Electrochemical models (EMs) can be used to accurately describe the phenomena occurring inside the cells composing a battery pack. However, the computational cost required for their numerical simulation grows exponentially with system size. The aim of this work is to propose a method based on the waveform relaxation framework able to provide a significant speed up in the simulation of large battery packs. The iterative approach we propose is based on the decomposition of the model in smaller submodels that are solved in parallel. The methodology is general and can be used in principle with any EM. A battery pack composed of parallel connected cells, modeled with a Single Particle Model with Electrolyte, Thermal and aging dynamics (SPMeT with aging), is used in this work as a proof of concept. Results show that using appropriate conditions, it is possible to obtain a significantly faster convergence than centralized methods to the solution of the original problem for realistic battery packs (e.g. Tesla Model S battery level — 74 cells in parallel) with a high level of precision.
A computationally efficient implementation of a battery pack electrochemical model using waveform relaxation
Ciaramella, Gabriele;
2022-01-01
Abstract
Electrochemical models (EMs) can be used to accurately describe the phenomena occurring inside the cells composing a battery pack. However, the computational cost required for their numerical simulation grows exponentially with system size. The aim of this work is to propose a method based on the waveform relaxation framework able to provide a significant speed up in the simulation of large battery packs. The iterative approach we propose is based on the decomposition of the model in smaller submodels that are solved in parallel. The methodology is general and can be used in principle with any EM. A battery pack composed of parallel connected cells, modeled with a Single Particle Model with Electrolyte, Thermal and aging dynamics (SPMeT with aging), is used in this work as a proof of concept. Results show that using appropriate conditions, it is possible to obtain a significantly faster convergence than centralized methods to the solution of the original problem for realistic battery packs (e.g. Tesla Model S battery level — 74 cells in parallel) with a high level of precision.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
