Toward Real-Time Simulation of Physics Based Lithium-Ion Battery Models

Abstract

Recent interest in lithium-ion batteries for electric and hybrid vehicles, satellite, defense, and military applications has increased the demand on the computational efficiency of lithium-ion battery models. This paper presents an effective approach to simulate physics based lithium-ion battery models in real-time ͑milliseconds͒ for simulation and control in hybrid environments. The battery model used for the simulation is derived from the first principles as an isothermal pseudo two-dimensional model with incorporation of concentrated solution theory, porous electrode theory, and due consideration for the variations in electronic/ionic conductivities and diffusivities using the Bruggmann coefficient.

Reduction of Model Order Based on Proper Orthogonal Decomposition for Lithium-Ion Battery Simulations

“…Consequently, several simplifications of their model have been published to reduce the computation time associated with diffusion of lithium ions in the solid phase. [8][9][10] Also, Subramanian et al 11 developed a real-time simulation model using a combination of perturbation techniques, volume averaging, and intuition-based simplications. Although they reported that the computational time for their real-time simulation model for a single process was around 100 ms, to derive the lower-order model by using this method one needs to carry out preprocessing and have a priori knowledge of the behavior of the system under different conditions, which makes their method less flexible than desired.…”

Towards an Ultimate Battery Thermal Management System: A Review

“…Safety concerns include safeguarding against fire, explosion, and the corrosive nature of the electrolyte, as well as the venting of toxic or flammable gases [48,85,[88][89][90][91][92][93][94][95][96][97][98]. All auxiliary equipment such as thermal sensors, thermostats, heaters and switching devices must be designed so that they cannot be the source of an explosion.…”

An IMEX scheme for reaction-diffusion equations: application for a PEM fuel cell model

“…These models are usually solved by using only a single numerical treatment, e.g., Runge-Kutta, Newton or Crank-Nicholson methods. Subramanian et al [14] developed a method to reduce the number of the governing equations of Li-ion battery simulation by using different mathematical techniques. The original problem with a proper discretization consists of 4800 equations which can be reduced to 49, and finally the simulation time of the discharge curve can be cut to 85 ms.…”