Parameter identification of a battery Equivalent Circuit Model (ECM) using Particle Swarm Optimization (PSO). The ECM includes a series resistance R0, an RC branch with R1 and C1, and a polynomial OCV-SOC relationship.
The equations that describe the Equivalent Circuit Model (ECM) are given:
The flow diagram takes you through the parameter optimization process.
The equations describing this optimisation are:
A drive cycle of current versus time was applied to a cell and the resultant terminal voltage was measured.
This current versus time profile was also applied to the cell ECM model and the terminal voltage predicted.
The measured versus ECM predicted cell terminal voltage shows a very good correlation.
The optimization explores different lower bounds for R0 and R1 and minimizes the RMSE between measured and simulated voltages.
The PSO algorithm iteratively updates particle positions and velocities based on cognitive and social components to find the global best parameters.
Results are presented of RMSE values and overlay plots of measured vs. predicted voltages for all lower bound combinations. This allows evaluation of the influence of parameter bounds on the model accuracy.
Krishna S Hanamaraddi has made the model available on GitHub
These approaches are not for real time implementations.