Erkan Deniz

Three-level Cascaded Inverter Based D-STATCOM Using Decoupled Indirect Current Control

Abstract In this paper, experimental setup of three-level cascaded inverter based 380V/±25kVAR D-STATCOM using decoupled indirect current control method (DICC) is realized. AC and DC side of D- STATCOM is modelled in dq-axis on account of D-STATCOM’s controlling. DICC is used for control of D-STATCOM’s dq-axis currents independently. Gate pulses for inverter are generated with multilevel sinusoidal pulse width modulation (SPWM) technique. In this study, controller card used for signal processing is dSPACEs DS1103. The control algorithm is prepared by the help of MATLAB/ Simulink® software. This algorithm is converted to C language by using Matlab/Real-Time Workshop and downloaded to DS1103’s program memory by dSPACE/Real-Time Interface. Implemented experimental setup is tested by changing reference reactive current (iqref) +20A to −20A and obtained results from this test are given.

DeepEMGNet: An Application for Efficient Discrimination of ALS and Normal EMG Signals

This paper proposes a deep learning application for efficient classification of amyotrophic lateral sclerosis (ALS) and normal Electromyogram (EMG) signals. EMG signals are helpful in analyzing of the neuromuscular diseases like ALS. ALS is a well-known brain disease, which progressively degenerates the motor neurons. Most of the previous works about EMG signal classification covers a dozen of basic signal processing methodologies such as statistical signal processing, wavelet analysis, and empirical mode decomposition (EMD). In this work, a different application is implemented which is based on time-frequency (TF) representation of EMG signals and convolutional neural networks (CNN). Short Time Fourier Transform (STFT) is considered for TF representation. Two convolution layers, two pooling layer, a fully connected layer and a lost function layer is considered in CNN architecture. The efficiency of the proposed implementation is tested on publicly available EMG dataset. The dataset contains 89 ALS and 133 normal EMG signals with 24 kHz sampling frequency. Experimental results show 96.69% accuracy. The obtained results are also compared with other methods, which show the superiority of the proposed method.

GA-based optimization and ANN-based SHEPWM generation for two-level inverter

Selective harmonic elimination (SHE) is a well-known PWM technique applied voltage source inverters (VSI) to control fundamental voltage and eliminate chosen harmonics. This technique requires the determination of optimum switching angles by solving the nonlinear equation set and a look-up table stored the switching times in a real-time application. This paper presents a hybrid genetic algorithm (GA) to optimize offline of optimum 11-switching angles for three-phase two-level inverter. In addition, the paper proposes two Artificial Neural Networks (ANN) based solution. The first ANN was trained by the data obtained from GA to calculate the switching angles without using a look-up table. Second ANN was trained by using these switching angles to generate PWM signals. GA and ANN are performed by using MATLAB environment. The ANN-based SHEPWM was designed to obtain inverter output voltage which has a bipolar waveform with quarter-wave symmetry. The waveforms of inverter output voltage and load current are analyzed with FFT for a RL load.

Design of a two-phase PMSM fed by an AC-AC converter

A two-phase permanent magnet synchronous motor is designed for using an AC-AC converter drive system fed from a single-phase source. This motor and drive system are proposed to replace with a traditional single-phase motors fed by electricity grid for improving the efficiency of the system. The proposed AC drive system can be connected directly between AC grid and motor without requiring any storage device such as DC-link large capacitors and rectification. The structure of AC-AC converter and modulation technique is presented with vector controlled two-phase PMSM. The converter output voltage and current waveforms and harmonic contents are analyzed with MATLAB simulation program. Additionally, responses of the motor speed and torque are presented for various conditions.

Transfer learning based histopathologic image classification for breast cancer detection

Breast cancer is one of the leading cancer type among women in worldwide. Many breast cancer patients die every year due to the late diagnosis and treatment. Thus, in recent years, early breast cancer detection systems based on patient’s imagery are in demand. Deep learning attracts many researchers recently and many computer vision applications have come out in various environments. Convolutional neural network (CNN) which is known as deep learning architecture, has achieved impressive results in many applications. CNNs generally suffer from tuning a huge number of parameters which bring a great amount of complexity to the system. In addition, the initialization of the weights of the CNN is another handicap that needs to be handle carefully. In this paper, transfer learning and deep feature extraction methods are used which adapt a pre-trained CNN model to the problem at hand. AlexNet and Vgg16 models are considered in the presented work for feature extraction and AlexNet is used for further fine-tuning. The obtained features are then classified by support vector machines (SVM). Extensive experiments on a publicly available histopathologic breast cancer dataset are carried out and the accuracy scores are calculated for performance evaluation. The evaluation results show that the transfer learning produced better result than deep feature extraction and SVM classification.