A novel Doyle-Fuller-Newman battery model formulation for online parameter estimation

This work proposes a new formulation of the Doyle-Fuller-Newman (DFN) model of an electrochemical cell by solving in an alternative manner the partial differential equations of the electrode potentials. This allows the user to run the model with shorter computational times and to capture the capacitive behavior of electrochemical cells by using a single simulation framework. Moreover, a novel parameter estimation method is proposed to enhance the accuracy of the model. Furthermore, the method allows extracting the meaningful information regarding the relationship between the relevant parameters of the model and the operating conditions of the cell. This work considers a Lithium-ion cell NMC111/Graphite pouch cell, and the accuracy of the approach is validated using laboratory galvanostatic measurement tests at different C-rate values and with a realistic Federal Urban Driving Schedule test (FUDS). The proposed method not only shows minimal computational times, but also displays high estimation accuracy. In fact, with the proposed online parameter estimator, the mean relative and mean absolute errors are around 0.1% and 6 – 12 mV , respectively, highlighting the robustness and high accuracy of the proposed model under a broad range of experimental conditions. • A novel manner of solving the boundary conditions of the Doyle-Fuller-Newman model. • The proposed model formulation exhibits small computational burdens, and it allows for detecting the electrical capacity behavior of the cell. • An online parameter estimation based on the square-root unscented Kalman filter is proposed. • The online parameter estimator enhances the accuracy of the model and allows one to extract the SoC-dependency nature of several model parameters.