Browsing by Author "Demirtas, M"

Now showing 1 - 3 of 3
  • No Thumbnail Available
    Item
    Optimal tuning of PI speed controller coefficients for electric drives using neural network and genetic algorithms
    Ustun, SV; Demirtas, M
    This paper presents a method of tuning Proportional Integral (PI) controller coefficients in the off-line control of a nonlinear system. In this method, the first step is the identification of the system via Artificial Neural Networks (ANNs), using maximum overshoot and settling time obtained from the application circuit for different K-p-K-i pairs. With this in mind, multi-layer ANN, which uses back-propagation of the error algorithm, was used as the learning algorithm. In the second step, the purpose is to find the optimum controller coefficients using the ANN model as the objective function via Genetic Algorithms (GAs). A Digital Signal Processor (DSP-TMS320C50) was used to carry out control applications. The C++ language was used for ANN and GA, and and the Assembly language was used for the DSP. It is determined that maximum overshoot and settling time are very small if the system is controlled by control parameters obtained from the optimization process that uses GA.
  • No Thumbnail Available
    Item
    HARMONIC ESTIMATION BASED SUPPORT VECTOR MACHINE FOR TYPICAL POWER SYSTEMS
    Özdemir, S; Demirtas, M; Aydin, S
    The power quality in electrical energy systems is very important and harmonic is the vital criterion. Traditionally Fast Fourier Transform (FFT) and Discrete Fourier Transform (DFT) have been used for the harmonic distortion analysis and in the literature harmonic estimations have been made using different methods. As an alternative method, this paper suggested using Support Vector Machine (SVM) for harmonic estimation. The real power energy distribution system has been examined and the estimation results have been compared with measured real data. The proposed solution approach was comparatively evaluated with the ANN and LR estimation methods. Comparison results show that THD estimation values that were obtained by the SVM method are close to the THD estimation values obtained from ANN (Artificial Neural Network) and LR (Linear regression) methods. The numerical results clearly showed that the SVM method is valid for THD estimation in the power system.
  • No Thumbnail Available
    Item
    Treatment of mild-to-moderate hypertension with calcium channel blockers: a multicentre comparison of once-daily nifedipine GITS with once-daily amlodipine
    Kes, S; Caglar, N; Canberk, A; Deger, N; Demirtas, M; Dortlemez, H; Kiliccioglu, B; Kozan, O; Ovunc, K; Turkoglu, C
    Background: Hypertension is one of the most important causes of cardiovascular disease, and treatment of hypertension leads to a significant reduction in cardiovascular mortality and morbidity. Although calcium channel blockers are regarded as an important part of the therapeutic armamentarium against cardiovascular diseases, and are among the most frequently prescribed anti hypertensive medications, concern has been aroused about these drugs, particularly the short-acting dihydropyridine derivatives. However, the value of nifedipine GITS (Adalat-Crono*), the long-acting dihydropyridine, is in need of being re-established. Objective: To compare the effectiveness, safety and tolerability of once-daily nifedipine and amlodipine treatment in patients with mild-to-moderate essential hypertension. Design: Randomised multicentre trial with an open comparison of treatments for 12 weeks, with a preceding placebo run-in period of 2 weeks (patients on beta-blockers at the time of enrolment entered a mandatory 2-week wash-out period before being allowed in the placebo run-in period; this wash-out period was one week for patients using any antihypertensive medication other than beta-blockers).