The Vehicle Maintenance System And Knowledge Base Construction Based On Service Chain Business Data Resources

With the rapid development of Chinese automobile manufacturing industry,Car ownership is growing rapidly.The huge vehicle population has brought huge market value to car after-sales service,but it also cause some problems.We have analysis the maintenance data in after-sale service system of the automotive industry chain collaboration platform and found that the service providers have uneven maintenance levels and some service providers have low one-time repair rate.There are many types of vehicles in the manufacturing plant,and the vehicle structure,the fault phenomena and reasons are complicated,which make it difficult to eliminate the fault when only depending on artificial experience.Because a huge number of vehicle maintenance cases cannot be used,resulting in increasing the information sharing barriers and the phenomenon of knowledge waste.These industrial data play an important role in the fields of intelligent service such as product production design,fault diagnosis,after sale service and so on.Therefore,according to the existing problems,a vehicle maintenance fault diagnosis system based on industrial chain collaborative service platform is designed and implemented in this thesis.This system will provide guidance and reference for vehicle maintenance.First of all,this thesis took the WP enterprise as an example,then analyzed its problems and repairing situation on the platform.And then the system requirements and system goals are proposed,and the overall solution and functional structure of the system are designed.This thesis proposes two kinds of fault diagnosis methods,one of which is fuzzy and other is precision.Fuzzy diagnosis is mainly used to analyze the fault phenomena when the vehicle failure happened.The Fuzzy diagnosis finds the vehicle failure according to the description of fault phenomena and the matching degree of vehicle information and case knowledge base.Accurate diagnosis is used for diagnosis of deterministic failures of vehicles.By selecting a certain certainty phenomenon,the reason of the failure is inferred from the rule base.Then three kinds of knowledge bases are constructed according to the proposed two diagnostic methods,which provide data support and knowledge sharing for fault diagnosis.Among them,the case knowledge base constructed through historical maintenance files of vehicles is the most abundant and most practical value;The rule knowledge base is a rule base formed by sorting out common vehicle faults and maintenance files,which can provide accurate reasoning;Knowledge sharing is the vehicle structure and maintenance guidance document maintained by the manufacturer to guide the service provider to repair the vehicle.The E-R model was analyzed by the three knowledge representations,and then the database table storage knowledge was designed.According to the fuzzy and precise methods used in diagnosis,the gray theory and the fuzzy case reasoning process based on grey correlation analysis are firstly analyzed,and then this paper studied the keyword extraction and semantic similarity based on How Net,the words that describe fault description and used in Reasoning process.And then this thesis analysis the reverse reasoning process based on the Petri network.And according the result from the reverse reasoning,process and fault phenomenon,this thesis analysis the primary fault reason based on the minimal cut set process.Finally,this thesis uses the three-tier architecture based on the B/S model to design the system according the actual situation of the distribution of service providers.This system has implemented functions which has been discussed in demands.This system also implemented the rapid vehicle fault diagnosis and vehicle maintenance knowledge sharing,which reduce maintenance time and improve effectiveness.