The Design Of Battery Management System For Electric Vehicles Based On ARM

With the growing global energy crisis and environmental pollution, electric vehicles, byadvantages of non-polluting and high-performance, are becoming into the developmentdirection of the contemporary car. The battery management system (BMS) is the keycomponent of the electric vehicle module, and batteries work properly with it. The state ofcharge of the battery (SOC) describes the remaining battery power, and is one of the mostimportant parameters of the battery in the use of the process. Accurately estimating the SOCcan prevent battery overcharge or over-discharge, effectively extend the service life of thebattery, and predict electric vehicle’s renewable mileage while traveling. This paper focuseson the main functions of the electric vehicle battery management system and SOC estimationmethod, puts up the SOC estimation method using adaptive fuzzy neural network, and designsan ARM7-based embedded battery management system.As SOC estimation is affected by temperature, aging, a charge-discharge rate,self-discharge and other factors, the battery is rendered as high nonlinearity in practicalapplication, and that brings great difficulties for the accurate estimation of the SOC. There islarge temperature difference between batteries when they are working,which is up to55℃.But most SOC estimation algorithm does not consider the influence of the temperature factors;otherwise, there is a nonlinear relationship between the SOC and its terminal voltage, butmany existing algorithms deal with it as a linear relationship directly, so that the SOCestimation error is large, which leads that the energy of the electric car cannot be managedaccurately, and it directly impacts on the economy and promotion of electric vehicles.The works of the Lithium-Ion battery has been introduced and collected the data of10AhLithium-Ion battery in this paper. The battery SOC estimation methods is proposed tocombine neural network self-learning ability and fuzzy logic reasoning ability adaptive fuzzy neural network and SOC estimation model based on ANFIS by analyzing the strengths andweaknesses and application occasions of the existing SOC estimation methods. The SOCestimation model on the power Lithium-Ion battery was established based on ANFIS. Thepaper made experiences and simulations on the MATLAB platform for the ANFIS powerLithium-Ion battery SOC estimation model, and used a large number of experimental datacollected to train the model, the results showed that the model could reflect the batterycharacteristics better and the error between the SOC obtained predicted value and themeasured value was low, the battery power can be used for the Lithium-Ion battery SOCestimation.Finally, LPC2210has been made as a master control system core chip and hardware partand software components of electric vehicle battery management system has been designedbased on the ARM7. The hardware part has included LPC2210minimum system, dataacquisition circuit, the protection circuit and the communication circuit. The software hasused design methods of modular program and given the flow chart of the main program andsubroutine and Focused on solving the SOC estimation method based on ANFIS.