• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2002.05a
• /
• pp.494-497
• /
• 2002

In this paper, we proposed a control method of an unknown nonlinear system using a dynamical neural network. The proposed method performs for a nonlinear system with unknown system, identification with using the dynamical neural network, and then a nonlinear adaptive controller is designed with these identified informations. In order to verify the effectiveness of the proposed method, we simulated one-link manipulator. The simulation results showed the effectiveness of using the dynamical neural network in the adaptive control of one-link manipulator.

• Journal of Navigation and Port Research
• /
• v.30 no.4
• /
• pp.259-265
• /
• 2006

This paper presents an improved adaptive neural network autopilot based on our previous study for track-keeping control of ships. The proposed optimal neural network controller can automatically adapt its learning rate and number of iterations. Firstly, the track-keeping control system of ships is described For the track-keeping control task, a way-point based guidance system is applied To improve the track-keeping ability, the off-track distance caused by external disturbances is considered in learning process of neural network controller. The simulations of track-keeping performance are presented under the influence of sea current and wind as well as measurement noise. The toolbox for track-keeping simulation on Mercator chart is also introduced.

• Proceedings of the KIEE Conference
• /
• 1995.07b
• /
• pp.798-800
• /
• 1995

The neural network predictive controller(NNPC) is proposed for the attempt to mimic the function of brain that forecasts the future. It consists of two loops, one is for the prediction of output(Neural Network Predictor) and the other one is for control the plant(Neural Network Controller). The output of NNC makes the control input of plant, which is followed by the variation of both plant error and prediction error. The NNP forecasts the future output based upon the current control input and the estimated control output. The method is applied to the control of temperature in boiler systems. The proposed NNPC is compared with the other conventional control methods such as PID controller, neural network controller with specialized learning architecture, and one-step-ahead controller. The computer simulation and experimental results show that the proposed method has better performances than the other methods.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2016.11a
• /
• pp.151-153
• /
• 2016

최근 스마트폰, PC, 태블릿 같은 전자기기들이 발전하면서 기계를 통해 소통하는 시대가 왔다. 기계와 소통하기 위해 우리가 사용하는 문자를 인식하는 것은 중요한 일이다. 이런 전자기기들이 문자, 영상인식을 해야 할 필요성이 더욱 증가함에 따라 머신러닝의 중요성이 대두되었다. 머신러닝은 컴퓨터의 학습을 위해 알고리즘과 기술을 개발하는 분야를 말한다. 머신러닝의 기법과 관련된 알고리즘의 종류는 수없이 많다. 그 중에서도 Neural Network는 사람의 뇌 신경구조를 토대로 착안하여 네트워크를 만들고 이를 학습에 이용한 머신러닝 기법이다. 이런 인공지능 알고리즘인 Neural Network 구조를 바탕으로 특징을 추출하여 학습을 하는 Convolution Neural Network 기법의 사용이 늘고 있다. 본 논문에서는 Neural Network와 Convolution Neural Network의 알고리즘을 이용한 필기체 인식 실험을 하고 그 내용을 비교하였다.

• Journal of the Korean Society of Manufacturing Technology Engineers
• /
• v.18 no.6
• /
• pp.681-689
• /
• 2009

Disk grinding was often applied to deburring process in order to enhance the final product quality. Inherent chamfering capability of the flexible disk grinding process in the early stage was analyzed with respect to various process parameters including workpiece length, wheel speed, depth of cut and feed. Initial chamfered edge defined as arc zone was characterized with local radius of curvature. Averaged radius and arc zone ratio was well evaluated using neural network system. Additional neural network analysis adding workpiece length showed enhance performance in predicting arc zone ratio and curvature radius with reduced error rate. A process condition design parameter was estimated using remaining input and output parameters with the prediction error rate lower than 2.0% depending on the relevant input parameter combination and neural network structure composition.

• Journal of the Korean Institute of Telematics and Electronics B
• /
• v.33B no.9
• /
• pp.135-143
• /
• 1996

In this paper, we popose a neural network that detects edges of moving objects in an image using a neural network of hierarchical spatial filter with orientation selectivity. We modify the temporal difference network by adding a self loop to each neuraon cell to reduce the problems of phantom edge detected by the neural network proposed by kwon yool et al.. The modified neural network alleviates the phantom edges of moving objects, and also can detect edges of miving objects even for the noisy input. By computer simulation with real images, the proposed neural netowrk can extract edges of different orientation efficiently and also can reduce the phantom edges of moving objects.

• Wind and Structures
• /
• v.15 no.1
• /
• pp.43-64
• /
• 2012

Many potential small wind turbine locations are near obstacles such as buildings and shelterbelts, which can have a significant, detrimental effect on the local wind climate. A neural network-based model has been developed which predicts mean wind speed and turbulence intensity at points in an obstacle's region of influence, relative to unsheltered conditions. The neural network was trained using measurements collected in the wakes of 18 scale building models exposed to a simulated rural atmospheric boundary layer in a wind tunnel. The model obstacles covered a range of heights, widths, depths, and roof pitches typical of rural buildings. A field experiment was conducted using three unique full scale obstacles to validate model predictions and wind tunnel measurements. The accuracy of the neural network model varies with the quantity predicted and position in the obstacle wake. In general, predictions of mean velocity deficit in the far wake region are most accurate. The overall estimated mean uncertainties associated with model predictions of normalized mean wind speed and turbulence intensity are 4.9% and 12.8%, respectively.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2002.12a
• /
• pp.179-182
• /
• 2002

In this paper, this system makes use of the analog sensor and converts the feature of fish analog signal when sensor is operating with CPU(80C196KC). Then, After signal processing, this feature Is classified a special feature and a outline of fish by using the neural network, one of the artificial intelligence scheme. This neural network classifies fish pattern of very simple and short calculation. This has linear activation function and the error backpropagation is used as a learning algorithm. And the neural network is learned in off-line process. Because an adaptation period of neural network is too long time when random initial weights are used, off-line learning Is induced to decrease the Progress time We confirmed this method has better performance than somewhat outdated machines.

• Proceedings of the KIEE Conference
• /
• 2001.07d
• /
• pp.2386-2389
• /
• 2001

This paper describes neural network steering controller for an AGV using MR sensor. The analytical magnetic fields model was compared with measured data and found to have less than 1 % difference. The neural network was also used to learn the steering behaviour of the AGV relative to the magnetic field values(Bx, By, Bz). A computer simulation of the AGV (including AGV's dynamics and steering) was used to verify the steering performance of the controller using the neural network. Good results were obtained. Also, the handmade AGV using neural network controller verified good results.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.58 no.10
• /
• pp.1968-1975
• /
• 2009

The paper proposes a novel control algorithm for tracking maximum power of PV generation system.. The maximum power of PV array is determinated by a insolation and temperature. Prior considered the term in PV generation system is how maximum power point(MPP) is accurately tracked.. The paper proposes a fuzzy neural network(FNN) control algorithm so as to accurately track those maximum power points. The proposed control algorithm comprises the antecedence part of fuzzy rule and clustering method, multi-layer neural network in the consequent part. FNN has the advantages which are depicted both high performance and robustness in fuzzy control and high adaptive control in neural network.. Specially, it can show the outstanding control performance for parameter variations appling to non-linear character of PV array. In this paper, the tracking speed and the accuracy prove the validity through comparing a proposed algorithm with a conventional one.