【作者】 李庭波；

【导师】 陈平留；

【作者基本信息】 福建农林大学 ， 森林经理学， 2009， 博士

【摘要】 森林资源经营决策知识库系统是近年来森林资源经营理念转变、计算机技术、人工智能应用等的发展和相互结合而形成的知识集成系统。森林资源经营具有生产周期长、干扰多、区域大、变异大、条件复杂等特点，知识库技术的研究以数据处理、知识自动获取、知识表示、决策预测等为目标，最终为实现森林资源的可持续经营提供智力支持。本文以知识库技术为主要线索，把森林资源经营决策的现实系统与本体知识库的结合点作为研究目标，采用基于本体的知识组织方式，运用语义语言、构建模型等方法，研究森林资源经营决策知识库的构成成与技术方法，以本体学习、OWL约束公理、实例推理等方面技术探讨森林资源经营决策知识的获取、表示、推理等方法及其在森林资源经营决策过程中的应用。本文探讨了森林资源经营决策数据、信息和知识的层次关系、森林资源经营决策知识库的构成。首次把本体概念引入森林资源经营决策知识库技术研究中，分析了本体在森林资源经营决策知识库中的意义。采用Prot(?)g(?)对基于本体的森林资源经营决策知识库的主要组成要素：类、属性、公理和实例进行了构建，形成以本体概念为知识组织方式的开放性森林资源经营决策知识库。首次提出从森林资源数据结构中获取森林资源核心本体的本体学习方法，定义了森林资源核心本体(Core Ontology)模型，并通过数据进行本体学习的应用。分析了网络本体语言(OWL)、Prot(?)g(?)本体编辑工具和约束公理的特点和组成，对开放性森林资源经营决策本体知识库应用所需要的直径结构规律和森林数量成熟规律进行OWL约束公理表示，同时以地理气候条件为主要内容，对主要的约束公理构造符进行运用。探讨了在Prot(?)g(?)中实现描述逻辑推理的方法，通过森林资源中的有林地和阔叶林地类的层次结构描述逻辑推理应用。根据基于模型的推理和基于实例的推理的优点以及森林资源经营决策知识库的应用特点，首次提出基于模型和实例的森林资源经营决策知识综合推理方法，阐明了综合推理在应用中的方法。由于森林资源经营决策知识库构成庞大，构建过程复杂而漫长，其应用一直以来都是森林资源经营决策知识库研究的关键问题，文中在Prot(?)g(?)中构建了杉木人工林成熟相关的开放性森林资源经营决策本体知识库的类、属性、公理和实例；建立了不同年龄杉木人工胸径结构模型，确定抚育间伐参数，获得杉木人工林成熟龄的判断指标；对杉木人工林新实例进行了模型与实例综合推理，并对推理结果进行了分析。本文的研究工作是本体概念在森林资源经营决策知识库技术研究和应用的一种尝试，对探讨林业领域的智能化问题是一种新方法，在森林资源经营决策知识库的构建、专家系统和决策支持系统等的研究中具有重要的意义。基于本体的知识库技术研究才刚刚开始，还有许多的工作有待于未来研究。更多还原

【Abstract】 Concepts of forest management are changed in recent years,technologies of computerscience and applications of artificial intelligence are developed rapidly,ontology-based forestresource management decision-making knowledge base (OFRMDKB abbr.)is a integrationsystem behind them.Forest resource management have the characters of long yield period,more disturbs,large region,much variation,complex environment.Data proceeding,knowledge auto-acquired,knowledge representation and prediction for decision-making arethe goal of research of OFRMDKB technologies,and then supply sustainable management offorest resource with intellective basics.The technologies of knowledge base are the main clueof this essay,and the connection between ontology knowledge base and the real naturesystem is the target of OFRMDKB.Methods of ontology-based knowledge managementframe,semantic language coding and establish models are applied in studying thecomponents and technologies of OFRMDKB.We also discussing knowledge acquired,representation,reasoning in the OFRMDKB with the technologies and methods of ontologylearning,constraint axioms of OWL (Web Ontology Language)and case-based reasoning.Technologies were applyed on throughout of forest resource management and decision-making.The hiberarchy of forest resouce management and decision-making (FRMD abbr.)data,information and knowledge,and the structure of FRMDKB are discussed in this essay.It isthe first time that study technologies of FRMDKB with the ontology concept.The maincostructure of OFRMDKB,include classes,properties,axioms and individuals,wereestablished.Knowledge were organised by the mode of ontology in oppening OFRMDKB.For the first time method of ontology learning that acquired core ontology from forestrecource data structure was proposed.We defined the core ontology of forest resource,andapplied the method to the forest resource data structure.Based on the analysis of characterand compose of web ontology language,prot(?)g(?)ontology editor and restriction axiom,diameter structure and quantitative maturity rules which are needed for the application ofopenning OFRMDKB represented by OWL restriction axiom.Construct symbols were usedin representing of geo-climate conditions of forest.We discussed the description logicreasoning in the tools of prot(?)g(?),and applied on the class hiberarchy of forest resource about forest covered land and broadleaved forest covered land.Characters of model baseedreasoning,case based reasoning and application of FRMDKB were analyzed in this essay,forthe first time we proposed the integration reasoning based on the forest management resourceand decision-making knowledge,and the appcation method of integration reasoning.It is main question that application of knowledge base within forest resourcemanagement and decision-making,due to the great component and long-range establishmentof FRMDKB.We created classes,properties,axiom and individuls of opening OFRMDKBwith protege,established diameter structure model of Chinese fir plantation in different age,acquired patameter of thinning and index of forest maturity estimateion,processed theintegrate reasoning with the new individuals of Chinese fir plantation,and analyzed the resultof reasoning.It is an attemption of research and application on forest resource management anddecision-making knowledge base with ontology concept,and is a new method of discussingproblems of intelligent in forest domain.There are signification of ontology amongestablishing FRMDKB,expert system and decision-making support system.It is thebeginning of research on OFRMDKB.There are further studies about ontology in the future.更多还原