Lityum-iyon pilin batarya yönetim sistemi için genişletilmiş kalman filtre kullanımı

There are numerous energy storage applications today, but batteries that store chemical and electromechanical storage are most common. Batteries are an essential energy source preferred in various fields where electronics are used and in renewable energy systems industries. Battery technology has an utmost significance in future and present technologies such as energy storage in smart grids, electric and hybrid vehicles, consumer electronics. Lithium-ion batteries are becoming more popular among the different battery technologies with their excellently solid performance characteristics, high energy density, low weight, and high galvanic potential. One or more battery cells require a Battery Management System (BMS) circuit of which voltage, current, and temperature can be controlled safely and reliably. The primary purposes of a BMS are to protect the cells or battery from damage, prolong the battery's life, and keep the battery in a condition that can meet specified operating parameters. Therefore, the BMS may include one or more of these functions. In this thesis, a BMS system is designed to control basic functional parameters such as voltage, current, and temperature during charging and discharging. Four different operating modes have been defined as Charge, Discharge, Normal Operation, and Protection Mode. The user has been allowed to monitor the State of Charge (SoC) value on the screen by taking the OCV values measured according to the open-circuit voltage as a reference. The BMS needs to be able to accurately predict battery states that cannot be directly measured, such as State of Charge (SoC). Within the scope of the study, the Dual Polarization Model was examined, and experimental tests and Extended Kalman Filter defined the parameters of the model. The developed Extended Kalman Filter for SoC estimation compares the measured cell voltage with the recorded current using the Coulomb Counting method with the cell voltage value predicted by the specified model. As a result, Coulomb Counting and Open Circuit Voltage (OCV) methods provide a corrected estimate for the SoC, considering the correlated and designed Extended Kalman filter measurement and model prediction accuracies.