Robust Motion Planning for Uncertain Systems With Disturbances Using the Invariant-Set Motion Planner

The invariant-set motion planner uses a collection of safe sets to find a collision-free path through an obstacle-filled environment [1] -[4] . This article extends the invariant-set motion planner to systems with persistently varying disturbances and parametric model uncertainty. This is accomplish...

Full description

Saved in:
Bibliographic Details
Published in:IEEE transactions on automatic control 2020-10, Vol.65 (10), p.4456-4463
Main Authors: Danielson, Claus, Berntorp, Karl, Weiss, Avishai, Cairano, Stefano Di
Format: Article
Language:English
Subjects:
Autonomous systems
autonomous vehicles
Barriers
Closed loop systems
Collision avoidance
Disturbances
Economic indicators
Feedback control
Heuristic algorithms
Invariants
Liapunov functions
Motion planning
path planning
Planning
robust control
Tracking
Trajectory
Uncertainty
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:The invariant-set motion planner uses a collection of safe sets to find a collision-free path through an obstacle-filled environment [1] -[4] . This article extends the invariant-set motion planner to systems with persistently varying disturbances and parametric model uncertainty. This is accomplished by replacing the previously used positive invariant sets with robust positive invariant sets. Since the model uncertainty obfuscates the relationship between the invariant sets in the state space, and the references and obstacles in the output space, we reformulate the dynamics in velocity form so that the system output appears directly in the modified system state. Since the persistently varying disturbances will prevent the closed-loop system from converging to the desired reference, we introduce a new robust connection rule where references are connected when the invariant set of one reference contains the minimal volume robust invariant-set of another. In addition, we bound the time required to transition between invariant sets to ensure safety when the obstacles are moving. By parameterizing the invariant sets using a precomputed input-to-state Lyapunov function, we reduce the real-time computational complexity of our motion planner. The robust invariant-set motion planner is demonstrated for an automated highway driving case study.
ISSN:0018-9286
1558-2523
DOI:10.1109/TAC.2020.3008126