In this thesis, a Kalman filter for estimating the state of charge (SoC) of a multicell
battery pack for an electric car was implemented.
To implement such an algorithm it was necessary to: measure the cell response
under different conditions, determine the characteristic parameters of the cell
from its responses, establish the Kalman filter model, prepare the program code
of the self-developed BMS, implement the SoC algorithm from the simulation to
the program code and validate the proper operation and accuracy of the algorithm.
After the implementation of the algorithm it was determined that it works reliably,
and, when compared to a current integration method, has an uncertainty
of about 5%. Although additional measurements and an upgraded algorithm
would be necessary to achieve better accuracy, it was determined that because
of the successful implementation and documentation of the algorithm, the purpose
of the thesis was successfully achieved, as it enables easier maintenance and
implementation of additional improvements.
|