Vehicle Fault Diagnosis System Based On Case Information Retrieval

With the continuous development of modern economy, people’s living standards improved continuously as well, automobiles have already become the most common traffic tools which many families have; at the same time, along with the deepening degree of modern industrialization and vigorous progress of information technology industry, more and more precision electronic module and computer technology is being used to the car. Except meet one of the basic functional requirements--riding instead of walking, safety, comfortable and entertaining has become the embodiment of automobile quality, which leads to the increasingly complex structure of the car,car’s maintenance and other problems have become increasingly prominent. How to prevent the possible failure of the car, to repair the troubles already occurred, has become the focus and difficult of current research on the automotive industry. Under such a background environment, a new subject--vehicle fault diagnosis technology emerge as the times require, establishing a easy-to-use vehicle diagnosis system for the automotive repair industry has become a urgent requirement.Vehicle fault diagnosis technology as a comprehensive technology, covers a number of areas of the discipline: the modern control theory, signal processing,artificial intelligence, applied mathematics, mathematical statistics and so on. With the breakthrough and development of these disciplines, automobile fault diagnosis technology has been from based on manual experience, chart analysis to diagnose by equipment and instrument, from the traditional "thirty percent by technology, seventy percent by tools" into "seven points diagnosis and three repair", comprehensive analysis of vehicle failure by detection and testing.This paper analyzes the reasons and types of automobile fault, deepen the understanding of the importance of automobile fault diagnosis. Order to basing on the case of automobile faults, establish the vehicle fault search engine. So as to achievethe target of improve the accuracy and speed of the system diagnosis for the vehicle maintenance. The main work done in this paper is as follow:1)Researched on the natural language processing technology, combined with the vehicle fault characteristics description of the system completed Chinese segmentation, and finished the fault cases segmentation according to the dictionary and corpus statistics;2)Established a case based vehicle fault search engine, to support users through the search for the way to get the vehicle fault diagnosis results, complete case matching;3) Through the study and Research on the topic model, realize the marking method of the device;4) Using the database management module management and maintenance,including the corpus, case library, component library and user operating records information data, to increase vehicle fault cases and system to further improve the provision of basic data.