• 대한화학회지
• /
• 제16권3호
• /
• pp.119-134
• /
• 1972

Various MO methods with differing degrees of sophistication are shown to yield qualitatively consistent results for methyl isoalloxazins. However, with crude methods such as the HMO and ${\omega}$-technique, the choice of Coulomb and resonance integralsis critical, in contrast with simpler molecular systems. The empirical value of ${\omega}$=0.5 appears to be more reasonable than 1.4. Methyl groups in these flaving are best treated by the group orbital approximation. The pseudo-heteroatom approximation overestimates methyl hyperconfiguration with the Pariser-Parrpole SCR MO method. siglet ${\pi}{\rightarrow}{\pi}^*$ transition energies are calculated by the P-P-P method and agree reasonably with the experimental values. 2- and 4-Thioisoalloxazine analogs are also treated. Reactivity indices of the flavin molecule are presented, includeing superdelocalizability. frontier orbital and radical densities. Various aspects of the applications of these indices of the methyl groups on dipolemoments, inozation potentialsm elctron affinities, and spectra are decribed in detail.

• 제어로봇시스템학회논문지
• /
• 제20권1호
• /
• pp.42-47
• /
• 2014

This paper investigates an adaptive approximation design problem for the tracking control of output-constrained non-affine pure-feedback systems. To satisfy the desired performance without constraint violation, we employ a barrier Lyapunov function which grows to infinity whenever its argument approaches some limits. The main difficulty in dealing with pure-feedback systems considering output constraints is that the system has a non-affine appearance of the constrained variable to be used as a virtual control. To overcome this difficulty, the implicit function theorem and mean value theorem are exploited to assert the existence of the desired virtual and actual controls. The function approximation technique based on adaptive neural networks is used to estimate the desired control inputs. It is shown that all signals in the closed-loop system are uniformly ultimately bounded.

• 한국전산구조공학회:학술대회논문집
• /
• 한국전산구조공학회 2003년도 봄 학술발표회 논문집
• /
• pp.201-208
• /
• 2003

Since the real steel box girder bridges have a large number of design variables and show complex structural behavior, it would be impractical to directly use the algorithm for its optimum design. Thus, in this study, for optimum design of real steel box girder bridge, approximated reanalysis using an higher-order Improved self-adjusted Convex Approximation (ISACA) which was newly proposed on a previous study by the author is applied for the numerical efficiency. To demonstrate the efficiency, robustness, and convergence of the approximated reanalysis technique using the ISACA, a real bridge having two continuous spans is used as an illustrative example. From the results of the numerical investigation, it may be positively stated that the efficiency, robustness, and convergence of the approximated reanalysis using an ISACA is superior compared with the previous approximated reanalysis techniques.

• 대한전자공학회논문지
• /
• 제20권1호
• /
• pp.7-14
• /
• 1983

직선 및 곡선과 같은 선형 특징은 영상 처리에 있어 중요한 특징중의 하나이다. 본 논문에서는 의미있는 선형 특징의 새로운 기법이 제안된다. 이 기법은 그래프 이론의 미니멀 스패닝 트리를 이용하여 경계점들을 연결하고 그 다음, 헤어(의미없는 잔가지)와 불합리한 선분을 제거한다. 이와 같이 추적된 선형 특징을 근사화 묘사하기 위하여 부분 선형 근사화를 수행한다. 본 논문에서 제안된 기법으로 실험을 수행하여 그 결과를 보여 주었다.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2008년도 제39회 하계학술대회
• /
• pp.1505-1506
• /
• 2008

In this paper, a robust adaptive backstepping control is developed for efficiency optimization of induction motors with uncertainties. The proposed control scheme consists of efficiency flux control(EFC) using a sliding mode adaptive flux observer and robust speed control(RSC) using a function approximation for mechanical uncertainties. In EFC, it is important to find the flux reference to minimize power losses of induction motors. Therefore, we proposed the optimal flux reference using the electrical power loss function. The sliding mode flux observer is designed to estimate rotor fluxes and variation of inverse rotor time constant. In RSC, the unknown function approximation technique employs nonlinear disturbance observer(NDO) using fuzzy neural networks(FNNs). The proposed controller guarantees both speed tracking and flux tracking. Simulation results are presented to illustrate the effectiveness of the approaches proposed.

• 말소리와 음성과학
• /
• 제2권1호
• /
• pp.87-95
• /
• 2010

In this paper, we address issues associated with mixed-lingual text-to-speech based on context-dependent HMMs, where there are multiple sets of HMMs corresponding to each individual language. In particular, we propose smoothing techniques of synthesis parameters at the boundaries between different languages to obtain more natural quality of speech. In other words, mel-frequency cepstral coefficients (MFCCs) at the language boundaries are smoothed by applying several linear and nonlinear approximation techniques. It is shown from an informal listening test that synthesized speech smoothed by a modified version of linear least square approximation (MLLSA) and a quadratic interpolation (QI) method is preferred than that without using any smoothing technique.

• 한국전산구조공학회:학술대회논문집
• /
• 한국전산구조공학회 2001년도 봄 학술발표회 논문집
• /
• pp.335-342
• /
• 2001

In this paper, an improved Element-Free Galerkin (EFG) method is proposed by adding enrichment function to the standard EFG approximation and a discontinuity function is implemented in constructing the shape function across the crack surface. In this method, the singularity and the discontinuity of the crack are efficiently modeled by using initial node distribution to evaluate reliable stress intensity factor, though the standard EFG method requires placing additional nodes near the crack tip. The proposed method enables the initial node distribution to be kept without any additional nodal d.o.f. and expresses the asymptotic stress field near the crack tip successfully. Numerical example verifies the improvement and the effectiveness of the method.

• 전자공학회논문지T
• /
• 제35T권1호
• /
• pp.33-39
• /
• 1998

본 논문에서는 다해상도 해석이 가능한 웨이브렛 시리즈 기법[4,5]을 이용하여 시변 시스템 임펄스 응답의 근사화를 이룬다. 웨이브렛 시리즈 표현이 행렬 연산의 형태로 표현되고 이러한 행렬 표현 형식은 망의 구성을 가능하게 하며 웨이브렛의 계수가 가중치 행렬로 나타난다. 웨이브렛 계수의 최적값을 구하기 위해 웨이브렛 망을 출력 오차로부터 계수 행렬을 적절한 값으로 수렴시키는 학습 법칙을 온-라인으로 학습하며 이의 결과가 시스템의 결과에 추정함을 보인다.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2007년도 심포지엄 논문집 정보 및 제어부문
• /
• pp.481-482
• /
• 2007

This paper describes the design of a l0-bit 100 KSample/S CMOS A/D Converter for biomedical applications such as pulse oximetry, body weight scale, ECG etc. We adopted an asynchronous architecture in the 10-b DAC design and hence reduces the number of switches by 11 and resistors by 64 compared with the conventional l0-b DAC. We also reduced the power consumption compare with the conventional architecture by 0.4mW. Output offset cancellation technique is applied to the design of comparator. The total power consumption of designed circuit is 190uW at the supply voltage of 1.8V with the 0.18um general CMOS technology.

• ETRI Journal
• /
• 제33권6호
• /
• pp.914-923
• /
• 2011

This paper concerns robust and reliable speaker model training for text-independent speaker verification. The baseline speaker modeling approach is the Gaussian mixture model (GMM). In text-independent speaker verification, the amount of speech data may be different for speakers. However, we still wish the modeling approach to perform equally well for all speakers. Besides, the modeling technique must be least vulnerable against unseen data. A traditional approach for GMM training is expectation maximization (EM) method, which is known for its overfitting problem and its weakness in handling insufficient training data. To tackle these problems, variational approximation is proposed. Variational approaches are known to be robust against overtraining and data insufficiency. We evaluated the proposed approach on two different databases, namely KING and TFarsdat. The experiments show that the proposed approach improves the performance on TFarsdat and KING databases by 0.56% and 4.81%, respectively. Also, the experiments show that the variationally optimized GMM is more robust against noise and the verification error rate in noisy environments for TFarsdat dataset decreases by 1.52%.