• 전자공학회논문지CI
• /
• 제46권3호
• /
• pp.1-7
• /
• 2009

본 논문은 효율적인 음악 데이터의 분류를 위한 방법으로 분산커널 기반의 퍼지 c-평균을 이용한 분류기 모델을 제안한다. 분산 커널 기반의 퍼지 c-평균은 주어진 오디오 데이터에서 추출된 특징벡터의 평균과 공분산 정보를 동시에 이용하여 기존의 평균값만을 사용하는 방식에 비해 성능을 월등히 향상시킬 수 있는 장점이 있다. 사용된 방식은 확률적 분포로 주어지는 데이터 사이의 거리를 분산거리척도로 측정하고, 복잡한 분류 경계를 단순화 시키는데 효율적인 커널 개념을 사용함으로서 분류의 정확도를 극대화 시킬 수 있는 장점이 있다. 제안하는 분류기의 성능을 평가하기 위하여 고전음악, 컨트리음악, 힙합, 재즈의 4개의 장르 음악데이터를 총 1200개 수집하여 실험을 진행하였다. 실험의 결과 제안된 분산커널 기반의 퍼지 c-평균을 이용하는 분류기는 기존의 방식과 비교하여 분류정확도에서 평균적으로 17.73%-21.84%의 성능향상을 보여준다.

• 한국통신학회논문지
• /
• 제38A권6호
• /
• pp.504-511
• /
• 2013

본 논문은 지적재산권 보호를 위한 방안으로 적응형 알고리즘 기반의 초소형 전자소자 탐지 기법을 제안한다. 전자소자를 탐지하는 기본 원리는 분류기의 송신기에서 특정 기본 주파수의 전파가 은닉된 물체로 전파되면, 물체로부터 반사되어 수신기로 들어오는 2차 및 3차 고조파의 크기를 분류기가 비교함으로써 판별하게 된다. 하지만, 측정 과정에서 발생하는 잡음 및 전자파의 간섭으로 인해 분류의 성능이 저하되므로, 이러한 환경에서도 은닉된 전자소자를 적응적으로 판별하기 위해 Fuzzy c-Means 클러스터링 알고리즘과 ${\kappa}$-Nearest Neighbor 분류 알고리즘을 복합적으로 이용하는 방안을 제시한다. 모의실험 결과, 제안 기법이 잡음 및 전자파 간섭 환경에서도 적응적으로 전자소자 잘 탐지할 수 있었으며, 이에 따라 지적재산권을 효율적으로 보호할 수 있을 것으로 기대된다.

• Journal of Power Electronics
• /
• 제13권4호
• /
• pp.516-527
• /
• 2013

This paper presents a new overall system for state-of-available-power (SoAP) prediction for a lithium-ion battery pack. The essential part of this method is based on an adaptive network architecture which utilizes both fuzzy model (FIS) and artificial neural network (ANN) into the framework of adaptive neuro-fuzzy inference system (ANFIS). While battery aging proceeds, the system is capable of delivering accurate power prediction not only for room temperature, but also at lower temperatures at which power prediction is most challenging. Due to design property of ANN, the network parameters are adapted on-line to the current battery states (state-of-charge (SoC), state-of-health (SoH), temperature). SoC is required as an input parameter to SoAP module and high accuracy is crucial for a reliable on-line adaptation. Therefore, a reasonable way to determine the battery state variables is proposed applying a combination of several partly different algorithms. Among other SoC boundary estimation methods, robust extended Kalman filter (REKF) for recalibration of amp hour counters was implemented. ANFIS then achieves the SoAP estimation by means of time forward voltage prognosis (TFVP) before a power pulse occurs. The trade-off between computational cost of batch-learning and accuracy during on-line adaptation was optimized resulting in a real-time system with TFVP absolute error less than 1%. The verification was performed on a software-in-the-loop test bench setup using a 53 Ah lithium-ion cell.

• International Journal of Fuzzy Logic and Intelligent Systems
• /
• 제5권4호
• /
• pp.286-290
• /
• 2005

We consider the poisson equation where the functions involved are periodic including the solution function. Let $R=[0,1]{\times}[0,l]{\times}[0,1]$ be the region of interest and let $\phi$(x,y,z) be an arbitrary periodic function defined in the region R such that $\phi$(x,y,z) satisfies $\phi$(x+1, y, z)=$\phi$(x, y+1, z)=$\phi$(x, y, z+1)=$\phi$(x,y,z) for all x,y,z. We describe a very simple method for solving the equation ${\nabla}^2u(x, y, z)$ = $\phi$(x, y, z) based on the cubic spline interpolation of u(x, y, z); using the requirement that each interval [0,1] is a multiple of the period in the corresponding coordinates, the Laplacian operator applied to the cubic spline interpolation of u(x, y, z) can be replaced by a square matrix. The solution can then be computed simply by multiplying $\phi$(x, y, z) by the inverse of this matrix. A description on how the storage of nearly a Giga byte for $20{\times}20{\times}20$ nodes, equivalent to a $8000{\times}8000$ matrix is handled by using the fuzzy rule table method and a description on how the shape preserving property of the Laplacian operator will be affected by this approximation are included.

• Journal of Biosystems Engineering
• /
• 제30권6호통권113호
• /
• pp.340-346
• /
• 2005

Major quality features of the beef carcass in most countries including Korea are size, marbling state of the lean tissue, color of the fat and lean tissue, and thickness of back fat of the 13th rib. To evaluate the beef quality, extracting loin parts from the sectional image of the 13th beef rib is crucial and is the first step. However, because of the inhomogeneous distribution and fuzzy pattern of the fat and lean tissues on the beef cut, it is difficult to extract automatically the proper contour of the lean tissue. In this paper, a prototype mobile beef quality measurement system, which can be implemented practically at the beef processing site was developed. The developed system was composed of the hand held image acquisition unit and mobile processing unit mounted with touch-pad screen. Algorithms to extract the boundary of the lean tissue and a proper tool to evaluate the marbling status have been developed using color image processing. The boundary extraction algorithm showed successful results for the beef cuts with simple and moderate patterns of the lean tissue and fat. However, it had some difficulty in eliminating complex pattern of the extraneous tissues adhered to the lean tissue in the boundary extraction. The developed algorithms were implemented to the prototype mobile processing unit.

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 2001년도 하계종합학술대회 논문집(4)
• /
• pp.85-88
• /
• 2001

Unlike rigid objects or This paper developes the algorithm for segmenting gaseous objects on an image plane. Unlike rigid objects or solid non-rigid objects, gaseous objects vary in density even within single-object regions and the edge intensity differs at different locations. So, an edge detector may detect only strong edges and detected edges may be an incomplete parts of an whole object's boundary. Due to this property of gaseous objects, it is not easy to distinguish the real edges of gaseous objects from the noisy-like edges such as leaves. Our algorithm uses two criteria of edge intensity and edge's line connectivity, then applies fuzzy set so as to obtain the proper threshold of the edge detector

• International Journal of Fuzzy Logic and Intelligent Systems
• /
• 제11권1호
• /
• pp.19-24
• /
• 2011

Edges represent significant boundary information between objects or classes. Various methods, which are based on differential operation, such as Sobel, Prewitt, Roberts, Canny, and etc. have been proposed and widely used. The methods are based on a linear convolution of mask with pre-assigned coefficients. In this paper, we propose an edge detection method based on Dempster-Shafer's evidence theory to evaluate edgeness of the given pixel. The effectiveness of the proposed method is shown through experimental results on several test images and compared with conventional methods.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2000년도 하계학술대회 논문집 A
• /
• pp.242-244
• /
• 2000

This paper presents a method for real-time estimation of TCSC quantity in order to enhance the power system transient stability energy margin using fuzzy neural network in multi-machine system. This paper has two parts, the first part is to estimate the energy margin. To set critical energy, we use the potential energy boundary surface(PEBS) method which one of the transient energy function(TEF) method. And the second is to determine the TCSC quantify and the line to be injected. In order to make training data in this step, we use genetic algorithm. The proposed method is applied to 6-bus, 7-line, 4-machine model system to show its effectiveness.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 1995년도 Proceedings of the Korea Automation Control Conference, 10th (KACC); Seoul, Korea; 23-25 Oct. 1995
• /
• pp.154-157
• /
• 1995

A magnetic levitation control system is nonlinear and very unstable. Thus there should be a stabilizing compensation network and a feedback path. Due to the levitation control a noncontact photoresistor sensor is generally used. One photocell provides a certain amount of variation in length by the ball shadow casted on the cell surface. Furthermore at the boundary of the cell, the linearity of sensitivity deteriorates severely. To overcome the constraints of the length and linearity, an efficient sensor array is deviced and applied in the feedback path of a large-gap magnetic levitation control system. A number of CdS photocells and a summing circuit of the sensor output signals are used for a sensor array. The levitation length of a ball and the transient performances are main objectives of the large-gap suspension system using the sensor array.

• International Journal of Fuzzy Logic and Intelligent Systems
• /
• 제2권3호
• /
• pp.227-230
• /
• 2002

A new iris segmentation and recognition method is described. Combining a statistical classification and elastic boundary fitting, the iris is first segmented robustly and accurately. Once the iris is segmented, one-dimensional signals are computed in the iris and decomposed into multiple frequency bands. Each decomposed signal is approximated by a piecewise linear curve connecting a small set of node points. The node points represent features of each signal. The similarity measture between two iris images is the normalized cross-correlation coefficients between simplified signals.