The Fault Diagnosis System Research Based On Hadoop

Fault diagnosis expert system technology is using the reasoning technique to fault diagnosis of advanced technology with the combination of expert knowledge and experience knowledge database.Because of the relatively good performance in system accuracy, timeliness and stability,the system is widely used in all fields. The paper mainly studies and designs the fault diagnosis expert system based on Hadoop, and its main work is as follows:In this paper, we first analyzes and compares with the existing fault diagnosis expert system, serious studys the machine learning, knowledge discovery,data mining technology and so on in detail, dive deeper into the Hadoop platform and Agent technology. The detailed improvement design and implementation are carried out on the fault diagnosis expert system based on Hadoop platform, According to the characteristics of its universal shortage and the lack of reasoning and judgment, in order to achieve better feedback results,the production expansion algorithm is adopted for the platform, the design and implementation of the improved scheme are carried out in this paper.Second,make comprehensive use of various methods and expert reasoning rule to reasoning judgment,designes the algorithm of generating extended rule by Agent,and improved the traditional fault diagnosis expert system in general shortcomings; search engine uses vertical search engine technology, making the system to complete collection is more effective on the URL; search results ranking by the improved PageRank algorithm,namely in the algorithm introduced a E vector, to avoid the suspension link; the knowledge base using split and overridden methods, reduces the computer memory and CPU overhead.Finally, through related experiments, analyzes and validates the performance of the systme in a variety of scenarios. The test results shows that the system in effectiveness,system versatility, speed has made good effect compared with the traditional system.