• International Journal of Air-Conditioning and Refrigeration
• /
• v.11 no.3
• /
• pp.125-131
• /
• 2003

The fault detection and diagnosis technology may be applied in order to decrease the energy consumption and the maintenance cost of the air-conditioning system. In this paper, two fault detection methods were considered. One is a generic neural network, and the other is an moving average neural network. In order to compare the performance of fault detection results from these methods, two different types of faults in an air-conditioning system were applied. These are the condenser 30% fouling and the evaporator fan 25% slowdown. Test results showed that the moving average neural network was more effective for the detection of partial faults in the air-conditioning system.

• Journal of the Korean Society for Industrial and Applied Mathematics
• /
• v.22 no.2
• /
• pp.115-123
• /
• 2018

This paper discusses extended Kalman filter method for solving learning problems of multilayered neural networks. A lot of learning algorithms for deep layered network are sincerely suffered from complex computation and slow convergence because of a very large number of free parameters. We consider an efficient learning algorithm for deep neural network. Extended Kalman filter method is applied to parameter estimation of neural network to improve convergence and computation complexity. We discuss how an efficient algorithm should be developed for neural network learning by using Extended Kalman filter.

• Journal of the Korean Society for Nondestructive Testing
• /
• v.25 no.6
• /
• pp.475-482
• /
• 2005

The iterative calculation by least square method was used to determine the source location of acoustic emission in rock, as so called "traditional method". The results were compared with source coordinates infered from the application of neural network system for new input data, as so called "new method". Input data of the neural network were based on the time differences of longitudinal waves arrived from acoustic emission events at each transducer, the variation of longitudinal velocities at each stress level, and the coordinates of transducer as in the traditional method. The momentum back propagation neural network system adopted to determine source location, which consists of three layers, and has twenty-seven input processing elements. Applicability of the new method were identified, since the results of source location by the application of two methods were similarly concordant.

• Journal of the Korean Society for Precision Engineering
• /
• v.12 no.12
• /
• pp.30-37
• /
• 1995

The mechanical seals, which are installed in rotating machines like pump and compressor, are gengrally used as sealing devices in the many fields of industries. The failure of mechanical seals such as leakage,fast and severe wear, excessive torque, and squeaking results in big problems. To monitor the failure of mechanical seals and to propose the proper monitoring techniques with artificial neural network, sliding wear experiments were conducted. Torque and temperature of the mechanical seals were measured during experiments. Optical microstructure was observed for the wear processing after every 10 minute sliding at rotation speed of 1750 rpm and scanning electron microscopy was also observed. During the experiment, the variation of torque and temperature that meant an abnormal phenomenon, was observed. That experimental data recorded were applied to the developed monitoring system with artificial neural network. This study concludes that torque and temperature of mechanical seals wil be used to identify and to monitor the condition of sliding motion of mechanical seals. An availability to monitor the mechanical seal failure with artificial neural network was confirmed.

• Phonetics and Speech Sciences
• /
• v.8 no.3
• /
• pp.31-38
• /
• 2016

Speech inversion (acoustic-to-articulatory mapping) is not a trivial problem, despite the importance, due to the highly non-linear and non-unique nature. This study aimed to investigate the performance of Deep Neural Network (DNN) compared to that of traditional Artificial Neural Network (ANN) to address the problem. The Wisconsin X-ray Microbeam Database was employed and the acoustic signal and articulatory pellet information were the input and output in the models. Results showed that the performance of ANN deteriorated as the number of hidden layers increased. In contrast, DNN showed lower and more stable RMS even up to 10 deep hidden layers, suggesting that DNN is capable of learning acoustic-articulatory inversion mapping more efficiently than ANN.

• Proceedings of the KIEE Conference
• /
• 2006.04a
• /
• pp.9-11
• /
• 2006

This paper presents JPEG quantization table design using RD optimization and neural network. Using R-D optimization, quantization table with good performance can be obtained. However, it is time-consuming and difficult to adopt to embedded systems. In this paper, a new quantization table design method is proposed using R-D optimization and neural network. Neural network learns the quantization table obtained from R-D optimization and produces a quantization table for the Images. The proposed system is applied to Yale face data. From the simulation results, it has been shown that the proposed codec has better performance than JPEG.

• Proceedings of the KIEE Conference
• /
• 1997.11a
• /
• pp.397-401
• /
• 1997

In this paper, we propose a neural-network that detects moving objects in an image using a diffusion neural network. The proposed neural network is improved by adding a self loop to diffusion layer to remove the noise in an image and to reduce the detection of phantom edge. Computer simulation with real images show that the proposed neural network can extract edges of moving object efficiently.

• Journal of Electrical Engineering and Technology
• /
• v.3 no.4
• /
• pp.468-475
• /
• 2008

This paper presents a method for estimating the transient stability status of the power system using radial basis function(RBF) neural network with a fast hybrid training approach. A normalized transient energy margin(${\Delta}V_n$) has been obtained by the potential energy boundary surface(PEBS) method along with a time-domain simulation technique, and is used as an output of the RBF neural network. The RBF neural network is then trained to map the operating conditions of the power system to the ${\Delta}V_n$, which provides a measure of the transient stability of the power system. The proposed approach has been successfully applied to the 10-machine 39-bus New England test system, and the results are given.

• Journal of the Korean Institute of Illuminating and Electrical Installation Engineers
• /
• v.17 no.5
• /
• pp.29-35
• /
• 2003

In this paper, a new optimum design method is proposed for the linear induction motor(LIM). The Genetic Neural Network(GNN) is introduced in the optimum design of LIM and the simulation result is compared with the Genetic Algorithm(GA) and Neural Network(NN). The maximum thrust and trust/weight are selected as the object functions. The comparison showed that the proposed method is better than GA and NN.

• Journal of Mechanical Science and Technology
• /
• v.19 no.spc1
• /
• pp.255-264
• /
• 2005

In this paper, a practical bushing model is proposed to improve the accuracy of the vehicle dynamic analysis. The results of the rubber bushing are used to develop an empirical bushing model with an artificial neural network. A back propagation algorithm is used to obtain the weighting factor of the neural network. Since the output for a dynamic system depends on the histories of inputs and outputs, Narendra algorithm of 'NARMAX' form is employed to consider these effects. A numerical example is carried out to verify the developed bushing model. Then, a full car dynamic model with artificial neural network bushings is simulated to show the feasibility of the proposed bushing model.