• Proceedings of the Korean Vacuum Society Conference
• /
• 2014.02a
• /
• pp.211.2-211.2
• /
• 2014

Epitaxial complex oxide thin film heterostructures have attracted a great attention for their multifunctional properties, such as ferroelectricity, and ferromagnetism. Two dimensional electron gas (2DEG) confined at the interface between two insulating perovskite oxides such as LaAlO3/SrTiO3 interface, provides opportunities to expand various electronic and memory devices in nano-scale. Recently, it was reported that the conductivity of 2DEG could be controlled by external electric field. However, the switched conductivity of 2DEG was not stable with time, resulting in relaxation due to the reaction between charged surface on LaAlO3 layer and atmospheric conditions. In this report, we demonstrated a way to control the conductivity of 2DEG in non-volatile way integrating ferroelectric materials into LAO/STO heterostructure. We fabricated epitaxial Pb(Zr0.2Ti0.8)O3 films on LAO/STO heterostructure by pulsed laser deposition. The conductivity of 2DEG was reproducibly controlled with 3-order magnitude by switching the spontaneous polarization of PZT layer. The controlled conductivity was stable with time without relaxation over 60 hours. This is also consistent with robust polarization state of PZT layer confirmed by piezoresponse force microscopy. This work demonstrates a model system to combine ferroelectric material and 2DEG, which guides a way to realize novel multifunctional electronic devices.

• Proceedings of the IEEK Conference
• /
• 2001.06d
• /
• pp.21-24
• /
• 2001

In this paper, we present a fast face detection algorithm by estimating the eye region using neural network. To implement a real time face detection system, it is necessary to reduce search space. We limit the search space just to a few pairs of eye candidates. For the selection of them, we first isolate possible eye regions in the fast and robust way by modified histogram equalization. The eye candidates are paired to form an eye pair and each of the eye pair is estimated how close it is to a true eye pair in two aspects : One is how similar the two eye candidates are in shape and the other is how close each of them is to a true eye image A multi-layer perceptron neural network is used to find the eye candidate region's closeness to the true eye image. Just a few best candidates are then verified by eigenfaces. The experimental results show that this approach is fast and reliable. We achieved 94% detection rate with average 0.1 sec Processing time in Pentium III PC in the experiment on 424 gray scale images from MIT, Yale, and Yonsei databases.

• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.28 no.10
• /
• pp.1541-1548
• /
• 2004

A nano-stage simulation tool is developed for an advanced E-beam lithography system. Even if piezo-actuators are believed to be compatible fer the E-beam lithograpy system it is difficult to predict their characteristics due to their nonlinearities such as hysteresis and creep. In this paper, the nonlinear properties are modeled for a piezo-actuator by considering the voltage range and speed variations. The hysteresis is described as the first order differential equation with 24 sets of parameters and the creep is modeled as a time-dependent logarithmic function with 2 sets of a parameter. A two-axis nano stage with piezo-actuators are investigated for realizing nano scale motions. The characteristics of flexure guide mechanisms are analyzed based on the finite element method using the ANSYS software. The simulation tool for the nano stage is constructed by using the RecurDyn software. The dynamic response of the nano stage is obtained in simulations and compared with the experimental data.

• Physical Therapy Korea
• /
• v.28 no.2
• /
• pp.101-107
• /
• 2021

Background: Studies suggest that induced emotional changes can affect the sensory-motor system involved in the practice of muscle activity and movement in physical aspects. Previous studies have shown focused on effects just feedback on muscle activity associated with emotions but rarely have focused induced emotional change on gross motor function such as muscle activity. Objects: The purpose of this study was to compare biceps activity and emotion that before and after viewing a video was induced positive or negative emotion. Methods: The study enrolled 34 healthy male and female who scored at normal points on the Center for Epidermiological Studies-Depression Scale. The study measured over two weeks, showing subjects pleasant and sad videos one by one in a week. We performed to measure the biceps brachii activity which is maximal voluntary isometric contraction (MVIC) and the visual analog mood scale (VAMS) scores before and after one week. The significance level was set to α = 0.05. Results: There was no significant difference in muscle activity of the biceps brachii before and after each video was viewed (p > 0.05). However, the visual analogue mood scale showed an increase in VAMS after viewing each video (p < 0.05). Conclusion: We figured out induced emotional changes are cause actual emotional changes but there are no differences in muscle activity. In this research, watching the video with a short time looks like insufficient to change muscle activity. Nevertheless, there might be different when we check various muscles with sufficient time for viewing the video. Further study is needed to measure a variety of muscles with more time for viewing the video.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.15 no.8
• /
• pp.2993-3010
• /
• 2021

The job shop scheduling problem (JSSP) plays a critical role in smart manufacturing, an effective JSSP scheduler could save time cost and increase productivity. Conventional methods are very time-consumption and cannot deal with complicated JSSP instances as it uses one optimal algorithm to solve JSSP. This paper proposes an effective scheduler based on deep learning technology named self-supervised long-short term memory (SS-LSTM) to handle complex JSSP accurately. First, using the optimal method to generate sufficient training samples in small-scale JSSP. SS-LSTM is then applied to extract rich feature representations from generated training samples and decide the next action. In the proposed SS-LSTM, two channels are employed to reflect the full production statues. Specifically, the detailed-level channel records 18 detailed product information while the system-level channel reflects the type of whole system states identified by the k-means algorithm. Moreover, adopting a self-supervised mechanism with LSTM autoencoder to keep high feature extraction capacity simultaneously ensuring the reliable feature representative ability. The authors implemented, trained, and compared the proposed method with the other leading learning-based methods on some complicated JSSP instances. The experimental results have confirmed the effectiveness and priority of the proposed method for solving complex JSSP instances in terms of make-span.

• Advances in aircraft and spacecraft science
• /
• v.8 no.1
• /
• pp.69-85
• /
• 2021

A modified fuzzy mechanical control of large-scale multiple time delayed dynamic systems in states is considered in this paper. To do this, at the first level, a two-step strategy is proposed to divide a large system into several interconnected subsystems. As a modified fuzzy control command, the next was received as feedback theory based on the energetic function and the LMI optimal stability criteria which allow researchers to solve this problem and have the whole system in asymptotically stability. Modeling the Fisher equation and the temperature gauge for high-speed aircraft and spacecraft shows that the calculation method is efficient.

• Journal of Institute of Control, Robotics and Systems
• /
• v.21 no.1
• /
• pp.1-6
• /
• 2015

This paper investigates the longitudinal flight dynamics and stability of flapping-wing micro air vehicles. Periodic external forces and moments due to the flapping motion characterize the dynamics of this system as NLTP (Non Linear Time Periodic). However, the averaging theorem can be applied to an NLTP system to obtain an NLTI (Non Linear Time Invariant) system which allows us to use a standard eigen value analysis to assess the stability of the system with linearization around a reference point. In this paper, we investigate the dynamics and stability of a hawkmoth-scale flapping-wing air vehicle by establishing an LTI (Linear Time Invariant) system model around a hovering condition. Also, a direct time integration of full nonlinear equations of motion of the flapping-wing micro air vehicle is conducted to see how the longitudinal flight dynamics appear in the time domain beyond the reference point, i.e. hovering condition. In the study, the flapping-wing air vehicle exhibited three distinct dynamic modes of motion in the longitudinal plane of motion: two stable subsidence modes and one unstable oscillatory mode. The unstable oscillatory mode is found to be a combination of a pitching velocity state and a forward/backward velocity state.

• Journal of the Institute of Electronics Engineers of Korea SC
• /
• v.46 no.1
• /
• pp.28-33
• /
• 2009

In the research of speech recognition, locating the beginning and end of a speech utterance in a background of noise is of great importance. The conventional methods for speech endpoint detection are based on two simple time-domain measurements-short-time energy, and short-time zero-crossing rate, which couldn't guarantee the precise results if in the low signal-to-noise ratio environments. This paper proposes a novel approach that finds the Lyapunov exponent of time-domain waveform. This proposed method has no use for obtaining the frequency-domain parameters for endpoint detection process, e.g. Mel-Scale Features, which have been introduced in other paper. Accordingly, this algorithm is low complexity and suitable for Digital Isolated Word Recognition System.

• KIPS Transactions on Computer and Communication Systems
• /
• v.8 no.2
• /
• pp.35-40
• /
• 2019

Traditional Malware Detection is susceptible for detecting malware which is modified by polymorphism or obfuscation technology. By learning patterns that are embedded in malware code, machine learning algorithms can detect similar behaviors and replace the current detection methods. Data must collected continuously in order to learn malicious code patterns that change over time. However, the process of storing and processing a large amount of malware files is accompanied by high space and time complexity. In this paper, an HDFS-based distributed processing system is designed to reduce space complexity and accelerate feature extraction time. Using a distributed processing system, we extract two API features based on filtering basis, 2-gram feature and APICFG feature and the generalization performance of ensemble learning models is compared. In experiments, the time complexity of the feature extraction was improved about 3.75 times faster than the processing time of a single computer, and the space complexity was about 5 times more efficient. The 2-gram feature was the best when comparing the classification performance by feature, but the learning time was long due to high dimensionality.

• KSBB Journal
• /
• v.6 no.2
• /
• pp.135-141
• /
• 1991

This study thus looks into two treatment processess : i) Anaerobic digester coupled with hollow fibre membrane unit. Treatment of starch waste with anaerobic digester-membrane system was studied. $0.17\m^2$ area of hollow fibre membrane unit of known pore size was immersed into laboratory-scale anaerobic digestion system. The pore size of membrane was varied from 0.03 to $\0.15mu$m. The hydraulic retention time of anaerobic digester was varied from 1.5 to 10 days. The effect of hydraulic retention time on treatment efficiency was significant while effect of membrane size was not significant. The gas production was about 0.74㎥/kg COD treated. The COD removal efficient was about 80-95% depending on the hydraulic retention time. ii ) Crossflow ultrafiltration as post treatment to anaerobic filter. The effluent from anaerobic filter, which had a total COD in the range of 4,500-5,200 mg/L was treated by crossflow ultrafiltration units. The study conducted with different membrane pore size indicated that membrace with 1,000,000 molecular weight cut-off size gave a higher COD removal efficiency in the range of 83-87% while giving a study flux of $120-130 L/\m^2$.h. A study was conducted to see the long term clogging effect of membrane also.