Mauridhi Hery Purnomo

Temporary short circuit detection in induction motor winding using combination of wavelet transform and neural network

Monitoring system for induction motor is widely developed to detect the incipient fault. Such system is desirable to detect the fault at the running condition to avoid the motor stop running suddenly. In this paper, a new method for detection system is proposed that emphasizes the fault occurrences as temporary short circuit in induction motor winding. The investigation of fault detection is focused on the transient phenomena during starting and ending points of temporary short circuit. The proposed system utilizes the wavelet transform for processing the motor current signal. Energy level of high frequency signal from wavelet transform is used as the input variable of neural network which works as detection system. Three types of neural networks are developed and evaluated including feed forward neural network (FFNN), Elman neural network (ELMNN) and radial basis functions neural network (RBFNN). The results show that ELMNN is the most simply and accurate system that can recognize all of unseen data test. Laboratory based experimental setup is performed to provide real-time measurement data for this research.

Early Detection on the Condition of Pancreas Organ as the Cause of Diabetes Mellitus by Real Time Iris Image Processing

Iridology is an alternative method in evaluating the condition of our internal organ by looking at the image of iris. Evaluating the iris is done by detecting the presence of some broken tissues in iris. In this paper, input image of the iris is taken by using a video camera in real time. The presence of broken tissues in iris in a certain area represents the condition of certain organ according to the chart of iris. The organ that is observed, in this research, is pancreas. Pancreass position in iris is at 07.15 - 07.45 when a circle of iris is divided by 120 points. Several image processing methods are used to enhance the quality of image of iris so that the broken tissues in area of pancreas can be detected clearly. Finally, the result of this detecting method is compared with the insulin normality test

Facial emotional expressions recognition based on Active Shape Model and Radial Basis Function Network

Facial emotional expressions recognition (FEER) is important research fields to study how human beings reflect to environments in affective computing. With the rapid development of multimedia technology especially image processing, facial emotional expressions recognition researchers have achieved many useful result. If we want to recognize the humans emotion via the facial image, we need to extract features of the facial image. Active Shape Model (ASM) is one of the most popular methods for facial feature extraction. The accuracy of ASM depends on several factors, such as brightness, image sharpness, and noise. To get better result, the ASM is combined with Gaussian Pyramid. In this paper we propose a facial emotion expressions recognizing method based on ASM and Radial Basis Function Network (RBFN). Firstly, facial feature should be extracted to get emotional information from the region, but this paper use ASM method by the reconstructed facial shape. Second stage is to classify the facial emotion expressions from the emotional information. Finally get the model which is matched with the facial feature outline after several iterations and use them to recognize the facial emotional expressions by using RBFN. The experimental result from RBFN classifiers show a recognition accuracy of 90.73% for facial emotional expressions using the proposed method.

Malaria parasite identification on thick blood film using genetic programming

Thin blood film is used to know type and phase of the malaria parasite, but which is widely used in Indonesia is the thick blood film. Therefore we need a method that can identify parasites in thick blood film image with a high percentage of accuracy. This research aims to establish a more objective classification system and reduce the subjective factors of medical personnel in diagnosing the type of malaria parasite include its phase. It has three main stages, there are preprocessing, feature extraction, and classification. Preprocessing aims to eliminate the noise, feature extraction using red-green-blue channel color histogram, hue channel HSV histogram, and hue channel HSI histogram, classification using Genetic Programming to identify parasites and also to detect type and phase of the parasite. Experiment was conducted on 180 thick blood film images that classiffied into two classes. The classification has an average accuracy of 95.49% for non-parasites and 95.58% for parasites. Meanwhile when system is used to classified into six classes, testing result have an average accuracy of 90.25% not parasites, 82.25% vivax thropozoit, 75.83% vivax schizont, 81.75% vivax gametocytes, 90.75% falciparum thropozoit, 86.75% falciparum gametocytes. This research confirm that identifying malaria parasite in thick blood film is possible.

Design and Development of Small Electric Vehicle using MATLAB/Simulink

The issue of the depletion of oil reserves in the world, and the problem of air pollution produced by motor vehicles, motivate many researchers to seek alternative energy sources to propel the vehicle. One promising way is to replace combustion motor with an electric motor, which is known as an electric vehicle. First stages of this research is to model the flow of power in the electric vehicle energy system to obtain its characteristics. Power flow efficiency in electric vehicle is very important because this type of vehicle is highly dependent on the limited electrical energy supplied by a battery. Therefore it should be managed properly. This study is to look into the power flow calculation so that the amount of electrical energy is in accordance with the needs of electric vehicle. The design of small electric vehicle model using MATLAB/Simulink software is to get the best power flow response to the electric vehicle energy system.

FFT-based features selection for Javanese music note and instrument identification using support vector machines

Most automatic music transcription research is related with Western music, and still less for the Javanese gamelan music. In this paper, we proposed a method for the features extraction, selection, and identification of gamelan note and the proper instrument. It was an approach based on Fast Fourier Transform (FFT), and support vector machines (SVMs) for note and instrument identification. We selected four spectral features (spectral centroid, two spectral rolloff, and fundamental frequency) as input for SVM. Experimental results show that fundamental frequency, spectral centroid, and spectral rolloff can be used to distinguish gamelan instrument with accuracy or recognition rate more than 95%.

The extraction of acoustic features of infant cry for emotion detection based on pitch and formants

In this paper, we present the development of a system for translating the normal infant cries, which come from pain, sadness, hunger, fear and anger cry sounds, of ages from one day up to nine months old. The aim of this research is to analyse the sound of the crying infant, and to derive the reason why the infant is crying. In this experiment we used acoustic features characteristic determined by pitch and formants. The acoustic feature vectors are then clustered using K-means algorithm to determine the class or the reason of the cry. The proposed system perform well with the maximum accuracy of 90%.

Digital overcurrent relay with conventional curve modeling using Levenberg-Marquardt backpropagation

Overcurrent relays (OCRs) play an important role in the protection component that requires high reliability to maintain high security for power systems. Modeling of the OCRcurve using methods like the direct data storage and curve fitting gave only approximate models. Therefore, in this paper proposes modeling of OCRs using Levenberg-Marquardt backpropagation (LMBP). An implementation of OCR in the digital OCR used ARM microcontroller STM32F407VGT6 is to improve performance of the relay significantly. LMBP is developed using different numbers of neurons. The current and opening time of the circuit breaker are used as input and output in the LMBP training. LMBP developed in the OCR curve model using sample data from protection coordination is implemented as real time in Hess Indonesia Corporation. The weights obtained by the LMBP are used to run the LMBP program in the digital OCR. The well known digital OCR product is used for comparison. The results show that this proposed method is accurate and encouraging with percentage error is 0.24% and very promising to be applied in the digital OCR.

Overcurrent relay curve modeling and its application in the real industrial power systems using adaptive neuro fuzzy inference system

Create an accurate model with over-current relays (OCRs) play an important role in the coordination of power system protection. Modeling of the OCR using methods like the direct data storage and software models gave only approximate models. Moreover, modeling based on mathematical models is not appropriate to deal with ill-defined and uncertain systems. Therefore, in this paper proposes modeling of OCRs using adaptive neuro fuzzy inference system (ANFIS). ANFIS is developed using different numbers and types of membership functions (MFs). Each MF is implemented using training and checking data. The load current and time of opening of the circuit breaker are used as input and output in the ANFIS training. ANFIS, which is developed in the OCR curve model using sample data from protection coordination, is implemented in Hess Indonesia Corporation. Different types of MFs are to obtain the optimal design of OCR curves. The result of ANFIS in the OCR curve modeling is accurate and encouraging; thus, the ANFIS model can be used in digital relays and applied successfully in the real systems. In all cases, ANFIS models using 30 Gbell-type MFs yields a very minimum average percentage error of 0.028419 %.

Photovoltaic module and maximum power point tracking modelling using Adaptive Neuro-Fuzzy Inference System

This paper proposes an intelligent control method using Adaptive Neuro-Fuzzy Inference System (ANFIS) for maximum power point tracking (MPPT) of PV module. The method is verified under several irradiance and temperature conditions. DC - DC boost converter is connected between the PV module and the load. Duty cycle of DC - DC boost converter is controlled by ANFIS in order to obtain the MPPT. The ANFIS directly takes operating power and voltage level as input. The proposed system is developed under Simulink-Matlab and the system of PV is simulated in PSIM to verify the effectiveness of method. The results show the proposed method can obtain the highest output power than Fuzzy Logic (FL) and Perturbation and Observation (P&O) method i.e., 30.893 and 42.973 for irradiance is 750W/m2 and 1000W/m2, respectively.