The Driving Range Estimation Of EV Based On Battery State-of -charge And Energy Consumption Prediction

With the increasingly serious environmental pollution and energy issues, people are increasingly concerned about energy-saving and environmental protection. Governments and vehicle manufacturers around the world are investing great efforts in the development of new energy vehicle for solving the environmental and energy issues. Battery electric vehicle (BEV), as one kind of new energy vehicles, has become the R& D hot spot in the automotive industry due to its obvious advantages. But there are still many problems in the BEV to be solved because many key technical problems have not been solved. The battery energy management is one of the main key technologies. The state-of-charge (SOC) estimation of the battery is one of the main functions of the battery management system. We can forecast the remaining driving range of EV based on the battery SOC. If we can make an accurate estimation of the driving range of EV, the users will become more confident when they drive an EV. Therefore, in order to promote the electric car, research efforts have to be made to develop an effective method for estimating the battery state of charge and remaining driving range of EV accurately.This dissertation studies on the estimation of SOC and the prediction of driving range of EV. We first discuss the estimation method of state-of charge. The validity of many estimation methods are just confirmed through single-cell experiment. But the battery pack used on the EV consists of many single cells. It is very important to develop a state estimation method for battery pack. In this dissertation, the battery pack is treated as one single cell with high voltage and large capacity, and then we make a research on the state estimation method. Firstly, we build a battery equivalent circuit model and identify the parameters of model using the recursive least squares based on the pack discharge experiment. Secondly, calculate the pack SOC based on particle filter (PF) algorithm. The state equation of battery system is established based on the battery equivalent circuit model, then complete the pack state estimation using PF method on account of the experiments data. We also estimate the pack state based on extended Kalman filter (EKF). The results of two estimation method show that the PF method has better performance on estimating the pack state-of-charge.Now we have gotten the residual capacity of the battery pack. In order to calculate the remaining driving range of EV further, we need to estimate the energy consumption of vehicles in the future driving condition. This dissertation puts forward a future energy consumption estimation method based on the vehicle energy consumption in the past. The relationship between the energy consumption difference and the changing in driving condition is analyzed based on the support vector machine regression theory. After determining the vehicle future energy needs, we can forecast the driving range of vehicle based on the energy state estimation of battery pack and vehicle energy consumption prediction. At last, through simulation experiment proves that the driving range estimation method is useful.