• Computers and Concrete
• /
• 제10권6호
• /
• pp.587-607
• /
• 2012

The purpose of this research is to propose a design technique of concrete mix proportions satisfying service life through genetic algorithm (GA) and neural network (NN). For this, thirty mix proportions and the related diffusion coefficients in high performance concrete are analyzed and fitness function for diffusion coefficient is obtained considering mix components like w/b (water to binder ratio), cement content, mineral admixture (slag, flay ash and silica fume) content, sand and coarse aggregate content. Through averaging the results of 10 times GA simulations, relative errors to the previous data decrease lower than 5.0% and the simulated mix proportions are verified with the experimental results. Assuming the durability design parameters, intended diffusion coefficient for intended service life is derived and mix proportions satisfying the service life are obtained. Among the mix proportions, the most optimized case which satisfies required concrete strength and the lowest cost is selected through GA algorithm. The proposed technique would be improved with the enhancement of comprehensive data set including wider the range of diffusion coefficients.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 1995년도 하계학술대회 논문집 B
• /
• pp.995-998
• /
• 1995

As the modern industrial processes become more complex, it is getting more difficult to model and control the processes. Naturally, an advanced type of DCS(Distributed Control System) with higher level functions is being sought Advanced DCS is a DCS with advanced functions such as fault diagnosis, GPC(Generalized Predictive Control), NN(Neural Network), and Fuzzy Control. In this thesis, we have studied a fuzzy control algorithm for realizing an advanced DCS. Its algorithm is implemented in a form of function code which is a process control language, being used by the industrial engineers. To verify the realized function code of the fuzzy control, the function code is applied to a continuous casting process of the Pohang Iron & Steel Works in Kwangyang. The rules of the fuzzy control were collected via interviews of the field operators and their operation documents. Finally, usability of the function code of the fuzzy control is shown via simulation for the continuous casting process model.

• International Journal of Fuzzy Logic and Intelligent Systems
• /
• 제8권3호
• /
• pp.185-191
• /
• 2008

In this paper, experimental studies of a neural network (NN) control technique for non-model based position control of the x-y table robot are presented. Decentralized neural networks are used to control each axis of the x-y table robot separately. For an each neural network compensator, an inverse control technique is used. The neural network control technique called the reference compensation technique (RCT) is conceptually different from the existing neural controllers in that the NN controller compensates for uncertainties in the dynamical system by modifying desired trajectories. The back-propagation learning algorithm is developed in a real time DSP board for on-line learning. Practical real time position control experiments are conducted on the x-y table robot. Experimental results of using neural networks show more excellent position tracking than that of when PD controllers are used only.

• 디지털콘텐츠학회 논문지
• /
• 제15권1호
• /
• pp.129-136
• /
• 2014

Wi-Fi fingerprinting 시스템은 실내에서 사용되는 위치 측위 방법이며 AP(Access Point)에서 발생하는 RSS(Received Signal Strength)에 의존한다. AP로부터의 RSS는 벽, 장애물 그리고 사람에 의한 간섭과 다중경로 페이딩 효과에 의하여 변할 수 있기 때문에 Wi-Fi fingerprinting 시스템은 낮은 측위 정확도를 갖는다. 또한, Wi-Fi 신호는 벽을 통과하기 때문에 기존의 시스템은 사용자가 현재 위치해 있는 층을 구분하기 어렵다. 이러한 단점을 극복하기 위하여 본 논문은 정확한 실내 측위를 위한 LED fingerprinting 시스템을 제안한다. 제안된 시스템은 LED로부터 발생하는 LED-ID와 LED의 광 파워를 사용한다. 본 시스템의 Training 단계에서는 각 장소에 해당하는 Fingerprinting을 데이터베이스에 기록한다. Serving 단계에서는 K-NN(K-Nearest Neighbor) 알고리즘을 적용하여 기존의 데이터와 새롭게 수신되는 사용자의 데이터를 비교한다. 컴퓨터 시뮬레이션을 통해 CDF(Cumulative Distribution Function) 형태로 본 시스템의 성능을 나타내었고, 시뮬레이션 결과로부터 제안된 시스템은 평균 측위 정확도보다 8.6% 높은 정확도를 얻을 수 있다.

• 융합신호처리학회논문지
• /
• 제5권3호
• /
• pp.190-197
• /
• 2004

본 연구는 신경망에 근거한 패턴매칭 방법을 사용하여 일시적 허혈 에피소드의 자동예측을 다루고 있다. 다층 신경망을 학습하기 위한 알고리즘은 수정된 역전파 알고리즘으로서 이 알고리즘은 학습속도를 향상시키기 위해 뉴런간의 연결계수 뿐만 아니라 뉴런내 비선형 함수의 변수들도 갱신한다. 제안된 방법의 성능은 MIT/BIH long-term 데이터베이스의 심전도(ECG) 신호를 사용하여 평가하였다. 총 15 레코드(237 허혈 에피소드)에 대한 실험결과에 의하면 허혈 에피소드 예측의 평균 sensitivity와 specificity 각각 85.71%와 71.11%이다. 또한 제안된 방법은 실제 허혈 에피소드로부터 평균 45.53초 이전에 예측하였다. 이러한 결과는 패턴매칭 분류기로서의 신경망 접근방법이 일시적 허혈 에피소드예측에 유용한 도구로 사용될 수 있음을 의미한다.

• 한국정밀공학회:학술대회논문집
• /
• 한국정밀공학회 2007년도 춘계학술대회 논문집
• /
• pp.797-798
• /
• 2007

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2002년도 하계학술대회 논문집 D
• /
• pp.2149-2151
• /
• 2002

In this paper, a novel speed estimation method of an induction motor using neural networks(NNs) is presented. The NN speed estimator is trained online by using the error backpropagation algorithm, and the training starts simultaneously with the induction motor working. The neural network based vector controller has the advantage of robustness against machine parameter variation. The simulation results using Matlab/Simulink verify the useful of the proposed method.

• The Journal of the Acoustical Society of Korea
• /
• 제17권3E호
• /
• pp.12-21
• /
• 1998

In this paper, we describe the one channel five-way, V/U/M/N/S (Voice/Unvoice/Nasal/Silent), classification algorithm for automatically classifying speech. The decision making process is viewed as a pattern viewed as a pattern recognition problem. Two aspects of the algorithm are developed: feature selection and classifier type. The feature selection procedure is studied for identifying a set of features to make V/U/M/N/S classification. The classifiers used are a vector quantization (VQ), a neural network(NN), and a decision tree method. Actual five sentences spoken by six speakers, three male and three female, are tested with proposed classifiers. From a set of measurement tests, the proposed classifiers show fairly good accuracy for V/U/M/N/S decision.

• Steel and Composite Structures
• /
• 제9권5호
• /
• pp.445-455
• /
• 2009

This study proposes a system identification technique for a fiber-reinforced polymer deck with neural networks. Neural networks are trained for system identification and the identified structure gives training data in return. This process is repeated until the identified parameters converge. Hence, the proposed algorithm is called an iterative neural network scheme. The proposed algorithm also relies on recent developments in the experimental design of the response surface method. The proposed strategy is verified with known systems and applied to a fiber-reinforced polymer bridge deck with experimental data.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제9권7호
• /
• pp.2667-2682
• /
• 2015

Due to the wide application of face recognition (FR) in information security, surveillance, access control and others, it has received significantly increased attention from both the academic and industrial communities during the past several decades. However, partial face occlusion is one of the most challenging problems in face recognition issue. In this paper, a novel method based on linear regression-based classification (LRC) algorithm is proposed to address this problem. After all images are downsampled and divided into several blocks, we exploit the evaluator of each block to determine the clear blocks of the test face image by using linear regression technique. Then, the remained uncontaminated blocks are utilized to partial occluded face recognition issue. Furthermore, an improved Distance-based Evidence Fusion approach is proposed to decide in favor of the class with average value of corresponding minimum distance. Since this occlusion removing process uses a simple linear regression approach, the completely computational cost approximately equals to LRC and much lower than sparse representation-based classification (SRC) and extended-SRC (eSRC). Based on the experimental results on both AR face database and extended Yale B face database, it demonstrates the effectiveness of the proposed method on issue of partial occluded face recognition and the performance is satisfactory. Through the comparison with the conventional methods (eigenface+NN, fisherfaces+NN) and the state-of-the-art methods (LRC, SRC and eSRC), the proposed method shows better performance and robustness.