Youngdae Kim
Ulsan National Institute of Science and Technology · Engineering
About the Lab
Professor Youngdae Kim's research lab specializes in computational science and engineering, with a focus on advancing autonomous systems and data-driven modeling for complex real-world applications. The lab develops cutting-edge reinforcement learning algorithms for autonomous underwater vehicles (AUVs), emphasizing dynamic environment navigation, obstacle avoidance, and optimal path planning. It also conducts large-scale human genetic studies to address health disparities through diverse population genomics, and contributes to high-performance computing by designing proxy applications for exascale computing initiatives. The integration of AI, systems biology, and computational modeling defines the lab’s interdisciplinary approach.
Research Overview
Research Output Trend
Figures are computed from collected data and may differ slightly.
Selected Papers
9One of the justifiable criticisms of human genetic studies is the underrepresentation of participants from diverse populations. Lack of inclusion must be addressed at-scale to identify causal disease factors and understand the genetic causes of health disparities. We present genome-wide associations for 2068 traits from 635,969 participants in the Department of Veterans Affairs Million Veteran Program, a longitudinal study of diverse United States Veterans. Systematic analysis revealed 13,672 ge
The FY22 Proxy App Suite Release milestone includes the following activities: Curate a collection of proxy applications that represents the breadth of ECP applications, including application domains, programming models, supporting libraries, numerical methods, etc. Identify gaps in coverage and work with application teams to commission or develop proxies to cover gaps. From within this collection, designate the ”ECP Proxy Application Suite” of 10–15 proxies that balance breadth of coverage with
해양 과학 및 군사 작전에 대한 수요가 증가함에 따라 자율 수중 차량(autonomous underwater vehicle, AUV)의 중요성이 점차 커지고 있다. AUV는 유연한 각도 변화를 바탕으로 다양한 임무에서 성공률 높은 제어를 달성할 수 있다. 그러나 대부분의 기존 연구는 주로 단순한 환경에서의 실험과 AUV 작전에서의 고유한 방향성을 간과한다. 이에 본 논문은 AUV의 방향 정책 강화학습 알고리즘(directional policy reinforcement learning, DPRL)을 통한 변침점 생성 알고리즘 기반의 AUV 임무 수행 전략을 제안한다. 이때 장애물의 크기 및 위치, 목표물의 위치, 그리고 목표물에 대한 접근 각도를 나타내는 충격 각도를 무작위로 설정한 동적인 환경에서의 실험 결과를 통하여 본 알고리즘의 우수성을 입증한다.
This study presents a fitting factor calibration method to enhance the accuracy and consistency of STANAG 4355 Modified Point-Mass Model trajectory simulations. Conventional approaches, typically based on Mach number or quadrant elevation, focus on matching only the impact range, whereas the proposed method fits the entire trajectory shape, including the maximum ordinate and its corresponding range at the trajectory apex. A new fitting scheme incorporating the yaw of repose and projectile veloci
본 논문은 동적 해양 환경에서 AUV (Autonomous Underwater Vehicle)의 자율 임무 수행 및 경로 최적화를 위해 AOPF (Autonomous Underwater Vehicle Optimal Path Finder) 알고리즘을 제안한다. 제안한 알고리즘은 PPO (Proximal Policy Optimization) 기반 강화학습 기법과 3-degree-of freedom (DOF) 모델을 적용하여 장애물 회피와 목표 도달의 균형을 달성하며, 기존 DDPG (Deep Deterministic Policy Gradient) 알고리즘 대비 빠른 수렴과 높은 임무 성능을 보인다. 실험 결과, 제안 기법은 안정적 학습과 효율적 경로 생성을 확인하였다. 본 연구는 동적 해양 환경에서 AUV의 자율 임무 수행을 위한 새로운 강화학습 기반 제어 방법을 제시하며, 실제 해양 적용 및 다중 AUV 협력 제어와 같은 확장 가능성에 관해 논의한다.
The FY21 Proxy App Suite Release milestone includes the following activities: Curate a collection of proxy applications that represents the breadth of ECP applications, including application domains, programming models, supporting libraries, numerical methods, etc. Identify gaps in coverage and work with application teams to commission or develop proxies to cover gaps. From within this collection, designate the "ECP Proxy Application Suite" of 12-15 proxies that balance breadth of coverage with
Research Areas
Dive deeper into Youngdae Kim's research on Nubint
Open this lab's papers in the app to read with AI, summarize, and cite in your writing.