On-line estimation of lithium-ion battery impedance parameters using a novel varied-parameters approach

• An efficient algorithm is proposed for the estimation of battery impedance parameters.
• The battery impedance is represented as a nonlinear equivalent circuit model.
• For the first time, current dependency of battery impedance is considered by parameter estimation.
• The proposed method can be used as a basis for an adaptive prediction of the battery power.

In this paper a novel technique for on-line estimation of impedance parameters of a battery equivalent circuit model (ECM) is presented. The ECM is often used as a basis of battery management algorithms to calculate the state of charge or the maximal available power. Especially the latter requires that the parameters of the ECM describe precisely the battery impedance. Unfortunately, the battery impedance changes over the battery lifetime due to aging. Therefore, the parameters of the ECM must be updated continuously. In this paper such an update is achieved using a novel approach that in addition allows considering current dependency of the battery impedance. The basic idea is that the change of the battery voltage under load is predicted using the ECM with different parameter sets during a short time period. The predicted voltage is then compared to the sensor data and the best parameter set is selected. It is then used as a basis in the next adaption step. A necessary strategy for building parameter sets is implemented for a fast and efficient convergence which results in an accurate parameter estimation. The algorithm runs on a standard 16 bit 80 MHz microcontroller at only 16% processor load.