Yazar "Kadem, Onur" seçeneğine göre listele

Listeleniyor 1 - 3 / 3
  • Küçük Resim Yok
    Öğe
    Co-Estimation of State of Health and State of Charge for Lithium-Ion Batteries via the Normalized State of Charge and Open Circuit Voltage Relationship
    (Wiley, 2025) Kadem, Onur
    The relationship between state of charge (SoC) and open circuit voltage (OCV) is fundamental to SoC estimation in equivalent circuit models (ECMs). While its dependency on temperature and aging is recognized, the influence of real-time capacity variations is often underexplored. This study investigates the impact of capacity degradation on the SoC-OCV relationship across different temperatures, aging levels, and OCV testing methods, using the CALCE and NASA battery datasets. Results show that when SoC is normalized by the degraded capacity, the SoC-OCV relationship remains nearly constant for SoC values above 20%. Leveraging this property, we propose a real-time algorithm capable of simultaneously estimating SoC and capacity throughout the battery lifecycle. The algorithm also estimates state of health (SoH) by independently quantifying resistance and capacity related degradation. A first-order ECM with a single resistor-capacitor branch models battery dynamics, while Kalman filtering enables real-time state updates. The method is validated under diverse conditions including partial and full discharges, varying temperatures, dynamic load profiles (e.g., US06, FUDS, BJDST, HPPC), and different aging states. Experimental results demonstrate robust performance, with SoC estimation errors within +/- 0.01 and capacity estimation errors within +/- 0.05 Ah, confirming the algorithm's effectiveness for real-world battery management system applications.
  • Küçük Resim Yok
    Öğe
    Data-Driven Kalman Filter with Maximum Incremental Capacity Measurement for Battery State-of-Health Estimation
    (Institute of Electrical and Electronics Engineers Inc., 2025) Kadem, Onur; Kim, Jongrae
    State of health (SoH) estimation is crucial for the reliable operation of battery management systems. While various SoH estimation approaches have been proposed, the integration of adaptive filtering with practical indirect health measurements remains insufficiently explored. This study introduces an online, data-driven SoH estimation framework that combines Gaussian process regression (GPR) with an extended Kalman filter (EKF). Moreover, the proposed method uses the normalized maximum incremental capacity measurement as a health indicator (HI). Extracting this HI during the constant current phase of the constant current constant voltage charging protocol enables online SoH estimation. The equivalent full cycle count is used for a priori prediction, while the HI is employed for a posteriori updates. Experimental validation of the method is carried out using three publicly available battery datasets, i.e., Dataset 1, Dataset 2, and Dataset 3, through a leave-one-battery-out cross-validation under varying operational conditions. The proposed GPR-EKF outperforms the EKF on Datasets 1-2 and is comparable on Dataset 3, while outperforming the Long Short-Term Memory-EKF across all datasets with average root mean square errors (RMSEs) of 1.88%, 0.45%, and 1.06%, respectively. Furthermore, the method exhibits robust SoH estimation, maintaining an RMSE of 0.91% even when the HI measurement is intermittently available with an 80% probability. These results highlight the potential of the proposed GPR-EKF method for accurate, robust, and online SoH estimation in practice. © 2015 IEEE.
  • Küçük Resim Yok
    Öğe
    Hybrid Deep Neural Network for Electric Vehicle State of Charge Estimation
    (Institute of Electrical and Electronics Engineers Inc., 2024) Kadem, Onur; Candan, Hasibe; Kim, Jongrae H.
    In electric vehicles (EVs), a battery management system plays a critical role in ensuring the reliable and safe operation of batteries. One of its main tasks is to display the battery state of charge (SoC), reflecting the battery's current available charge level. However, the accuracy of SoC estimation is a formidable challenge due to the intricate nature of battery modelling. To overcome this challenge, data-driven methods have recently emerged as the dominant approach for SoC estimation. Considering the SoC estimation problem as a time series problem, we propose a hybrid deep neural network (DNN) that eliminates the need for feature engineering or adaptive filtering. The proposed DNN incorporates a convolutional layer, a long short-term memory layer, and a dense layer. The DNN was trained using data collected from benchmark EV driving cycles (DST, BJDST, and FUDS drive cycles) within a temperature range of 0 °C to 50 °C. The performance evaluation of the trained DNN has been carried out using another standard EV driving cycle (US06 drive cycle) at various operating temperatures. The results demonstrate that the trained DNN effectively captures the dynamic behaviour of the battery under various operational conditions, yielding a maximum percentage SoC estimation error of approximately 3%. Furthermore, the results indicate that the proposed DNN technique is capable of generalising the battery's dynamic response to unseen data. Overall, our findings show that the proposed technique is promising for EV applications in which battery operating conditions exhibit significant variability. © 2024 IEEE.