Python Implementation of Reciprocal Velocity Obstacle (RVO) for Multi-agent Systems
@ARTICLE{8361450,
author={M. {Guo} and M. M. {Zavlanos}},
journal={IEEE Transactions on Robotics},
title={Multirobot Data Gathering Under Buffer Constraints and Intermittent Communication},
year={2018},
volume={34},
number={4},
pages={1082-1097},
doi={10.1109/TRO.2018.2830370}}
This package contains a plug-and-play Python package for collision-avoidance in multi-agent system, based on reciprocal velocity obstacles (RVO) and hybrid reciprocal velocity obstacles (HRVO).
It has minimal impact on your control objective and requires minimal integration.
- Takes a 2D workspace with any number of non-overlaping circular or square obstacles
- Any number of dynamic agents with non-zero volume.
- Allow the choice of VO, RVO, HRVO.
- Direct plug-and-play and fully integrate-able with your control objective, i.e., the output velocity is a minimal modification of the desired velocity.
from your_module import compute_desired_V, Update_V
from RVO import RVO_update
# your control objective here
V_desired = compute_desired_V(X, control_objective, V_max)
# plug in the RVO controller from this package
V = RVO_update(X, V_desired, workspace_model)
# let the robot move
X = Update_X(X, V, step)
- Scalable and fast, see examples below.
- See example.py for test run. [Video1], [Video2]
- Papers on RVO, HRVO
- There are Python bindings of the C++ implementation from the algorithm developers. For my purposes, the formality is too heavy to be integrated into my own project, so I have my own try.
- This package does not depend on the Clearpath geometric package to compute velocity obstacles.
- For very clustered workspace with a large number of robots, you may need to limit the
maximal velocity
and use verysmall step size
. - You may add additional constraints in
RVO_update
such as the change rate ofV
, the lower bound ofV
. - When applying this module to experimental robot control, you may need to set the step size higher due to hardware constraints.
- In most practical experiments, this scheme should still work by limiting the maximal velocity.