Yazar "Lokman, Gurcan" seçeneğine göre listele

Listeleniyor 1 - 7 / 7
  • [ X ]
    Öğe
    A new method for Anomaly Detection and Target Recognition
    (Ieee, 2014) Lokman, Gurcan; Yilmaz, Guray
    Use of unmanned Aerial Vehicles (UAVs) has gained significant importance in the recent years because they are capable of to be used in in civilian and military purposes for reconnaissance, surveillance, disaster relief, among other tasks. In this paper we present new automated anomaly detection and target recognition methodology that can be used on such a UAV. The standard paradigm for anomaly detection and target recognition in hyperspectral imagery (HSI) is to run a detection or recognition algorithm, typically statistical in nature, and visually inspect each high-scoring pixel to decide whether it is an anomaly or background data. A new method of anomaly detection and target recognition in HSI was studied based on a Neural Network (NN). Two multi-layered neural networks are used for anomaly detection and target recognition. The first phase of the model is used to detect anomalies in HSI. The second phase of the model is to use determine whether the anomaly is a predefined target or not. Both networks are trained in accordance with its intended purpose, so increase in performance is provided. This method can be a suitable solution for applications where the unmanned aerial vehicles used.
  • [ X ]
    Öğe
    Anomaly Detection and Target Recognition With Hyperspectral Images
    (Ieee, 2014) Lokman, Gurcan; Yilmaz, Guray
    In this study, a new method that can perform anomaly detection and target recognition gradually with hyperspectral images (HSI) is introduced. This study constitutes the lower step a system that can process the HSI obtained by unmanned aerial vehicle (UAV) quickly and can enable the anomaly detection and target recognition process done in the mission time of UAV. In the proposed model, firstly the unusual pixels in the HSI data are detected, and then it is investigated whether these pixels are belonging to a target defined. These two phases in the model are carried out using artificial neural network that is a technique of the artificial intelligence. In the model, Parallel programming methods that run with CUDA platform on the GPU are proposed for the processing power requirement resulting from size of HSI.
  • [ X ]
    Öğe
    Designing a fuzzy logic controller with particle swarm optimisation algorithm
    (Inst Engineering Tech-Iet, 2018) Lokman, Gurcan; Topuz, Vedat; Baba, Ahmet Fevzi
    In this study, we designed a trajectory-tracking fuzzy logic controller (TTFLC) for the TRIGA Mark-II Training and Research Reactor, which is located at the Istanbul Technical University. The designed fuzzy logic controller (FLC) is based on the zero-order Sugeno method. The parameters of the FLC membership functions and the action weights of the 15 rules in the rule base are optimised by using the particle swarm optimisation (PSO) algorithm. The objective of this study is to control the TRIGA Mark-II reactor using the designed PSO-tuning TTFLC in a simulator. We used a simulation code from the literature called 'YAVCAN' for studying the non-linear behaviour of the core of the TRIGA Mark-II reactor. To select the best parameters of the PSO algorithm for this system, we conducted some experiments. After selecting the best PSO parameters, the algorithm was started a number of times to determine the optimal parameters of the designed FLC and the optimal parameters of the controller. After determining these parameters, the performance of the designed controller was tested for various initial and desired power levels and under conditions of disturbance. The simulation results showed that the proposed controller could control the reactor power successfully, and it could ensure that the reactor tracks the desired trajectory power within the acceptable error tolerance. Therefore, the PSO algorithm is suitable for finding the optimal parameters of the FLC.
  • [ X ]
    Öğe
    Hyperspectral Image Classification Based on Multilayer Perceptron Trained with Eigenvalue Decay
    (Taylor & Francis Inc, 2020) Lokman, Gurcan; Celik, Hasan Huseyin; Topuz, Vedat
    Hyperspectral Images (HSI) require sufficient labeled samples and a complex classifier to identify an area. Support Vector Machine (SVM) is one of the most competent algorithms in this field. Neural Networks (NN) is another approach used for classification problems, and both have been widely proposed in the literature. The Convolutional Neural Network (CNN) method has also received significant attention in the deep learning field recently. Nevertheless, during NN training, the overfitting problem may cause continuous dragging of the algorithm toward larger error. In this case, a regularization technique is needed to constitute the most useful decision boundary. The Eigenvalue Decay method is one of the regularization techniques that may be applied for HSI. This study investigates the performance of Multilayer Perceptron trained with an Eigenvalue Decay (MLP-ED) algorithm for HSI classification. The SVM, CNN with Pixel-Pair and CNN-Ensemble methods are used as comparison algorithms for MLP-ED performance assessment. All methods were tested with 3 different high-resolution HSI datasets. While SVM is one of the classic classifiers, and the 2 new CNN algorithms show high performance, the proposed MLP-ED method has more computational efficiency and achieves higher success than the others do.
  • [ X ]
    Öğe
    Hyperspectral Image Classification Using Support Vector Neural Network Algorithm
    (Ieee, 2015) Lokman, Gurcan; Yilmaz, Guray
    With the developing technology, Hyperspectral images can be obtained with the satellites, aircraft and even unmanned aerial vehicles. Therefore, the classification applications made on the HSI are becoming increasingly important. In particular, fast and reliable classification algorithms are needed. The basic principle in classification algorithms is using characteristics of the data to find classification function that separate the data from each other. Neural Networks are among the non-linear classification method that can perform with high success. But, syntactic classifier has some problems that occur during training. One of this problems is called over-fitting. In many cases, especially in hyperspectral images, regularization is required for preventing the learning algorithm from over fitting the training data. In this study, a regularization scheme that named eigenvalue decay is used to make to this regularization in the training phase of networks. A training method that uses such a regularization scheme provides a margin maximization as in SVM for NNs. The two well-known data sets that are AVIRIS image of the Salinas Valley in California and image of Okavango Delta in Botswana acquired by The Hyperion sensor on NASA EO-1 satellite are used to test this classifier. The effectiveness of this algorithm on the HSI is evaluated using a series of experiments.
  • [ X ]
    Öğe
    Power loss and voltage stability optimization with meta-heuristic algorithms in power system
    (Pamukkale Univ, 2021) Iscan, Serkan; Kaplan, Orhan; Lokman, Gurcan
    Power flow, which is one of the most prominent problems in the field of power system, is the calculation of the voltage amplitudes and phase angles of each bus and the power losses by using the bus data with known steady state voltage amplitudes and power values. Increasing demand and the connection of decentralized energy sources to the power system at various points make more complicated power flow problem. The power flow problem is of great importance for both electricity generation and transmission. Planning new loads that can be connected to the system in the future and using the existing transmission lines at full capacity are based on the solution of the power flow problem. Power flow, which is a nonlinear problem, has traditionally been solved using numerical methods such as Newton-Raphson and Gauss Seidel. However, the success of classical solution algorithms decreases depending on the conditions of the power system. Meta-heuristic optimization techniques and search algorithms developed in recent years show that better results can be obtained in solving the power flow problem. In this study, Artificial Bee Colony (ABC), Gray Wolf (GWO), Particle Swarm Optimization (PSO) and Newton Raphson algorithms have been applied to optimize the power flow problem in the IEEE-14 bus test power system created using Matlab software. The performance of the algorithms has been compared by considering the voltage amplitudes, voltage deviation, phase angles, power losses and calculation times obtained from the model power system.
  • [ X ]
    Öğe
    Target Detection in Hyperspectral Images Using Support Vector Neural Networks Algorithm
    (Ieee, 2015) Lokman, Gurcan; Yilmaz, Guray
    In this study, the use of Support Vector Neural Network (SVNN) algorithm is offered for target detection process in HSI. The basic principle in classification algorithms is using characteristics of the data to find classification function that separate the data from each other. Neural Networks are among the non-linear classification method that can perform with high success. The classification success depends on the training data and the training algorithm that are used. SVNN Algorithm is one of the methods used to increase the classification margin of the NNs. In this algorithm is provided a training method that used eigenvalue decay that provides a margin maximization as in SVM for NNs. In this context, a minimization problem that provide margin maximization for target detection in Hyperspectral images is defined and this problem is solved by Genetic Algorithms. In this way an algorithm that has high classification performance arises.