Study of Parameters Identification Method of Li-Ion Battery Model for EV Power Profile Based on Transient Characteristics Data

Power simulation of lithium ion battery through battery model is of great significance for dynamic response simulation, heat generation calculation and charge-discharge strategy development. The accuracy and applicability of the model become crucial. In order to demonstrate the battery transient characteristics more effectively, a novel identification method for parameters of the 2nd order RC equivalent circuit model was proposed. Based on the derived evolution law of battery transient characteristics under the continuous pulse excitation, four feature points are extracted for parameter identification in each cycle. The proposed method reduced the time cost of identification from 11796.88s to 0.06s while ensuring that the error of voltage doesn’t exceed 2.2mV. In order to verify the power profiles applicability of the proposed method, applicability analysis of power profile for different identification methods was carried out including the methods using different amount of data (4N points, 200 points, 6000 points) under unidirectional current pulse excitation (UCPE), bidirectional current pulse excitation (BCPE) and unidirectional voltage pulse excitation (UVPE). It was illustrated that the identification process using data of multiple cycles could significantly reduce errors, including maximum error and average error. What’s more, the proposed method under UCPE had the lowest maximum error of 0.420% in voltage simulation and -0.421% in the current simulation of power profiles. Compared with the conventional method (using 200 points of single pulse data for parameter identification), the proposed method can reduce the average voltage error and the maximum error by 62.5% and 11.8% respectively under the DST power profile.


  • English

Media Info

Subject/Index Terms

Filing Info

  • Accession Number: 01766347
  • Record Type: Publication
  • Files: TLIB, TRIS
  • Created Date: Jan 5 2021 2:01PM