Estimation of state of charge considering impact of vibrations on traction battery pack

Interest towards electric vehicle adoption is on the rise due to the lower running and maintenance cost it offers, along with zero tailpipe emissions. Range anxiety is one of the only concern that affects the adoption of electric vehicles. The state of charge of the traction battery pack has to be accurately determined and provided to the user to avoid range anxiety. Minute battery parameters has to be considered to improve the accuracy of the state of charge determination. In order to overcome the problem of range anxiety, an innovative strategy that takes into account how vibrations affect the performance of EV batteries is developed in this research. By doing this, the state of charge estimation precision is improved and thereby raises the driver’s faith in electric vehicles. The impacts and vibrations felt on the traction battery pack during driving would lead to heat generation. The heat generated is found to be highest when the vibrations resonate at the natural frequencies of the traction battery pack. The natural frequency of the battery pack is considered when the battery is kept in the battery chamber of the two-wheeler electric vehicle. The vibrations at natural frequency produces heat which is accounted for when the state of charge is determined. To obtain accurate state of charge estimation, a Kalman filter-based approach is used. The Kalman filter-based estimation uses the conventional methods which are the open circuit voltage method and the Coulomb counting method to improve the estimation process along with the consideration of the heat component due to vibrations and impact. The vibration analysis is performed using MATLAB, while the state of charge determination is implemented in hardware and the Kalman estimation done using Python. The system is modelled on an electric two-wheeler platform and the testing is done to compare the state of charge accuracy of the open circuit voltage method, the Coulomb counting method and the Kalman filter-based estimation approach. The inclusion of the vibrational heat analysis for State of Charge estimation in the hardware testing of the electric two-wheeler provides an accurate state of charge value.