문화기술의 융합 (The Journal of the Convergence on Culture Technology)

  • 제10권6호
  • /
  • Pages.597-604
  • /
  • 2024
  • /
  • 2384-0358(pISSN)
  • /
  • 2384-0366(eISSN)
국제문화기술진흥원 (The International Promotion Agency of Culture Technology)
권호 표지
DOI QR코드

DOI QR Code

역기구학과 강화 학습을 활용한 리드 클라이밍 루트 파인딩 시뮬레이션 개발

Development of a Lead Climbing Route-Finding Simulation Using Inverse Kinematics and Reinforcement Learning

  • 노승현 (성결대학교 미디어소프트웨어학과 XICOM LAB) ;
  • 진성아 (성결대학교 미디어소프트웨어학과 XICOM LAB)
  • 투고 : 2024.08.12
  • Accepted : 2024.11.01
  • Published : 2024.11.30
PDF 다운로드
⟨ 이전 논문 다음 논문 ⟩

Abstract

본 연구는 올림픽 정식 종목 중 하나인 리드 클라이밍에 초점을 맞춰서 루트 파인딩 시뮬레이션을 개발하는것을 목표로 한다. Unity를 활용하여 장면에 오브젝트를 배치하고 변수들을 정의하여 클라이밍 환경을 구성하고, 클라이밍 자세에 따른 상태를 생성하여 클라이머 모델의 관절 위치를 FABRIK 알고리즘으로 계산하였다. 그리고 ML-Agents를 활용하여 벡터 관측 선택과 이산적인 행동을 정의하고, CCW 알고리즘을 이용한 안정적인 자세를 정의하여 행동 마스킹과 에이전트의 행동 선택에 따른 처리 함수를 구현하였다. 모의실험 결과, 탑 홀드까지 루트 파인딩이 가능함을 확인하였고, 이러한 시뮬레이션이 리드 클라이밍 훈련을 진행하는 선수들에게 많은 도움을 줄 것으로 기대된다.

This study aims to develop a route-finding simulation focused on lead climbing, an official Olympic discipline. Objects were strategically using Unity placed and variables defined to construct a realistic climbing environment. Various climbing postures were generated, and the joint positions of the climber model were calculated using the FABRIK algorithm. ML-Agents were utilized to define vector observations and discrete actions, with stable postures determined using the CCW algorithm. Functions for behavior masking and agent action selection were implemented. Simulation results confirmed the feasibility of route-finding to the top hold, showing promise in enhancing training for lead climbers.

Keywords

References

  1. S.K. Lee. B.M KIM, and H. Heon, "Development of Control Method for Self-Driving Roller Conveyor Based on 3D Simulation Self-Driving Roller Conveyor, Manufacturing Automation, Simulation, Unity," The Journal of the Convergence on Culture Technology, Vo1. 10, No. 3, pp. 861-864, 2024, DOI: 10.17703/JCCT.2024.10.3.861
  2. S.H. Lee, K,H. KIM and B.S. Jeong, "Design and Simulation of RFID Tag for Container-Grown Seedlings System," The International Journal of Advanced Culture Technology, Vol.10 No.2, pp. 292-299, 2022, DOI: 10.17703/IJAC T.2022.10.2.292
  3. KAFTV, "Sports Climbing Basics - Understanding Combined Competition Format(Speed, Bouldering, Lead)," YouTube, February 7, 2020. https://youtu.be/4sbSO5yzC4c
  4. H.K. Kim, and Y.K. Lee, "Route-finding simulator for Sports Climbing Education," Proceedings of the Korean Association of Computer Education Conference, Vol. 17, No. 2, pp. 273-277, 2013
  5. K. Naderi, J. Rajamaki, and P. Hamalainen, "Discovering and synthesizing humanoid climbing movements," ACM Transactions on Graphics (TOG), Vol. 36, No. 4, pp. 1-11, 2017. DOI:10.1145/3072959.3073707
  6. K. Naderi, A. Babadi, S. Roohi, and P. Hamalainen, "A reinforcement learning approach to synthesizing climbing movements," 2019 IEEE Conference on Games (CoG), pp. 1-7, 2019. DOI:10.1109/CIG.2019.8848127
  7. S.R. Buss, "Introduction to Inverse Kinematics with Jacobian Transpose, Pseudoinverse and Damped Least Squares methods," Technical report, Univ. of California, 2009. https://scholar.google.co.kr/scholar?q=Introduction+to+Inverse+Kinematics+with+Jacobian+Transpose,+Pseudoinverse+and+Damped+Least+Squares+methods&hl=ko&as_sdt=0&as_vis=1&oi=scholart
  8. A. Aristidou, and J. Lasenby, "FABRIK: A fast, iterative solver for the Inverse Kinematics problem," Graphical Models, Vol. 73, No. 5, pp. 243-260, 2011. DOI:10.1016/j.gmod.2011.05.003
  9. T.H. Cormen, C.E. Leiserson, R.L. Rivest, and C. Stein, Introduction to algorithms, MIT press, pp. 1014-1017, 2009. https://books.google.co.kr/books?hl=ko&lr=&id=RSMuEAAAQBAJ&oi=fnd&pg=PR13&dq=introduction+to+algorithms&ots=a3i0Y00DVN&sig=b3aSr9LEnO4kz6KYm5WelcozqUo#v=onepage&q=introduction%20to%20algorithms&f=false
  10. Unity Technologies, "Unity ML-Agents Toolkit," GitHub, October 29, 2023. https://github.com/Unity-Technologies/ml-agents/blob/643df3599f81ad472a1f7b46e8d59984ef62c39a/docs/Readme.md