• Journal of the Korea Computer Graphics Society
• /
• v.29 no.4
• /
• pp.7-16
• /
• 2023

For decades, creating a desired locomotive motion in a goal-oriented manner has been a challenge in character animation. Data-driven methods using generative models have demonstrated efficient ways of predicting long sequences of motions without the need for explicit conditioning. While these methods produce high-quality long-term motions, they can be limited when it comes to synthesizing motion for challenging novel scenarios, such as punching a random target. A state-of-the-art solution to overcome this limitation is by using a GAN Discriminator to imitate motion data clips and incorporating reinforcement learning to compose goal-oriented motions. In this paper, our research aims to create characters performing combat sports such as boxing, using a novel reward design in conjunction with existing GAN-based approaches. We experimentally demonstrate that both the Adversarial Motion Prior [3] and Adversarial Skill Embeddings [4] methods are capable of generating viable motions for a character punching a random target, even in the absence of mocap data that specifically captures the transition between punching and locomotion. Also, with a single learned policy, multiple task controllers can be constructed through the TimeChamber framework.

• Journal of the Korean Society of Environmental Restoration Technology
• /
• v.13 no.3
• /
• pp.73-83
• /
• 2010

A technology using continuous fiber soil reinforcement system for the creation of ecological restoration in a damaged area has been developed and introduced. The continuous fiber soil reinforcement system (Geofiber system) is an environmentally friendly slope protection technique that continuous fiber soil reinforced layers are constructed with green plantation on cut slope. The characteristics of this system in terms of the strength and hydraulic performance, and the vegetation were investigated in this study. The main objectives of this comparative study was to quantify the potential contribution of geofiber system for the revegetation on the cut slope in a damaged area. A Geofiber system was constructed to reinforce the lower layer of slopes and revegetation methods including wood chips were carried out on the upper layer by machineries. The results of monitoring during 3 years on cut slopes were as follows : 1) All the quadrat existed in the proper range for vegetation. 2) Species richness was 4.4 (site-1) and 18.5 (site-2) respectively. 3) The averaged coverage rates of quadrats was 90%. It is remarkable that the continuous fiber soil reinforcement system would be capable of applying to a damaged area and also would serve maintaining a healthier environment for floras. However, it behooves to continue monitoring on succession of vegetation for ecological restoration.

• Journal of Korean Society of Steel Construction
• /
• v.10 no.2 s.35
• /
• pp.303-315
• /
• 1998

Four types of repair procedures are applied to the fatigue cracked crane runway gilders, which are stop-holes as crack arrester stop-holes reinforced with high strength bolts, welding repair and reinforcement with high strength bolted splices. The fatgiue cracks are reinitiated at the region where stop-holes and weld repairments are applied, while none of the cracks are observed in the cases of stop-holes reinforcement and reinforcement with high strength bolted splices. When using stop-holes and hole-reinforcement all repaired regions show a same fatigue strength to the one before the repairments. The experiments also reveal that the proper weldment is an essential factor when applying the welding repairement as a properly welding produces the same level of fatigue strength after the repairement. When the situation permits to use reinforcement with high strength boilted splices, the experiments shows the repairment is the best possible method among the procedures available.

• Journal of Korean Society of Steel Construction
• /
• v.20 no.4
• /
• pp.539-548
• /
• 2008

This study aims to determine the applicability of the previously published T-type modular profile beam in the manner of producing specimens designed specially for the said purpose, determining their bending and shear behaviors depending on the presence of shear reinforcement, and analyzing the results in comparison with the theoretical equation of plastic deformation. The modular profile beam contributes to bending and shear resistance with the addition of the profile to the form function, and enhances the molding performance through the modular concept. The experimental results showed that the TS series specimens with shear reinforcement have bending behaviors superior to those of the T series specimens without shear reinforcement, which suggests that the used shear reinforcement appropriately bears the shear force. However, it was considered that all the specimens except for the T1-1 specimen failed to have adequate bending performance because of the intermodular slipping caused by the shear failure of the bolts. It is expected that further studies on the T-type modular profile beam, in which shear connectors will be considered as a variable,be performed to develop optimal intermodular connection methods.

• KIPS Transactions on Software and Data Engineering
• /
• v.10 no.4
• /
• pp.143-152
• /
• 2021

It is very important to learn behavioral policies that allow multiple agents to work together organically for common goals in various real-world applications. In this multi-agent reinforcement learning (MARL) environment, most existing studies have adopted centralized training with decentralized execution (CTDE) methods as in effect standard frameworks. However, this multi-agent reinforcement learning method is difficult to effectively cope with in a dynamic environment in which new environmental changes that are not experienced during training time may constantly occur in real life situations. In order to effectively cope with this dynamic environment, this paper proposes a novel multi-agent reinforcement learning system, C-COMA. C-COMA is a continual learning model that assumes actual situations from the beginning and continuously learns the cooperative behavior policies of agents without dividing the training time and execution time of the agents separately. In this paper, we demonstrate the effectiveness and excellence of the proposed model C-COMA by implementing a dynamic mini-game based on Starcraft II, a representative real-time strategy game, and conducting various experiments using this environment.

• Journal of The Geomorphological Association of Korea
• /
• v.25 no.4
• /
• pp.37-47
• /
• 2018

Coastal sand dunes have been regarded as natural defenses to protect hinterland from disasters such as storm surge and typhoons. However, many dunes are not well-deserved in South Korea because of imprudent land development or inappropriate measures after coastal erosion. Lately, beach nourishment and dune reinforcement are emphasized as the effective and environmentally sustainable solution for the coastal protection. They are regarded good strategies to keep landscapes for a time, with little side effects. However, there is little knowledge on the construction methods including proper design and time plans for the best results.In addition, the effects of dune reinforcement in the field should be tested.In thisstudy, we performed sand filling in an eroded dune scarp and surveyed topographic changes in the beach-dune system, which is located along Dasa-ri coast, Chungnam Province, South Korea. Using a network RTK-GPS and drone-based aerial photographs, we analyzed the temporal and spatial changes in the area, before and after the reinforcement. As a result, the dune reinforcement seems to be helpful to mitigates the coastal erosion and to prevent the coastline retreat at least for one year.

• Journal of the Korea Society of Computer and Information
• /
• v.26 no.3
• /
• pp.9-17
• /
• 2021

In this paper, an autonomous multi-agent path planning using reinforcement learning for monitoring of infrastructures and resources in a computationally distributed system was proposed. Reinforcement-learning-based multi-agent exploratory system in a distributed node enable to evaluate a cumulative reward every action and to provide the optimized knowledge for next available action repeatedly by learning process according to a learning policy. Here, the proposed methods were presented by (a) approach of dynamics-based motion constraints multi-agent path-planning to reduce smaller agent steps toward the given destination(goal), where these agents are able to geographically explore on the environment with initial random-trials versus optimal-trials, (b) approach using agent sub-goal selection to provide more efficient agent exploration(path-planning) to reach the final destination(goal), and (c) approach of reinforcement learning schemes by using the proposed autonomous and asynchronous triggering of agent exploratory phases.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.14 no.11
• /
• pp.4268-4289
• /
• 2020

Advanced traffic signal timing method plays very important role in reducing road congestion and air pollution. Reinforcement learning is considered as superior approach to build traffic light timing scheme by many recent studies. It fulfills real adaptive control by the means of taking real-time traffic information as state, and adjusting traffic light scheme as action. However, existing works behave inefficient in complex intersections and they are lack of feasibility because most of them adopt traffic light scheme whose phase sequence is flexible. To address these issues, a novel adaptive traffic signal timing scheme is proposed. It's based on actor-critic reinforcement learning algorithm, and advanced techniques proximal policy optimization and generalized advantage estimation are integrated. In particular, a new kind of reward function and a simplified form of state representation are carefully defined, and they facilitate to improve the learning efficiency and reduce the computational complexity, respectively. Meanwhile, a fixed phase sequence signal scheme is derived, and constraint on the variations of successive phase durations is introduced, which enhances its feasibility and robustness in field applications. The proposed scheme is verified through field-data-based experiments in both medium and high traffic density scenarios. Simulation results exhibit remarkable improvement in traffic performance as well as the learning efficiency comparing with the existing reinforcement learning-based methods such as 3DQN and DDQN.

• Geomechanics and Engineering
• /
• v.33 no.6
• /
• pp.553-565
• /
• 2023

As the structure of broken rock mass is complex, with obvious discontinuity and anisotropy, it is generally necessary to reinforce broken rock mass using grouting in underground construction. The purpose of this study is to experimentally investigate the mechanical properties of broken rock mass after grouting reinforcement with consideration of the characteristics of broken rock mass (i.e., degree of fragmentation and shape) and a range of reinforcement methods such as relative strength ratio between the broken rock mass and cement-based grout stone body (λ), and volumetric block proportion (VBP) representing the volumetric ratio of broken rock mass and the overall cement grout-broken rock mass mixture after the reinforcement. The experimental results show that the strength and deformation of the reinforced broken rock mass is largely determined by relative strength ratio (λ) and VBP. In addition, the enhancement in compressive strength by grouting is more obvious for broken rock mass with spherical shape under a relatively high strength ratio (e.g., λ=2.0), whereas the shape of rock mass has little influence when the strength ratio is low (e.g., λ=0.1). Importantly, the results indicate that columnar splitting failure and inclined shear failure are two typical failure modes of broken rock mass with grouting reinforcement.

• Journal of the Computational Structural Engineering Institute of Korea
• /
• v.36 no.5
• /
• pp.283-293
• /
• 2023

With the steady increase in the annual number of earthquakes in South Korea, the need to apply seismic reinforcement on public facilities has recently increased. To reinforce seismic capacity, spaced full-column-height steel bars are attached to column faces. In this study, nonlinear finite element analysis was conducted to analyze the effect of external reinforcement steel bars on the seismic capacity of RC columns with a square or rectangular cross-section. For verification, the analysis results were compared with test results. Results showed that the finite element analysis reasonably predicted the actual structural behavior of RC columns with steel bars. In addition, both the analysis and the test results showed that the failure mode was converted from brittle failure to ductile fracture, owing to the external reinforcement steel bars. Both loading capacity and ductility were increased as well. Therefore, the external reinforcement steel bar can effectively enhance the seismic capacity of existing RC columns. This study is expected to contribute to relevant research areas such as the development of design methods.