Research And Implementation Of Real Time Vehicle Monitoring And Diagnosis System Based On Android Client

The great popularity of vehicles increases tremendous pressure on vehicles after-sales service and fault repair industry.With the application of electronic technology in vehicles,the vehicle has greatly improved in these aspects,safety,power,flexibility and artificial intelligence.However,this application also causes a series of problems,such as multiple kinds of vehicle faults and missed diagnosis.Therefore,effective diagnosis and repair are necessary conditions for the long-term development of the vehicle industry.In this paper,we firstly investigate the current research status of the vehicle fault diagnosis system at home and abroad.When analyzing the existing fault diagnosis schemes,we find the current schemes have the following shortcomings,large size of PC and embedded devices,high cost and inflexibility.To cope with these drawbacks,based on B/S and C/S mixed structure,a new system structure has been proposed,which combines the real-time monitoring in Android terminal with the remote monitoring in Web terminal.In detail,via Bluetooth,we use smart phones and OBD interfaces to transmit data to remote server,which aims to achieve real-time monitoring and remote diagnostics.Then,via wireless network,we connect the data collection,data transmission and remote diagnosis,which complete the integration of vehicles,users and repairs.Secondly,a new vehicle fault diagnosis model has been proposed.Due to the randomness and intermittency of vehicle faults,the complexity of vehicle system and the difficulty of data collection in short-term,we combine Fuzzy rules with Neural network to diagnose the vehicle faults.Specifically,without depending on the precise mathematical model of the vehicle system,we introduce Fuzzy rules to accelerate the learning speed of Neural networks,which can take advantages of language knowledge and supervised learning simultaneously and lead to continuous improvement of diagnostic results.Moreover,we use MVP model to achieve realtime monitoring based on Android terminal.To be more specific,using mobile phone Bluetooth to communicate with OBD-II interface ELM327,which based on K-bus and CAN-bus,fault code and vehicle real-time status can be transmitted.Analyzing the functional requirements of clients,we design and implement seven functional modules,i.e.,device connection,data collection,fault diagnosis,location acquisition,upload data,account management and system settings and tackle many technical problems,such as geographic coordinates conversion,no response of mobile phone,screen adaptation and network communications.Thirdly,we apply B/S model to design and realize the remote monitoring of vehicle fault diagnosis.The user group is divided into two levels: ordinary owners and internal staffs of maintenance companies,which are assigned with different operating authority.Adopting modular design ideas,we divide the system into five functional modules,namely,Android communication module,database access interface module,browser-based management module and car owner module.The communications with server can be established by AJAX,which can transmit user instructions and collect new data.What's more,the maintenance staff can login in remotely.With the assistance of designed the vehicle fault diagnosis model,they can help users achieve remote diagnosis and repair.To improve browser compatibility,we use the BootStrap framework to achieve response layout and fit into different screens.Finally,the testing tools,such as Mocha,Zombie and Load Test,are used to analyze the compatibility,security and function of Android-based real-time monitoring terminal and Webbased remote monitoring terminal.The analysis results demonstrate that the system is stable,safe and reliable,which verifies the feasibility of the designed scheme and achieves the expected goal.