• 한국항해학회지
• /
• 제21권1호
• /
• pp.109-118
• /
• 1997

In this paper, applications of multilayer neural networks to control of flexible robot beam are considered. The multilayer nerual networks can be used to approximate any continuous function to a desired degree of accuracy and the weights are updated by Gradient Method. When a flexible beam is rotated by a motor through the fixed end, transverse vibration may occur. The motor torque should be controlled insuch a way that the motor rotates by a specified angle, while simultaneously stabilizing vibration of the flexible manipulators so that is arrested as soon as possbile at the end of rotation. Accurate control of lightweight beam during the large changes in configuration common to robotic tasks requires dynamic models that describe both rigid body motions, as well as the flexural vibrations. Therefore, a linear dynamic state-space model of for a single link flexible robot beam is derived and PD controller, LQP controller, and inverse dynamical neural networks controller are composed. The effectiveness the proposed control system is confirmed by computer simulation.

• 한국전기전자재료학회:학술대회논문집
• /
• 한국전기전자재료학회 2003년도 춘계학술대회 논문집 기술교육전문연구회
• /
• pp.90-95
• /
• 2003

For the safety driving of an automobile which is become individual requisites, a new Neural Network algorithm which recognized the load vehicles in real time is proposed. The proposed neural network technique is the real time computation method through the inter-node diffusion. In the network, a node corresponds to a state in the quantized input space. Each node is composed of a processing unit and fixed weights from its neighbor nodes as well as its input terminal. The most reliable algorithm derived for real time recognition of vehicles, is a dynamic programming based algorithm based on sequence matching techniques that would process the data as it arrives and could therefore provide continuously updated neighbor information estimates. Through several simulation experiments, real time reconstruction of the nonlinear image information is processed. 1-D LIPN hardware has been composed and various experiments with static and dynamic signals have been implemented.

• 한국지능시스템학회:학술대회논문집
• /
• 한국퍼지및지능시스템학회 1998년도 춘계학술대회 학술발표 논문집
• /
• pp.107-111
• /
• 1998

This paper presents a now approach to the design of intelligent contorl system for track vehicle system using fuzzy logic based on neural network. The proposed control scheme uses a Gaussian function as a unit function in the neural network-fuzzy, and back propagation algorithm to train the fuzzy-neural network controller in the framework of the specialized learning architecture. Moreover, We develop a Windows 95 version dynamic simulator which can simulate a track vehicle model in 3D graphics space. It is proposed a learning controller consisting of two neural network-fuzzy based of independent reasoning and a connection net with fixed weights to simply the neural networks-fuzzy. The dynamic simulator for track vehicle is developed by Microsoft Visual C++. Graphic libraries, OpenGL, by Silicon Graphics, Inc. were utilized for 3D Graphics. The performance of the proposed controller is illustrated by simulation for trajectory tracking of track vehicle speed.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 1998년도 하계학술대회 논문집 G
• /
• pp.2189-2192
• /
• 1998

The objective of this paper is, based upon the principles of artificial life, to induce emergent behaviors of multiple autonomous mobile robots which form from simple local rules to complex global intelligence. Here, we propose an architecture of neural network learing with reinforcement signals which perceives the neighborhood information and decides the direction and the velocity of movement as mobile robots navigates in a group. As results of the simulations, the optimum weights are obtained in real time, which not only prevent from the collisions between agents and obstacles in the dynamic environment, but also have the mobile robots move and keep in various patterns.

• 한국전산구조공학회:학술대회논문집
• /
• 한국전산구조공학회 2000년도 가을 학술발표회논문집
• /
• pp.331-338
• /
• 2000

In this paper, a non-linear finite element formulation for the spatial cable-net structures is simulated and using this formulation, the characteristics of structural behaviors for the elastic catenary cable are examined In the simulating procedure for the elastic catenary cable, nodal forces and tangential stiffness matrices are derived using catenary parameters of the exact solutions by a governing differential equation of catenary cable, cable self-weights and unstressed cable length. Dynamic Relaxation Method that considers kinetic damping is used for the structure analysis and Newton Raphson Method is used to verify the accuracy of solutions. In the analysis of two dimensional cable, the results obtain from the elastic catenary elements are shown more accurate than does of truss elements and in the case of spatial cable-net structures, Dynamic Relaxation Method is more stable to be converged than Newton Raphson Method.

• 한국전기전자재료학회:학술대회논문집
• /
• 한국전기전자재료학회 2001년도 추계학술대회 논문집
• /
• pp.46-51
• /
• 2001

For Fast Real-time Recognition of Nonlinear Error Data, a new Neural Network algorithm which recognized the map in real time is proposed. The proposed neural network technique is the real time computation method through the inter-node diffusion. In the network, a node corresponds to a state in the quantized input space. Each node is composed of a processing unit and fixed weights from its neighbor nodes as well as its input terminal. The most reliable algorithm derived for real time recognition of map, is a dynamic programming based algorithm based on sequence matching techniques that would process the data as it arrives and could therefore provide continuously updated neighbor information estimates. Through several simulation experiments, real time reconstruction of the nonlinear map information is processed.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 1996년도 하계학술대회 논문집 B
• /
• pp.939-942
• /
• 1996

This paper presents a stable learning algorithm for diagonal recurrent neural network(DRNN). DRNN is applied to a problem of controlling nonlinear dynamical systems. A architecture of DRNN is a modified model of the Recurrent Neural Network(RNN) with one hidden layer, and the hidden layer is comprised of self-recurrent neurons. DRNN has considerably fewer weights than RNN. Since there is no interlinks amongs in the hidden layer. DRNN is dynamic mapping and is better suited for dynamical systems than static forward neural network. To guarantee convergence and for faster learning, an adaptive learning rate is developed by using Lyapunov function. The ability and effectiveness of identifying and controlling a nonlinear dynamic system using the proposed algorithm is demonstrated by computer simulation.

• 한국운동역학회지
• /
• 제25권1호
• /
• pp.85-94
• /
• 2015

Purpose : The purpose of this study was to investigate the influence on the ground reaction force parameters according to wearing positions of backpack for during stair ascending and descending. Methods : Participants selected as subject were consisted of young female(n=10) and performed stairs walks(ascending and descending) with 2 types of wearing position(front of trunk[FT], rear of trunk[RT]). Passive(Fz 1) and active(Fz 2) forces of the vertical GRF were determined from time function and frequency domain. Also shear forces(Fx, Fy 1, Fy 2), dynamic postural stability index(MLSI, APSI, VSI, DPSI), loading rate and center of pressure (${\Delta}COPx$, ${\Delta}COPy$, COP area) were calculated from time function and frequency domain. Results : Fx, Fy 1, Fy 2, and Fz 1 in GRF didn't show significant differences statistically according to the wearing positions of backpack(p>.05), but stair descending showed higher forces than that of stair ascending. Particularly, Fz 2 of stair ascending showed higher forces than that of stair descending(p<.001), RT types showed higher than that of FT types(p<.05). MLSI, APSI, VSI, and DPSI of stair descending showed the increased stability index than that of stair ascending(p<.05), MLSI of RT types showed the decreased stability index than that of FT types(p<.05). Loading rate didn't show significant differences statistically according to the wearing positions of backpack(p>.05), but stair descending showed higher loading rate than that of stair ascending(p<.001). Also, ${\Delta}COPx$ in stair descending showed the increased movement than that of stair ascending(p<.05). Conclusions : A backpack of 10 kg(10 kg(ratio of body weights $17.61{\pm}1.17%$) showed significantly change GRF parameters according to wearing positions during stair ascending and descending. If possible, we suggest that the dynamic stability, in case of stairs walking with a smaller weights can be further improved.

• 한국정보과학회논문지:소프트웨어및응용
• /
• 제32권3호
• /
• pp.141-152
• /
• 2005

본 논문에서는 새로운 감성기반 영상검색방법을 제안한다. 서로 다른 색상, 명도, 도트크기를 나타내는 30개의 랜덤 패턴이 제시될 때 인간이 느끼는13가지 감성("like", "beautiful", "natural", "dynamic", "warm", "gay", "cheerful", "unstable", "light", "strong", "gaudy", "hard", "heavy") 평가 데이타로부터 질의 칼라코드와 질의 그레이코드로 명명한 질의 감성 표시자를 설계한다. 감성영상검색을 위해서 질의 감성을 선택하면 질의를 표현하는 칼라코드와 그레이코드가 선택되고 데이타베이스의 영상의 색상 정보를 나타내는 DB 칼라코드와 명도와 도트크기 정보를 나타내는 DB그레이코드값을 추출하여, 칼라코드간의 매칭과 그레이 코드간의 매칭을 통해 유사도를 판단한다. 또한 검색과정에 사용자의 의도를 반영하여 질의 칼라코드와 질의 그레이코드사이의 가중치와 칼라코드내의 가중치를 자동적으로 갱신하는 새로운 유사도 피드백 방법을 제안한다. 430개의 영상에 대해 실험한 결과 최초 질의에 대해 적합한 영상이 부적합한 영상보다 많았으며 유사도 피드백을 사용함에 따라 적합한 영상의 개수가 증가하였다.

• 한국정보통신학회논문지
• /
• 제26권10호
• /
• pp.1531-1536
• /
• 2022

본 논문은 직선상에 n개의 점들과 m개의 구간들이 주어 질 때, 모든 점들을 포함하는 구간들의 집합을 구하는 문제를 다룬다. 이러한 구간들의 집합을 점들의 구간 커버(interval cover)라고 부른다. 이 문제는 NP-hard 문제로 잘 알려진 집합 커버(set cover)의 특별한 경우이다. 이 문제의 최적화 기준으로 커버하는 구간 개수의 최소화, 점을 커버하는 구간이 1개인 점들의 개수 최대화 등을 생각할 수 있다. 본 논문에서는 구간에 가중치가 주어지는 경우, 각 점을 커버하는 구간들의 가중치 합을 그 점의 맴버쉽으로 정의한다. 그리고 점들의 맴버쉽의 최대값을 최소화하는 구간 커버를 찾는 문제를 연구한다. 동적계획법 설계를 이용하여, 이전 연구의 시간 복잡도 O(nm log n)를 개선하는 O(m²)시간 알고리즘을 제안한다.