CONTINUOUS MOTION PLANNING IN COMPLEX AND DYNAMIC UNDERWATER ENVIRONMENTS
Hui Yu, Anwen Shen, and Yansen Su
Keywords
Hybrid search algorithms, path planning, Fast Marching (FM),autonomous underwater vehicles (AUVs)
Abstract
In this paper, we present a new continuous smooth motion planning
algorithm for autonomous underwater vehicles (AUVs) used in
complex and dynamic underwater environments. Existing motion
planning algorithms in artificial intelligence are not designed to deal
with complex and dynamic continuous underwater environments
prone to currents. We present a novel Fast Marching (FM) method
to address the following issues. First, we develop an algorithm
called Hybrid Search Fast Marching (HSFM) method, which gives
consideration to both total path cost and computational time, to
efficiently generate a 3D continuous smooth path from a discrete
representation of underwater environments. Second, we develop a
novel anisotropic FM algorithm to deal with underwater currents
based on the geometric relationships between characteristic lines
and gradient lines in speed profiles. Third, we introduce multiple
decision criteria and constraints in complex and dynamic underwater
environments such as currents, dynamic obstacles, reefs, shoals and
navigation rules in motion planning process. Extensive random
simulations for an AUV sail planning task show that the HSFM
algorithm meets the time constraints of online replanning and finds
continuous smooth paths of better cost in shorter time compared to
the MSA and A algorithms.
Important Links:
Go Back