Browsing by Author "Ucuncu, Murat"

Filter results by typing the first few letters
Now showing 1 - 2 of 2
  • No Thumbnail Available
    Item
    RadGT: Graph and Transformer-Based Automotive Radar Point Cloud Segmentation
    (2023) Sevimli, Rasim A.; Ucuncu, Murat; Koc, Aykut; 0000-0002-2113-1398; KDO-6837-2024
    The need for visual perception systems providing situational awareness to autonomous vehicles has grown significantly. While traditional deep neural networks are effective for solving 2-D Euclidean problems, point cloud analysis, particularly for radar data, contains unique challenges because of the irregular geometry of point clouds. This letter proposes a novel transformer-based architecture for radar point clouds adapted to the graph signal processing (GSP) framework, designed to handle non-Euclidean and irregular signal structures. We provide experimental results by using well-established benchmarks on the nuScenes and RadarScenes datasets to validate our proposed method.
  • No Thumbnail Available
    Item
    State of Charge (SOC) Estimation for Lithium-Ion Battery Cell Using Extended Kalman Filter
    (2019) Ucuncu, Murat; Altindag, Arda; GYI-8414-2022
    The State of Charge (SOC) estimation is an essential part of a Battery Management System. Nevertheless, because the discharge and charging of battery cells requires complicated chemical operations, therefore, it is hard to determine the state of charge of the battery cell. In this paper, a lithium iron phosphate battery cell with 8 Ah capacity and 3.2 Volt rated voltage was studied. The battery is modelled to reflect the dynamic of the battery encompassing mainly four elements; an Open Circuit Voltage (OCV) source, two RC network, and one resistor. The model parameters are identified by using Forgetting Factor Recursive Least Mean Squares and time domain extraction method. Parameters are converged to their real values and these values are used to estimate the state of charge of the cell using Extended Kalman Filter algorithm based on battery model dynamic. The results in our study show that SoC which is estimated by the implemented Extended Kalman Filter converges to battery's real SoC value. It is also shown that the Extended Kalman Filter update and prediction stages iterations move forward to minimize the error between real SoC and estimated SoC.