SABER: Data-Driven Motion Planner for Autonomously Navigating Heterogeneous Robots

被引:9
作者
Schperberg, Alexander [1 ]
Tsuei, Stephanie [2 ]
Soatto, Stefano [2 ]
Hong, Dennis [1 ]
机构
[1] Univ Calif Los Angeles, Robot & Mech Lab, Dept Mech & Aerosp Engn, Los Angeles, CA 90095 USA
[2] Univ Calif Los Angeles, UCLA Vis Lab, Dept Comp Sci, Los Angeles, CA 90095 USA
关键词
Motion and path planning; multi-robot systems; deep learning methods; optimization and optimal control; SLAM;
D O I
10.1109/LRA.2021.3103054
中图分类号
TP24 [机器人技术];
学科分类号
080202 ; 1405 ;
摘要
We present an end-to-end online motion planning framework that uses a data-driven approach to navigate a heterogeneous robot team towards a global goal while avoiding obstacles in uncertain environments. First, we use stochastic model predictive control (SMPC) to calculate control inputs that satisfy robot dynamics, and consider uncertainty during obstacle avoidance with chance constraints. Second, recurrent neural networks are used to provide a quick estimate of future state uncertainty considered in the SMPC finite-time horizon solution, which are trained on uncertainty outputs of various simultaneous localization and mapping algorithms. When two or more robots are in communication range, these uncertainties are then updated using a distributed Kalman filtering approach. Lastly, a Deep Q-learning agent is employed to serve as a high-level path planner, providing the SMPC with target positions that move the robots towards a desired global goal. Our complete methods are demonstrated on a ground and aerial robot simultaneously (code available at: https://github.com/AlexS28/SABER).
引用
收藏
页码:8086 / 8093
页数:8
相关论文
共 25 条
[1]   CasADi: a software framework for nonlinear optimization and optimal control [J].
Andersson, Joel A. E. ;
Gillis, Joris ;
Horn, Greg ;
Rawlings, James B. ;
Diehl, Moritz .
MATHEMATICAL PROGRAMMING COMPUTATION, 2019, 11 (01) :1-36
[2]  
Arias F. F., P IEEE INT C ROB AUT, P2021
[3]  
Bajcsy A, 2019, IEEE INT CONF ROBOT, P936, DOI [10.1109/ICRA.2019.8794457, 10.1109/icra.2019.8794457]
[4]  
Baker B., 2020, P INT C LEARN REPR
[5]   Chance-Constrained Optimal Path Planning With Obstacles [J].
Blackmore, Lars ;
Ono, Masahiro ;
Williams, Brian C. .
IEEE TRANSACTIONS ON ROBOTICS, 2011, 27 (06) :1080-1094
[6]   An algorithmic framework for convex mixed integer nonlinear programs [J].
Bonami, Pierre ;
Biegler, Lorenz T. ;
Conna, Andrew R. ;
Cornuejols, Gerard ;
Grossmann, Ignacio E. ;
Laird, Carl D. ;
Lee, Jon ;
Lodi, Andrea ;
Margot, Francois ;
Sawaya, Nicolas ;
Wachter, Andreas .
DISCRETE OPTIMIZATION, 2008, 5 (02) :186-204
[7]  
Bouffard P., 2012, THESIS U CAL BERKELE
[8]  
Chiang H.-T. L., 2020, ALGORITHMIC FDN ROBO, VIII, P52
[9]  
Fankhauser P, 2016, 2016 IEEE/RSJ INTERNATIONAL CONFERENCE ON INTELLIGENT ROBOTS AND SYSTEMS (IROS 2016), P2859, DOI 10.1109/IROS.2016.7759443
[10]   Visual-Inertial Object Detection and Mapping [J].
Fei, Xiaohan ;
Soatto, Stefano .
COMPUTER VISION - ECCV 2018, PT XI, 2018, 11215 :318-334