Electrical Method for Battery Chemical Composition Determination

Creative Commons License

Dikmen İ. C., Karadağ T.

IEEE Access, no.10, pp.6496-6504, 2022 (SCI-Expanded)

  • Publication Type: Article / Article
  • Publication Date: 2022
  • Doi Number: 10.1109/access.2022.3143040
  • Journal Name: IEEE Access
  • Journal Indexes: Science Citation Index Expanded (SCI-EXPANDED), Scopus, Compendex, INSPEC, Directory of Open Access Journals
  • Page Numbers: pp.6496-6504
  • Inonu University Affiliated: Yes

Abstract

Storage of electrical energy is one of the most important technical problems in terms of today’s technology. The increasing number of high-capacity high-power applications, especially electric vehicles and grid scale energy storage, points to the fact that we will be faced with a large number of batteries that will need to be recycled and separated in the near future. Additionally multi-chemistry battery management systems that enables the collective use of superior features of different batteries with different chemical composition. Here, battery chemical composition determination emerges as a technical problem. In this study, an alternative method to the currently used methods for categorizing batteries according to their chemistry is discussed. As the foundation, batteries with four different chemical composition including Lithium Nickel Cobalt Aluminium Oxide, Lithium Iron Phosphate, Nickel Metal Hydride, and Lithium Titanate Oxide aged with a battery testing hardware. Fifth, is Lithium Sulphur battery which is simulated. Brand new and aged batteries are used in experimental setup that is consist of a programmable electronic DC load and a software developed to run the algorithm on it. According to the algorithm, batteries are connected to two different loads one by one and voltage-current data are stored. Collected data are pre-processed by framing them and framed data are processed with a separation function. Eventually, the determination problem is converted to a classification problem. In order to solve this, artificial neural network and classification tree algorithms are applied. Because the artificial neural network algorithm is applied in previous studies and the high computational cost of it is presented; classification tree algorithm is concluded to be more applicable especially on low-power microcontroller applications. Consequently, 100% accuracy for battery chemical composition determination is achieved and results are presented comparatively.