• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.4 no.1
• /
• pp.145-154
• /
• 2000

In this paper we proposed the orthogonal neural network(ONN) to control and identify the unknown controlled system. The proposed ONN used the buffer layer in front of the hidden layer and the hidden layer used the sigmoid function and its derivative a derived RBF that is a derivative of the sigmoid function. In order to verify the property of the proposed, it is examined by simulation results of the Narendra model. Controlled system is composed of ONN and confirmed its usefulness through simulation and experimental results.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 1999.11a
• /
• pp.214-217
• /
• 1999

In this paper we proposed a heterogeneous hidden layer consisting of both sigmoid functions and RBFs(Radial Basis Function) in multi-layered neural networks. Focusing on the orthogonal relationship between the sigmoid function and its derivative, a derived RBF that is a derivative of the sigmoid function is used as the RBF in the neural network. so the proposed neural network is called ONN's feasibility Neural Network). Identification results using a nonlinear. function confirm both the ONN's feasibility and characteristics by comparing with those obtained using a conventional neural network which has sigmoid function or RBF in hidden layer.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.8 no.3
• /
• pp.33-40
• /
• 1998

This paper presents an Modified Orthogonal Neural Network(MONN), new modified model of Orthogonal Neural Network(0NN) based on orthogonal functions, and applies it to nonlinear system approximator. ONN proposed by Yang and Tseng, doesn't have the problems of traditional multilayer feedforward neural networks such as the determination of initial weights and the numbers of layers and processing elements. And tranining of ONN converges rapidly. But ONN cannot adapt its orthogonal functions to a given system. The accuracy of ONN, in terms of the minimal possible deviation between system and approximator, is essentially dependent on the choice of basic orthogonal functions. In order to improve ability and effectiveness of approximate nonlinear systems, MONN has an input transformation layer to adapt its basic orthogonal functions to a given nonlinear system. The results show that MONN has the excellent performance of approximate nonlinear systems and the input transfnrmation makes the ability of MONN better than one of ONN.

• Proceedings of the Korea Association of Information Systems Conference
• /
• 1997.10b
• /
• pp.207-213
• /
• 1997

본 논문은 논리함수의 Module-2 방식에 따라 신경회로망을 구성하고, 구성된 신 경회로망에 의해서 논리함수 XOR을 예를 들어 표현하였다. 논리함수에서 사용된 Module-2 방식은 2치자료를 처리하는 것으로, 논리함수의 대표적인 것이 부울함수(Boole function)이 다. 신경회로망의 입력층에서는 논리함수의 값을 입력하고, 중간층에서는 논리함수의 직교 전개를 이용하고, 출력층에서도 직교전개등을 통하여 최종적인 결과값을 얻는 것이다. 이를 위해서 내적(inner product) 및 norm등을 정의하고, 직교(orthogonal)에 대한 정의를 하였 다. 또한, 부울함수(Boole Function)의 미분에 대한 정의 및 부울함수의 Maclaurin 전개에 대해서도 정의하였다. 본 논문에서 논리함수의 최종적인 출력값은 부울함수의 Maclaurin 전 개에 의해서도 실제적인 값이 출력될 수 있다는 것을 나타내었다.

• 전기의세계
• /
• v.10
• /
• pp.10-18
• /
• 1963

본 논문은 Laguerre함수의 Fourier변환에 의한 Wiener와 Lee의 회로망구성법을 일반화한 것이다. 즉 Laplace변환을 사용하여 시간영역에서 임의의 과도특성을 나타내는 회로를 구성하는 문제를 복소주파수면상의 직교함수계에 의한 근사문제로 귀착시켰으며 이때 얻어진 근사함수의 물리적실현성을 확인하고 실제구성예를 들었다. 이때 근사함수로는 복소극이 포함되었을 때도 적용할 수 있도록 이론과 방법이 일반화되었다.

• Proceedings of the KIEE Conference
• /
• 2006.10c
• /
• pp.279-281
• /
• 2006

적분연산행렬은 구간연속 직교함수들이 시스템 동정, 해석, 제어기 설계 등의 분야에 널리 이용될 수 있는 계기를 제공하였다. 특히 블럭펄스 함수는 연산이 간단하기 때문에 선형 시변계와 비선형계 등의 제어문제 둥에 널리 이용되어 오고 있다. 본 논문은 기존의 블럭펄스 함수 적분 연산행렬과 비교했을 때 적분오차를 줄이는 적분연산행렬을 소개하였으며, 이를 이용하여 고차 시스템의 응답에 가장 최적한 응답을 갖는 저차 시스템의 응답을 갖도록 최적응답 방법에 적용하여 대수적인 방법으로 저차 시스템의 파라메터를 구하는 알고리즘을 제시함으로서 유용성을 확인하였다.

• The KIPS Transactions:PartB
• /
• v.8B no.4
• /
• pp.395-402
• /
• 2001

본 논문에서는, 1차원 및 2차원 웨이브렛 함수를 전역통과필터(APF)와 지연기의 병렬접속에 위한 재귀형(IIR) 디지털 필터로 구성하는 방법을 제안한다. Mallat에 의하여 웨이브렛 변환과 필터뱅크가 밀접한 관계에 있다는 것이 알려졌고, 완전 재구성 필터뱅크로부터 웨이브렛 함수를 도출하는 다양한 방법이 알려져 있다. 그러나, 이러한 방법의 대부분은 비재귀형(FIR) 디지털 필터에 근거를 두는 것으로, 재귀형 디지털 필터에 의한 방법은 거의 제안되어 있지 않다. 재귀형 필터를 이용하는 장점은 비재귀형에 비하여 낮은 차수로 표현되는 점이다. 또 직교 웨이브렛 함수를 끌어내기 위한 직교조건을 용이하게 만족시킬 수 있다. 본 논문에서는 웨이브렛 함수에 요구되는 레귤레리티(Regularity)조건을 만족시키기 위하여, 최대 평탄성(Maximally Flat)을 부가한 새로운 1차원 및 2차원 재귀형 웨이브렛 함수의 도출법을 보인다.

• Journal of Korea Multimedia Society
• /
• v.3 no.4
• /
• pp.380-388
• /
• 2000

In computing the optical flow, Horn and Schunck's method which is a representative algorithm is based on differentiation. Therefore it is difficult to estimate the velocity for a large displacement by this algorithm. In this paper, we propose a method for estimating nonuniform motion from sequential images which is based on integral brightness constancy constraints. The equations which transform a source image to a target image are expressed as a function of the displacement field. If marginal effects can be neglected, the form of the transformation integral transform or orthogonal expansion can be determined from the expansion coefficients of the two images. The apparent displacement field is then computed iteratively by a projection method which utilities the functional derivatives of the linearized moment equations. We demonstrate that the performance of the orthogonal function transform on the data set of large motion.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.6 no.4
• /
• pp.43-52
• /
• 1986

With increasing of computer use, a least squares method is now widely used in the regression analysis of various data. Unreliable results of regression coefficients due to the floating point of computer and problems of ordinary least squares method are described in detail. To improve these problems, a least squares method using orthogonal function is developed. Also, Comparison and analysis are performed through an example of numerical test, and re-orthogonalization method is used to increase the accuracy. As an example of application, the optimum order of AR process for the time series of monthly flow at the Pyungchang station is determined using Akaike's FPE(Final Prediction Error) which decides optimum degree of AR process. The result shows the AR(2) process is optimum to the series at the station.

• Korean Journal of Fisheries and Aquatic Sciences
• /
• v.30 no.2
• /
• pp.188-202
• /
• 1997

The seasonal variability of the sea surface winds over the last Sea of Korea (Japan Sea) is investigated by means of empirical orthogonal function (EOF) analysis. The combined representation of fields of three climatic variables by empirical orthogonal functions is discussed. The eigenvectors are derived from daily sea level pressure, wind speed and 10-day mean sea surface temperature (SST) during 15 years $(1978\~1992)$. The spatial patterns of the mean pressure are characterized by the high pressure in the western part and the low pressure in the eastern part. The spatial distribution of the standard deviation (SD) of pressure are characterized by max SD of 6.6 mb near the Vladivostok, and minima along the coast of the Japan. In Vladivostok, the maxima of SD of SST and south-north wind (WV) were also occurred. The representation of fields of individual meteorological variables by EOF shows that the first mode of the west-east wind (WU) explain over $47.3\%$ of the variance and the second mode of WU represents $30\%$. Especially, the first mode of the WV explain $70.9\%$ of the variance and their time series coefficients show 1-cpy, 0.5-cpy frequency spectrum. The spatial distribution of the first mode eigenvectors of SST are characterized by maximum near Vladivostok. The combined representation of fields of several variables (pressure, wind, SST) reveals that the first mode magnitudes of the variance of the combined eigenvectors (WU-PR) are increased. By means of this result, the 1-year peak and the 6-months peak are remarkable. In the three combined patterns (wind, pressure, SST), the second mode of the eigenvector (wind) is affected by the SST. Their time coefficients of the first mode show noticeable 1-year peak. The spectral analysis of the second mode shows broad seasonal signal with the period of 4-months and a significant peak of variability at 3-month period.