The Optimization And Composite Reasoning Of Knowledge Base System Of The Intelligent Design System Of Sugarcane Harvester

This thesis is based on the item subsidized by Natural and Scientific foundation named as "the Research on key technique of intelligent design system about sugarcane harvester based on knowledge". This thesis gives the general structure of intelligent design system of sugarcane harvester and the main function of key component of sugarcane harvester on the basis of studying and dissertating of the development of intelligent design system and the methods of optimization and maintenance of knowledge base system.The intelligent design system of sugarcane harvester has two different repositories, it analyzes the characteristic of sugarcane harvester's design knowledge and differences between basic knowledge base and extended knowledge base's knowledge representation, constructs an interface module between two knowledge base, the knowledge can be analyzed and transformed ,then putted into basic knowledge base.It analyzes the defect of manual maintenance of knowledge base system, constructs condition aggregate and conclusion aggregate, compares the conclusion of the rule, then puts forward an effect arithmetic to check the rules of redundance, rules of antinomy, and gives the check arithmetic of cycle rules, then achieves the consistent check of knowledge base system. It gives the notion of selection of rules, constructs the gathers between selections to get the goal of check the rules which be not allowed to reach, omitted and dead, then accomplishes the integrality check of knowledge base system.This paper discusses the advantages and disadvantages of rules of reasoning and neural network, the differences between them. It gives sugarcane cutter's composite reasoning and puts forward concrete structure, it discusses unitary processing of knowledge, neural network learning algorithms, how to decide the conditions and conclusions of composite reasoning, it adopts S function to improve neural network, and gives detailed flow of composite reasoning process.