• Proceedings of the KIEE Conference
• /
• 2004.11c
• /
• pp.693-695
• /
• 2004

Among modem control method, the observer is being used widely because it has the advantage of the guarantee of reliability on financial problem, over heating, and physical shock. However, an existing state observer and a sliding observer have such problems that an experimenter needs to know dynamics and parameters of the system. And also, the high gain observer has such a problem that it has transient state at the beginning of the observation. In this paper, the neuro observer is proposed to improve these problems. The proposed observer complement a problem that occur from increase of gain of High-gain observer in proportion to the square number of observable state variables. And also, the proposed observer can tune the gain obtained by differentiating observational error at transient state automatically by using the backpropagation training method to stabilize the observational speed. To prove a performance of the proposed observer, it is simulated that the comparison between the state estimate performance of the proposed observer and that of sliding, high gain observer is made by using a sinusoidal input to the observer which consists of four layers in stable 2nd order system.

• The Transactions of the Korean Institute of Electrical Engineers B
• /
• v.53 no.3
• /
• pp.188-195
• /
• 2004

This paper presents neural load torque observer that is used to deadbeat load torque observer and gain compensation by parameter estimator As a result, the response of the PMSM(permanent magnet synchronous motor) follows that nominal plant. The load torque compensation method is composed of a neural deadbeat observer To reduce the noise effect, the post-filter implemented by MA(moving average) process, is adopted. The parameter compensator with RLSM (recursive least square method) parameter estimator is adopted to increase the performance of the load torque observer and main controller The parameter estimator is combined with a high performance neural load torque observer to resolve the problems. The neural network is trained in on-line phases and it is composed by a feed forward recall and error back-propagation training. During the normal operation, the input-output response is sampled and the weighting value is trained multi-times by error back-propagation method at each sample period to accommodate the possible variations in the parameters or load torque. As a result, the proposed control system has a robust and precise system against the load torque and the Parameter variation. A stability and usefulness are verified by computer simulation and experiment.

• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.24 no.4 s.175
• /
• pp.858-865
• /
• 2000

It is well recently recognized that quench is one of the serious problems for the integrity of superconducting magnets, which is mainly attribute to the rapid temperature rising in the magnet due to some extrinsic factors such as conductor motion, crack initiation etc. In order to apply acoustic emission(AE)echnique effectively to monitor and diagnose superconducting magnets, it is essential to identify the sources of acoustic emission. In this paper, an acoustic emission technique has been used to monitor and diagnose quenching phenomenon in racetrack shaped superconducting magnets at cryogenic environment of 4.2K. For these purposes special attention was paid to detect AE signals associated with the quench of superconducting magnets. The characteristics of AE parameters have been analyzed by correlating with quench number, winding tension of superconducting coil and charge rate by transport current. In addition, the source location of quench in superconducting magnet was also discussed on the basis of correlation between magnet voltage and AE energy.

• Journal of the Korean Society of Physical Medicine
• /
• v.15 no.1
• /
• pp.1-9
• /
• 2020

PURPOSE: Many patients with stroke have difficulties in walking with foot-drop. Various types of ankle-foot orthoses (AFOs) have been developed, but their weight needs to be reduced with the assistance of the ankle dorsiflexor. Therefore, an elastic AFO (E-AFO) was devised that not only improves the stability and flexibility of the ankle but also assists with ankle dorsiflexion while walking. This study examined the effects of an E-AFO, on the walking patterns of foot-drop patients with stroke. METHODS: Fourteen patients walked with and without an E-AFO, and the gait parameters were assessed using the GAITRite system. The spatiotemporal data on the gait patterns of stroke patients with foot-drop were compared using paired t-tests; the level of statistical significance was set to α<.05. RESULTS: No significant differences were observed in the velocity (p=.066) and affecte+d step length (p=.980), but the affected and less-affected stance (p=.022, p=.002) and swing time (p=.012, p=.005) were significantly different. The E-AFO produced a significant difference in the less-affected step length (p=.032). CONCLUSION: The E-AFO has a significant effect on the walking patterns of individuals with foot-drop and stroke. The E-AFO could be a useful assistive device for gait training in stroke patients.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.8 no.8
• /
• pp.1201-1206
• /
• 2013

Active Appearance Model(AAM) is one of the facial feature extraction techniques. In this paper, we propose the Multiple Active Appearance Model(MAAM). Proposed method uses two AAMs. Each AAM trains using different training parameters. It causes that each AAM has different strong points. One AAM complements the weak points in the other AAM. We performed the facial feature extraction on the 100 images to verify the performance of MAAM. Experiment results show that MAAM gives more accurate results than AAM with less fitting iteration.

• Journal of the Korean Institute of Illuminating and Electrical Installation Engineers
• /
• v.25 no.4
• /
• pp.45-56
• /
• 2011

This paper is proposed artificial neural network(ANN) rotor resistance estimation of induction motor drive controlled by multi-adaptive fuzzy learning controller(AFLC). A simple double layer feedforward ANN trained by the back-propagation technique is employed in the rotor resistance identification. In this estimator, double models of the state variable estimations are used; one provides the actual induction motor output states and the other gives the ANN model output states. The total error between the desired and actual state variables is then back propagated to adjust the weights of the ANN model, so that the output of this model tracks the actual output. When the training is completed, the weights of the ANN correspond to the parameters in the actual motor. The estimation and control performance of ANN and multi-AFLC is evaluated by analysis for various operating conditions. Also, this paper is proposed the analysis results to verify the effectiveness of this controller.

• Journal of Navigation and Port Research
• /
• v.32 no.7
• /
• pp.521-527
• /
• 2008

The Istanbul Strait is one of the important waterways in the world. And its southern entrance has a highly congested local traffic. Till now there are several studies regarding how the Istanbul Strait is dangerous to navigate and how those dangers can be mitigated. But there is no study regarding local traffic which is posing great collision risk. In a certain traffic area, marine traffic safety assessment parameters are traffic volume, frequency of collision avoidance maneuver, traffic density, traffic flow and potential encounter, In this paper local traffic volume, traffic flow and potential encounter number of local traffic vessels and possibility of collision are investigated in order to find degree of danger at the southern entrance of the Istanbul Strait. Finally by utilizing those, risky areas are determined for southern entrance of the Istanbul Strait. Results have been compared to a previous study regarding risk analysis at congested areas of the Istanbul Strait (Aydogdu, 2006) and consistency of the results were presented.

• Journal of the Korean Society of Manufacturing Process Engineers
• /
• v.12 no.3
• /
• pp.7-12
• /
• 2013

In Die casting process, the problem of die degradation is often issued. In oder to increase of die life the material degradation of die steel was investigated using test core pins. Three test core pins were positioned in front of the gate entry and observed washout and soldering resistance during Mg die casting process. The test parameters are set as different commercial die materials, coatings condition and hardness of die surface. Usign 220t magnesium die casting machine was employed to cast AZ91 magnesium alloys. After 150 shots, macroscopic observation of die surface was carried out. Additional 50 cycles later, test pins were chemically cleaned with 5% HCl aqueous solution to find out the existence of washout and soldering layers. Microstructural characterization of die surface and the die roughness measurement were performed together. Computational simulation using AnyCasting program was also beneficial to correlate the extent of die damage with the position of test pin inside die cavity. As results, the optimal combination of die steel with productive coating as well as its hardness was drawn out. it will be helpful to decide the material and condition considering increasing of tool life.

• Journal of Biomedical Engineering Research
• /
• v.28 no.1
• /
• pp.29-35
• /
• 2007

This study was designed to investigate the feasibility to utilize the electromyogram (EMG) for estimating the muscle torque. The muscle torque estimation plays an important role in functional electrical stimulation because electrical stimulation causes muscles to fatigue much faster than voluntary contraction, and the stimulation intensity should then be modified to keep the muscle torque within the desired range. We employed the neural network method which was trained using the major EMG parameters and the corresponding knee extensor torque measured and extracted during isometric contractions. The experimental results suggested that (1) our neural network algorithm and protocol was feasible to be adopted in a real-time feedback control of the stimulation intensity, (2) the training data needed to cover the entire range of the measured value, (3) different amplitudes and frequencies made little difference to the estimation quality, and (4) a single input to the neural network led to a better estimation rather than a combination of two or three. Since this study was done under a limited contraction condition, the results need more experiments under many different contraction conditions, such as during walking, for justification.

• Computers and Concrete
• /
• v.10 no.6
• /
• pp.649-662
• /
• 2012

This paper presents the results of an investigation on the compressive strength and weight loss of mortars containing three types of fillers as cement replacements; Limestone Filler (LF), Silica Fume (SF) and Trass (TR), subjected to elevated temperatures including $400^{\circ}C$, $600^{\circ}C$, $800^{\circ}C$ and $1000^{\circ}C$. Results indicate that addition of TR to blended cements, compared to SF addition, leads to higher compressive strength and lower weight loss at elevated temperatures. In order to model the influence of the different parameters on the compressive strength and the weight loss of specimens, artificial neural networks (ANNs) were adopted. Different diagrams were plotted based on the predictions of the most accurate networks to study the effects of temperature, different fillers and cement content on the target properties. In addition to the impressive RMSE and $R^2$ values of the best networks, the data used as the input for the prediction plots were chosen within the range of the data introduced to the networks in the training phase. Therefore, the prediction plots could be considered reliable to perform the parametric study.