The Research And Implementation On Extension Strategy Generation System Based On Agents

With the development of the social economy and information technology, decision making systems become more and more complicated. There are too many parameters to consider, and too many strategies to choose. So it is hard to generate strategies and make decision only by human brain activities. Artificial intelligence is in a period of vigorous development, it closely integrated with the computer for people dealing with the study of knowledge. This thesis introduces the theory of Extension and Agent to make up for intelligent decision-making system and the existing system deficiencies.At present, the basic-element models of contradiction problems and the extension methods of strategy generation to solve contradiction problems has been established. The work of this thesis is introducing the extension theory and extension engineering methods into computer systems. Its aim is according to the research results of extension strategy generating logic basis and the Extenics theoretical results to study the strategy generating system that operates on computers.Based on analyzing extension theory and knowledge representations, this thesis sets up the knowledge model with basic elements, achieving to describe information and knowledge formally, and researches the methods to solve problems in practice on this model.After that, combining to computer technology, the thesis sets up knowledge database in a established extension model. It processes the problems related with current knowledge. Researching the method in computer system to get knowledge by extension inference and extension transformation. It generates strategies to solve present problems and synthetically evaluate strategies with extension methods.The creative work in this thesis includes:1. The use of Extension methods to solve contradiction problems and overcome the drawback of computer system in the past. 2. The use of extension of the basic-element model to describe the objective world of things and the laws of reasoning, the use of extension inference and extension transformation to build strategy generation inference rule, the use of extension set and dependent function as a quantitative evaluation tool for strategy generation and strategy evaluation, so that provide new approach for computer to help decision.3. In order to meet the dynamic needs of complex systems, in the realization process, we introduce the agent technology, let the system of problem solving knowledgebase perform according to changes in the environment and changes in user requirements and update independently related parameters and rules. This thesis is supported by the Guangdong Provincial Natural ScienceFoundation——Based on the extension of the information - knowledge - Formalsystem of Research.