• Journal of the Korean Institute of Electrical and Electronic Material Engineers
• /
• v.33 no.2
• /
• pp.105-108
• /
• 2020

In this study, experiments and simulations were conducted for a 1,200-V-class trench Si insulated-gate bipolar transistor (IGBT) with a small cell pitch below 2.5 ㎛. Presently, as a power device, the 1,200-V-class trench Si IGBT is used for automotives including electric vehicles, hybrid electric vehicles, and industrial motors. We obtained a breakdown voltage of 1,440 V, threshold of 6 V, and state voltage drop of 1.75 V. This device is superior to conventional IGBTs featuring a planar gate. To derive its electrical characteristics, we extracted design and process parameters. The cell pitch was 0.95 ㎛ and total wafer thickness was 140 ㎛ with a resistivity of 60 Ω·cm. We will apply these results to achieve fine-pitch gate power devices suitable for electrical automotive industries.

• The Journal of Korea Institute of Information, Electronics, and Communication Technology
• /
• v.5 no.1
• /
• pp.33-44
• /
• 2012

A class of recurrent neural networks is proposed and proven to be capable of identifying any discrete-time dynamical system. The applications of the proposed network are addressed in the encoding, identification, and extraction of finite state automata. Simulation results show that the identification of finite state automata using the proposed network, trained by the hybrid greedy simulated annealing with a modified error function in the learning stage, exhibits generally better performance than other conventional identification schemes.

• Journal of the Korean Society of Propulsion Engineers
• /
• v.14 no.6
• /
• pp.1-8
• /
• 2010

In this paper, we present some results of technical surveys, power analyses, and weight estimation on electric propulsion systems for 1-ton class Personal Air Vehicles(PAV) applications. When hybrid electric propulsion is adopted, its power performance using fuel cells and batteries in inferior to that of internal combustion engines. However, hybrid electric propulsion systems may replace IC engines when energy density and power density reach 0.75 kW$^*$hr/kg and 2.5 kW/kg, respectively.

• Journal of Satellite, Information and Communications
• /
• v.11 no.1
• /
• pp.1-5
• /
• 2016

In this paper, High Power Pulsed Amplifier with wide-band characteristic is implemented for L-band Navigational Aid(NAVAID). Due to the characteristics of L-Band NAVAID, implemented SSPA is demanded characteristics of high RF power, high linearity and high efficiency. Therefore, in this paper, efficiency characteristic is improved by modified class F technique. And linearity characteristic is improved by balance structure using hybrid coupler, $2^{nd}$ & $3^{rd}$ harmonic trap and anti-phase technique using non-linear characteristics of drive amplifier. Implemented SSPA shows that bandwidth of 300MHz, RF Output power of 1.5KW and efficiency of 55%.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2000.05a
• /
• pp.249-255
• /
• 2000

HRIV(Hybrid Rule-Interval Variation) method is presented to stabilize a class of nonlinear systems, where SMC(Sliding Mode Control) and ADC (ADaptive Control) schemes are incorporated to overcome the unstable characteristics of a conventional FLC(Fuzzy Logic Control). HRIV method consists of two modes: I-mode (Integral Sliding Mode PLC) and R-mode(RIV method). In I-mode, SMC is used to compensate for MAE(Minimum Approximation Error) caused by the heuristic characteristics of FLC. In R-mode, RIV method reduces interval lengths of rules as states converge to an equilibrium point, which makes the defined Lyapunov function candidate negative semi-definite without considering MAE, and the new uncertain parameters generated in R-mode are compensated by SMC. In RIV method, the overcontraction problem that the states are out of a rule-table can happen by the excessive reduction of rule intervals, which is solved with a dynamic modification of rule-intervals and a transition to I-mode. Especially, HRIV method has advantages to use the analytic upper bound of MAE and to reduce Its effect in the control input, compared with the previous researches. Finally, the proposed method is applied to stabilize a simple nonlinear system and a modified inverted pendulum system in simulation experiments.

• Korean Journal of Remote Sensing
• /
• v.37 no.4
• /
• pp.719-731
• /
• 2021

This study proposes a two-stage hybrid classification model for crop classification using multi-temporal remote sensing images; the model combines feature embedding by using an autoencoder (AE) with a convolutional neural network (CNN) classifier to fully utilize features including informative temporal and spatial signatures. Long short-term memory (LSTM)-based AE (LAE) is fine-tuned using class label information to extract latent features that contain less noise and useful temporal signatures. The CNN classifier is then applied to effectively account for the spatial characteristics of the extracted latent features. A crop classification experiment with multi-temporal unmanned aerial vehicle images is conducted to illustrate the potential application of the proposed hybrid model. The classification performance of the proposed model is compared with various combinations of conventional deep learning models (CNN, LSTM, and convolutional LSTM) and different inputs (original multi-temporal images and features from stacked AE). From the crop classification experiment, the best classification accuracy was achieved by the proposed model that utilized the latent features by fine-tuned LAE as input for the CNN classifier. The latent features that contain useful temporal signatures and are less noisy could increase the class separability between crops with similar spectral signatures, thereby leading to superior classification accuracy. The experimental results demonstrate the importance of effective feature extraction and the potential of the proposed classification model for crop classification using multi-temporal remote sensing images.

• The KSFM Journal of Fluid Machinery
• /
• v.15 no.4
• /
• pp.50-54
• /
• 2012

Wind-power generation, which is recently drawing attention as one of renewable energies across the world, has been developed mainly by Europe. As the demand for the wind-power generation rose and the amount of wind-power generation increased, the studies on megawatt-class wind-power system have been active, and the use of composite with such properties as less weight, more strength, anti-corrosion and environment-friendliness has required gradually. In other word, wind turbine tower will be required to be lighter, more reliable and more consistent. Therefore it is necessary to lose weight of the wind turbine tower. This points squarely toward hybrid/composite tower production growing. It is important to note however that hybrid/composite tower production as it is today is flawed and that there are ways to improve greatly on the performance of these towers in manufacturing process and in their in-service performance. Through this, we have some detail on the current process and its advantage of cost and weight of towers.

• Journal of electromagnetic engineering and science
• /
• v.15 no.3
• /
• pp.185-193
• /
• 2015

In this paper, a hybrid multiple input multiple output (MIMO) antenna for eight-band mobile handsets is designed and implemented. For the MIMO antenna, two hybrid antennas are laid symmetrically and connected by an interconnection tie, thereby enabling complementary operation. The tie affects both the impedance and radiation characteristics of each antenna. Further, printed circuit board (PCB) embedded type is applied to the antenna design. To verify the results of this study, we designed eight bands-LTE class 12, 13, and 14, CDMA, GSM900, DCS1800, PCS, and WCDMA-and implemented them on a bare board the same size as the real board of a handset. The voltage standing wave ratio (VSWR) is within 3:1 over the entire design band. Antenna isolation is less than -15 dB at the lower band, and -12 dB at the WCDMA band. Envelope correlation coefficient (ECC) of 0.0002-0.05 is obtained for all bands. The average gain and efficiency are measured to range from -4.69 dBi to -2.88 dBi and 33.99% to 51.5% for antenna 1, and -4.74 dBi to -2.97 dBi and 33.45% to 50.49% for antenna 2, respectively.

• Proceedings of the Korean Society for Noise and Vibration Engineering Conference
• /
• 2011.04a
• /
• pp.224-229
• /
• 2011

In this paper, in order to improve safety of hybrid electric vehicle a fault detection algorithm is introduced. The proposed algorithm uses SVDD techniques. Two methods for learning a lot of data are used in this technique. One method is to learn the data incrementally. Another method is to remove the data that does not affect the next learning. Using lines connecting support vectors selection of removing data is made. Using this method, lot of computation time and storage can be saved while learning many data. A battery data of commercial hybrid electrical vehicle is used in this study. In the study fault boundary via SVDD is described and relevant algorithm for virtual fault data is verified. It takes some time to generate fault boundary, nevertheless once the boundary is given, fault diagnosis can be conducted in real time basis.

• Proceedings of the Korean Society of Propulsion Engineers Conference
• /
• 2008.11a
• /
• pp.251-254
• /
• 2008

Ground firing test facility and test field for firing test of hybrid rocket engine were constructed. Ground firing test facility were composed of hybrid rocket engine, thrust stand, oxidizer storage/supply system, control system and data acquisition system. Firing tests of thrust 50 kgf class were conducted. Stable performance data was obtained and operational reliability of ground firing test facility were found.