• The Transactions of the Korea Information Processing Society
• /
• v.6 no.5
• /
• pp.1303-1311
• /
• 1999

In order to learn in dynamic environments, reinforcement learning algorithms like Q-learning, TD(0)-learning, TD(λ)-learning have been proposed. however, most of them have a drawback of very slow learning because the reinforcement value is given when they reach their goal state. In this thesis, we have proposed a reinforcement learning method that can approximate fast to the goal state in maze environments. The proposed reinforcement learning method is separated into global learning and local learning, and then it executes learning. Global learning is a learning that uses the replacing eligibility trace method to search the goal state. In local learning, it propagates the goal state value that has been searched through global learning to neighboring sates, and then searches goal state in neighboring states. we can show through experiments that the reinforcement learning method proposed in this thesis can find out an optimal solution faster than other reinforcement learning methods like Q-learning, TD(o)learning and TD(λ)-learning.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.31 no.4A
• /
• pp.311-321
• /
• 2011

To analysis the effect of reinforcement ratio and impact velocity on local damage, a series of impact analyses are performed to predict local effects. According to these results, the reinforcement ratio has no effect on the penetration depth and perforation thickness, but notable change to the scabbing area were observed. The higher the missile velocity becomes, the greater the degree of local damage to the reinforced concrete slabs is. Analysis results will be useful in the impact-resistance design of containment buildings and structures.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.9 no.6
• /
• pp.650-656
• /
• 1999

In this paper, we study on the convergence of the evolution strategies by introducing the Lamarckian evolution and the Baldwin effect, and propose a random local searching and a reinforcement local searching methods. In the random local searching method some neighbors generated randomly from each individual are med without any other information, but in the reinforcement local searching method the previous results of the local search are reflected on the current local search. From the viewpoint of the purpose of the local search it is suitable that we try all the neighbors of the best individual and then search the neighbors of the best one of them repeatedly. Since the reinforcement local searching method based on the Lamarckian evolution and Baldwin effect does not search neighbors randomly, but searches the neighbors in the direction of the better fitness, it has advantages of fast convergence and an improvement on the global searching capability. In other words the performance of the evolution strategies is improved by introducing the learning, reinforcement local search, into the evolution. We study on the learning effect on evolution strategies by applying the proposed method to various function optimization problems.

• Structural Engineering and Mechanics
• /
• v.38 no.4
• /
• pp.529-547
• /
• 2011

This paper presents the global-local finite cover method (GL-FCM) that is capable of analyzing structures involving local heterogeneities and propagating cracks. The suggested method is composed of two techniques. One of them is the FCM, which is one of the PU-based generalized finite element methods, for the analysis of local cohesive crack growth. The mechanical behavior evaluated in local heterogeneous structures by the FCM is transferred to the overall (global) structure by the so-called mortar method. The other is a method of mesh superposition for hierarchical modeling, which enables us to evaluate the average stiffness by the analysis of local heterogeneous structures not subjected to crack propagation. Several numerical experiments are conducted to validate the accuracy of the proposed method. The capability and applicability of the proposed method is demonstrated in an illustrative numerical example, in which we predict the mechanical deterioration of a reinforced concrete (RC) structure, whose local regions are subjected to propagating cracks induced by reinforcement corrosion.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.35 no.2
• /
• pp.265-274
• /
• 2015

The structural safety of nuclear power plant against impact from aircraft crash has been performed so far in two viewpoints such as local behavior and global behavior, and the local behavior has been evaluated using local damage evaluation formulas suggested based on the results of experimental data of RC (Reinforcement Concrete) wall. However, few data have been collected from recent research to evaluate the local behavior and damage of SC (Steel plate reinforced Concrete) wall, which is recently applied to the newly designed nuclear power plant. In this study, local damages of SC wall and RC wall against an idealized aircraft engine projectile impact are evaluated through FE simulation analyses with various wall thicknesses and steel ratio. Through analysis of local collision simulation results of SC and RC wall, the penetration depth of SC wall and RC wall are compared.

• Proceedings of the Korean Geotechical Society Conference
• /
• 2005.10a
• /
• pp.210-220
• /
• 2005

We have analyzed the stability for cut-slopes at main street 3-2 line section in Pa-ju local industrial complex. After studying an additional boring test, laboratory test and face mapping etc., we have determined the extent of reinforcement, slope inclinations and soil strength parameter from the analysis of test results. After changing the inclination of slopes for ground limit and carrying out the analysis of slope stability, we applied the Mass Nailing Method to the site because of need for reinforcement to soil and weathered rock slopes. In slope for soft and hard rock sections, we also reinforced the sections that are difficult to obtain the safety without reinforcement in alteration zone.

• Korean Journal of Materials Research
• /
• v.34 no.3
• /
• pp.125-137
• /
• 2024

Composite laminates are used in a wide range of applications including defense, automotive, aviation and aerospace, marine, wind energy, and recreational sporting goods. These composite beams still exhibit problems such as buckling, local deformations, and interlaminar delamination. To overcome these drawbacks, a novel viscoelastic autoclave bonding with tapered epoxy reinforcement polyurethane films is proposed. In existing laminates, compression face wrinkling and interlaminar delamination is caused in the sandwich beam. The unique viscoelastic autoclave spunbond interlayer bonding is designed to prevent face wrinkling and absorb and distribute stresses induced by external loads, thereby eliminating interlaminar delamination in the sandwich beam. Also, the existing special reinforcement causes stress concentrations, and the core is not effectively connected, which directly affects the stiffness of the beam. To address this, a novel tapered epoxy polyurethane reinforcement adhesive film is proposed, whose reinforcement thickness gradually tapers as it enters the core material. This minimizes stress concentrations at the interface, preventing excessive adhesive squeeze-out during the bonding process, and improves the stiffness of the beam. Results indicate the proposed model avoids the formation of micro cracks, interlaminar delamination, buckling, and local deformations, and effectively improves the stiffness of the beam.

• International Journal of Fuzzy Logic and Intelligent Systems
• /
• v.1 no.1
• /
• pp.125-130
• /
• 2001

In this paper, we propose a new type of evolution strategies combined with reinforcement learning. We use the variances of fitness occurred by mutation to make the reinforcement signals which estimate and control the step length of mutation. With this proposed method, the convergence rate is improved. Also, we use cauchy distributed mutation to increase global convergence faculty. Cauchy distributed mutation is more likely to escape from a local minimum or move away from a plateau. After an outline of the history of evolution strategies, it is explained how evolution strategies can be combined with the reinforcement learning, named reinforcement evolution strategies. The performance of proposed method will be estimated by comparison with conventional evolution strategies on several test problems.

• Steel and Composite Structures
• /
• v.11 no.2
• /
• pp.93-114
• /
• 2011

Flange and web local buckling in beam plastic hinge regions of steel moment frames can prevent beam-column connections from achieving adequate plastic rotations under earthquake-induced forces. This threat is especially valid for existing steel moment frame buildings with beams that lack adequate flange/web slenderness ratios. As the use of fiber reinforced polymers (FRP) have increased in strengthening and repair of steel members in recent years, using FRPs in stabilizing local instabilities have also attracted attention. Previous computational studies have shown that longitudinally oriented glass FRP (GFRP) strips may serve to moderately brace beam flanges against the occurrence of local buckling during plastic hinging. An experimental study was conducted at Izmir Institute of Technology investigating the effects of GFRP reinforcement on local buckling behavior of existing steel I-beams with flange slenderness ratios (FSR) exceeding the slenderness limits set forth in current seismic design specifications and modified by a bottom flange triangular welded haunch. Four European HE400AA steel beams with a depth/width ratio of 1.26 and FSR of 11.4 were cyclically loaded up to 4% rotation in a cantilever beam test set-up. Both bare beams and beams with GFRP sheets were tested in order to investigate the contribution of GFRP sheets in mitigating local flange buckling. Different configurations of GFRP sheets were considered. The tests have shown that GFRP reinforcement can moderately mitigate inelastic flange local buckling.

• Magazine of the Korea Concrete Institute
• /
• v.3 no.2
• /
• pp.115-121
• /
• 1991

Slippage of beam longitudinal reinforcement at beam-column connections is an important cause of damage to reinforced concrete frames under static and dynamic loads, This paper summarizes the results of an experimen¬tal study on the effects of confinements and compressive strength of concrete on the local bond stress-slip cha¬racteristics of deformed bars. I t is concluded from experimental results that, as far as the bond splittmg cracks are restrained by the vertical column reinforcement, confinement of concrete by transverse reinforcement has insignigicant direct effect on the local bond behavior. The ultimate bond strength, however, Increases pro¬portionally with the square root of concrete compressive strength. An empirical model was developed for local bond st ressslip relationslip of deformed bars in confined concrete of different compressive strengths.