Research On Life Of Battery Of The Electric Vehicle Based On Hidden Markov Model

With the rapid development of world’s economy, the production and sales ofdomestic and car have maintained a high growth rate and the number of cars hasgrowth substantially. With the world’s energy being shortage and environmentalpollution being increasingly serious, electric vehicles, this kind of car is lowenergy consumption and less pollution, go into everyone’s life. But the batteriesof electric vehicle have some problems just like its high cost, short mileage andshort life. These problems restrict the development of electric vehicle. Therefore,various types of battery management system（BMS） detect the fade of batteryas an important parameter.This article proposed a new method that could distinguish the fade mode ofthe batteries of electric vehicle and predicted future fade model and remaininguseful life. By this way users can learn their own life of battery and therebyadjusting their drive model so that improve efficiency of battery.Firstly, the article has completed the analysis of using model of the batteryof electric vehicle. Then it used a curve clustering algorithm to classfy thedifferent using modes and using modes belong to the same fade mode would beclassification.The article proposed a new concept-fade point to evaluate the health ofbattery instead of capacity. Then the article used the classification above tobuild the model based on Hidden Markov Model to predict the future fade model.With the history of battery’s fade, we predict remaining useful life of battery.Finally, we designed a new charging and discharging experiments of battery,built experimental platform and finished the experiment. We analyzed theexperimental data and used data to build the model to verify the correctness ofthe algorithm.