• Journal of Internet Computing and Services
• /
• v.21 no.5
• /
• pp.1-7
• /
• 2020

Deep neural networks(DNN), which are used as approximation functions in reinforcement learning (RN), theoretically can be attributed to realistic results. In empirical benchmark works, time difference learning (TD) shows better results than Monte-Carlo learning (MC). However, among some previous works show that MC is better than TD when the reward is very rare or delayed. Also, another recent research shows when the information observed by the agent from the environment is partial on complex control works, it indicates that the MC prediction is superior to the TD-based methods. Most of these environments can be regarded as 5-step Q-learning or 20-step Q-learning, where the experiment continues without long roll-outs for alleviating reduce performance degradation. In other words, for networks with a noise, a representative network that is regardless of the controlled roll-outs, it is better to learn MC, which is robust to noisy rewards than TD, or almost identical to MC. These studies provide a break with that TD is better than MC. These recent research results show that the way combining MC and TD is better than the theoretical one. Therefore, in this study, based on the results shown in previous studies, we attempt to exploit a random balance with a mixture of TD and MC in RL without any complicated formulas by rewards used in those studies do. Compared to the DQN using the MC and TD random mixture and the well-known DQN using only the TD-based learning, we demonstrate that a well-performed TD learning are also granted special favor of the mixture of TD and MC through an experiments in OpenAI Gym.

• Journal of the Korea institute for structural maintenance and inspection
• /
• v.16 no.3
• /
• pp.31-40
• /
• 2012

The irregularity of railway track affects not only the comfort of ride such as noise or vibration but also the safety of train operation. For this reason, it is an interesting research area to design a reliable and sustainable railway track system and to analyze the train movement mechanism based on systematic approaches considering reasons of track irregularity possible in a specific local environment. Irregularity data inspected by EM-120, an railway inspection system in Korea includes unavoidable incomplete and erratic information, so it is encountered lots of problem to analyse those data without appropriate pre-data-refining processes. In this research, for the efficient management and maintenance of railway system, progress rate of standard deviation of irregularity is quantified. During the computation, some important components of railways such as rail joint, ballast, roadbed, and fastener have been considered. Probabilistic distributions of irregularity growth with respect to time are computed to predict the remaining service life of railway track and to be adapted for the safety assessment.

• Journal of the Korean Institute of Traditional Landscape Architecture
• /
• v.33 no.3
• /
• pp.95-102
• /
• 2015

This study conducted the survey on the resources of traditional villages based on AHP in the subjects with the specialists in Korea and China to seek the resource evaluation standards to apply the preservation and development of traditional villages, and the differences of the importance on the resources among the specialists in both countries. We classified three levels of evaluation items to aim the deductions of the importance and priority in the resources of traditional villages. Upon the analysis results, natural resources were important in the level 1; environmental, historical, facility resources were important in the level 2; and the factors such as air, topography, traditional houses, agricultural landscape, shared community facilities, interchanges between urban and rural areas, family activities, and so on were important in the level 3. The factors that both Korean and Chinese groups evaluated as the most important ones were the same. In terms of overall importance by evaluation items, the factors such as air, water quality, noise, traditional houses, topography, shared community facilities, and so on were rated as relatively important in both Korean and Chinese groups. That is, the traditional villages have the necessity to preserve the cultural resources like their duties, however, it is required to control the natural environment with good quality preferentially. This study results can compare the importance on the resources of traditional village between Korea and China. Moreover, with calculation of the priority and scores for the preservation and management of traditional villages, they are expected to be used as the tool to apply the quantitative data in the evaluation process of traditional village resources in both countries.

• Journal of Convergence for Information Technology
• /
• v.10 no.7
• /
• pp.33-40
• /
• 2020

This work tackles the problem of conventional reinforcement learning scheme which has a relatively long training time to reduce energy consumption in underwater network. The enhanced scheme adjusts the learning range of reinforcement learning based on distance estimation. It can be reduce the scope of learning. To take account the fact that the distance estimation may not be accurate due to the underwater wireless network characteristics. this research added noise in consideration of the underwater environment. In simulation result, the proposed AQ-NAV scheme has completed learning much faster than existing method. AQ-NAV can finish the training process within less than 40 episodes. But the existing method requires more than 120 episodes. The result show that learning is possible with fewer attempts than the previous one. If AQ-NAV will be applied in Underwater Networks, It will affect energy efficiency. and It will be expected to relieved existing problem and increase network efficiency.

• The Bulletin of The Korean Astronomical Society
• /
• v.39 no.1
• /
• pp.49.1-49.1
• /
• 2014

The NISS (Near-infrared Imaging Spectrometer for Star formation history) onboard NEXTSat-1 is the near-infrared instrument onboard NEXTSat-1 which is being developed by KASI. The imaging low-resolution spectroscopic observation in the near-infrared range for nearby galaxies, low background regions, star-forming regions and so on will be performed on orbit. After the System Requirement Review, the optical design is changed from on-axis to the off-axis telescope which has a wide field of view (2 deg. ${\times}$ 2 deg.) as well as the wide wavelength range from 0.95 to $3.8{\mu}m$. The mechanical structure is considered to endure the launching condition as well as the space environment. The design of relay optics is optimized to maintain the uniform optical performance in the required wavelength range. The stray light analysis is being made to evade a light outside a field of view. The dewar is designed to operate the infrared detector at 80K stage. From the thermal analysis, we confirmed that the telescope can be cooled down to around 200K in order to reduce the large amount of thermal noise. Here, we report the current status of the NISS development.

• The Journal of Korean Institute of Electromagnetic Engineering and Science
• /
• v.9 no.4
• /
• pp.427-438
• /
• 1998

The bit error rate(BER) performance of Trellis Coded QPSK signal in the presence of cochannel interference (CCI) and Rician fading is investigated. The trellis coded QPSK system adopts Maximum Ratio Combining (MRC) diversity reception and Reed-Solomon code to enhance system performance. Using the derived error probability equation, the error performance of trellis coded QPSK system has been evaluated and shown in figures to discuss as a function of signal power to noise power ratio (SNR), signal power to interference power ratio(SIR), direct to indirect signal power ratio ($K_R$), the number of diversity branch (M), the frame length of Reed-Solomon code (n), the number of error correction symbol (t), and the number of state of trellis encoder. From the results, we know that proposed system is affected by cochannel interference and fading in microcell environment. Also, BER performance of Trellis Coded QPSK system can be improved as increasing either the power of desired signal or the value of SIR. And the BER floor in microcellular system is caused by the cochannel interference and it occurs at high BER when SIR is low. And Reed-Solomon code (n=15, t=2) is more effective to restrain the affection of CCI and fading than MRC diversity reception (M=2).

• Geophysics and Geophysical Exploration
• /
• v.18 no.1
• /
• pp.31-34
• /
• 2015

A gravity meter has been used for exploring subsurface mineral resources and monitoring long-period events such as Earth tides. Recently, researchers found several other intriguing features that we could even detect large teleseismic earthquakes and monitor seismic ambient noise using gravimeters. The zero-length spring suspension technology gives the gPhone (Micro-g LaCoste) excellent low frequency sensitivity, which may have implications for investigating much longer-period natural events (e.g., Earth's hum, tsunami waves, etc.). In this study, we present preliminary results through temporary operation of the gPhone at Geumsan in South Korea for 9 months (Nov. 2008-Jul. 2009). The gPhone successfully recorded large teleseismic events and showed a clear seasonal variation of the Double frequency microseisms during its operation period.

• Geophysics and Geophysical Exploration
• /
• v.13 no.4
• /
• pp.336-348
• /
• 2010

The Single-channel seismic survey with the source of bubble pulser and drilling survey was carried out in 2008 and 2009 for the site survey of Daecheon-Wonsando area, which was a proposed area of Korea-China subsea tunnel. The goal of this study is to analyze the depth and characteristics of acoustic basement for the stability assessment and tunnel design in this proposed area through combining drilling data with this single-channel seismic data after detailed processing. For this purpose, among the data processing schemes which are usually applied to multi-channel seismic data, we applied the F-K filtering to eliminate the AC(alternating current) noise and the post-stack depth migration to produce depth section. As a result, we verified that the improved depth section could be obtained from single-channel seismic data, and the distribution and characteristics of basement could be analyzed in survey area through the combined analysis with drilling data. However, we could not interpret the detailed structures, fault and fracture zone, due to the quality of bubble pulser source and single-channel data. We expect that those detailed structures can be analyzed when higher resolution seismic data is provided. Therefore, we recommend some items for future seismic survey of subsea tunnel to obtain the high resolution seismic data.

• The Journal of the Acoustical Society of Korea
• /
• v.39 no.4
• /
• pp.351-363
• /
• 2020

Launch vehicles are subject to airborne acoustic loads during atmospheric flight and these effects become pronounced especially in transonic region. As the vibration due to the acoustic loads can cause malfunction of payloads, it is essential to predict and reduce the acoustic loads. In this study, a complete process has been developed for predicting airborne vibro-acoustic environment inside the payload pairing and subsequent noise reduction procedure employing acoustic blankets and Helmholtz resonators. Acoustic loads were predicted by Reynolds-Averaged Navier-Stokes (RANS) analysis and a semi-empirical model for pressure fluctuation inside turbulent boundary layer. Coupled vibro-acoustic analysis was performed using VA One SEA's Finite Element Statistical Energy Analysis (FE-SEA) hybrid module and ANSYS APDL. The process has been applied to a hammerhead launch vehicle to evaluate the effect of acoustic load reduction and accordingly to verify the effectiveness of the process. The presently developed process enables to obtain quick analysis result with reasonable accuracy and thus is expected to be useful in the initial design phase of a launch vehicle.

• Geophysics and Geophysical Exploration
• /
• v.5 no.4
• /
• pp.299-308
• /
• 2002

We have derived an analytical expression for the sensitivity of the frequency domain small-loop electromagnetic (EM) surveys over a two-layer earth in order to estimate the depth of investigation with an instrument having the source-receiver separation of about 2 m. We analyzed the sensitivities to the lower layer normalized by those to the upper half-space and estimated the depth of investigation from the sensitivity analyses and the mutual impedance ratio. The computational results showed that the in-phase components of the sensitivity to the lower layer dominates those to the upper layer when the thickness of the upper layer is less than 20 m, while the quadrature components are not sensitive to the lower layer over the entire frequency range. Hence we confirmed that the accurate measurement of the in-phase component is essential to increase the depth of investigation in the multi-frequency small-loop EM survey. When conductive basement of 10 ohm-m underlies the upper layer of 100 ohm-m, an accurate measurement of the in-phase components ensures the depth of the investigation more than 10 m even accounting a noise effect, from which we conclude that the small-loop EM survey is quite effective in imaging the conductive plume down to a considerable depth. On the other hand, in the presence of the resistive basement of 1,000 ohm-m, the depth of investigation may not exceed 5 m considering the instrumental accuracy, which implies that the application of the small-loop EM survey is not recommended over the resistive environment other than detecting the buried conductor.