논문 강의자료만 소개하면 교류가 안일어나지 않을까요? ㅎㅎ 이동윤님이 운영하시면 페이지는 로보틱스 특화 연구를 교류하기엔 약간 눈치(?)보이기도 하구요. 아래는 최근에 읽은 논문들입니다. 보시면 알겠지만 대부분 path planning쪽인데 RRT랑은 궤를 조금 다르게 생각하고 있어요. 현재 관심있는 연구 분야는 임의의 점들이 주어졌을 때, 이 점들을 잇는 path에 대한 분포가 있다고 할 때, 이걸 어떠 써먹을지에 대한 것입니다. 물론 sampling도 가능합니다. 1. Connors, John, and Gabriel Elkaim. Analysis of a spline based, obstacle avoiding path planning algorithm. Vehicular Technology Conference, VTC2007. IEEE, 2007. Cubic spline based approach to find collision free path. The midpoint of the collision is moved out until a clear path is found Only simulaions.. 2. Richter, Charles, John Ware, and Nicholas Roy. High-Speed Autonomous Navigation of Unknown Environments using Learned Probabilities of Collision., ICRA 2014 Sampling based (I guess) receding horizon control The cost function consists of two terms Graph based distance to goal point based on constructed map at each time step Probability of collision learned from simulations Specifically, they used logistic regression where input is the featre of current state and output is the probability of collision Feature: Dist to obs / Dist to unkown /Vel twd obs / Vel twd unknown / Speed 3. Tay, Meng Keat Christopher, and Christian Laugier. Modelling smooth paths using gaussian processes. International Conference on Field and Service Robotics. Springer Berlin Heidelberg, 2007. Modelling path using GP Closely resemble to GP based motion model (not ours) Use kmeans to cluster existing paths with GP likelihood model 4. Saska, Martin, et al. Robot path planning using particle swarm optimization of Ferguson splines. Emerging Technologies and Factory Automation, 2006. ETFA06. IEEE Conference on. IEEE, 2006. Optimize in the space of parameters of cubic splines using PSO. Objective is to find a collision free path.. Other spline based methods 5. Sprunk, Christoph, Boris Lau, and Wolfram Burgard. Improved non-linear spline fitting for teaching trajectories to mobile robots. Robotics and Automation (ICRA), 2012 6. Connors, John, and Gabriel Elkaim. Manipulating B-Spline based paths for obstacle avoidance in autonomous ground vehicles. Proceedings of the ION National Technical Meeting. 2007. 7. Pedrosa, Diogo PF, Adelardo AD Medeiros, and Pablo J. Alsina. Point-to-point paths generation for wheeled mobile robots. Robotics and Automation, 2003. Proceedings. ICRA03. 8. Howard, Thomas M., Colin J. Green, and Alonzo Kelly. Receding horizon model-predictive control for mobile robot navigation of intricate paths. Field and Service Robotics. Springer Berlin Heidelberg, 2010. 9. Otte, Michael, and Emilio Frazzoli. RRT X: Real-Time Motion Planning/Replanning for Environments with Unpredictable Obstacles. Propose an efficient way of replanning given expanded trees where each node has its neighbor informations. Suitable for dynamically changing environment where the operation area is fixed and partial information is given. But not appropriate for situations where a robot constantly encounters new environments as RRT X requires some time for expanding its tree. 10. Pan, Jia, et al. Collision-free and curvature-continuous path smoothing in cluttered environments. RSS, 2012. Smoothing collision free path using spline interpolations by recursively add knots to existing path. 11. Barfoot, Tim, Chi Hay Tong, and Simo Sarkka. Batch continuous-time trajectory estimation as exactly sparse gaussian process regression. RSS ,2014 Proposed a batch state estimation through the lens of GPR Similar to what I try to do but they rather focused on estimating the state (position) of moving object given (time, state) data. Try to solve mapping problem
Posted on: Tue, 30 Dec 2014 01:52:56 +0000
Trending Topics
Recently Viewed Topics
© 2015