Study On Prediction Of The Residual Life Of Battery By Improving The Initial Value ? Hidden Markov Model

In recent years,with the rapidly development of the global economy,the production of cars has increased greatly.As more and more fuel cars exhaust emissions during driving,environmental is becoming more and more serious.In order to improve environmental problems,the electric vehicles are gradually accept and used by people.But as the only source of the car power battery,the problems such as high manufacturing cost,short cycle usage and residual life hard to detect,it fundamentally restricts the promotion of electric vehicles.An improved prediction model was proposed,that be used to predict the performance of the recession of the battery life and the remaining life of the battery,so that people more clearly understand and rationalize the use of batteries in the electric vehicles.Based on Lithium-ion battery as the research object,firstly,the main factors that affecting battery recession life was analyzed by the experimental results,then used the influencing factors to establish the battery cycle experiment,in order to represent the instantaneous signal of battery recession,the discharge platform point and the concept of threshold voltage range was proposed by the discharge voltage curve.The set of discharge curves produced by the experiment is classified and analyzed,the potential declining performance is used to indicate the characteristics of battery life attenuation.Then through the improvement of the elements of the hidden Markov model,the improved initial value ? hidden Markov model is obtained,combined with the daily driving of electric vehicles and the influencing factors,battery charge and discharge experimental program was designed,the experimental platform was build and complete the test.An improved model was established based on experimental data to predict the subsequent recession performance,and then according to the battery capacity of the recession,the recession performance coefficient was putted forward to indicate battery capacity attenuation degree of correlation.Finally in the experimental database,the cumulative capacity of the battery was used to directly predict the residual service life of the battery.Through the experimental results of the comparison among the improved model,the original model and the real measured value can be found that the model can predict the residual life of batteries,to verify the accuracy of the algorithm.