节点文献

速生丰产林病虫害防治专家系统的研究与实现

Research and Implementation of Disease and Insect Prevention and Cure for Fast-growing and High-yield Forest Expert System

【作者】 姜玥

【导师】 王霓虹;

【作者基本信息】 东北林业大学 , 计算机应用技术, 2008, 硕士

【摘要】 本课题源于国家支撑科技计划课题子课题——东北地区速生丰产林培育作业远程咨询系统。这个项目是面向广大林农和造林企业,利用现代信息技术集成林木培育与管护、病虫害防治、林木收获利用等整个森林的经营管理的全过程中的相关技术手段和知识,研制速生丰产林培育技术远程咨询服务系统和速生丰产林生产(培育)、经营过程智能决策支持系统。其目的是针对速生丰产林生产经营全过程提供专业技术支持,以改变落后的粗放的经营管理模式,提高土地生产力,为广大农户、林业企业和林场提高经济收入水平服务。本文主要针对其中的速生丰产林病虫害防治专家系统(以下简称系统)的实现方法做研究和讨论。速生丰产林病虫害防治专家系统是一个集病虫害诊断、预报、防治、林木与病虫害基本信息查询以及常见问题问答于一体的综合系统。其中病虫害诊断、预报、防治系统是基于产生式规则的推理系统,而常见问题问答系统则是检索系统。不同的系统功能是建立在不同的数据库结构之上的,而建立这些系统所需数据库的平台只有一个——关系数据库,如何表达不同系统所需要的知识,开发相应的推理机制是本文讨论的重点。本文介绍了速生丰产林病虫害防治系统与诊断系统的设计与实现,包括诊断与防治知识的表达,专家系统推理机的设计以及解释功能的实现。在病虫害诊断推理过程中本文提出了一种采用向量空间模型,通过改写文本分类中TF-IDF公式计算权重,以向量相似度作为规则选择标准的规则选取方法。另外将检索技术引进专家系统功能模块是本系统对突破传统专家系统开发模式的一个尝试。系统研究过程中采用了B/S结构,B/S结构前后台较强的独立性为知识库的设计打下良好基础。后台关系数据库软件采用了著名开源数据库软件MYSQL5.0,前台开发采用当今流行的JSP技术,JSP技术对后台数据库强大的连接、访问与控制能力也为知识库建立提供了有力支持。

【Abstract】 This article is from the country propping up science and technology plan problem the key technology studies and applies of fast-growing and high-yield timber forests manage process. This project is to be geared to the needs of extensive forest farmer and forestation enterprise , the integrated forest tree making use of modern information technology cultivates and the tube protects the intelligence decision-making hold out system making use of the process waiting for relevance technology means and knowledge in the entire proceeding of entire forest management control , developing fast-growing , high-yield timber forests cultivating long-range consultancy service of technology system and fast-growing , high-yield timber forests giving birth to a child (cultivate) , managing , pest control , forest tree gain. Whose purpose is to give birth to a child specifically for fast-growing , high-yield timber forests manage the entire proceeding provides the special field technology hold out, take changing the rough management control pattern falling behind , improving land productivity as, the extensive peasant household , forestry enterprise and forest farm that improve economy income level service.The system is a integrative system that contains aspects as plant disease and insect pests prevention, cure and forecast, basic information of forest, disease and insect query and FAQ. Plant disease and insect pests prevention, cure and forecast system are ruling systems. Basic information of forest, disease and insect query system is general querying system. FAQ system is based on searching. Different systems are based on different structures of database and the only one thing to discuss is RDS. How to express knowledge for different systems and save it in RDS is the main point to discuss.In this paper, fast growing forest pest and disease control systems and diagnostic systems design and implementation, including the expression of knowledge, design of reasoning machine and the implement of the explain function. Pest and disease diagnosis of reasoning which in the course of this paper, a method of rule selecton is put foword. This method is based on vector space model, rewrites the TF-IDF fomular which is ususlly used in test classification to compute the weight and chooses the rule according to the similarity of vectors.In addition to the introduction of search technology expert system modules of the system is a try to break the traditional development model of expert system.The system adopts the B/S structure which provides independence between proscenium and background. The RDS here, is the famous open code database software MYSQL5.0 and the exploitation technology is JSP whose strong ability of connecting, accessing and controlling of database provides powerful support for knowledge database.

节点文献中: 

本文链接的文献网络图示:

本文的引文网络