Core Temperature Observer Design and Model Parameter Uncertainty Analysis for a Lithium Battery

摘要

Temperature monitoring for Lithium ion batteries is an important factor for its properties. Since the core temperature is difficult to measure, a thermal model is used for estimation. In this paper, an offline parameter identification procedure is applied, by which the parameters of a three state lumped thermal model can be identified. Based on the model a Luenberger observer is designed with the strength of compensating variations of the model parameters. Only the easily accessible and measurable temperature of the copper pole is fed back to the observer. The paper deals with the robustness analysis of the proposed observer in that the actual plant differs from the model considered in the observer. The capability of the approach is beneficial as an online (and computational power consuming) parameter identification is no longer required. Instead, the thermal model is obtained offline and the observer is robust against model parameter variations, e. g. ageing, so that a continuous update of the model is not required.