IAES International Journal of Artificial Intelligence (IJ-AI)
Not a member yet
1769 research outputs found
Sort by
Adaptive Neural Subtractive Clustering Fuzzy Inference System for the Detection of High Impedance Fault on Distribution Power System
Get PDFHigh impedance fault (HIF) is abnormal event on electric power distribution feeder which does not draw enough fault current to be detected by conventional protective devices. The algorithm for HIF detection based on the amplitude ratio of second and odd harmonics to fundamental is presented. This paper proposes an intelligent algorithm using an adaptive neural- Takagi Sugeno-Kang (TSK) fuzzy modeling approach based on subtractive clustering to detect high impedance fault. It is integrating the learning capabilities of neural network to the fuzzy logic system robustness in the sense that fuzzy logic concepts are embedded in the network structure. It also provides a natural framework for combining both numerical information in the form of input/output pairs and linguistic information in the form of IF–THEN rules in a uniform fashion. Fast Fourier Transformation (FFT) is used to extract the features of the fault signal and other power system events. The effect of capacitor banks switching, non-linear load current, no-load line switching and other normal event on distribution feeder harmonics is discussed. HIF and other operation event data were obtained by simulation of a 13.8 kV distribution feeder using PSCAD. The results show that the proposed algorithm can distinguish successfully HIFs from other events in distribution power systemDOI: http://dx.doi.org/10.11591/ij-ai.v1i2.42
Hybrid Genetic Algorithms for Solving Winner Determination Problem in Combinatorial Double Auction in Grid
Get PDFNowadays, since grid has been turned to commercialization, using economic methods such as auction methods are appropriate for resource allocation because of their decentralized nature. Combinatorial double auction has emerged as a major model in the economy and is a good approach for resource allocation in which participants of grid, give their requests once to the combination of resources instead of giving them to different resources multiple times. One problem with the combinatorial double auction is the efficient allocation of resources to derive the maximum benefit. This problem is known as winner determination problem (WDP) and is an NP-hard problem. So far, many methods have been proposed to solve this problem and genetic algorithm is one of the best ones. In this paper, two types of hybrid genetic algorithms were presented to improve the efficiency of genetic algorithm for solving the winner determination problem. The results showed that the proposed algorithms had good efficiency and led to better answers. DOI: http://dx.doi.org/10.11591/ij-ai.v1i2.44
A Semi-Automated Lyrics Generation Tool for Mauritian Sega
Get PDFIn this paper, we give an overview of how Sega lyrics, in Mauritian Creole language, are being written by Mauritian Lyricists and a tool which has been developed to automatically generate Sega lyrics. Research shows that song writing is not always an easy task. Someone cannot be told exactly how to write lyrics, but that does not mean there are not ways in which he/she can learn to do it better. In-depth analysis has been carried out on Natural Language Processing, Text Mining, Machine Learning and existing Sega lyrics to consolidate the foundation of the project. Interviews have been done with a domain expert to learn the process of conventional song writing. Thus a tool, Paroles Sega Morisien, was developed. Paroles Sega Morisien enables users to generate Sega lyrics from randomly selected Mauritian Creole keywords. It is the first time that such a tool has been developed. An evaluation, consisting of a comparability study, was carried out to compare existing lyrics against lyrics generated by the tool. The result obtained was favorable.DOI: http://dx.doi.org/10.11591/ij-ai.v1i4.80
Implementation of Artificial Bee Colony Algorithm
Get PDFEvolutionary algorithm is a stochastic search method that mimics the natural biological evolution and the social behavior of species. Artificial bee colony algorithm is also a kind of evolutionary algorithm which was proposed by Dervis karaboga in 2005.Such algorithms have been developed to arrive at near-optimum solutions of multimodal optimization problems, which may not be possible with traditional algorithms. This paper describes implementation of ABC algorithm on complex benchmark functions like rastrigin, rosenbrock; sphere and schwefel the analysis of the performance of ABC algorithm were compared for the optimization of above benchmark functions with Partical Swarm Optimization (PSO). The ABC algorithm was successfully implemented in software tool ‘c’.DOI: http://dx.doi.org/10.11591/ij-ai.v1i3.58
Support Vector Machines Regression for MIMO-OFDM Channel Estimation
Get PDFIn this paper, we propose a robust highly selective nonlinear channel estimator for Multiple -Input Multiple-Output (MIMO) Orthogonal Frequency Division Multiplexing (OFDM) system using complex Support Vector Machines Regression (SVR) and applied to Long Term Evolution (LTE) downlink under high mobility conditions .The new method uses the information provided by the pilot signals to estimate the total frequency response of the channel in two phases: learning phase and estimation phase. The estimation algorithm makes use of the reference signals to estimate the total frequency response of the highly selective multipath channel in the presence of non-Gaussian impulse noise interfering with pilot signals. Thus, the algorithm maps trained data into a high dimensional feature space and uses the Structural Risk Minimization (SRM) principle to carry out the regression estimation for the frequency response function of the highly selective channel. The simulations show the effectiveness of the proposed method which has good performance and high precision to track the variations of the fading channels compared to the conventional LS method and it is robust under high mobility conditions.DOI: http://dx.doi.org/10.11591/ij-ai.v1i4.183
Estimating Processed Cheese Shelf Life with Artificial Neural Networks
Get PDFCascade multilayer artificial neural network (ANN) models were developed for estimating the shelf life of processed cheese stored at 7-8oC.Mean square error , root mean square error,coefficient of determination and nash - sutcliffo coefficient were applied in order to compare the prediction ability of the developed models.The developed model with a combination of 5à16à16à1 showed excellent agreement between the actual and the predicted data , thus confirming that multilayer cascade models are good in estimating the shelf life of processed cheese.DOI: http://dx.doi.org/10.11591/ij-ai.v1i1.33
Design and Implementation of Fuzzy Position Control System for Tracking Applications and Performance Comparison with Conventional PID
Get PDFThis paper was written to demonstrate importance of a fuzzy logic controller in act over conventional methods with the help of an experimental model. Also, an efficient simulation model for fuzzy logic controlled DC motor drives using Matlab/Simulink is presented. The design and real-time implementation on a microcontroller presented. The scope of this paper is to apply direct digital control technique in position control system. Two types of controller namely PID and fuzzy logic controller will be used to control the output response. The performance of the designed fuzzy and classic PID position controllers for DC motor is compared and investigated. Digital signal Microcontroller ATMega16 is also tested to control the position of DC motor. Finally, the result shows that the fuzzy logic approach has minimum overshoot, and minimum transient and steady state parameters, which shows the more effectiveness and efficiency of FLC than conventional PID model to control the position of the motor. Conventional controllers have poorer performances due to the non-linear features of DC motors like saturation and friction.DOI: http://dx.doi.org/10.11591/ij-ai.v1i1.40
Performance Analysis Of Clustering Protocol Using Fuzzy Logic For Wireless Sensor Network
Get PDFIn order to gather information more efficiently, wireless sensor networks are partitioned into clusters. The most of the proposed clustering algorithms do not consider the location of the base station. This situation causes hot spots problem in multi-hop wireless sensor networks. In this paper, we analyze a fuzzy clustering algorithm which aims to prolong the lifetime of wireless sensor networks. This algorithm adjusts the cluster-head radius considering the residual energy and the distance to the base station parameters of the sensor nodes. This helps decreasing the intra-cluster work of the sensor nodes which are closer to the base station or have lower battery level. In this paper fuzzy logic is utilized for handling the uncertainties in cluster-head radius estimation. We compare this algorithm with LEACH according to first node dies, half of the nodes alive and energy-efficiency metrics. Our simulation results show that the fuzzy clustering approach performs better than LEACH. Therefore, the fuzzy clustering algorithm is a stable and energyefficient clustering algorithm.DOI: http://dx.doi.org/10.11591/ij-ai.v1i3.125
Ontology-based Social Recommender System
Get PDFKnowledge sharing is vital in collaborative work environments.People working in the same environment aid better communication due to sharing information and resources within a contextual knowledge structure constructed based on their scope. Social networks play important role in our daily live as it enables people to communicate, and share information. The main idea of social network is to represent a group of users joined by some kind of voluntary relation without considering any preference. This paper proposes a social recommender system that follows user’s preferences to provide recommendation based on the similarity among users participating in the social network. Ontology is used to define and estimate similarity between users and accordingly being able to connect different stakeholders working in the community field such as social associations and volunteers.This approach is based on integration of major characteristics of content-based and collaborative filtering techniques. Ontology plays a central role in this system since it is used to store and maintain the dynamic profiles of the users which is essential for interaction and connection of appropriate knowledge flow and transaction.DOI: http://dx.doi.org/10.11591/ij-ai.v1i3.77