State of charge estimation of lithium titanate oxide batteries with artificial neural networks

Lithium-based batteries, which have different types according to the purpose of use and need, are gaining more and more importance day by day according to the needs of developing technology with their high energy density and low weight. Lithium-based batteries are used to store and benefit from energy in devices we use in many areas of our lives, such as electronic devices, medical applications, electric vehicles, energy storage, and the aviation industry. Due to the non-linear characteristics of batteries, which are one of today's important materials, many studies are carried out on battery life, charge status and health status monitoring systems and many models are developed using different methods. In this study, definitions and explanations of battery terminology that should be known in the correct selection of lithium-based batteries are made. The development processes of rechargeable lithium batteries, the definitions of existing and still under development lithium batteries, their characteristics, advantages and weaknesses arising from their different chemistries are explain and summarize. The types of estimation methods developed for the state of charge values required for efficient, healthy, and reliable operation of lithium-based batteries, optimizing their life and performance, and providing charge/discharge control are examined and summarized. Lithium titanate oxide battery is discharged with different discharge currents at room temperature. Through the experiments, the discharge capacity, current, voltage and temperature values of the lithium titanate oxide battery are recorded. Using the artificial neural network method, the aim is to determine battery charge values, which will give an idea about the charging/discharging strategies of lithium titanate oxide batteries and accurately determine the usable energy, with the least error. Keywords: Lithium titanate oxide battery, battery terminology, state of charge estimation methods, artificial neural networks, lithium-based batteries.