• 제어로봇시스템학회:학술대회논문집
• /
• 2000.10a
• /
• pp.494-494
• /
• 2000

This paper demonstrates that the largest Lyapunov exponent $\lambda$ of recurrent neural networks can be controlled by a gradient method. The method minimizes a square error $e_{\lambda}=(\lambda-\lambda^{obj})^2$ where $\lambda^{obj}$ is desired exponent. The $\lambda$ can be given as a function of the network parameters P such as connection weights and thresholds of neurons' activation. Then changes of parameters to minimize the error are given by calculating their gradients $\partial\lambda/\partialP$. In a previous paper, we derived a control method of $\lambda$via a direct calculation of $\partial\lambda/\partialP$ with a gradient collection through time. This method however is computationally expensive for large-scale recurrent networks and the control is unstable for recurrent networks with chaotic dynamics. Our new method proposed in this paper is based on a stochastic relation between the complexity $\lambda$ and parameters P of the networks configuration under a restriction. Then the new method allows us to approximate the gradient collection in a fashion without time evolution. This approximation requires only $O(N^2)$ run time while our previous method needs $O(N^{5}T)$ run time for networks with N neurons and T evolution. Simulation results show that the new method can realize a "stable" control for larege-scale networks with chaotic dynamics.

• Proceedings of the KIEE Conference
• /
• 1996.11a
• /
• pp.179-181
• /
• 1996

In conventional hydrothermal coordination problem, the lambda-gamma iteration method is generally used for generation schedule. The procedure of classical lambda-gamma iteration method consists of 3 main loops and it is very complex. Therefore, it needs many iterative calculations. This paper proposes an advanced hydrothermal algorithm based on newly developed lambda-gamma iteration method. As lambda calculation loop is removed in the newly developed iteration method, iterative calculations are reduced and whole procedure is simplified. The proposed algorithm is verified on simple system.

• Journal of Korean Ophthalmic Optics Society
• /
• v.8 no.2
• /
• pp.163-168
• /
• 2003

The eye sensitivity in the difference conditions of a light source intensity consists of two functions by the receptor of cone and rod according to a wavelength. We derived a distribution function of $P{\lambda}=A{\cdot}e^{-({\lambda}-{\lambda}_u)^2/2W^2}$ using a respond probability of the receptor of cone-rod for a photon. It was well applied for a CIE eye's sensitivity curve of a wavelength. When there is lens In front of eye, luminous efficiency should be corrected. Transmission light which permeate to depends on absorption wavelength, and relationship of final luminous efficiency's estimation method is expressed by multiplication of luminous efficiency and transmittance intensity of lens. $$Pf({\lambda})=T({\lambda}){\cdot}P({\lambda})$$. The theory was applied to photopic and scopic vision with brown color lens.

• Journal of Institute of Control, Robotics and Systems
• /
• v.14 no.6
• /
• pp.587-594
• /
• 2008

In this paper, an efficient integer ambiguity resolution method for GNSS attitude determination system is described. The proposed method solves the integer least-squares with quadratic equality constraints(ILSQE) problem and shows an expansion of the LAMBDA method can be used to solve it. The solution of ILSQE is shown and an efficient implementation with a LAMBDA based method is given. The method is compared with some other methods. The results of static and dynamic tests show the dramatic improvement of the success rates of integer ambiguity resolution.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2009.10a
• /
• pp.179-180
• /
• 2009

We propose the method of analysis, separation, and express from normalized and standard method to reduce the problem with the existing scene change detection and the complicated and variety of images of real change. In this paper, we propose a new local ${\lambda}^2$-test which was combined the ${\lambda}^2$-existing test methods and local color histograms. This method is robust method of abrupt and gradual detection and solve the problem of traditional method.

• ETRI Journal
• /
• v.31 no.1
• /
• pp.68-70
• /
• 2009

In this letter, a parametric analysis of the Fresnel-field antenna measurement method is performed for a square aperture. As a result, the optimum number of Fresnel fields for one far-field point is guided as $M_{opt}=N_{opt}=D^2/{\lambda}R+5$, where D is the antenna diameter, ${\lambda}$ is the wavelength, and R is the distance between the source antenna and the antenna under test. For the aperture size 5 ${\leq}$ $L_x/{\lambda}$ ${\leq}$ 20, the tolerable distances for gain errors of 0.5 dB and 0.2 dB can be guided as $R_{0.5\;dB}$ ${\approx}$ $1.2Lx/{\lambda}$ and $R_{0.2\;dB}$ ${\approx}$ $2.0L_x/{\lambda}$, where $L_x$ is the lateral length of the square aperture. The tolerable distances for 20 ${\leq}$ $L_x/{\lambda}$ ${\leq}$ 200 are also proposed. This measurement guideline can be fully utilized when performing the Fresnel-field antenna measurement method.

• The Journal of Korean Institute of Electromagnetic Engineering and Science
• /
• v.16 no.5 s.96
• /
• pp.465-471
• /
• 2005

A method of calculation fur propagating and focusing of focused beams generated in antenna arrays, using BPM(Beam Propagation Method), is presented in this paper. Based on the diffraction theory, the beam focusing and Propagation is studied specially for the case of the antenna way fed by the Rotman lens that is able to focus microwave power on its focal arc or generate multiple beams. There are difficulties in performing a full-wave simulation using a commercial EM simulation tool for propagating and focusing of beams because of the structural complexity and the feeding assignment of the antenna array. Therefore, as an alternative solution, the BPM is presented to calculate the beam propagation from the aperture-type antennas. From the point of view of optics, the propagations of the lens have been simplified from the Fresnel diffraction integral to the Fourier transform. Using Fourier Transform, a beam propagation method is developed to show improvement of the resolution by controlling the wavefront of wave Propagating from an aperture-type antenna array. The beam width(or spot size) and the intensity are calculated for a focused beam propagating from an array having $10\lambda$ of its size. For the beams with $20\lambda,\;30\lambda$, and $50\lambda$ of geometrical focal length, the half-power beam widths(or spot size) are about 1.1\lambda,\;1.3\lambda,\;and\;1.9\lambda$ respectively.

• Proceedings of the KSRS Conference
• /
• 2009.03a
• /
• pp.280-285
• /
• 2009

1998년 8월부터 2005년 6월까지 한반도 주변 해역에서 현장관측한 해수의 고유 광특성(IOPs)과 외형적 광특성(AOPs) 자료들을 이용하여 원격반사도$(R_{rs}(\lambda))$와 성분별 흡광계수의 총 합 $(\alpha(\lambda)=\alpha_w(\lambda)+\alpha_{ph}(\lambda)+\alpha_{ss}(\lambda)+\alpha_{dom}(\lambda))$의 상관관계를 분석하고, $R_{rs}(\lambda)$ 밴드비를 이용하여 흡광계수 산출 알고리즘을 개발하였다. 파장에 따른 $R_{rs}(\lambda)$와 총합 $\alpha(\lambda)$의 상관관계는 반비례적인 관계를 보였고, 파장 443 nm일 때 상관도$(R^2)$는 0.717이다. $\alpha_{ph}(\lambda)$ 산출알고리즘은 엽록소의 흡광과 관련된 파장 490 nm와 부유물의 산란과 관련된 파장 555 nm의 $R_{rs}(\lambda)$ 밴드비의 함수 형태로 구성하였고, 파장 443 nm일 때 RMS 값은 0.223이다. $\alpha_{ss}(\lambda)$과 $\alpha_{dom}(\lambda)$ 산출 알고리즘은 용존유기물의 흡광과 관련된 파장 412 nm와 부유물의 산란과 관련된 파장 555 nm의 $R_{rs}(\lambda)$ 밴드비의 함수 형태로 구성하였고, 파장 412 nm일 때 RMS 값은 각각 0.324와 0.230이다. $\alpha_{ph}(\lambda),\;\alpha_{ss}(\lambda),\;\alpha_{dom}(\lambda)$ 산출 알고리즘들은 대체적으로 현장값보다 높게 추정하였고 스펙트럼들은 잘 재현해냈다. 추후 이에 대한 개선과 알고리즘의 검보정이 요구된다.

• Journal of Communications and Networks
• /
• v.3 no.4
• /
• pp.361-364
• /
• 2001

Over an additive abelian group G of order g and for a given positive integer $\lambda$, a generalized Hadamard matrix GH(g, $\lambda$) is defined as a gλ$\times$gλ matrix[h(i, j)], where 1 $\leq i \leqg\lambda and 1 \leqj \leqg\lambda$, such that every element of G appears exactly $\lambd$atimes in the list h($i_1, 1) -h(i_2, 1), h(i_1, 2)-h(i_2, 2), …, h(i_1, g\lambda) -h(i_2, g\lambda), for any i_1\neqi_2$. In this paper, we propose a new method of expanding a GH(g^m, \lambda_1) = B = [B_{ij}] over G^m$ by replacing each of its m-tuple B_{ij} with B_{ij} + GH(g, $\lambda_2) where m = g\lambda_2. We may use g^m/\lambda_1 (not necessarily all distinct) GH(g, \lambda_2$)s for the substitution and the resulting matrix is defined over the group of order g.

• Journal of the Korean Mathematical Society
• /
• v.56 no.4
• /
• pp.985-1000
• /
• 2019

Let k be an integer with $k{\geq}3$. Define $h(k)=[{\frac{k+1}{2}}]$, ${\sigma}(k)={\min}\(2^{h(k)-1},\;{\frac{1}{2}}h(k)(h(k)+1)\)$. Suppose that ${\lambda}_1,{\ldots},{\lambda}_5$ are non-zero real numbers, not all of the same sign, satisfying that ${\frac{{\lambda}_1}{{\lambda}_2}}$ is irrational. Then for any given real number ${\eta}$ and ${\varepsilon}>0$, the inequality $${\mid}{\lambda}_1p^2_1+{\lambda}_2p^2_2+{\lambda}_3p^2_3+{\lambda}_4p^2_4+{\lambda}_5p^k_5+{\eta}{\mid}<({\max_{1{\leq}j{\leq}5}}p_j)^{-{\frac{3}{20{\sigma}(k)}}+{\varepsilon}}$$ has infinitely many solutions in prime variables $p_1,{\ldots},p_5$. This gives an improvement of the recent results.