花生虫害管理的计算机专家系统

AN EXPERT SYSTEM APPLIED TO PEANUT PEST MANAGEMENT

专家系统是用来模拟人类在某些领域内解决问题的计算机程序。它在农业中的应用比较广泛。一个用于花生害虫——非洲蔗螟(LCB)管理的专家系统,已在美国阿拉巴马州建立。该系统由知识库,推理机及用户接口三个部分组成。在这个专家系统模型中用了以下信息:花生播种地点,日期,土壤类型以及历年来花生种植地区的非洲蔗螟发生情况等。一旦虫害突然蔓延,专家系统中非洲蔗螟数量生态动态模型会作出判断。有关非洲蔗螟的知识被编写成一条条规则存贮在专家系统的规则库内。推理机根据用户提供的信息及知识库中的规则进行推理,从而使专家系统获得正确判断。本文叙述了专家系统的定义,组成部分及其功能。同时还介绍了建立专家系统的方法步骤。本专家系统的外壳程序是用BASIC语言写成,它可在IBM PC微型计算机及其兼容机上运行。

<Abstract >Wang, Y. B. and Mack, T. P., An expert system applied to peanut pest mana- gement. Expert systems are computer programs designed to simulate the problem-solving ability of human experts in certain domains. The range of opportunities for the application of an expert system is vast in agricultural fields such as integrated pest management An expert system model for the lesser cornstalk borer (LCB) has been developed for conventionally planted runner peanuts in Alabama, U. S. A. The prototype expert system model consists of an inference engine, a knowledge base, and a user interface. The model uses information on field location, previous history of lesser cornstalk borer damage, soil type, environmental events, and a LCB population dynamics model to determine if a population outbreak will occur. Expert knowledge of the LCB is translated into rules required by the expert system. This paper describes the definition elements and their functions of the LCB expert system. The construction of a small expert system is also detailed. The shell program for the LCB expert system is written in BASIC and runs on IBM PC microcomputer and 100% compa- tible machines.