• 전산구조공학
• /
• 제1권1호
• /
• pp.87-94
• /
• 1988

반복 비선형 계획법의 하나인 선형화 기법을 절대수렴의 전제아래 합성 구조물의 최적 설계에 응용한다. 선형화 기법은 설계문제의 제약조건을 선형화된 등가 제약조건으로 변형시키며 active-set 정책을 구사한다. 결과, 매 설계단계에서 풀어야 할 상태 및 수반 방정식의 수를 줄임으로써 실질적인 계산의 절감을 기한다. 기둥으로 받쳐진 판-보 구조물은 최적화 기법의 능력을 시험키 위한 합성구조물의 좋은 예로서, 설계결과 선형화 기법은 만족할만한 수렴치로써 최적해를 산출함을 알 수 있고 나아가 이 방법은 모든 종류의 최적화 문제에 적용될 수 있을 것으로 보인다.

• 한국통신학회논문지
• /
• 제21권6호
• /
• pp.1533-1545
• /
• 1996

In this paper, the design and implementation of echo canceller with new variable step size algorithm is discussed. The method used in the new algorithm is to periodically adopt the test function which helps an optimal coefficient tracking. This algorithm outperforms LMS and VS algorithms in convergence speed and steady state error. As the period of test function is decreased, the speed of convergence is improved, but the number of calculation is increased, then the trade off between these parameters must be considered. Simulation results show new algorithm outperforms LMS and VS algorithms in convergence rate. For the design of hardware, circuit is designed with VHDL, and synthesized with Act1 withc is a FPGA library of ActelTM in use of synovation of InterGraph$^{TM}$. Verification of the synthesized circuit is carried out with simulator DLAB. The circuit based on the algorithm which is suggested in this paper calculated 7 radix places of inary number. A simulation data for the verification is based on the data of algorithm simulation. When the same input data is applied to the both simulation, output results of circuit simulation had slight difference in compare with that of algorithm simulation. The number of used gate is about 5,500 and We have 5.53MHz in maximum frequency.y.

• 한국전자통신학회논문지
• /
• 제8권10호
• /
• pp.1451-1456
• /
• 2013

본 논문은 고유값분석을 이용하여 주파수영역 독립성분석(FDICA)의 수렴속도를 증가시키기 위한 기법을 제안한다. 소나 시스템 등에서는 지연정보를 획득하여 간섭신호원을 빠른 속도로 제거하는 알고리즘이 중요하다. 두 개의 독립신호가 $2{\times}2$ 지연 혼합된 경우에 대한 고유값 분석을 통하여, 초기값 파라미터로 사용할 지연정보를 획득할 수 있다. 본 알고리즘을 검증하기 위한 컴퓨터 시뮬레이션은 수렴속도와 잡음제거에서의 성능향상을 보여준다. 초기조건이 목표값에 근접하고 있기 때문에 3회 정도의 반복수렴 실험으로도 잡음제거에 상당한 성능을 나타낸다. 수렴 후에도 기존 알고리즘보다 1~3dB 더 나은 잡음제거 성능을 볼 수 있다.

• Journal of applied mathematics & informatics
• /
• 제25권1_2호
• /
• pp.1-15
• /
• 2007

Consider a Rayleigh distribution with $$pdf\;p(x/{\theta})\;=\;2x{\theta}^{-1}\;{\exp}\;({-x^2}/{\theta})$$ and mean lifetime ${\mu}\;=\;\sqrt{\pi\theta}/2$. We study the two-action problem of testing the hypotheses $H_{0}\;:\;{\mu}{\leq}{\mu}_{0}$ against $H_{1}\;:\;{\mu}\;>\;{\mu}_{0}$ using a linear error loss of ${\mid}{\mu}\;-\;{\mu}_{0}{\mid}$ via the empirical Bayes approach. We construct a monotone empirical Bayes test ${\delta}^{*}_{n}$ and study its associated asymptotic optimality. It is shown that the regret of ${\delta}^{*}_{n}$ converges to zero at a rate $\frac{{\ln}^{2}n}{n}$, where n is the number of past data available when the present testing problem is considered.

• International journal of advanced smart convergence
• /
• 제6권1호
• /
• pp.18-25
• /
• 2017

Recently, VR HMD (virtual reality head mounted display) has been utilized for virtual training, entertainment, vision therapy, and optometry. In particular, virtual reality contents are increasingly used for vision therapy and optometry. Accordingly, high-quality virtual reality contents such as a natural vision of life is required. Therefore, it is necessary to study the content production according to the optical characteristics of the VR HMD. The purpose of this paper is to suggest a proper FOV (field of view) of contents according to the distortion rate. We produced virtual reality contents and obtained distorted images by virtual camera. The distortion rate is calculated by using the distorted image. It is proved that the optimal FOV of the VR content with the minimum distortion is $90{\sim}100^{\circ}$. The results of this study are expected to be applied to the production of high quality contents.

• 한국전기전자재료학회:학술대회논문집
• /
• 한국전기전자재료학회 2004년도 하계학술대회 논문집 Vol.5 No.2
• /
• pp.1234-1238
• /
• 2004

A Controlled neural networks are proposed in order to measure nonlinear environments in adaptive and in realtime. The structure of it is similar to recurrent neural networks: a delayed output as the input and a delayed error between tile output of plant and neural networks as a bias input. In addition, we compute the desired value of hidden layer by an optimal method instead of transfering desired values by backpropagation and each weights are updated by RLS(Recursive Least Square). Consequently, this neural networks are not sensitive to initial weights and a learning rate, and have a faster convergence rate than conventional neural networks. This new neural networks is Error Estimated Neural Networks. We can estimate nonlinear models in realtime by the proposed networks and control nonlinear models. To show the performance of this one, we have various experiments. And this controller call prove effectively to be control in the environments of various systems.

• 한국물환경학회지
• /
• 제33권1호
• /
• pp.8-14
• /
• 2017

Microcystis aeruginosa (M. aeruginosa) is a cyanobacterium species that can form harmful algal blooms in freshwater bodies worldwide. The use of pine needle extract (PNE) to control nuisance algae by allelopathic inhibition will be environmentally friendly and promising. PNE removed successfully upto 98% of M. aeruginosa at the following optimal conditions: pH 7, $25^{\circ}C$ of temperature, 100 rpm of mixing rate, 5 min of mixing time. These results was indicated that the amount of 1 g/L PNE was removed 1g dryweight/L of M. aeruginosa. The kinetic data showed substrate inhibition kinetics and maximum growth rate was obtained when the M. aeruginosa was grown in medium containing 0.5 g/L of initial concentration of PNE. Different substrate inhibition models were fitted to the kinetic data and found the Luong model was best. The model predicted kinetic parameters were in agreement with the experimental findings. The natural extract, PNE, can be a promising inhibition due to its high efficiency and low dose requirements.

• 한국전기전자재료학회:학술대회논문집
• /
• 한국전기전자재료학회 2008년도 하계학술대회 논문집 Vol.9
• /
• pp.525-526
• /
• 2008

For measuring nonlinear Vehicle Modeling based on 7-Depth Sensor, the neural networks are proposed m adaptive and in realtime. The structure of it is similar to recurrent neural networks; a delayed output as the input and a delayed error between the output of plant and neural networks as a bias input. In addition, we compute the desired value of hidden layer by an optimal method instead of transfering desired values by backpropagation and each weights are updated by RLS(Recursive Least Square). Consequently, this neural networks are not sensitive to initial weights and a learning rate, and have a faster convergence rate than conventional neural networks. This new neural networks is Error Estimated Neural Networks. We can estimate nonlinear models in realtime by the proposed networks and control nonlinear models.

• Journal of applied mathematics & informatics
• /
• 제23권1_2호
• /
• pp.153-165
• /
• 2007

This paper deals with the empirical Bayes two-action problem of testing $H_0\;:\;{\theta}{\leq}{\theta}_0$: versus $H_1\;:\;{\theta}>{\theta}_0$ using a linear error loss for some discrete nonexponential families having probability function either $$f_1(x{\mid}{\theta})=(x{\alpha}+1-{\theta}){\theta}^x\prod\limits_{j=0}^x\;(j{\alpha}+1)$$ or $$f_2(x{\mid}{\theta})=[{\theta}\prod\limits_{j=0}^{x-1}(j{\alpha}+1-{\theta})]/[\prod\limits_{j=0}^x\;(j{\alpha}+1)]$$. Two empirical Bayes tests ${\delta}_n^*\;and\;{\delta}_n^{**}$ are constructed. We have shown that both ${\delta}_n^*\;and\;{\delta}_n^{**}$ are asymptotically optimal, and their regrets converge to zero at an exponential decay rate O(exp(-cn)) for some c>0, where n is the number of historical data available when the present decision problem is considered.

• Journal of information and communication convergence engineering
• /
• 제19권1호
• /
• pp.1-7
• /
• 2021

Reinforcement learning is an area of machine learning that studies how an intelligent agent takes actions in a given environment to maximize the cumulative reward. In this paper, we propose a new MAC protocol based on the Q-learning technique of reinforcement learning to improve the performance of the IEEE 802.11 wireless LAN CSMA/CA MAC protocol. Furthermore, the operation of each access point (AP) and station is proposed. The AP adjusts the value of the contention window (CW), which is the range for determining the backoff number of the station, according to the wireless traffic load. The station improves the performance by selecting an optimal backoff number with the lowest packet collision rate and the highest transmission success rate through Q-learning within the CW value transmitted from the AP. The result of the performance evaluation through computer simulations showed that the proposed scheme has a higher throughput than that of the existing CSMA/CA scheme.