Research On Remaining Driving Range Estimation Of Battery Electric Vehicle

The existing driving range estimation method of Electric vehicles is relatively simple and there is a big gap compared with the actual mileage, this directly causes the users to worry about the current power can not ensure that the vehicle arrived at the expected location, resulting in "range anxiety" and reducing users confidence in electric vehicle purchases. Thus, In order to improve the popularity of electric vehicles and using convenience, to a certain extent, it is an important method to improve the estimation accuracy of the remaining range, but also it is the purpose of this research, the core of research reads as follows:(1) Based on MATLAB/Simulink establishing of vehicle dynamics, driver, motor, battery and vehicle energy consumption model, the model is established to pave the way for the subsequent chapters of the remaining estimated driving range of simulation models.(2)Detailed study of three remaining driving range estimation method: first, the traditional fuel vehicles estimated remaining driving range method is applied in pure electric vehicles, namely the average energy consumption, the method can be used to calculate the energy consumption by obtaining the voltage and current of the battery, but it is relatively rough; then leads to driving condition recognition method, which automatically determine the current consumption according to the current driving conditions, and add the following three points based on the working conditions of traditional identification methods:(1) Establishing fuzzy rules between the characteristic parameters of driving conditions and energy consumption, that is, the identification of the condition is not corresponding to a single energy consumption, but according to different characteristic parameter values of different working conditions, the corresponding energy consumption is different.(2) Unit energy consumption driving range optimization, according to the practical experience, the linear relationship between the unit energy consumption and the remaining energy consumption is established and the remaining mileage is reduced to the optimization trend.(3) Using Kalman filter to remaining driving range which is the output to the instrument, making the remaining driving range more realistic; the last, the new condition prediction method is on the base of the condition recognition, the method will be combined with BP neural network and Markov chain together, achieve forecasting in a certain time range of driving condition, thus once again improve remaining driving range estimation accuracy.(3) Due to the impact of air conditioning on the remaining driving range, aiming at the opening and closing of air conditioning, used another average energy consumption estimation program separately: average energy consumption method based on air conditioning power and time of opening.(4) In order to compare the advantages and disadvantages of the three programs in detail, the advantages and disadvantages of the remaining range were evaluated from two aspects of the driving range and energy consumption respectively, simulation results show the superiority of the condition prediction method.(5) The rapid prototype development platform was used in this paper. Firstly, based on D2P-Motohawk and HIL simulation platform to build a hardware in the loop simulation test environment, demonstrate the feasibility and real-time of the algorithm model proposed; Then test remaining driving range by tracking NEDC on the drum test bench, By the running mileage and energy consumption, compared the pros and cons of the three programs, the experimental results show that the combination of conditions identification and prediction method can meet the requirements of actual control, and improve the accuracy of remaining driving range estimation.