Font Size: a A A

The Research Of SOC Estimation Methods Based On Equivalent Circuit Model Of Li-ion Battery

Posted on:2018-05-21Degree:MasterType:Thesis
Country:ChinaCandidate:D LuoFull Text:PDF
GTID:2382330596965617Subject:Power Machinery and Engineering
Abstract/Summary:PDF Full Text Request
With the emerging issues of environmental pollution and fuel shortage,the problems of resources and environment have been increasingly concerned by governments.EV,seen as the breakthrough solution,has gained tremendous progress under various supporting policies in recent years.Power battery and the manage system are the core components of an EV.Accurate estimation of SOC makes senses for improvements in battery's energy efficiency,security and service life.Three SOC estimation methods have been studied according to a certain type of NCM battery,with the finding showed as follows.a)The battery equivalent circuit model with temperature factor was built and its feasibility was verified.Data of different charge/discharge condition and SOC were collected through VPRT tests under different temperature.Parameters of the equivalent circuit model were identified by writing MATLAB script.Temperature factor was utilized to simplify parameters of models at different temperature points,then a battery model with temperature factor was proposed,whose accuracy was validated in Simulink.The simulation result showed that the battery model was of high precision in the temperature range of 0 ~ 40? with the largest RMS less than 25 mV and the overall estimation error less than 5% when applied to SOC estimation of NCM.b)The robustness of the 3 estimation methods was analyzed according to measurement error and modeling error.On the basis of EKF,UKF and WRLS theories,simulation models of the 3 estimation methods were established in Simulink,whose accuracy were verified.Simulation analysis of the models robustness were conducted on condition of comparable model accuracy and results showed:(1)All methods performed well with RMS less than 3% when the model parameters are precise and using NEDC and US06 as the test cycle.(2)WRLS was more dependent on the OCVSOC curve than UKF and EKF,however,the estimating accuracy of which was independent of battery ohmic resistance and polarization resistance.(3)SOC estimating error caused by battery modelling error was not additive.The battery model with temperature factor and the comparative robustness analysis of the 3 SOC estimating methods can provide references for practical engineering.
Keywords/Search Tags:Li-ion battery, SOC estimation, Kalman Filter, WRLS
PDF Full Text Request
Related items