Research On SOC Estimation Algorithm For Lithium Battery Based On AEKF

Now the earth's energy consumption will be exhausted.In order to improve the utilization of green energy and reduce environmental pollution,so many scientists have begun to explore the use of green energy to replace the traditional fossil fuels.At present,the major automobile manufacturers are also developing new energy vehicles to reduce the damage to the earth and the resources requirement.Battery is one of the key components of electric vehicles.Due to the advantages of low cost,high stability,high energy ratio,long service life,environmental protection and so on,lithium iron phosphate batteries are widely used in electric vehicles.It is of great significance to estimate the state of charge(SOC)using the estimation algorithm in the design of the battery management system(BMS).It not only greatly improves the efficiency of energy use can avoid the overcharge or overdischarge of battery caused irreversible damage.Through the rational utilization of battery energy,it can also achieve the ultimate purpose of improving the battery life.We choose a more appropriate estimation algorithm-AEKF algorithm.According to the relevant data analysis shows that the extended Kalman filter method can accurately estimate the battery SOC estimation in the state,But when using the EKF method needs accurately equivalent model and stable noise.In fact these conditions basically are difficult to achieve.In final based on the shortcomings of the EKF algorithm and choose the lithium iron phosphate as the experimental samples.Selecting and using AEKF algorithm as the feasibility of battery state estimation methods to verify the effectiveness and feasibility of the AEKF algorithm.The paper aims at the following: the Li+ battery equivalent model selection method,equivalent model parameters identification algorithm,the feasibility of the model validation,the battery AEKF estimation algorithm to verify the feasibility,and combining the equivalent model and AEKF estimation algorithm the advantages of the targeted research.The main research contents are as follows :(1)Research on equivalent model of Li+ battery.By analyzing a large number of experimental data and considering the complexity of the model.The equivalent circuit model of PNGV is selected as the equivalent model of Li+ power battery in view of the contradiction between the accuracy and complexity of the model.The advantage of this model is that it not only takes into account the internal resistance of the ohmic resistance,but also takes into account the polarization of the battery.(2)According to the equivalent circuit model of battery are selected.The state equation and output equation are established.Based on the mixed pulse power test and the least square identification method,the parameters of the PNGV circuit model are identified.The feasibility of the selected PNGV circuit model is verified by HPPC experiment.Based on the analysis and processing of the experimental data by Matlab software the feasibility and effectiveness of the proposed equivalent model is validated which lays the foundation for the further study on the estimation of the state of battery charge.(3)The Matlab software is used to verify the effectiveness and feasibility of the two selected SOC estimation methods of EKF and AEKF algorithm under the combined pulse condition.(4)In final the selected equivalent circuit model of battery is combined with the EKF(extend Kalman filter)and AEKF(adaptive extend Kalman filter)estimation algorithm.Simulation of SOC estimation of lithium ion power battery is executed by Matlab under UDDS and NYCC conditions.By comparing the simulation results of the two algorithms,we can see that the two selected estimation methods are useful to estimate the battery SOC.The results show that the AEKF algorithm compared to EKF algorithm is not only effective.It also can significantly reduce the prediction error and improve the estimation accuracy.