ALCGEIS知识获取与推理机设计

Research on Knowledge Acquisition and Design of Inference Engine for ALCGEIS

运用专家系统的知识获取和推理机设计原理与方法,探讨了ALCGEIS系统的基本结构、知识分级、知识获取过程和实现程序,推理机的基本原理和控制策略及其程序实现的算法,以及推理知识的搜索策略;论述了 AIEGEIS推理规则的设计方法及其推理算法。

With the establishment of market economic system and the practice of using land institution, integrated evaluation of the quality grade of agricultural land has been a very important task in land administration at present and in the future. Classifying and grading of agricultural land is a very comprehensive task, the results of it not only connect with many natural factors, but also relate to the economic factors, society factors and many other factors. Moreover, in practice it involves analysis and handling huge amount of data and pictures, so the calculation task is very gigantic. Agricultural Land Classifying ＆ Grading is a problem, which can not be described with mathematics completely, but needs to be resolved with experience and knowledge. Using a great deal of data accumulated in the past reasonably and integrating the expert knowledge adequately to classify ＆ grade agricultural land, setting up Agricultural Land Classifying ＆ Grading Expert Information System （ALCGEIS） with intellectualized analysis and decision function to improve the precision of classifying and grading results of agricultural land and to raise the efficiency of work are very necessary. Building a system based on Artificial Intelligence （AI） is also the goal and developing orientation in the future research. This paper presents knowledge acquisition and design approaches and principles of inference engine for Expert System （ES）, and probes basic structure of ALCGEIS, knowledge classification, knowledge acquisition and algorithms of program realization, basic control strategies of inference engine and algorithms of program realization as well as search strategies of reasoning knowledge. In addition, the author designs the inference engine of ALCGEIS. And taking the structure of inference engine as an example, the reasoning algorithms, and elaborates organization structures author designs methods of reasoning rules as well as of reasoning knowledge.