• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2022.10a
• /
• pp.9-11
• /
• 2022

Applying Reinforcement Learning in everyday applications and varied environments has proved the potential of the of the field and revealed pitfalls along the way. In robotics, a learning agent takes over gradually the control of a robot by abstracting the navigation model of the robot with its inputs and outputs, thus reducing the human intervention. The challenge for the agent is how to implement a feedback function that facilitates the learning process of an MDP problem in an environment while reducing the time of convergence for the method. In this paper we will implement a reward shaping system avoiding sparse rewards which gives fewer data for the learning agent in a ROS environment. Reward shaping prioritizes behaviours that brings the robot closer to the goal by giving intermediate rewards and helps the algorithm converge quickly. We will use a pseudocode implementation as an illustration of the method.

• Steel and Composite Structures
• /
• v.8 no.6
• /
• pp.457-473
• /
• 2008

Yielding of the internal steel reinforcement is an important mechanism that influences the Intermediate Crack-induced debonding (IC debonding) behavior in FRP-strengthened RC members since the FRP is required to carry additional forces beyond the condition of steel yielding. However, rational design practice dictates an appropriate limit state is defined when steel yielding is assured prior to FRP debonding. This paper proposes a criterion which correlates the occurrence of IC debonding to the formulation of a critical steel yielding length. Once this length is exceeded the average bond stress in the FRP/concrete interface exceeds its threshold value, which proves to correlate with the average bond resistance in an FRP/concrete joint under simple shear loading. This proposed IC debonding concept is based on traditional sections analysis which is conventionally applied in design practice. Hence complex bond stress-slip analyses are avoided. Furthermore, the proposed model incorporates not only the bond properties of FRP/concrete interface but also the beam geometry, and properties of steel and FRP reinforcement in the analysis of IC debonding strength. Based upon a solid database, the validity of the proposed simple IC debonding criterion is demonstrated.

• Steel and Composite Structures
• /
• v.48 no.5
• /
• pp.599-610
• /
• 2023

To avoid debonding failure, a novel type of hybrid anchorage (HA) is proposed in this study that uses a slotted plate to lock the ends of the fiber-reinforced polymer (FRP) sheet in addition to the usual bonding over the substrate of the strengthened member. An experimental investigation was performed on three groups of RC beams, which differed from one another in either concrete strength or steel reinforcement ratio. The test results indicate that the end self-locking of the CFRP sheet can improve the failure ductility, ultimate capacity of the beams and its utilization ratio. Although intermediate debonding occurred in all the strengthened beams, it was not a fatal mode of failure for the three specimens with end anchorage. Among them, FRP rupture occurred in the beam with higher concrete strength and lower steel reinforcement ratio, whereas the other two failed by concrete crushing. The beam strengthened by HA obtained a relatively high percentage of increase in ultimate capacity when the rebar ratio or concrete strength decreased. The expressions in the literature were inspected to calculate the critical loads at intermediate debonding, FRP rupturing and concrete crushing after debonding for the strengthened beam. Then, the necessity of further research is addressed.

• Magazine of the Korea Concrete Institute
• /
• v.7 no.4
• /
• pp.169-179
• /
• 1995

The purpose of this study is to spread beam plastic hinging zone of the high-strength($f'_c=700kg/cm^2$) reinforced concrete beam-column joints away from the column face by vertically anchored intermediate reinforcements. The newly proposed intermediate reinforcements which are vertically anchored by interlinking each intermediate rebars are tested to insure the ductile behavior of R /C beam-column joins. Main variable is the shape of intermediate reinforcements. From the test results, the newly proposed intermediate rebar detail can move arid expand the beam plastic hinging zone about 1.Od from column face and can delay the strength decay of the high-strength R /C beam-column joint. Also energy dissipation capacity of specimen IV-1.OD10 which is reinforced by vertically anchored intermediate rebars about 1.0d is 1.6 times as high as the specimen CM-STAN which is designed by ACI318-89.

• Journal of the Korea Concrete Institute
• /
• v.17 no.1 s.85
• /
• pp.51-58
• /
• 2005

Moment frames have been widely used in building construction. In current design codes, concrete moment frames are classified into ordinary, intermediate, and special moment resisting concrete frames (OMRCF, IMRCF, SMRCF)). The objective of this study is to investigate the seismic behavior of columns in ordinary moment resisting concrete frames (OMRCF) and intermediate moment resisting concrete frames (IMRCF). For this purpose 3 story OMRCF and IMRCF buildings were designed and detailed in compliance to ACI 318 (2002) and KCI (1999). In this study the buildings were assumed to be located in seismic zone 1 classified by UBC (1997). This study considered the columns in the 1st story since these columns shall resist the largest axial and lateral forces during an earthquake. Eight 2/3 scale column specimens were made for representing the upper part and lower part of exterior and interior columns of the OMRCF and the IMRCF Quasi-static reversed cyclic loading was applied to each specimen with a constant or varying axial load. Test results show that seismic behaviors of columns are influenced by existence of lap splices, axial force levels, and lateral reinforcement at possible plastic hinging region. However, the effect of such variables strongly co-related to each other.

• Journal of the Computational Structural Engineering Institute of Korea
• /
• v.20 no.6
• /
• pp.699-708
• /
• 2007

Reinforced concrete deep beams are general structural members used as transfer-girder, pile cap, foundation wall and so on. They have a complex stess formation. Generally, failure mechanisms differ from either continuous deep beams or simple supported deep beams. In continuous deep beams, a negative moment is occurred over intermediate support and the location of maximum moment coincide with high shear force. Therefore, failure usually occurs at this region. While on the other hand, in simple supported deep beam, the region of high shear coincides with the region of low moment. The web opening of deep beams for accepting a facility makes shear behaviors of deep beams more complex and gives rise to an expansion of crack around the opening and a decline of shear capacity of deep beams. Therefore, Engineers must determine a delicate reinforcement method to control a crack and increase a shear capacity. The purpose of this report is a computation of an effective reinforcement method through non-linear finite element method by means of adopting various reinforcement method as variables and a computation of shear capacity formula taking an effectiveness of reinforcement into consideration.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.23 no.8
• /
• pp.1011-1017
• /
• 2019

This paper proposes reinforcement learning to solve the node-disjoint path problem which establishes multipath for reliable data transmission in wireless ad-hoc networks. The node-disjoint path problem is a problem of determining a plurality of paths so that the intermediate nodes do not overlap between the source and the destination. In this paper, we propose an optimization method considering transmission distance in a large-scale wireless ad-hoc network using Q-learning in reinforcement learning, one of machine learning. Especially, in order to solve the node-disjoint path problem in a large-scale wireless ad-hoc network, a large amount of computation is required, but the proposed reinforcement learning efficiently obtains appropriate results by learning the path. The performance of the proposed reinforcement learning is evaluated from the viewpoint of transmission distance to establish two node-disjoint paths. From the evaluation results, it showed better performance in the transmission distance compared with the conventional simulated annealing.

• Journal of the Korea Concrete Institute
• /
• v.26 no.2
• /
• pp.181-188
• /
• 2014

In this study, an interactive analysis considering column-pile interaction is performed on the basis of an equivalent base spring model for supplementing virtual fixed point design of bent pile structures. Through this analytical method, the application of the minimum steel reinforcement ratio of the pile (0.4%) is analyzed by taking into account the major influencing parameters. Furthermore, the limit depth for steel reinforcement ratio is proposed through the relationships between column and pile conditions. To obtain the detailed information, it is found that an interactive analysis is intermediate in theoretical accuracy between the virtual fixed point model analysis and full-modeling analysis. Base on this study, it is also found that the maximum bending moment is located within cracking moment of the pile when material nonlinearity is considered. Therefore, the minimum steel reinforcement ratio is appropriately applicable for the optimal design of bent pile structures. Finally, the limit depth for steel reinforcement ratio ($L_{As=x%}$) is proposed by considering the field measured results. It is shown that the normalized limit depth ratio for steel reinforcement ratio ($L_{As=x%}/L_P$) decreases linearly as the length-diameter ratio of pile ($L_P/D_P$) increases, and then converges at a constant value.

• Land and Housing Review
• /
• v.1 no.1
• /
• pp.67-73
• /
• 2010

Method of protruding steel bar embedded in PHC pile for connecting with foundation plate is an intermediate form of fixed and hinged connection and has often been used in architectural structures such as apartment complex. However, mechanical properties of this method have not been proved and its construction process is not simple. In this study, therefore, by analyzing previous research and by considering ratio of steel bar and concrete in PHC pile, which is minimum reinforcement of rebar, the newly optimized method of reinforcing joint of PHC pile and foundation plate is suggested with respect to PHC pile type (PHC 450, PHC 500, and PHC 600). To assess mechanical properties (ultimate tensile and shear strength) of joint of PHC pile and foundation plate, full scale experimental tests are performed. As a result, all cases are satisfied with required design criteria and can be practically applied. Our results indicate that reduction of rebar reinforcement compared to previous method would lead cost saving in PHC pile construction.

• International Journal of Highway Engineering
• /
• v.19 no.3
• /
• pp.31-43
• /
• 2017

PURPOSES : The objective of this study is to verify the effect of fiber grid reinforcement on the long-term performance of asphalt pavement overlaid on old concrete pavement by performing field investigation, laboratory test, and finite element analysis. METHODS : The reflection cracking, roughness, and rutting of fiber grid reinforced overlay sections and ordinary overlay sections were compared. Cores were obtained from both the fiber grid reinforced and ordinary sections to measure bonding shear strength between the asphalt intermediate and asphalt overlay layers. Fracture energy, displacement after yield, shear stiffnesses of the cores were also obtained by analyzing the test results. Finite element analysis was performed using the test results to validate the effect of the fiber grid reinforcement on long-term performance of asphalt pavement overlaid on the old concrete pavement. The fatigue cracking and reflection-cracking were predicted for three cases: 1) fiber grid was not used; 2) glass fiber grid was used; 3) carbon fiber grid was used. RESULTS : The reflection-cracking ratio of fiber grid reinforced sections was much smaller than that of ordinary sections. The fiber grid reinforcement also showed reduction effect on rutting while that on roughness was not clear. The reflection-cracking was not affected by traffic volume but by slab deformation and joint movement caused by temperature variation. The bonding shear strength of the fiber grid reinforced sections was larger than that of the ordinary sections. The fracture energy, displacement after yield, and shear stiffnesses of the cores of the fiber grid reinforced sections were also larger than those of the ordinary sections. Finite element analysis results showed that fatigue cracking of glass or carbon fiber grid reinforced pavement was much smaller than that of ordinary pavement. Carbon fiber grid reinforcement showed larger effect in elongating the fatigue life of the ordinary overlay pavement compared to glass fiber grid reinforcement. The binder type of the overlay layer also affected the fatigue life. The fiber grid reinforcement resisted reflection-cracking and the carbon fiber grid showed the greater effect. CONCLUSIONS :The results of field investigation, laboratory test, and finite element analysis showed that the fiber grid reinforcement had a better effect on improving long-term performance of asphalt pavement overlaid on the old concrete pavement.