Yazar "Kesilmis, Zehan" seçeneğine göre listele

Listeleniyor 1 - 6 / 6
  • [ X ]
    Öğe
    A manhattan metric based perturb and observe maximum power point tracking algorithm for photovoltaic systems
    (Taylor & Francis Inc, 2022) Kesilmis, Zehan
    Maximum power point tracking (MPPT) requires the use of fast and efficient algorithms that can efficiently discover maximum power point (MPP) even under rapidly changing atmospheric conditions. The conventional perturb and observe (P&O) is a commonly utilized MPPT algorithm due to its parameter-independent and straightforward nature. Despite these merits, P&O suffers from low tracking efficiency, power fluctuations around the MPP, and drift. Various adaptive P&O algorithms have been proposed to reduce these drawbacks. In this paper, a novel Manhattan distance-metric-based adaptive P&O (MPO) algorithm is proposed to deal with problems of the P&O algorithm. The MPO algorithm excels with increasing the MPP convergence rate, reducing the convergence time, and decreasing the power fluctuations. The proposed method has been validated using PSIM simulations and experimental studies under constant and rapidly changing irradiance conditions. Experimental verifications were carried out using the experimental setup containing a step-down converter and ATmega328p microcontroller. In these experiments, the success of the MPO algorithm is compared with P&O and adaptive P&O algorithms. The results show that the MPO algorithm successfully tracks the MPP in steady and rapidly changing irradiance conditions when others fail occasionally. On top of that, over 99% MPPT efficiency is achieved, and the convergence time is also improved.
  • [ X ]
    Öğe
    A Novel MPPT Method Based on Inflection Voltages
    (Elsevier Sci Ltd, 2020) Kesilmis, Zehan; Karabacak, Mustafa Alpaslan; Aksoy, Murat
    This study presents a novel maximum power point tracking method named inflection voltage method for photovoltaic systems. Inflection voltage method aims to find the global maximum power point using inflection voltages. It is known that multiple peaks occur in the power/voltage curves of photovoltaic arrays under partial shading conditions. Traditional methods such as perturb and observe and fractional open circuit voltage methods cannot perform MPPT with high performance in case of partial shading conditions. The inflection voltage method provides solutions to the problems of algorithms like fractional open circuit voltage and perturb and observe, such as failure to work under partial shading conditions and oscillation of output power. The success of inflection voltage method was tested first in the PSIM environment and then in the experimental setup developed. Computer simulations and experimental results showed that the proposed algorithm outperforms both fractional open circuit voltage, perturb and observe, particle swarm optimization under different partial shading conditions. (C) 2020 Elsevier Ltd. All rights reserved.
  • [ X ]
    Öğe
    A novel two-stage MPPT method based on inflection voltages
    (Taylor & Francis Inc, 2022) Karabacak, Mustafa Alpaslan; Kesilmis, Zehan; Aksoy, Murat
    The electrical characteristic of a photovoltaic system is directly dependent on atmospheric conditions such as temperature, solar insolation, and even shading conditions. Hence, utilizing a maximum power point tracking hardware with a fast and accurate algorithm is crucial to obtaining the highest possible photovoltaic power. Several algorithms have been proposed in the literature to obtain maximum photovoltaic power, such as particle swarm optimization and perturb & observe variants. However, these algorithms provide less performance than optimal under partial shading. This study proposes a novel two-stage maximum power point tracking algorithm that equipped inflection voltages method and conventional perturb & observe algorithms. The proposed algorithm displays superior properties of both the inflection voltages method and the perturb & observe algorithms, such as predicting the location of maximum power points and high-speed climbing to these power points. A set of computer simulations are conducted in the PSIM environment to compare the performance of the proposed algorithm with particle swarm optimization, perturb & observe, and inflection voltages algorithms. Furthermore, a DC/DC converter-based experimental setup was produced and connected to a BK precision solar array simulator for experimental verification. The analysis results have shown that the proposed algorithm outperforms perturb & observe, inflection voltages, and particle swarm optimization methods in convergence speed and rate. Experimental and simulation results also show that the proposed algorithm successfully tracks the maximum peak in uniform and non-uniform irradiance conditions when others fail occasionally. On top of that, over 99% tracking efficiency is achieved, and the convergence time is also under 1 second.
  • [ X ]
    Öğe
    Investigation of MPPT Algorithms under Partial Shading and Changing Atmospheric Conditions in PSIM Environment
    (Institute of Electrical and Electronics Engineers Inc., 2019) Kesilmis, Zehan; Karabacak, M. Alpaslan; Aksoy, Murat
    In this study, computer simulations were performed in PSIM environment in order to compare the success of Stochastic Beam Search, Perturb Observe, Fractional Open Circuit and Random Perturb Observe algorithms in partial shading and monitoring maximum power point under changing atmospheric conditions. The computer simulation model consists of 4 series connected solar panels, Buck type DC/ DC converter and C block which serves as micro controller. The simulation model provides a realistic and controlled environment for testing algorithms. The results of the study showed that the Stochastic Beam Search algorithm successfully performed the MPPT function under different shading conditions and changing atmospheric conditions. © 2019 IEEE.
  • [ X ]
    Öğe
    Investigation of the Relationship between Vibration Signals Due to Oil Impurity and Cavitation Bubbles in Hydraulic Pumps
    (Mdpi, 2022) Yilmaz, Ozgur; Aksoy, Murat; Kesilmis, Zehan
    Although hydraulic pumps are frequently used in daily life, improper use due to oil analysis or oil contamination is ignored. There is no instantaneous inspection; instead, the oil is changed periodically at certain times, whether it is contaminated or not. Hydraulic systems operate based on Pascal's law, which states that the fluid will distribute the pressure equally to every point in a closed area. The fluid oil taken from an oil reservoir is moved into the pump by engine power. During this movement, as it passes through different pressure areas and different sections, undesirable events such as viscosity change and gas formation occur in the hydraulic oil. These formations collide with the outer walls and cause cavitation with respect to unwanted oil impurities. This cavitation causes unwanted vibration signals to occur in the normal working order of the system. As a result of cavitation, the particles that affect the lubricity and fluidity of the oil in the oil are mixed into the liquid and circulate freely. At the connection points, the blockage caused by the liquid in the pump cylinder block or the valve plate and the collisions of particles is effective. As a result, it creates vibrations of different frequencies. The frequency and amplitudes of these vibrations differ according to the degree of oil contamination. A method has been developed to find the degree of contamination of the oil circulating in the pump by looking at the amplitude and frequency of these vibrations measured from the motor body. There exist standards about the pollution of hydraulic fluid. With these standards, the maximum number of particles allowed for a given pollution level is defined. This topic is discussed in the conclusion to this study. This method has also been proven experimentally. Error and vibration analysis studies on pumps using a different approach are available in the literature. In these studies, pressure variation, total energy transmission, or artificial intelligence models were used to detect anomalies in the pump. In this study, the impurity rate of the oil was set at five different levels and the operating regime of the pump at each level was investigated experimentally. Rayleigh-Plesset and Zwart-Gerber-Belamri models, which are the most common cavitation models, were used to explain the bubble formation in the moving oil and the relationship of these bubbles with vibration. Frequency components were examined by the Discrete Fast Fourier Analysis method, where the operation of the pump was affected by the increase in oil impurity.
  • [ X ]
    Öğe
    LOW COST AND LABORATORY SCALE NIR SPECTROSCOPY FOR QUALITY EVALUATION OF FRUITS AND VEGETABLES
    (Univ Agricultural Sciences & Veterinary Medicine Bucharest, 2018) Kesilmis, Zehan; Vursavus, Kubilay K.
    NIR spectroscopy has proved to be one of the efficient and easy tools to monitor the quality of agricultural products. NIR spectrometers are versatile devices to monitor the ripeness or quality parameters of the fruits. We demonstrate a low-cost spectrometer design that is produced with off the shelf components. In this work, the development, characterization and validation of a prototype is discussed. The proposed device has a dedicated user interface on the PC to plot and analyze spectral data. The performance of the proposed spectrometer is comparable to existing laboratory scale spectrometers in terms of stability and resolution. The spectral resolution and response range of the proposed spectrometer are 20 nm and 640-1050nm, respectively. Proposed device consists of MEMS based Hamamatsu spectrometer sensor (C11708MA), microcontroller (Arduino) and IR light source. Roles of the Arduino are generating essential control signals and sampling output of the C11708MA. These spectral response data have a huge advantage in generating data sets that may be useful in building machine learning based models.