• 대한전기학회논문지:시스템및제어부문D
• /
• 제49권5호
• /
• pp.269-277
• /
• 2000

We propose two new types multi-stage blind clustering equalizers for QAM demoulation, which are called a complex classification algorithm(CCA) and a radial basis function algorithm(RBFA). The CCA uses a clustering technique based on the joint gaussian probability function and computes separately the real part and imaginary part for simple implementation as well as less computation. In order to improve the performance of CCA, the Dual-Mode CCA(DMCCA) incorporates the CCA tap-updating mode with the decision-directed(DD) mode. The RBFA reduces the number of cluster centers through three steps using the classification technique of RBF and then updates the equalizer taps for QAM demodulation. Test results on 16-QAM confirm that the proposed algorithms perform better the conventional multi-state equalizers in the senses of SER and MSE under multi-path fading channel.

• 한국지리정보학회지
• /
• 제5권2호
• /
• pp.16-24
• /
• 2002

원격탐사 응용분야 중 토지피복 분류를 통한 지구환경의 원격탐지기법은 환경 관리, 도시계획 및 지리정보시스템의 응용분야에 광범위하게 사용되고 있는 접근방식이다. 본 연구는 다목적 실용위성(Korea Multi-Purpose Satellite : KOMPSAT)의 전자광학카메라(electro-optical camera : EOC)를 통해 취득한 영상의 토지피복 정보를 추출하는 방안을 제시하였다. 사용영상은 다중 분광정보를 보유하고 있는 공간해상도 30m의 Landsat TM과 6.6m의 공간해상도와 단일밴드로 구성되어 있는 KOMPSAT EOC영상이며, 연구 대상지역은 청주시 미호천 수계이다. 영상합성은 IHS(intensity hue saturation), HPF(high pass filtering), CN(color normalization), 그리고 Wavelet 변환방식을 적용하여 결과를 비교하였다. 합성된 영상은 RBF-NN(radial basis function neural network)과 ANN(artificial neural network)법을 이용하여 피복분류를 실시하였으며, 이상의 과정을 통해 최적 결과를 도출하는 영상합성 및 분류기법을 제시하였다.

• Natural Product Sciences
• /
• 제24권3호
• /
• pp.164-170
• /
• 2018

Rapid geographical classification of Bupleuri Radix is important in quality control. In this study, near infrared spectroscopy (NIRS) combined with supervised pattern recognition was attempted to classify Bupleuri Radix according to geographical origins. Three supervised pattern recognitions methods, partial least square discriminant analysis (PLS-DA), quadratic discriminant analysis (QDA) and radial basis function support vector machine (RBF-SVM), were performed to establish the classification models. The QDA and RBF-SVM models were performed based on principal component analysis (PCA). The number of principal components (PCs) was optimized by cross-validation in the model. The results showed that the performance of the QDA model is the optimum among the three models. The optimized QDA model was obtained when 7 PCs were used; the classification rates of the QDA model in the training and test sets are 97.8% and 95.2% respectively. The overall results showed that NIRS combined with supervised pattern recognition could be applied to classify Bupleuri Radix according to geographical origin.

• 한국소음진동공학회논문집
• /
• 제12권10호
• /
• pp.799-807
• /
• 2002

The vibration signal can give an indication of the condition of rotating machinery, highlighting potential faults such as unbalance, misalignment and bearing defects. The features in the vibration signal provide an important source of information for the faults diagnosis of rotating machinery. When additional training data become available after the initial training is completed, the conventional neural networks (NNs) must be retrained by applying total data including additional training data. This paper proposes the fault diagnostics algorithm using the ART-Kohonen network which does not destroy the initial training and can adapt additional training data that is suitable for the classification of machine condition. The results of the experiments confirm that the proposed algorithm performs better than other NNs as the self-organizing feature maps (SOFM) , learning vector quantization (LYQ) and radial basis function (RBF) NNs with respect to classification quality. The classification success rate for the ART-Kohonen network was 94 o/o and for the SOFM, LYQ and RBF network were 93 %, 93 % and 89 % respectively.

• 한국콘텐츠학회논문지
• /
• 제1권1호
• /
• pp.8-14
• /
• 2001

본 논문에서는 의료영상에서 특정 장기를 추출하여 질환 부위를 인식하는 알고리즘을 제안한다. 의료영상이 추출되어진 장기 부위에서 질환을 인식하기 위하여 단일 신경회로망을 이용하면 신경회로망의 학습 능력과 일반화 능력이 한정적이므로 성능개선에 많은 문제가 있다. 따라서 추출된 장기로부터 질환부위를 인식하는 것은 신경회로망을 복합적인 방법, 즉 RBF (Radial Basis Function), BP (Back Propagation)로 구성하여 단일 신경회로망의 단점을 극복하였다. 본 논문에서 제안하는 알고리즘은 입력 의료영상의 다양한 형태 변화에 적응력이 뛰어남을 실험결과로 알 수 있었다. 그리고, 전체 알고리즘의 수행시간이 장기추출 알고리즘을 포함하여 일반적으로 10초 이내에 수행됨을 실험 결과 알 수 있었다. 제안된 알고리즘은 실시간으로 의료영상의 질환부위를 인식하여 판별 자동화를 통해 원격의료에 사용 되어 질 수 있다.

• 한국가시화정보학회지
• /
• 제16권2호
• /
• pp.23-30
• /
• 2018

The present study compares flow fields reconstructed by data assimilation method with different combinations of parameters. As a data assimilation method, Vortex-in-Cell-sharp (VIC#), which supplements additional constraints and multigrid approximation to Vortex-in-Cell-plus (VIC+), is used to reconstruct flow fields from scattered particle tracks. Two parameters, standard deviation of Gaussian radial basis function (RBF) and grid spacing, are mainly tested using artificial data sets which contain few particle tracks. Consequent flow fields are analyzed in terms of flow structure sizes. It is demonstrated that sizes of the flow structures are proportional to an actual scale of the standard deviation of RBF. It implies that a combination of larger grid spacing and smaller standard deviation which preserves the actual standard deviation is able to save computational resources in case of a low track density. In addition, a simple comparison using an experimental data filled with dense particle tracks is conducted.

• 한국지능시스템학회논문지
• /
• 제21권5호
• /
• pp.595-601
• /
• 2011

본 논문에서는 영상 정보를 이용하여 로보커 팔위의 역진자의 밸런싱 제어를 한다. 로봇 팔위에 놓인 역진자의 각도는 카메라로 검출하고 검출된 각도 값은 제어기로 귀환되어 오차를 생성한다. 따라서 전체 제어루프는 폐회로 루프를 형성한다. 제어 성능을 높이기 위해 기존 선형제어기에 신경망 제어기를 더하였다. RBF 네트워크의 학습 알고리즘은 FPGA에 설계된 부동소수점 연산이 가능한 디지털 제어기에 의해 수행된다. 실험을 통하여 전체 시스템 성능을 검증하였다.

• JSTS:Journal of Semiconductor Technology and Science
• /
• 제16권4호
• /
• pp.436-442
• /
• 2016

An electroencephalogram (EEG)-connectome processor to monitor and diagnose mental health is proposed. From 19-channel EEG signals, the proposed processor determines whether the mental state is healthy or unhealthy by extracting significant features from EEG signals and classifying them. Connectome approach is adopted for the best diagnosis accuracy, and synchronization likelihood (SL) is chosen as the connectome feature. Before computing SL, reconstruction optimizer (ReOpt) block compensates some parameters, resulting in improved accuracy. During SL calculation, a sparse matrix inscription (SMI) scheme is proposed to reduce the memory size to 1/24. From the calculated SL information, a small world feature extractor (SWFE) reduces the memory size to 1/29. Finally, using SLs or small word features, radial basis function (RBF) kernel-based support vector machine (SVM) diagnoses user's mental health condition. For RBF kernels, look-up-tables (LUTs) are used to replace the floating-point operations, decreasing the required operation by 54%. Consequently, The EEG-connectome processor improves the diagnosis accuracy from 89% to 95% in Alzheimer's disease case. The proposed processor occupies $3.8mm^2$ and consumes 1.71 mW with $0.18{\mu}m$ CMOS technology.

• 한국지능시스템학회논문지
• /
• 제20권5호
• /
• pp.652-658
• /
• 2010

본 논문에서는 밸런싱 메커니즘의 두 바퀴 형태의 이동로봇(Two Wheeled Mobile Robot:TWMR)을 구현하고 제어한다. TWMR은 역진자 시스템과 이동로봇을 합친 모바일 역진자 구조로 기존의 막대 형태의 진자대신 두 팔 달린 로봇 형태를 나타낸다. 각도와 위치에 대한 동시제어에 있어 외란에 대한 강건성을 부여하기 위해 RBF 신경회로망 제어 방식을 사용한다. 신경회로망 제어 방식으로는 입력보상 방식(RCT)을 사용하여 제어기의 성능을 실험을 통해 검증한다. 또한 원격으로 제어가능하게 하도록 시스템을 구현하여 실험하였다.

• 센서학회지
• /
• 제13권1호
• /
• pp.41-46
• /
• 2004

There is currently much interest in the development of instruments that emulate the senses of humans. Increasingly, there is demand for mimicking the human sense of smell, which is a sophisticated chemosensory system. An electronic nose system is applicable to a large area of industries including environmental monitoring. We have designed a protable electronic nose system using an array of commercial chemical gas sensors for recognizing and analyzing the various odours. In this paper, we have implemented a portable electronic nose system using an array of gas sensors for recognizing and analyzing VOCs (Volatile Organic Compounds) in the field. The accuracy of a portable electronic nose system may be lower than an instrument such as GC/MS (Gas Chromatography/Mass Spectrometer). However, a portable electronic nose system could be used on the field and showed fast response to pollutants in the field. Several different algorithms for odours recognition were used such as BP (Back-Propagation) or LM-BP (Levenberq-Marquardt Back-Propagation). We applied RBF (Radial Basis Function) Network for recognition and quantifying of odours, which has simpler and faster compared to the previously used algorithms such as BP and LM-BP.