Research And Implementation Of The System Of Network Test And Intelligent Judgment Based On Expert System And Evidence Theory

With the rapid development and application of information technology represented by computer science, we human beings have already entered the information society. Computer network influences almost every aspect of human manufacture and human life, lots of our habits and behavior have changed because of it. It is impossible for human to live without water and food, so does network. Computer network has become one of the most important infrastructure of the modern society. The rapid expansion of network scale and the appearance of different types and the topology structures make the difficulty of network management grows exponentially, which makes network test become a research hotspot and difficulty. The most difficult part of network test is how to analysis and process the complicated network test data automatically and efficiently. However there is no complete theoretical basis or a feasible method to deal with the test data till now. In some complex network test scenarios, the huge amount of test data consist different parameter types of which its meaning differs a lot. The traditional method which depends on experts’ manual procession cost huge manpower and material resources, its efficiency is very low and the result is not very ideal. For this problem, this paper research on the automatic processing and intelligent analysis method based on expert system and evidence theory, relying on actual project of scientific research institutes.Firstly, this paper explains the concept, characteristics and different classification of knowledge in detail and then introduces several common methods of knowledge representation and comparative analysis of their advantages and disadvantages. According to the principle of knowledge representation should respect, this paper designs a new knowledge representation method based on XML to meet the demands of judging information and relevant knowledge. According to the principle and method of how to construct a knowledge system, this paper designs the instrument library and rule library of the system. Through the comparative analysis of different information storage structures’ advantages and disadvantages, the paper chooses the tree structure to store the data and information of this system. This paper gives different processing method corresponding to different topological structure evaluation models. Through the study and analysis of different reasoning methods and strategies, this paper designs aneffective reasoning strategy, controlling strategy and reasoning process corresponding to the characteristics of the knowledge library.Then the whole design scheme of the system is given.At last the system which based on expert system and evidence theory is implemented and system testing is performed under different scenes. The test result shows that this system can work normally. Specialist Network evaluation and the corresponding suggestions can be got in very short time.