Yazar "Tutsoy, Onder" seçeneğine göre listele

Listeleniyor 1 - 3 / 3
  • [ X ]
    Öğe
    A comprehensive benchmark of machine learning-based algorithms for medium-term electric vehicle charging demand prediction
    (Springer, 2025) Tolun, Omer Can; Zor, Kasim; Tutsoy, Onder
    The current difficulties faced by evolutionary smart grids, as well as the widespread electric vehicles (EVs) into the modernised electric power system, highlight the crucial balance between electricity generation and consumption. Focusing on renewable energy sources instead of fossil fuels can provide an enduring environment for future generations by mitigating the impacts of global warming. At this time, the popularity of EVs has been ascending day by day due to the fact that they have several advantages such as being environmentally friendly and having better mileage performance in city driving over conventional vehicles. Despite the merits of the EVs, there are also a few disadvantages consisting of the integration of the EVs into the existing infrastructure and their expensiveness by means of initial investment cost. In addition to those, machine learning (ML)-based techniques are usually employed in the EVs for battery management systems, drive performance, and passenger safety. This paper aims to implement an EV monthly charging demand prediction by using a novel technique based on an ensemble of Pearson correlation (PC) and analysis of variance (ANOVA) along with statistical and ML-based algorithms including seasonal auto-regressive integrated moving average with exogenous variables (SARIMAX), convolutional neural networks (CNNs), extreme gradient boosting (XGBoost) decision trees, gated recurrent unit (GRU) networks, long short-term memory (LSTM) networks, bidirectional LSTM (Bi-LSTM) and GRU (Bi-GRU) networks for the Eastern Mediterranean Region of T & uuml;rkiye. The performance and error metrics, including determination coefficient (R 2 \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$<^>2$$\end{document} ), mean absolute percentage error (MAPE), mean absolute error (MAE), and mean absolute scaled error (MASE), are evaluated in a benchmarking manner. According to the obtained results, in Scenario 1, a hybrid of PC and XGBoost decision trees model achieved an R 2 \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$<^>2$$\end{document} of 96.21%, MAPE of 5.52%, MAE of 6.5, and MASE of 0.195 with a training time of 2.08 s and a testing time of 0.016 s. In Scenario 2, a combination of ANOVA and XGBoost decision trees model demonstrated an R 2 \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$<^>2$$\end{document} of 96.83%, a MAPE of 5.29%, a MAE of 6.0, and a MASE of 0.180 with a training time of 1.62 s and a testing time of 0.012 s. These findings highlight the superior accuracy and computational efficiency of the XGBoost models for both scenarios compared to others and reveal XGBoost's suitability for EV charging demand prediction.
  • [ X ]
    Öğe
    An experimental vision-based integrated guidance and control strategy for autonomous landing of a faulty UAV
    (SAGE Publications Ltd, 2025) Khaneghaei, Mohammad; Asadi, Davood; Zahmatkesh, Mohsen; Tutsoy, Onder
    Uncrewed Aerial Vehicles (UAVs) have emerged as a transformative asset in surveillance, mapping, and delivery tasks since they have sophisticated autonomous capabilities. This paper develops a practical vision-based optimal flight planning strategy for autonomous safe landing and control of a multirotor UAV using a low-cost monocular camera in the presence of a motor fault. An optimal integrated guidance and control strategy is developed by utilizing an innovative discrete system model and a state observer from the triggering point to the identified landing position. Additionally, compatible image processing techniques and UAV kinematics are integrated to detect the suitable landing site and translate its location into desired attitude inputs to the controller. This approach empowers the UAV to autonomously land in no-GPS environments, relying solely on camera data. Initially, vision-based sensors, image processing techniques, and the developed guidance and control algorithms undergo initial evaluation in Software in the Loop (SIL) simulations using the Robot Operating System (ROS) and Gazebo simulation environments. The efficacy of the proposed framework is then assessed through experimental flight tests across various landing scenarios, accounting for local wind conditions and motor faults.
  • [ X ]
    Öğe
    Big data dimensionality reduction-based supervised machine learning algorithms for NASH diagnosis
    (BMC, 2025) Tutsoy, Onder; Ozturk, Huseyin Ali; Sumbul, Hilmi Erdem
    BackgroundIdentifying the Non-Alcoholic Steatohepatitis (NASH) that can cause liver failure-based morbidity remains a challenging research problem since there is no confirmed and effective approach for its early and accurate diagnosis yet. A large amount of medical data is collected to diagnose the NASH where the majority of them are redundant.MethodsThis paper initially focuses on selecting the most informative blood test data among the collected big data with the Pearson correlation statistical approach and modified Particle Swarm Optimization with Artificial Neural Networks (PSO-ANN) machine learning algorithm. Then, a gradient based Batch Least Squares (BLS) and a search-based Artificial Bee Colony (ABC) machine learning algorithms are implemented to optimize the NASH prediction models. Confirmed operational NASH diagnosis supervise the statistical and machine learning algorithms to develop accurate prediction models.ResultsTwo machine learning algorithms were trained and also validated with the varying number of selected input features. The results yielded that the trained BLS machine learning model is able to diagnose benign and malignant cases with 100% and 98% accuracies, respectively. The trained ABC machine learning algorithm diagnoses the benign and malignant cases with 90.5% and 94.3% accuracies, respectively.