• Korean Journal of Materials Research
• /
• v.22 no.3
• /
• pp.155-158
• /
• 2012

In this work, AlON-$Al_2O_3$ coatings were prepared on Al2021 alloy by the electrolytic plasma processing (EPP) method. The experimental electrolytes include: 2 g/l NaOH as the electrolytic conductive agent, 10 g/l $Na_2AlO_2$ as the alumina formative agent, and 0.5 g/l $NaNO_2$, $NaNO_3$, and $NH_4NO_3$ as the nitride inducing agents. The effects of different nitrogen inducing agents were studied by a combined compositional and structural analyses of the ceramic coatings carried out by Xray diffractometry (XRD) and scanning electron microscopy (SEM) for the specimens EPP-treated at room temperature for 15 min under a hybrid voltage of 260 DC along with an AC 50 Hz power supply (200 V). Microhardness tests and wear tests were carried out to correlate the evolution of the microstructure and the resulting mechanical properties. Potentiodynamic polarizations and immersion corrosion tests were carried out in 3.5wt% NaCl water solutions under static conditions in order to evaluate the corrosion behavior of the coated samples. The results demonstrate that $NaNO_2$ is proven to be a good nitrogen inducing agent to produce high quality AlON-$Al_2O_3$ ceramic coatings.

• Journal of Intelligence and Information Systems
• /
• v.5 no.1
• /
• pp.103-123
• /
• 1999

Input filtering as a preprocessing method is so much crucial to get good performance in time series forecasting. There are a few preprocessing methods (i.e. ARMA outputs as time domain filters, and Fourier transform or wavelet transform as time-frequency domain filters) for handling time series. Specially, the time-frequency domain filters describe the fractal structure of financial markets better than the time domain filters due to theoretically additional frequency information. Therefore, we, first of all, try to describe and analyze specially some issues on the effectiveness of different filtering methods from viewpoint of the performance of a neural network based forecasting. And then we discuss about neural network model architecture issues, for example, what type of neural network learning architecture is selected for our time series forecasting, and what input size should be applied to a model. In this study an input selection problem is limited to a size selection of the lagged input variables. To solve this problem, we simulate on analyzing and comparing a few neural networks having different model architecture and also use an embedding dimension measure as chaotic time series analysis or nonlinear dynamic analysis to reduce the dimensionality (i.e. the size of time delayed input variables) of the models. Throughout our study, experiments for integration methods of joint time-frequency analysis and neural network techniques are applied to a case study of daily Korean won / U. S dollar exchange returns and finally we suggest an integration framework for future research from our experimental results.

• Journal of the Korea Institute of Military Science and Technology
• /
• v.15 no.3
• /
• pp.250-256
• /
• 2012

Unmanned ground vehicles may be operated by remote control unit through the wireless communication or autonomously. However, the autonomous technology is still challenging and not perfectly developed. For some reason or other, the wireless communication is not always available. If wireless communication is abruptly disconnected, the UGV will be nothing but a lump of junk. What was worse, the UGV can be captured by enemy. This paper suggests a method, autonomous return technology with which the UGV can autonomously go back to a safer position along the reverse path. The suggested autonomous return technology for UGV is based on multi-correlated information based DB creation and matching. While SUGV moves by remote-control, the multi-correlated information based DB is created with the multi-sensor information; the absolute position of the trajectory is stored in DB if GPS is available and the hybrid MAP based on the fusion of VISION and LADAR is stored with the corresponding relative position if GPS is unavailable. In multi-correlated information based autonomous return, SUGV returns autonomously based on DB; SUGV returns along the trajectory based on GPS-based absolute position if GPS is available. Otherwise, the current position of SUGV is first estimated by the relative position using multi-sensor fusion followed by the matching between the query and DB. Then, the return path is created in MAP and SUGV returns automatically based on the MAP. Experimental results on the pre-built trajectory show the possibility of the successful autonomous return.

• Journal of the Korean Society for Precision Engineering
• /
• v.29 no.11
• /
• pp.1190-1198
• /
• 2012

A hot forming of large thick Al plate using a grid-type hybrid die is a process to make a shell plate for the production of a spherical LNG tank. This process is characterized by using a grid-typed die with an additional air cooling system for reducing the cooling time of the heated plate after hot forming. The process consists of the plate's feeding, heating, forming and cooling in detail and each of them is continuously performed along the rail. This paper was designed to propose the analytical and experimental methods for determining the convection and interfacial heat transfer coefficients required in hot forming analysis of Al plate. These values in the analysis are to reproduce numerically the cooling performance of grid-typed die and cooling device. Interfacial heat transfer was obtained from the heat transfer experiments for different pressures and inverse analysis method. To verify the efficiency of the coefficient values obtained from above methods, FE analysis and experiment of the hot spherical-forming process were conducted for a small-scaled model. The convection coefficient was also calculated from flow analysis of air released by cooling device within grid-typed die using ANSYS-CFX.

• Proceedings of the Korean Society for Noise and Vibration Engineering Conference
• /
• 2012.04a
• /
• pp.667-671
• /
• 2012

Low frequency noises (up to about 200 Hz) mainly occur due to particular modes, resulting in booming noises, and in general the solutions may be found based on mode controls where conventional methods such as FEM can be used. However, at higher frequencies between 0.3~ 1 kHz, as the number of modes rapidly increase, radiation characteristics from structures, performances of damping sheets and sound packages may be more crucial rather than particular modes, and consequently the conventional FEM may be less practical in dealing with this kinds of structure-borne problems. In this context, so-called 'mid-frequency simulation model' based on FE-SEA hybrid method is studied and validated. Energy Transmission loss (i.e. air borne noise) is also studied. A dash panel component is chosen for this study, which is an important path that transfers both structure-borne and air borne energies into the cavity. Design modifications including structural modifications, attachment of damping sheets and application of different sound packages are taken into account and the corresponding noise characteristics are experimentally identified. It is found that the dash member behaves as a noise path. The damping sheet or sound packages have similar influences on both sound radiation and transmission loss. The comparison between experiments and simulations shows that this model could be used to predict the tendency of noise improvement.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.17 no.5
• /
• pp.157-163
• /
• 2017

The binary and interpolation search algorithms are the most famous among search area algorithms, the former running in $O(log_2n)$ on average, and the latter in $O(log_2log_2n)$ on average and O(n) at worst. Also, the interpolation search use only the probability of key value location without priori information. This paper proposes another search algorithm, which I term a 'hybrid block and interpolation search'. This algorithm employs the block search, a method by which MSB index of a data is determined as a block, and the interpolation search to find the exact location of the key. The proposed algorithm reduces the search range with priori information and search the reduced range with uninformed situation. Experimental results show that the algorithm has a time complexity of $O(log_2log_2n_i)$, $n_i{\simeq}0.1n$ both on average and at worst through utilization of previously acquired information on the block search. The proposed algorithm has proved to be approximately 10 times faster than the interpolation search on average.

• Journal of the Institute of Electronics Engineers of Korea SD
• /
• v.44 no.4
• /
• pp.1-9
• /
• 2007

Linearity is the one of the most important features for analog-optic transmission system. In our research, the available bandwidth for the feed-forward compensation circuit is enhanced by using a 180 hybrid coupler in the circuit. The bandwidth having the decreased 3rd-order intermodulation distortion(IMD3) over 10 dB is extended over 200 MHz with the center frequency of 1.6 GHz. We performed an efficient bandwith measurement for the feed-forward compensation system, which uses the network analyzer instead of the traditional measuring system that uses two RF signal generators and the spectrum analyzer. We identify the usefulness of this method from experimental results. In this study, we used cheap digital-purpose laser diodes for economical aspect, which proves the efficiency of the proposed analog system. The spurious-free dynamic range is improved about 6 dB/Hz.

• Convergence Security Journal
• /
• v.15 no.4
• /
• pp.43-55
• /
• 2015

The effective cyber capability assessment methodology does not exist, it's difficult to check the current state of the our country's cyber capabilities and to establish cyber capability enhancement plan based on the result of assessment. The cyber capability assessment is necessary in order to determine the current level of the country, establish policy directions and provide the basis for effective budgeting. But assessment of national cyber capability is multi-criteria decision makink problem. In this paper develops an assessment model based on the AHP and TOPSIS. AHP is used to determine weight of the criteria and TOPSIS method is used to obtain final ranking. We also introduce the result of four major nations cyber capability assessment using the proposed methodology. The experiment used the virtual experimental data. And the result show that the proposed methodology can expand the number of criteria and alternatives for assessment and provide more sophisticated mathematical base for objective assessment comparing methodology using AHP only.

• Composites Research
• /
• v.23 no.3
• /
• pp.13-18
• /
• 2010

One of the biggest challenges in the nano-field is how to effectively disperse nano-scale particles, especially CNTs, which are strongly agglomerated by intermolecular van der Waals forces. This study suggests a new method, discharge flocking, in order to disperse nano-scale particles effectively, which combines corona discharge phenomenon and a traditional electrostatic flocking process. In order to evaluate the discharge flocking process, composite specimens were fabricated by the process and RFI(resin film infusion) process, and then the mechanical and electrical properties of the specimens were measured and compared. Moreover, the evaluation of gas discharge effect on the CNTs and epoxy was performed to compare the mechanical and electrical properties of the composite specimens including the plasma treated CNTs. The experimental results showed that the electrical and mechanical properties of the specimens fabricated by the discharge flocking process were similar to those of the RFI process. In addition, plasma treated CNTs were not affected by gas discharge during the discharge flocking process.

• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.20 no.1
• /
• pp.115-126
• /
• 1996

In the time optimal control problem, bang-bang control has been used becaese it is the theoretical time minimum solution. However, to improve tracking speed performance in the time optimal control, it is important to select a switching point accurately which makes the velocity zero near the target track. But it is not easy to select the swiching point accurately because of the damping coefficient variation and uncertainties of modeling an actual system. The Adaptive model following control(AMFC) is implemented to relieve the difficulty and inconvenience of this task. The AMFC and make the controlled plant follow as closely as possible to a desired reference model whose switching point can be calculated easily and accurately, assuring the error between the states of the reference model and those of the controlled plant appoaches zero. The hybrid control method composed of AMFC and PID is applied to a tracking actuator of the magneto optical disk drive(MODD) in random access devices to improve its slow tracking performance. According to the simulaion and experimental results, the average tracking time as small as 20ms is obtained for a 3.5 magneto-optical disk drive. The AMFC also can be applied for other random access devices to improve the average tracking performance.