|Inaccurate estimation of the state-of-charge (SOC) and the state-of-health (SOH) may lead to safety issues. Currently, estimations of SOC and SOH that based on electric equivalent circuit models (ECMs) have been widely applied. ECMs consist of electrical elements, which do not represent electrochemical behaviors, so they lack the ability of long-term prediction of battery lives. Alternatively, an electrochemical-based reduced-order model (ROM) can predict internal states, such as concentrations and potentials, which are directly related to the SOC. Moreover, degradation mechanisms can be modeled and integrated into the ROM to predict change aging-related parameters such as loss of lithium ions, loss of active materials, consumption of the electrolyte solvent and increase of internal resistances, which are related to the SOH.
In this work, we propose an SOC and SOH estimator in conjunction with the beginning-of-life (BOL) and aging parameter estimation method based on a ROM. At first, a procedure for automatic estimation of parameters of a lithium-ion battery at the BOL is developed, where a two-step sensitivity analysis is designed to group the parameters and divide them by the best SOC window. Then, the genetic algorithm is applied to minimize a multi-objective function that is the mean square error of voltage for each SOC window.
Based on the ROM with accurately estimated parameters, an SOC estimation algorithm is developed using an adaptive square-root sigma-point Kalman filter (ASR-SPKF) with equality state constraints. Equality state constraints derived from the principle of charge conservation are introduced to improve the accuracy of both anode and cathode SOC estimations. Because of its fast convergence speed, the cathode SOC is used to represent the bulk SOC. Approaches used to adaptively update the covariance matrices of the filter based on the covariance matching method are also incorporated. As a result, the covariance matrix of process noise is adjusted automatically. Comparative studies of three nonlinear filters concerning estimation accuracy, error bounds, recovery time from an initial offset, and computational time reveal that the ASR-SPKF has the most outstanding performance.
Finally, an online estimator for SOH and aging parameters is developed using a high-fidelity, reduced-order physics-based life model, where a pseudo-two-dimensional model coupled with degradation model of two types of side reactions—solid electrolyte interphase (SEI) layer formation and lithium plating—is used for the negative electrode, and a single particle model is used for the positive electrode to increase computational efficiency. A control-oriented incremental aging model is developed with the employment of a particle filter, from which the SOH and aging parameters are continuously monitored from real-time current and terminal voltage measurements. Finally, the developed method is tested in a battery-in-the-loop test station with a large format, 42 Ah, lithium-ion battery with Li(NiMnCo)O2/Carbon electrode chemistry.