• Journal of the Korean Institute of Intelligent Systems
• /
• v.19 no.4
• /
• pp.574-579
• /
• 2009

There have been much difficulties to construct an optimized neural network in complex nonlinear regression problems such as selecting the networks structure and avoiding overtraining problem generated by noise. In this paper, we propose a stepwise constructive method for neural networks using a flexible incremental algorithm. When the hidden nodes are added, the flexible incremental algorithm adaptively controls the number of hidden nodes by a validation dataset for minimizing the prediction residual error. Here, the ELM (Extreme Learning Machine) was used for fast training. The proposed neural network can be an universal approximator without user intervene in the training process, but also it has faster training and smaller number of hidden nodes. From the experimental results with various benchmark datasets, the proposed method shows better performance for real-world regression problems than previous methods.

• Journal of the Society of Naval Architects of Korea
• /
• v.33 no.1
• /
• pp.40-53
• /
• 1996

In this paper, the wave drift damping is studied. An approximate method is adopted to calculate the wave drift damping for the sake of practical applications. By assuming the ship's forward speed to be low, the Green function and the velocity potential are expanded asymptotically with respect to the Brard number(${\tau}$) and terms up to the first order of ${\tau}$ are retained. Mean wave drift forces are computed straightforwardly. The wave drift damping is estimated as the change rate of the mean wave drift force with respect to the ship's speed. In order to validate the present method, Series 60(Cb=0.7) ship is exemplified for forward speed of Fn=0, 0.02 and 0.04. To predict the wave drift damping experimentally, three geosym models of the Esso-Osaka tanker are used. Also the effect of drift angle on the wave drift damping is also considered. Comparisons between numerical and experimental results show reasonable agreements.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.50 no.11
• /
• pp.90-98
• /
• 2013

The stereo system based on local matching is very popular due to its algorithmic simplicity, however it is limited to apply to various applications because it shows poor quality with low matching rates. In this paper, we propose and design a realtime stereo system based on an adaptive support-weight and the system shows low error rates and realtime performance. Generally, in the adaptive support-weight algorithm the intermediate computing results can not be reused to reduce the number of computations. In this research we modify the scheduling to reuse the intermediate results for the better performance by processing rows and columns separately. The nonlinear functions such as exponential or arc tangent have been designed with piecewise linear and step functions by empirical simulations and error analysis. The proposed architecture is composed of 9 processing elements for realtime performance. The proposed stereo system has been designed and synthesized using Donbu Hitek 0.18um standard cell library and can run up to 350Mhz operation frequency (33 frames per second) with 424K gates.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.27 no.7A
• /
• pp.711-723
• /
• 2002

A finer granular scalable (FGS) version of ISO/IEC MPEG-4 video streaming is investigated in this work with the prioritized stream delivery over loss-rate differentiated networks. Our proposed system is focused on the seamless integration of rate adaptation, prioritized packetization, and simplified differentiation for the MPEG-4 FGS video streaming. The proposed system consists of three key components: 1) rate adaptation with scalable source encoding, 2) content-aware prioritized packetization, and 3) loss-based differential forwarding. More specifically, a constant-quality rate adaptation is first achieved by optimally truncating the over-coded FGS stream based on the embedding rate-distortion (R-D) information (obtained from a piecewise linear R-D model). The rate-controlled video stream is then packetized and prioritized according to the loss impact of each packet. Prioritized packets are transmitted over the underlying network, where packets are subject to differentiated dropping and forwarding. By focusing on the end-to-end quality, we establish an effective working conditions for the proposed video streaming and the superior performance is verified by simulated MPEG-4 FGS video streaming.

• Journal of the Korean Institute of Telematics and Electronics
• /
• v.27 no.11
• /
• pp.49-54
• /
• 1990

In this paper, a simple model is presented for the gate-voltage dependence of the parasitic resistance in MOSFETs with the lightly-doped drain (LDD) structure. At the LDD region located under the gate electrode, an accumulation layer is formed due to the gate voltage. The parasitic resistance of the source side LDD in the channel is treated as a parallel combination of the resistance of the accumulation layer and that of the bulk LDD, which is approximated as a spreading resistance from the end of the channel inversion layer to the ${n^+}$/LDD junction boundary. Also the effects of doping gradients at the junction are discussed. As result of the model, the LDD resistance decreases with increasing the gate voltage at the linear regime, and increase quasi-linearly with the gate voltage at the saturation regime, considering th velocity saturation both in the channel and in the LDD region. The results are in good agreement with experimental data reported by others.

• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.37 no.12
• /
• pp.1529-1534
• /
• 2013

A voice coil actuator with a rotating moving magnet has been developed for a miniaturized mobile robot. The actuator has simple structure comprising a magnet, a coil, and a yoke. Actuator performance is predicted using a linearized theoretical model, and dynamic performance based on the air-gap between the magnet and the coil is predicted using motor constant and restoring constant obtained through finite element simulations. The theoretical model was verified using a prototype with 60 Hz resonance and 80 Hz bandwidth. We found that an input of 1.5 V can make the actuator rotate by $20^{\circ}$ statically. The driving configuration of the proposed actuator can be simplified because of its implementation of open-loop control.

• Journal of Biomedical Engineering Research
• /
• v.22 no.5
• /
• pp.403-412
• /
• 2001

The Present study analyzed the pressure-flow characteristics of a Korean shunt valve. Changes in the characteristic currie depending on the design parameters were also investigated. The Korean shunt valve used in the present study was constant pressure type and our analyses were validated through experiments. We applied fluid-structure interaction to solve the flow dynamic Problem because the small diaphragm in the valve was made from flexible silicone elastomers. Considering the material nonlinearity of the hyper-elastic material. the Mooney-Rivlin approximation was employed. The results of the numerical analyses were close to the experimental results The major Pressure drop was observed to happen in the small diaphragm. The slope of the pressure-flow characteristic curve was computed to be 0.37mm$H_2O$.hr/cc, which was similar to the average value of commercial shunt valves. 0.40mm$H_2O$.hr/cc. Therefore. our valves analyzed in the Present study showed a Proper Pressure control characteristics of the constant pressure type shunt valves. The opening pressure could be controlled by adjusting the amount of predeflection of the valve diaphragm. In order to obtain opening pressures of 25mm$H_2O$ and 80mm$H_2O$, respectively, and the required predeflection was found to be 10.2$\mu$m and 35.3$\mu$m. The flow orifice size was found to be within 10$\mu$m during valve operation Therefore, Precision design and manufacturing techniques are necessary for successful operations of the shunt valve. The study indicated the amount of predeflection as well as the magnitude of corner rounding of the diaphragm edge are important design parameters to influence the slope of the pressure-flow characteristic curve.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.15 no.7
• /
• pp.846-851
• /
• 2005

In this paper, we propose Genetic Algorithms(GAs)-based Optimized Polynomial Neural Networks(PNN). The proposed algorithm is based on Group Method of Data Handling(GMDH) method and its structure is similar to feedforward Neural Networks. But the structure of PNN is not fixed like in conventional neural networks and can be generated in a dynamic manner. As each node of PNN structure, we use several types of high-order polynomial such as linear, quadratic and modified quadratic, and it is connected as various kinds of multi-variable inputs. The conventional PNN depends on the experience of a designer that select the number of input variables, input variable and polynomial type. Therefore it is very difficult to organize optimized network. The proposed algorithm leads to identify and select the number of input variables, input variable and polynomial type by using Genetic Algorithms(GAs). The aggregate performance index with weighting factor is proposed as well. The study is illustrated with tile NOx omission process data of gas turbine power plant for application to nonlinear process. In the sequel the proposed model shows not only superb predictability but also high accuracy in comparison to the existing intelligent models.

• Korean Journal of Soil Science and Fertilizer
• /
• v.17 no.3
• /
• pp.212-217
• /
• 1984

Spatial variabilities of particle size distribution of 96 samples from Hwadong SiCL and Jungdong Sl were studied by using geostatistical concepts. The measurement was made at the nodes of the regular grid consisting of 12 rows and 8 columns. Sample spacing within rows and columns was 3 and 2 meters, respectively. The results are summarized as follows. 1. Variograms of Hwadong SiCL were fitted for the linear model and those of Jungdong SL for the spherical model. 2. Variograms of properties for Hwadong and clay for Jungdong showed the pure nugget effect. Those of silt and clay for Jungdong, however, appeared the nugget effect. 3. The minimum number of samples necessary to reproduce results similar to the true mean of the 96 measured values was approximately estimated. The minimum sample sizes of silt, clay, and sand in Hwadong SiCL were 27, 13, and 6, respectively. And the minimum sample size of clay in Jungdong SL was 17. 4. The approximate number of samples required to detect the difference of 5% of the true mean with 0.95 confidence level was estimated. The resulting number of samples for silt and sand in Jungdong was 14, and 26, respectively.

• Journal of Korean Society of Transportation
• /
• v.18 no.6
• /
• pp.89-99
• /
• 2000

To cope with the limits of conventional O-D trip matrix collecting methods, several approaches have been developed. One of them is bilevel Programming method Proposed by Yang(1995), which uses Sensitivity Analysis Based(SAB) algorithm to solve Generalized Least Square(GLS) problem. However, the SAB a1gorithm has revealed two critical short-comings. The first is that when there exists a significant difference between target O-D matrix and true O-D matrix, SAB algorithm may not produce correct solution. This stems from the heavy dependance on the historical O-D information, in special when gravel Patterns are dramatically changed. The second is the assumption of iterative linear approximation to original Problem. Because of the approximation, SAB algorithm has difficulty in converging to Perfect Stackelberg game condition. So as to avoid the Problems. we need a more robust and stable solution method. The main purpose of this Paper is to show the problem of the dependency of Previous models and to Propose an alternative solution method to handle it. The Problem of O-D matrix estimation is intrinsically nonlinear and nonconvex. thus it has multiple solutions. Therefore it is necessary to require a method for searching globa1 solution. In this paper, we develop a solution algorithm combined with genetic algorithm(GA) , which is widely used as probabilistic global searching method To compare the efficiency of the algorithm, SAB algorithm suggested by Yang et al. (1992,1995) is used. From the results of numerical example, the Proposed algorithm is superior to SAB algorithm irrespective of travel patterns.