Thermal management plays a critical role in battery operations to improve safety and prolong battery life, especially in high power applications such as electric vehicles. A lumped parameter (LP) battery thermal model (BTM) is usually preferred for real-time thermal management due to its simple structure and ease of implementation. Considering the time-varying model parameters (e.g., the varying convective heat dissipation coefficient under different cooling conditions), an online parameter estimation scheme is needed to improve modelling accuracy. In this paper, a new formulation of adaptive LP BTM is proposed. Unlike the conventional LP BTMs that only consider convection heat transfer, the radiative heat transfer is also considered in the proposed model to better approximate the physical heat dissipation process, which leads to an improved modelling accuracy. On the other hand, the radiative heat transfer introduces nonlinearity to the BTM and poses challenge to online parameter estimation. To tackle this problem, the simplified refined instrumental variable approach is proposed for real-time parameter estimation by reformulating the nonlinear model equations into a linear-in-the-parameter manner. Finally, test data are collected using a Li ion battery. The experimental results have verified the accuracy of the proposed BTM and the effectiveness of the proposed online parameter estimation algorithm.
CORE (COnnecting REpositories)
CORE (COnnecting REpositories)