System Identification of Battery Single Particle Model Parameters Using New Data Optimization Approach

Identification of model parameters is a critical topic in battery modeling research, as the accuracy of parameters determines the efficacy of the widely used model-based battery engineering. The data-based system identification approach, which only uses easily available input and output measurement data, has been widely adopted due to its convenience and noninvasiveness. However, such approach faces two major challenges, namely how to manage the large parameter set of the model and generate the optimal/adequate data (input excitation) for accurate estimation. To overcome the challenges, we first propose a new parameter ranking and grouping method based on the Bode plots of the derived parameter sensitivity transfer functions to categorize parameters into different groups, which would enable sequential estimation of smaller parameter groups. Subsequently, we formulate a new data optimization approach to generate the optimal current excitations for estimating each group of parameters, which considers both the traditional metric of Fisher information and the data structures for mitigating the impacts of system uncertainties, which are not addressed in current practice. The proposed methodology is applied to estimating the electrochemical parameters of a single particle lithium-ion battery model in experiments, showing improved estimation and voltage prediction accuracy compared with the traditional approach.