• Proceedings of the IEEK Conference
• /
• 1999.06a
• /
• pp.482-485
• /
• 1999

In this paper we proposed a heterogeneous hidden layer consisting of both sigmoid functions and RBFs(Radial Basis Function) in multi-layered neural networks. Focusing on the orthogonal relationship between the sigmoid function and its derivative, a derived RBF that is a derivative of the sigmoid function is used as the RBF in the neural network. so the proposed neural network is called ONN(Orthogonal Neural Network). Identification results using a nonlinear function confirm both the ONN＇s feasibility and characteristics by comparing with those obtained using a conventional neural network which has sigmoid function or RBF in hidden layer

• Journal of Sensor Science and Technology
• /
• v.20 no.3
• /
• pp.151-155
• /
• 2011

A variety of pattern recognition algorithms including neural networks may be applicable to the identification of odors. In this paper, an identification technique for an electronic odor sensing system applicable to wound state monitoring is presented. The performance of the radial basis function(RBF) network is highly dependent on the choice of centers and widths in basis function. For the fine tuning of centers and widths, those parameters are initialized by an ill-conditioned genetic fuzzy c-means algorithm, and the distribution of input patterns in the very first stage, the stochastic gradient(SG), is adapted. The adaptive RBF network with singular value decomposition(SVD), which provides additional adaptation capabilities to the RBF network, is used to process data from array-based gas sensors for early detection of wound infection in burn patients. The primary results indicate that infected patients can be distinguished from uninfected patients.

• Journal of Electrical Engineering and Technology
• /
• v.3 no.4
• /
• pp.468-475
• /
• 2008

This paper presents a method for estimating the transient stability status of the power system using radial basis function(RBF) neural network with a fast hybrid training approach. A normalized transient energy margin(${\Delta}V_n$) has been obtained by the potential energy boundary surface(PEBS) method along with a time-domain simulation technique, and is used as an output of the RBF neural network. The RBF neural network is then trained to map the operating conditions of the power system to the ${\Delta}V_n$, which provides a measure of the transient stability of the power system. The proposed approach has been successfully applied to the 10-machine 39-bus New England test system, and the results are given.

• Proceedings of the KIEE Conference
• /
• 2008.07a
• /
• pp.1889-1890
• /
• 2008

본 연구에서는 퍼지 추론 메커니즘에 기반 한 Polynomial RBF Neural Network(p-RBFNN)를 설계하고 얼굴인식 문제로 적용하여 분류기로서의 성능을 분석한다. 제안된 p-RBFNN 구조는 FCM 클러스터링에 기반 한 분할 함수를 활성 함수로 사용하며, 다항식 함수로 구성된 연결가중치를 사용함으로서 기존 신경회로망 분류기의 선형적인 특성을 개선한다. p-RBFNN 구조는 언어적 해석관점에서 "If-then"의 퍼지 규칙으로 표현되며 퍼지 추론 메커니즘에 의해 구동된다. 즉 조건부, 결론부, 추론부 세 가지의 기능적 모듈로 나뉘어 네트워크 구조가 형성된다. 조건부는 FCM 클러스터링을 사용하여 입력 공간을 분할하고, 결론부는 분할된 로컬 영역을 다항식 함수로 표현한다. 마지막으로, 네트워크의 최종출력은 추론부의 퍼지추론에 의한다. 또한 제안된 p-RBFNN을 얼굴인식 문제로 적용하여 성능을 분석한다.

• Journal of information and communication convergence engineering
• /
• v.8 no.6
• /
• pp.651-654
• /
• 2010

In this paper, we propose two power randomization schemes for the random beamforming (RBF) based MIMO systems in cellular downlink. In the proposed system, a BS randomizes not only the pre-coding matrix but also the power allocation matrix, while the conventional RBF system allocates an equal power to each transmit stream. The proposed water-filling based power randomization scheme (Scheme-I) is proper in the low SNR values and the proposed random-power based randomization scheme (Scheme-II) is proper in the high SNR values. The proposed system with the power randomization outperforms the conventional RBF system which allocates the same power for each data stream.

• Proceedings of the Korean Information Science Society Conference
• /
• 2006.10b
• /
• pp.485-488
• /
• 2006

본 논문에서는 두 대의 카메라를 직각으로 배치하여 얻은 동영상을 통해 3차원 인체 동작을 추정하는 방법을 제안한다. 제안된 시스템은 실루엣에서 전역 특징과 지역 특징을 추출하여 이 특징들을 정적 특징과 동적 특징으로 다시 나눈다. 모든 실루엣 특징은 RBF 신경망의 입력으로 이용되어 동작을 분류한다. 본 논문에서 제안된 신경망 동작 추정 시스템은 유아들의 동작 교육에 적용되었다. 동작 교육을 위해 제시되는 기본 동작은 걷기, 뛰기, 앙감질 등의 이동 동작과 구부리기, 뻗기, 균형잡기, 회전하기 등 비 이동 동작으로 구분되고, 이 7 가지 기본 동작은 성공적으로 추정되었다.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2003.05a
• /
• pp.141-146
• /
• 2003

본 논문에서는 문자열 영역 추출을 위한 3배 축소 명함 영상, 개별 문자 추출을 위한 2배 축소 명함 영상, 정확한 인식을 위한 원본 영상으로 명함 영상을 분리하고, 분리된 영상들을 대상으로 각 영상 크기에 적합한 처리를 수행하고 각각의 결과들을 이용하여 정확한 문자를 추출할 수 있는 방법을 제안한다 그리고 추출된 개별 문자들의 인식을 위해서 ART1을 적용한 개선된 RBF 네트워크를 제안하여 적용한다 제안된 명함 추출 방법은 원 영상을 각각의 처리 방법에 적합하도록 하기 위해서 다해상도로 분리한다. 문자열의 추출은 문자들의 간격을 축소 시켜서 블록을 추출하기 쉬운 적절한 최소 크기의 영상에서 수행하고, 개별 문자의 추출은 문자들의 간격을 분리할 수 있는 적절한 영상의 크기에서 수행한다 개별 문자 인식은 문자의 형태학적 특성을 잘 나타내기 위해서 원본 영상에 적용한다 본 논문에서 제안한 추출 방법은 문자를 정확히 추출할 수 있으며 병렬 처리가 가능하여 처리시간을 단축할 수 있는 장점을 가진다. 그리고 정확히 추출된 개별 문자들을 개선된 R8F 네트워크를 이용하여 인식률을 향상시킨다. 제안된 명함 추출 및 인식 방법의 성능을 확인하기 위해서 실제 영문 명함 영상을 대상으로 실험한 결과, 기존의 방법보다 명함 추출 및 인식에서 우수한 성능이 있음을 확인하였다.

• 제어로봇시스템학회:학술대회논문집
• /
• 2000.10a
• /
• pp.199-199
• /
• 2000

Most of linear time-varying(LTV) systems except special cases have no general solution for the dynamic equations. Thus, it is difficult to design time-varying controllers in analytic ways, and other control design approaches such as robust control have been applied to control design for uncertain LTI systems which are the approximation of LTV systems have been generally used instead. A robust control method such as quantitative feedback theory(QFT) has an advantage of guaranteeing the stability and the performance specification against plant parameter uncertainties in frozen time sense. However, if these methods are applied to the approximated linear time-invariant(LTI) plants which have large uncertainty, the designed control will be constructed in complicated forms and usually not suitable for fast dynamic performance. In this paper, as a method to enhance the fast dynamic performance, the approximated uncertainty of time-varying parameters are reduced by the proposed QFT parameter-scheduling control design based on radial basis function (RBF) networks for LTV systems with bounded time-varying parameters.

• Journal of Environmental Impact Assessment
• /
• v.18 no.1
• /
• pp.31-39
• /
• 2009

A lot of data which are used in environment analysis of air pollution have characteristics that are distributed continuously in space. In this point, the collected data value such as precipitation, temperature, altitude, pollution density, PM10 have spatial aspect. When geostatistical data analysis are needed, acquisition of the value in every point is the best way, however, it is impossible because of the costs and time. Therefore, it is necessary to estimate the unknown values at unsampled locations based on observations. In this study, spatial interpolation method such as local trend surface model, IDW(inverse distance weighted), RBF(radial basis function), Kriging were applied to PM10 annual average concentration of Seoul in 2005 and the accuracy was evaluated. For evaluation of interpolation accuracy, range of estimated value, RMSE, average error were analyzed with observation data. The Kriging and RBF methods had the higher accuracy than others.

• Transactions of the Korean Society of Machine Tool Engineers
• /
• v.18 no.1
• /
• pp.110-115
• /
• 2009

In general, the radial error motion of a machine tool spindle system is effected on the accuracy of the parts to be made. This paper presents in milling process an investigation into spindle rotational accuracy effects on surface roughness of processing parts. We experimented the effects on spindle rotational accuracy in milling process by cutting AL 7075 workpiece at various rotational speed. In order to analyze the effects of rotational accuracy on surface roughness, we proposed the method using iSIGHT's RBF Approximation. The proposed method can be used fur anticipating the surface roughness when some spindle rotational accuracy experiments could be done in milling process.