Indonesian Journal of Electrical Engineering and Informatics (IJEEI)
Not a member yet
776 research outputs found
Sort by
An Ensemble Classifier Based on Individual Features for Detecting Microaneurysms in Diabetic Retinopathy
Get PDFIndividuals with diabetes are more likely to develop Diabetic Retinopathy (DR), a chronic ailment that can lead to blindness if left undiagnosed. Early-stage Diabetic Retinopathy (DR) is characterized by Microaneurysms (MA), which appear as tiny red lesions on the retina. This paper investigates a unique approach for the automated early identification of microaneurysms in eye images. A unique ensemble classifier technique is suggested in this work. Classifiers like SVM, KNN, Decision Tree, and Naïve Bayes are chosen in this study for building an ensemble model. After preprocessing the image, certain common image characteristics such as shape and intensity features were retrieved from the candidate. The mean absolute difference of each feature is computed. Based on mean ranges that would give improved classification results, an expert classifier is chosen and trained. The outputs of the classifiers are integrated for each of the distinct characteristics, and the number of categories that have been most frequently repeated is utilized to reach a final decision. The process has been comprehensively validated using two available open datasets, like e-ophtha and DIARETDB1. On the e-ophtha and DIARETDB1 datasets, the ensemble model achieved an AUC of 0.928 and 0.873, Sensitivity of 90.7% and 85%, Specificity of 90% and 91% respectively
Sentiment Analysis in Karonese Tweet using Machine Learning
Get PDFRecently, many social media users expressed their conditions, ideas, emotions using local languages on social media, for example via tweets or status. Due to the large number of texts, sentiment analysis is used to identify opinions, ideas, or thoughts from social media. Sentiment analysis research has also been widely applied to local languages. Karonese is one of the largest local languages in North Sumatera, Indonesia. Karo society actively use the language in expression on twitter. This study proposes two things: Karonese tweet dataset for classification and analysis of sentiment on Karonese. Several machine learning algorithms are implemented in this research, that is Logistic regression, Naive bayes, K-nearest neighbor, and Support Vector Machine (SVM). Karonese tweets is obtained from timeline twitter based on several keywords and hashtags. Transcribers from ethnic figures helped annotating the Karo tweets into three classes: positive, negative, and neutral. To get the best model, several scenarios were run based on various compositions of training data and test data. The SVM algorithm has highest accuracy, precision, recall, and F-1 scores than others. As the research is a preliminary research of sentiment analysis on Karonese language, there are many feature works to improvement
Fuzzy-based Nutrient System for Chili Cultivation in Urban Area
Get PDFThe right level of nutrients is crucial for chilli cultivation as the crop requires different nutrient levels at different growth stages. The current fertiliser supply needs many human resources, which is time-consuming. Thus, an automatic nutrient controlling system giving the exact amount of fertiliser based on Fuzzy logic and IoT technology is proposed in this paper. The proposed system uses Hostinger platform to monitor water level, electrical conductivity (EC) and pH values in real-time. Fuzzy membership functions and rules decide the precise amount of nutrients to chilli plants based on the EC value and water level at each growth stage. The Fuzzy membership functions are designed according to the nutrients requirement in each chilli’s phenological stage. The proposed system results are compared with the traditional approach, where fertilisers are supplied manually every week. The experiment results showed that the proposed system could meet the precise and automatic fertiliser addition requirement, eliminate human intervention and ensure the plants grew well.
Application of CaTiO3:Pr3+ Phosphor for Enhancing the Hue Standard of WLEDs with Double-Film Distant Phosphor Structure
Get PDFDue to its great thermal stability, the WLEDs (short for white-light diodes), which are made of PiG (short for phosphor-in-glass), appear to be an optical source most effective at generating potent white illumination. However, the actual applications they offer are limited by their poor color rendering and color uniformity. To improve the color rendering and uniformity, this study suggested utilizing a configuration involving a PiG-RPL (short for PiG integrated with a lens of phosphor in red). The phosphor YAGG (also known as Y3Al3.08Ga1.92O12:Ce3+) and borosilicate glass powders were printed, then sintered to generate the green PiG, and then using an inverted dispensing approach, we applied the CASN phosphor silicone in red color (also known as CaAlSiN3:Eu2+) to the said PiG. As a result, with a current of 350 mA, the WLEDs made of PiG-RPL exhibit highly remarkable chromatic performance for the CRI (also known as color rendering index) values determined as Ra = 95.6 with R9 = 95.2 and fidelity/gamut values determined as Rf = 92 with Rg = 99.2. The color quality of the PiG-RPL based WLEDs is steady over a wide range of currents, from 100 to 1000 mA. Moreover, PiG-RPL based WLEDs have better color uniformity than typical PiG based WLEDs. The PiG-RPL structure was proposed in this paper to improve the WLEDs’ chromatic generation as well as chromatic homogeneity. We made the PiGRPL via placing a lens of CASN phosphor above a PiG in green color made of YAGG phosphor. For WLEDs, the PiG-RPL provides extremely remarkable chromatic generation. Furthermore, the WLEDs that are made of PiG-RPL have better color uniformity than those made of PiG and PiG-RPP. The PiG-RPL color converters are considered to be promising for high-power WLEDs, offering great color rendering and color uniformity
Customer Churn Prediction in Telecommunication Industry Using Classification and Regression Trees and Artificial Neural Network Algorithms
Get PDFCustomer churn is a serious problem, which is a critical issue encountered by large businesses and organizations. Due to the direct impact on the company's revenues, particularly in sectors such as the telecommunications as well as the banking, companies are working to promote ways to identify the churn of prospective consumers. Hence it is vital to investigate issues that influence customer churn to yield appropriate measures to diminish churn. The major objective of this work is to advance a model of churn prediction that helps telecom operatives to envisage clients that are most probable to be subjected to churn. The experimental approach for this study uses the machine learning procedures on the telecom churn dataset, using an improved Relief-F feature selection algorithm to pick related features from the huge dataset. To quantify the model's performance, the result of classification uses CART and ANN, the accuracy shows that ANN has a high predictive capacity of 93.88% compared to the 91.60% CART classifie
An Extended Kalman Filter for Nonsmooth Attitude Control Design of Quadrotors using Quaternion Representation
Get PDFThis paper proposed Extended Kalman Filter specifically designed for nonlinear and nonsmooth control system applied in Autonomous Quadrotor Control such as sliding mode control. Many controllers focused on global stability usually consider exact parameters through measurements. Such assumptions are not always possible due to the unavailability of sensors or unmeasurable state in real-life condition. In this paper, we consider only the angular velocity is possible for measurement, i.e., only gyroscope measurement is available. This condition is known as omega-state-measurement (OSM). Without loss of generality, for theoretical simplification beside gyroscope measurement, we assume the orientation measurement represented in quaternion is also available. Additive random gaussian noise is included to the measurement model to be used in Kalman Filter. Finally the proposed Extended Kalman Filter implemented in a PD Sliding Mode controller is simulated using many scenarios to verify its effectiveness. The Kalman Filter works well in spite of model error and disturbance
Bifacial Vertical Photovoltaic System Design for Farming Irrigation System
Get PDFThe increase in population leads to an upsurge in food demands which also expend the agriculture activities. A wide range of electrically powered machines is essential for the success of modern agriculture setup. Farming required electrical power for numerous activities such as irrigations and electric tractors. A large number of farms are located in remote areas where access to electricity could be costly. Also, farms that are located within the electrical grid suffer from the cost of electricity bills. In line with the United Nations' recommendations to deploy renewable energy sources for electrical power generations, photovoltaic systems are installed for farming activities across countries. A photovoltaic system converts solar radiation into electrical power and with the use of advanced power electronics devices, PV technologies become very attractive to farmers. The work in this paper capture the PV system operation for farming purposes. The contents cover standard PV panels and their current deployment layout for farms. The paper introduces the bifacial panel's concept and its novel layout. Furthermore, the paper proposed a novel installation layout for bifacial panels to support the farm electrical demands. The case study, which is based on three-phase irrigation pumps, explains and verifies the advanced role that the proposed layout of bifacial panels over the standard one. The advantages of the proposed installation layout are also include
Swarm Intelligence Autotune For Differential Drive Wheeled Mobile Robot
Get PDFDifferential Drive Wheeled Mobile Robot (DDWMR) is a nonholonomic robot with constrained movement. Such constraint makes robot position control more difficult. A closed-loop control system such as PID can control robot position. However, DDWMR is a Multiple-Input-Multiple-Output system. There will be many feedback gains to be tuned, and the wrong value will make the system unstable. Therefore this research proposes an offline autotune method to choose optimal feedback gain that minimizes a fitness function. The fitness function uses Integral Absolute Error (IAE) and Integral Time Absolute Error (ITAE). These works propose to autotune feedback gain for DDWMR Jetbot, which implements a PI control system with six feedback gains. The methods used to tune the feedback gain are Particle Swarm Optimization (PSO) and Bird Swarm Algorithm (BSA). There are four different scenarios to do the autotune. The autotune result performance shows that those two methods can find an optimal gain to make the robot follow four different continuous trajectories without much trajectory deformation. PSO and BSA can do an autotune PI gain with six variables to minimize the Integral Absolute Error (IAE) and Integral Time Absolute Error (ITAE)
Multi-objective Predictive Control of 3L-NPC Inverter Fed Sensorless PMSM Drives for Electrical Car Applications
Get PDFThis paper proposes a multi-objective FS-MPC approach based on three-step optimization for a surface-mounted PMSM fed by a 3L-NPC inverter. It helps to significantly reduce torque ripples, current harmonics while controlling the inverter's neutral point voltage. To overcome the drawbacks of using mechanical sensors, a sliding mode observer is used to estimate the machine speed and rotor angular position. Compared to existing works, the proposed control method is implemented using the proportionality between the electromagnetic torque and the current component on the q-axis to eliminate the computational redundancy related to the current and torque control. To further reduce torque ripples and current harmonics, a 3L-NPC inverter is used. Compared to other types of three-level inverters, it uses less power semiconductors and attenuates the problem of voltage fluctuation at the neutral point and current harmonics. Matlab/Simulink simulations of the proposed approach yield a current THD of 1.69 %
Dynamic Security Assessment For Power System Using Attribute Selection Technique
Get PDFThe evaluation of the dynamic security of the electrical power system after the occurrence of disturbances in the network is one of the most important tools that the control center uses to maintain the system in a safe operating mode, as well as prevent cases of system out of control and cases of complete shutdown. With the annual increase in the size of the electrical system and its distribution over a very wide geographical area, this led to a new challenge to assess dynamic security assessment (DSA), which is dealing with a huge and varied amount of data that requires processing in a very short time. To address these challenges, this study presented a new technique of artificial intelligence, which is the attribute selection technique, to reduce the size of this data and thus improve the accuracy and speed of results. This method relied on the combination of decision tree algorithms and a technique (Attribute selection) in the data obtained from the test system (IEEE-30Bus). The results of this method showed a significant reduction in the number of data used, which amounted to (45.55%) of the total data, Which led to an improvement in the classification accuracy, as the classification accuracy reached (97.27%). This reduction is very important when dealing in the online operating environment, as it saves the time necessary to reach the most accurate evaluation decision and thus issue gives a greater opportunity to take the appropriate decision in the event of disturbances and keep the electrical system in a secure state