• Structural Engineering and Mechanics
• /
• 제52권4호
• /
• pp.843-855
• /
• 2014

Concrete infill and reinforcement are one of the most well-known strengthening methods of structural elements. This study investigated flexural performance of concrete infill composite PHC pile (ICP pile) reinforced by infill concrete and longitudinal rebars in hollow PHC pile. A total four series of pile specimens were tested by four points bending method under simply supported conditions and investigated bending moment experimentally and analytically. From the test results, it was found that although reinforcement of infilled concrete on the pure bending moment of PHC pile was negligible, reinforcement of PHC pile using infilled concrete and longitudinal rebars increase the maximum bending moment with range from 1.95 to 2.31 times than that of conventional PHC pile. The error of bending moment between experimental results and predicted results by nonlinear sectional analysis on the basis of the conventional layered sectional approach was in the range of -2.54 % to 2.80 %. The axial compression and moment interaction analysis for ICP piles shows more significant strengthening effects of infilled concrete and longitudinal rebars.

• 한국게임학회 논문지
• /
• 제19권3호
• /
• pp.5-14
• /
• 2019

게임 서버는 분산 서버를 기본으로 하고 있다. 분산 게임서버는 서버의 작업 부하를 분산하기 위한 일련의 알고리즘에 의해 각 게임 서버의 부하를 일정하게 나누어서 클라이언트들의 요청에 대한 서버의 응답시간 및 서버의 가용성을 효율적으로 관리한다. 본 논문에서는 시뮬레이션 환경에서 기존 연구 방식인 Greedy 알고리즘과, Reinforcement Learning의 한 줄기인 Policy Gradient 중 PPO(Proximal Policy Optimazation)을 이용한 부하 분산 Agent를 제안하고, 시뮬레이션 한 후 기존 연구들과의 비교 분석을 통해 성능을 평가하였다.

• 한국군사과학기술학회지
• /
• 제24권1호
• /
• pp.123-132
• /
• 2021

In order for the underwater vehicle to perform various tasks, it is important to control the depth, course, and roll of the underwater vehicle. To design such a controller, it is necessary to construct a dynamic model of the underwater vehicle and select the appropriate hydrodynamic coefficients. For the controller design, since the dynamic model is linearized assuming a limited operating range, the control performance in the steady state is well satisfied, but the control performance in the transient state may be unstable. In this paper, in order to overcome the problems of the existing controller design, we propose a A2C(Advantage Actor-Critic) based roll controller for underwater vehicle with stable learning performance in a continuous space among reinforcement learning methods that can be learned through rewards for actions. The performance of the proposed A2C based roll controller is verified through simulation and compared with PID and Dueling DDQN based roll controllers.

• Structural Monitoring and Maintenance
• /
• 제9권3호
• /
• pp.271-287
• /
• 2022

This study investigates the bending behavior of a composite concrete slab roof with different methods of externally strengthing using steel plates and carbon fiber reinforced polymer (CFRP) strips. First, the concrete slab model which was reinforced with CFRP strips on the bottom surface of it is validated using experimental data, and then, using numerical modeling, 7 different models of square-shaped composite slab roofs are developed in ABAQUS software using the finite element modeling. Developed models include steel rebar reinforced concrete slab with variable thickness of CFRP and steel plates. Considering the control sample which has no external reinforcement, a set of 8 different reinforcement states has been investigated. Each of these 8 states is examined with 6 different uncertainties in terms of the properties of the materials in the construction of concrete slabs, which make 48 numerical models. In all models loading process is continued until complete failure occurs. The results from numerical investigations showed using the steel plates as an executive method for strengthening, the bending capacity of reinforced concrete slabs is increased in the ultimate bearing capacity of the slab by about 1.69 to 2.48 times. Also using CFRP strips, the increases in ultimate bearing capacity of the slab were about 1.61 to 2.36 times in different models with different material uncertainties.

• 한국군사과학기술학회지
• /
• 제25권4호
• /
• pp.390-400
• /
• 2022

Military deception is an action executed to deliberately mislead enemy's decision by deceiving friendly forces intention. In the lessons learned from war history, deception appears to be a critical factor in the battlefield for successful operations. As training using war-game simulation is growing more important, it is become necessary to implement military deception in war-game model. However, there is no logics or rules proven to be effective for CGF(Computer Generated Forces) to conduct deception behavior automatically. In this study, we investigate methodologies for CGF to learn and conduct military deception using Reinforcement Learning. The key idea of the research is to define a new criterion called a "deception index" which defines how agent learn the action of deception considering both their own combat objectives and deception objectives. We choose Korea Marine Corps Amphibious Demonstrations to show applicability of our methods. The study has an unique contribution as the first research that describes method of implementing deception behavior.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제16권9호
• /
• pp.3104-3123
• /
• 2022

The development of Vehicular Ad hoc Network (VANET) has greatly improved the efficiency and safety of social transportation, and the routing strategy for VANET has also received high attention from both academia and industry. However, studies on dynamic matching of routing policies with the message types of VANET are in short supply, which affects the operational efficiency and security of VANET to a certain extent. This paper studies the message types in VANET and fully considers the urgency and reliability requirements of message forwarding under various types. Based on the diversified types of messages to be transmitted, and taking the diversified message forwarding strategies suitable for VANET scenarios as behavioral candidates, an adaptive routing method for the VANET message types based on reinforcement learning (RL) is proposed. The key parameters of the method, such as state, action and reward, are reasonably designed. Simulation and analysis show that the proposed method could converge quickly, and the comprehensive performance of the proposed method is obviously better than the comparison methods in terms of timeliness and reliability.

• 인간식물환경학회지
• /
• 제24권6호
• /
• pp.651-662
• /
• 2021

Background and objective: This study identifies whether children's planning-organizing executive function can be significantly classified and predicted by home environment quality and wealth factors. Methods: For empirical analysis, we used the data collected from the 10th Panel Study on Korean Children in 2017. Using machine learning tools such as support vector machine (SVM) and random forest (RF), we evaluated the accuracy of the model in which home environment factors classify and predict children's planning-organizing executive functions, and extract the relative importance of variables that determine these executive functions by income group. Results: First, SVM analysis shows that home environment quality and wealth factors show high accuracy in classification and prediction in all three groups. Second, RF analysis shows that estate had the highest predictive power in the high-income group, followed by income, asset, learning, reinforcement, and emotional environment. In the middle-income group, emotional environment showed the highest score, followed by estate, asset, reinforcement, and income. In the low-income group, estate showed the highest score, followed by income, asset, learning, reinforcement, and emotional environment. Conclusion: This study confirmed that home environment quality and wealth factors are significant factors in predicting children's planning-organizing executive functions.

• Structural Engineering and Mechanics
• /
• 제82권6호
• /
• pp.757-770
• /
• 2022

For the design of flexural and shear crack control for reinforced concrete (RC) beams related to serviceability and reparability ensuring, eight simply-supported normal-strength reinforced concrete (NSRC) beam specimens are tested and the existing high-strength reinforced concrete (HSRC) experimental data are included in the investigation of this work. According to the investigation results of flexural and shear cracks, this works modifies the existing design formulas to determine the spacing of the tensile reinforcement for the flexural crack control of a HSRC/NSRC beam design. Additionally, for a specified shear crack width of 0.4 mm, the allowable stresses of the shear reinforcement are also identified. For the serviceability and reparability ensuring of HSRC/NSRC beams, this works proposes the relationship curves between the maximum flexural width and allowable stress of the tensile reinforcement, and the relationship curves between the shear crack width and allowable shear force that can be used to do the crack width control directly.

• Geomechanics and Engineering
• /
• 제28권1호
• /
• pp.1-10
• /
• 2022

In Korea, accidents related to the collapse of deteriorated aging reservoirs occur every year. The grouting method is generally applied to reinforce an aging reservoir. However, when using this method, different reinforcing effects appear depending on the ground conditions. Thus, new construction methods and materials capable of providing consistent reinforcing effects are required. In this study, the direct shear test (DST), model test, and simulation analysis were performed to evaluate the impact of surface stabilizers, generally used to reinforce roads, rivers, and slopes of roads, applied using surface curtain walls on aging reservoirs. The DST results indicate that when the surface stabilizer was mixed with in-situ soil, the increase in cohesion was the highest at a mixing ratio of 9%. No changes in the friction angle were evident; therefore, 9% was determined to be the optimal mixing ratio. In addition, the model test and simulation analysis showed that when 9% of the surface stabilizer was mixed and applied to the aging reservoir, the seepage quantity of water and the saturated area were reduced by approximately 42% and 73%, respectively. Moreover, the comprehensive analysis of results showed that the grouting method could be completely replaced by surface stabilizers applied through surface curtain walls because the technique could secure stability by decreasing the seepage in the aging reservoir.

• 로봇학회논문지
• /
• 제17권4호
• /
• pp.455-462
• /
• 2022

Since electric connectors such as power connectors have a small assembly tolerance and have a complex shape, the assembly process is performed manually by workers. Especially, it is difficult to overcome the assembly error, and the assembly takes a long time due to the error correction process, which makes it difficult to automate the assembly task. To deal with this problem, a reinforcement learning-based assembly strategy using contact states was proposed to quickly perform the assembly process in an unstructured environment. This method learns to generate a search trajectory to quickly find a hole based on the contact state obtained from the force/torque data. It can also learn the stiffness needed to avoid excessive contact forces during assembly. To verify this proposed method, power connector assembly process was performed 200 times, and it was shown to have an assembly success rate of 100% in a translation error within ±4 mm and a rotation error within ±3.5°. Furthermore, it was verified that the assembly time was about 2.3 sec, including the search time of about 1 sec, which is faster than the previous methods.