| Autor: |
E. Bekaert, L. Oca, M. Scott Trimboli, Gregory L. Plett, U. Iraola, E. Miguel |
| Rok vydání: |
2021 |
| Předmět: |
|
| Zdroj: |
Journal of Energy Storage. 44:103388 |
| ISSN: |
2352-152X |
| DOI: |
10.1016/j.est.2021.103388 |
| Popis: |
Electrochemical models are an incipient technique for estimation of battery cells internal variables, useful for cells design or state of function optimization. One of the non-trivial procedures that allow the use of this type of models is the estimation of model parameter values. This paper presents a review of the existing computational parameter estimation methods for rocking chair batteries electrochemical models, a crucial step for real case implementation. Physics-based models cannot reach accurate predictions if the parameters are not properly estimated, what highlights the necessity of reviewing the validity of these protocols, that are not extensively treated within literature. The gathered methods are explained and analyzed taking into account the accuracy and extent of the presented results, to give the most objective overview of their applicability within real case scenarios. The methods are classified into two different groups: single optimization analysis (using only one optimization procedure to estimate parameters) and multiple optimization analysis (methods using multiple optimizations). In addition, the need for at least some amount of physico-chemical characterization is analyzed as a common procedure for all the parameter estimation methods. The accuracy of each method is determined, taking as reference the best achievements found in literature. The results show that it is possible to estimate parameters with a high accuracy using non-invasive parameter estimation methods. Finally the potential of mixed (non invasive and physico-chemical based) methodologies is presented. These type of estimation procedures can potentially increase the accuracy of the procedures by lightening up the optimizations involved in the processes, and increasing the ability to estimate values for insensitive parameters. These mixed methods could achieve faster and cheaper estimation protocols, making them more efficient in general. |
| Databáze: |
OpenAIRE |
| Externí odkaz: |
|
Full Text from ScienceDirect