从而为砀山酥梨的营养诊断提供一个更为全面合理的工具。 为了对不同的情形做出相应的处理,本文采用了三种不同营养诊断的方法:树体土壤营养诊断法,盈亏指数PLI和营养综合指数DRIS诊断结合法,以及Fuzzy模糊诊断法。树体土壤营养诊断法结合了树相诊断法和土壤诊断方法,简单易行。树相诊断主要通过树体表现的外部营养失调的症状进行初步的判断,树体缺什么营养元素。为了进一步确定缺素的原因,需通过对叶片和树体所在的土壤做相应的分析处理。营养临界值PLI和营养综合指数DRIS诊断结合法是从营养元素平衡的理论出发,对树体营养状况进行整体判断,过程复杂,计算量大。临界值理论以元素浓度表示营养状况,没有考虑到元素间平衡;而DRIS理论则是更多地考虑元素间相对平衡,没有反映元素浓度的具体指标,因而这种平衡可能是高水平的也可能是低水平的营养平衡。本系统将基于PLI与DRIS两种理论结合起来考虑,进行砀山酥梨营养诊断和矫治,以弥补两者各自的不足。模糊营养诊断法对临界值划分标准的不同而引起营养元素浓度交叉重叠做出了很好的处理。模糊数学对事物从差异的一方到另一方,中间经历了一个从量变到质变的连续过渡过程而造成的不确定性给出了定量处理方法,该方法切实可靠。 该系统的研制是基于面向对象的构件技术和雄风专家系统开发工具的集成环境。为了降低了系统的复杂性,提高了系统的可维护性、可扩展性,系统采用面向对象的构件技术,利用Visual C++6.0语言开发了基于PLI与DRIS结合的营养诊断构件子系统、模糊营养诊断构件子系统。为了提高了知识表达能力,增强了推理机能,知识库的编写是采用人工智能专家系统的产生式规则表示方法。该方法具有自然性、一致性和完备性。 系统经过测试,运行准确稳定,人机交互界面友好,可视化的操作简单易学,达到了设计要求,与同类的专家系统相比,具有一定的优越性。

【Abstract】 Dangshansu pear is especially famous fruit throughout the country in china, originating from Dangshan County in Anhui province. It has been cultivated for over 2500 years. The plantation region for Dangshansu pear is located at the old site of Huanhe River. Its soil structure is not good, and it contains less organic matter in soil. Fruit farmers use a large number of fertilization on the soil and pesticide on the fruit tree in order to excessively pursuit for the output, which destroys the local ecosystem to some extent with pest dieing out, yield improved and benefit increasing. Such as soil pollution, destroy of the soil structure, nutrition disorder on the tree, deficiency of the nutrition element, and the quality dropping of the Dangshansu pear especially mentioned. With the china entry into WTO, the fruit produce in china will face severe challenge. If china wanted to hold a position about Chinese fruit in the world, the most primary problem would be improving the fruit quality.The paper systematically integrates the different diagnosis based on the principle of the fruit mineral nutriology. It provides a more comprehensive and reasonable diagnosis means and tool for Dangshansu pear to deal with the phenomena mentioned above last paragraph.In order to correspondingly deal with different status, three measures are used in the paper as following: the combination nutrition diagnosis through tree body and soil, the combination nutrition diagnosis method based on the Profit and Loss Index (PLI) and the Diagnosis and Recommendation Integrated System (DRIS), and Fuzzy nutrition diagnosis. The combination nutrition diagnosis through tree body and soil provides a primary judgment which nutrition element the tree is lacking through the dystrophy syndrome presenting outside. It is simple and easy to operate. In order to conform the cause of lack element, it needs to deal with and analyze the leaf and its located soil. The combination nutrition diagnosis method based on the Profit and Loss Index (PLI) and the Diagnosis and Recommendation Integrated System (DRIS) wholly judge the tree nutrition status from the nutrition element balance theory aspect. Its process is complicated and calculation is very large. Nutrition status is showed by element concentration by critical value theory. It is no consideration about the balance among nutrition elements. Compared to critical value theory, DRIS theory thinks more relative balance among nutrition elements, but it does not show the detail indexes about nutrition element concentration. Maybe the balance is high or low. In order to make up the shortcoming respectively, the paper takes two theories into consideration to provide diagnosis and correction advice for Dangshansu pear. Fuzzy nutrition diagnosis very well deals with the crossing overlap of the nutrition element because of the different criteria of critical value. As to the uncertainty that is caused by the continuous process existing from quantitative change to qualitative change when things change from discrepant one side to another side, the fuzzy mathematics can deal with the problem by quantified. It accords with the facts and has credibility.The development of the system uses Object-Oriented component technique and integrated environment of XiongFeng expert system developing tool. In order to decrease the complication and improve the maintenance and expansibility of the system using Object-Oriented component technique, Component subsystems based on the combination of PLI and DRIS nutrition diagnosis and based on fuzzy nutrition diagnosis are developed by Visual C++ 6. 0 tool. In order to improve the express capability and enhance the reasoning function, the knowledge base on basis of tree body and soil nutrition diagnosis and integrating with component nutrition diagnosis subsystem is expressed by produce rule belonging to artificial intelligence expert system. The method has nature^ consistency and maturity character.After being tested, the system has achieved its designing aim. It runs stably and u

  • 【分类号】S661.2
  • 【被引频次】3
  • 【下载频次】329
节点文献中: 

本文链接的文献网络图示:

本文的引文网络