Analysis and evaluation of electric batteries model
Loading...
Date
2023-10-10
Authors
Χορνός, Κυριάκος
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Batteries nowadays are used in a big variety of applications, from small household electric devices to aerospace applications. Understanding the batteries’ behavior under loads is essential since in some cases the type of battery and its characteristics must be determined by evaluating various parameters. Environmental conditions, loads profiles, safety parameters and many more must be considered in the initial designs of the systems. To achieve a successful design and to predict a battery’s behavior, extensive simulations and experiments must be conducted. Considering that, models are developed with every one of them having characteristics to properly fit every application’s needs. As a result, models need to be robust so that successful simulations and predictions can be achieved. The current literature is focused on analyzing and evaluating a proposed battery model, that would enhance the simulation capabilities of a battery cell. To do so, multiple experiments were conducted to evaluate this model in every aspect; from extracting the model’s parameters from experimental data, to evaluating the final results of the proposed model. Along with this, new techniques are proposed to provide proof-of-concept techniques on batteries’ simulations. These techniques focus on predicting the external voltage of the cell under an unknown load but also to predict the parameters of battery models. To achieve this, Neural Networks were used and trained with experimental data of one cell. As a result of these trained Neural Networks, multiple systems were constructed which were able to predict external voltages of cells and the parameters of the models, providing a proof of concept of systems that in the future may minimize the necessary resources to achieve a successful simulation of a battery.
Description
Keywords
Batteries modeling, Lithium-ion cells, Neural networks