• Journal of the Korean Institute of Intelligent Systems
• /
• v.13 no.2
• /
• pp.200-208
• /
• 2003

This paper proposes a separation and feature extraction of image signals using the independent component analysis(ICA) based on neural networks of efficient learning rule. The proposed learning rule is a hybrid fixed-point(FP) algorithm based on secant method and momentum. Secant method is applied to improve the performance by simplifying the 1st-order derivative computation for optimizing the objective function, which is to minimize the mutual informations of the independent components. The momentum is applied for high-speed convergence by restraining the oscillation in the process of converging to the optimal solution. The proposed algorithm has been applied to the composite images generated by random mixing matrix from the 10 images of $512\times512$-pixel. The simulation results show that the proposed algorithm has better performances of the separation speed and rate than those using the FP algorithm based on Newton and secant method. The proposed algorithm has been also applied to extract the features using a 3 set of 10,000 image patches from the 10 fingerprints of $256\times256$-pixel and the front and the rear paper money of $480\times225$-pixel, respectively, The simulation results show that the proposed algorithm has also better extraction speed than those using the another methods. Especially, the 160 basis vectors(features) of $16\times16$-pixel show the local features which have the characteristics of spatial frequency and oriented edges in the images.

• Physical Therapy Korea
• /
• v.3 no.2
• /
• pp.95-105
• /
• 1996

Type I, II, III are regarded as "true" joint receptors, type IV is considered a class of pain receptor. Type I, II and III mechanoreceptors, via static and dynamic input, signal joint position, intraarticular pressure changes, and the direction, amplitude, and velocity of joint movements. Type I mechanoreceptor subserve both static and dynamic physiologic functions. Type I are found primarily in the stratum fibrosum of the joint capsule and ligaments. Type I receptors have a low threshold for activation and are allow to adapt to changes altering their firing frequency. Type II receptors have a low threshold for activation. These dynamic receptors respond to joint movement. Type II receptors are thus termed rapidly adapting. Type II joint receptors are located at the junction of the synovial membrane and fibrosum of the joint capsule and intraarticular and extraarticular fat pads. Type III receptors have been found in collateral ligaments of the joints of the extremities. Morphologically similar to Golgi tendon organ. These dynamic receptors have a high threshold to stimulation and are slowly adating. Type IV receptors possess free nerve ending that have been found in joint capsule and fat pads. They are not normally active, but respond to extreme mechanical deformation of the joint as well as to direct chemical or mechanical irritation. Small amplitude oscillatory and distraction movements(joint mobilization) techniques are used to stimulate the mechanoreceptors that may inhibit the transmission of nociceptors stimuli at the spinal cord or brain stem levels.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.3 no.3
• /
• pp.27-37
• /
• 1983

A method of dynamic analysis is developed for torsional free vibrations of elliptical-box girder type or stiffening truss system suspension bridge. In this study, the method based on a finite element technique using a digital computer, is illustrated by two numerical examples, the Namhae Bridge which is located in Kyungsang nam-do opened on June, 1973, and the Mt. Chunma Bridge is simple span pedestrian's suspension bridge with lateral bracing system in Mt. Chunma, Kyungki-do, are used. In general, dynamic modes of complex suspension bridges are three-dimensional in form, i.e., coupling between vertical and torsional motions. However, introduced that amplitudes of oscillation are infinitesimal for coincidence with the purpose of it's use, thereupon, the torsional vibration analyses are treated without coupling terms. A sufficient number of natural frequencies and mode shapes for torsional free vibration are presented in this paper. In the case of Mt. Chunma Bridge, the natural frequencies and periods are computed with and without reinforcement, respectively, and compared their discrepancies. The influence of the auxiliary reinforcing cables is prevailing in the first few modes, namely, 1st and 2nd in symmetric and 1st, 2nd and 3rd in antisymmetric vibration, and conspicuous in the symmetric compared with the antisymmetric motion, but in the higher modes, this kind of simple accessory elucidates rether converse effects. In the Namhae Bridge, the results are compared with the Manual's obtained by wind tunnel test. It reveals commendable agreement.

• KEPCO Journal on Electric Power and Energy
• /
• v.3 no.2
• /
• pp.79-88
• /
• 2017

In this paper, the wind tunnel test for the measurement of aerodynamic characteristics of transmission conductors with asymmetric sections is described. A single conductor model and bundled conductor models with ice accreted shapes are tested both in steady and turbulent flow, and the aerodynamic coefficients are acquired. Transmission conductor galloping is a kind of wind-induced vibration which is characterized by primarily vertical oscillation with a very low frequency and a high amplitude. It is well known that transmission conductor galloping is generally caused by moderately strong, steady winds when a transmission conductor has an asymmetric cross-section shaped by accreted ice. Galloping should be considered from the design stage of overhead lines because it can cause severe wear and fatigue damage to attachments as well as transmission conductors. It is reported that there have been normally 20 events of galloping per year in Korea, which may be followed by serious consequences in the electric power system. Therefore, this research is performed to measure aerodynamic characteristics of ice accreted transmission conductors to understand and control transmission conductor galloping so that it would help to prevent unexpected failures and reduce the maintenance costs caused by galloping.

• Progress in Superconductivity
• /
• v.8 no.1
• /
• pp.33-39
• /
• 2006

We have developed a second-order double relaxation oscillation SQUID(DROS) gradiometer with a baseline of 35 mm, and constructed a poorly magnetically-shielded room(MSR) with an aluminum layer and permalloy layers for magnetocardiography(MCG). The 2nd-order DROS gradiometer has a noise level of 20 $fT/{\surd}Hz$ at 1 Hz and 8 $fT/{\surd}Hz$ at 200 Hz inside the heavily-shielded MSR with a shielding factor of $10^3$ at 1 Hz and $10^4-10^5$ at 100 Hz. The poorly-shielded MSR, built of a 12-mm-thick aluminum layer and 4-6 permalloy layers of 0.35 mm thickness, is 2.4mx2.4mx2.4m in size, and has a shielding factor of 40 at 1 Hz, $10^4$ at 100 Hz. Our 64-channel second-order gradiometer MCG system consists of 64 2nd-order DROS gradiometers, flux-locked loop electronics, and analog signal processors. With the 2nd-order DROS gradiometers and flux-locked loop electronics installed inside the poorly-shielded MSR, and with the analog signal processor installed outside it, the noise level was measured to be 20 $fT/{\surd}Hz$ at 1 Hz and 8 $fT/{\surd}Hz$ at 200 Hz on the average even though the MSR door is open. This result leads to a low noise level, low enough to obtain a human MCG at the same level as that measured in the heavily-shielded MSR. However, filters or active shielding is needed fur clear MCG when there is large low-frequency noise from heavy air conditioning or large ac power consumption near the poorly-shielded MSR.

• Journal of the Korean Society for Aeronautical & Space Sciences
• /
• v.44 no.4
• /
• pp.343-351
• /
• 2016

Cryogenic spray characteristics of a nitrogen swirl injector operating in supercritical environment have been numerically investigated. By comparing the equation of states(EOS) used for supercritical condition, SRK EOS was applied to predict the nitrogen thermodynamic property under supercritical environment. A Chung's method was implemented for the calculation of viscosity and conductivity and Takahashi's correlation based on Fuller's Theorem was implemented for the calculation of diffusion coefficient. By injecting the nitrogen with 5 bar differential pressure into 50 bar chamber filled with nitrogen, numerical simulation has been conducted. The dynamic Smagorinsky sub-grid scale (SGS) model has been compared with the algebraic Smagorinsky SGS model using FFT frequency analysis. The instability at the liquid film and gas core inside injector and the propagation of pressure oscillation into the injector has been investigated. The spreading angle of swirl injector obtained by numerical calculation has been validated with experimental result.

• Journal of the Korean GEO-environmental Society
• /
• v.15 no.5
• /
• pp.23-29
• /
• 2014

So far, the grouting using pressure injection has been extensively used to avoid adverse effects such as soil disturbance. Whereas, the pressure injection to the limitations of the diffusion range, so that the kinks would last injection of cement particles by introducing a frequency oscillation effect improved injection method have been recently developed. In this study, a pilot test was performed to compare injection effects of the both methods. The injections using both methods were tested on the embankment which consists of core clay and weathered soil. Subsequently, the injected volume, SPT N values, in-situ permeability and electrical resistivity were measured to compare their effects. The vibration method showed more effective permeation comparing with the pressure method. Also, it showed more homogeneously improved ground than the existing method. For SPT results, the vibration method presented increase of mean N value as much as 17.4 % comparing with the conventional method. Higher electrical resistivity was presented in case of injecting with vibration method and it indicated the injection was extensively completed. Finally, it is expected that the economic feasibility will be improved by decrease of drilling spacing, when the existing method is replaced with vibration method.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.33 no.2
• /
• pp.585-594
• /
• 2013

In this study, laboratory experiments and numerical simulations were conducted to test the performance of resonance power buoy system proposed by Kweon et al.(2010). The system is composed of a linear generator and a mooring buoy. The mover of the linear generator mainly has heave motion driven by vertical oscillation of the buoy. In this system, the velocity discrepancy between the mover and the buoy makes electricity. However, ocean wave energy as a natural resource around Korean peninsula is comparatively small and the driving force for producing electricity is not enough for commercialization. Therefore, it is necessary that the buoy motion be amplified by using resonance characteristics. In order to verify the resonance effects on the test power buoy, the experimental investigations were conducted in the large wave flume (length of 110 m, width of 8 m, maximum depth of 6 m) equipped with regular and random plunger wave generator. The resonance draft of test power buoy is designed for the corresponding period of incident wave, 1.96 sec. Regular wave test results show that the heave response amplitude operator(RAO) by a test buoy has the amplification of 5.66 times higher compared to the wave amplitude at the resonance period. Test results of random waves show that the buoy has the largest spectrum area of 20.73 times higher at the point of not the resonance period but the shorter one of 1.85 sec. Therefore this study suggests the resonance power buoy for wave power generation for commercial application in the case of the coastal and oceanic area with smaller wave energy.

• The Journal of Korean Academy of Orthopedic Manual Physical Therapy
• /
• v.2 no.1
• /
• pp.9-19
• /
• 1996

Type I, II, III are regarded as "true" joint receptors, type IV is considered a class of pain receptor. Type I, II and III mechanoreceptors, via static and dynamic input, signal joint position, intraarticular pressure changes, and the direction, amplitude, and velocity of joint movements. Type I mechanoreceptor subserve both static and dynamic physiologic functions. Type I are found primarily in the stratum fibrosum of the joint capsule and ligaments. Type I receptors have a low threshold for activation and are allow to adapt to changes altering their firing frequency. Type II receptors have a low threshold for activation. These dynamic receptors respond to joint movement. Type II receptors are thus termed rapidly adapting. Type II joint receptors are located at the junction of the synovial membrane and fibrosum of the joint capsule and intraarticular and extraarticular fat pads. Type III receptors have been found in collateral ligaments of the joints of the extremities. Morphologically similar to Golgi tendon organ. These dynamic receptors have a high threshold to stimulation and are slowly adating. Type IV receptors possess free nerve ending that have been found in joint capsule and fat pads. They are not normally active, but respond to extreme mechanical deformation of the joint as well as to direct chemical or mechanical irritation. Small amplitude oscillatory and distraction movements(joint mobilization) techniques are used to stimulate the mechanoreceptors that may inhibit the transmission of nociceptors stimuli at the spinal cord or brain stem levels.

• The Journal of Korean Institute of Electromagnetic Engineering and Science
• /
• v.22 no.3
• /
• pp.399-405
• /
• 2011

In this paper, we present the design of Q-band LC VCO and injection locking buffer for 77 GHz automotive radar sensor using 130 nm RF CMOS process. To improve the phase noise characteristic of LC tank, the transmission line is used. The negative resistance by the active device cross-coupled pair of buffer is used for high output power, with or without oscillation of buffer. The measured phase noise is -102 dBc/Hz at 1 MHz offset frequency and tuning range is 34.53~35.07 GHz. The output power is higher than 4.1 dBm over entire tuning range. The fabricated chip size is $510{\times}130\;um^2$. The power consumption of LC VCO is 10.8 mW and injection locking buffer is 50.4 mW from 1.2 V supply.