The Design And Implementation Of System Of Algorithm Management For EV’s Battery Monitoring Platform

The development of new energy vehicle industry is a major strategy for China to achieve energy saving and emission reduction and "overtaking at corners" in the automobile industry.Electric vehicles(EV)are the main representatives of new energy vehicles at present,and ensuring the safety of automotive power batteries has become one of the topics of general concern.The market needs to accurately estimate and detect the safety state of batteries.It is an important measure to develop a battery monitoring platform to monitor the battery state of the vehicle and send an alarm before the battery safety failure or failure occurs.Battery monitoring platform is based on a large number of battery data and requires the system to provide a large number of algorithm model services.However,algorithm developers cannot manage a large number of algorithm tasks manually,so the algorithm management system,Proxima,is urgently needed.The system provides functions of multi-service management,algorithm metadata,parameter,scheduling,monitoring and result management.This thesis introduces the project development background and related research and application status quo,clarifies the necessity of system development,and briefly introduces the technology used in system development.In system analysis and overall design,the article illustrates the system overall plan,define the system boundary,this thesis briefly introduces the project plays the role of the battery monitoring platform,analyzes the system requirements,refined out of the business management,algorithm of metadata management,parameter management,scheduling and monitoring as well as the algorithm results management,and other functions.This thesis describes the working principle of the system,through the structure design,communication design,functional composition design,deployment design and other methods,the overall design of the system,analysis and design of the system database.Using class diagram,time sequence diagram and other description methods,the key functions of each function module are introduced in detail and the specific implementation details and results are given.The system is tested,and the test results meet expectations.As an important part of the battery monitoring platform,this system has been put into use,and the application results are good.It has better solved the problems such as irregular algorithm online process,difficult algorithm data management,low efficiency of algorithm service scheduling and scattered algorithm running results,and provided a good basic service for the battery business of enterprises.