Robust Lattice-based Motion Planning

TL;DR

This paper introduces a robust lattice-based motion planning method for nonlinear systems with bounded disturbances, ensuring safety and convergence by using tube-based motion primitives and graph search.

Contribution

It presents a novel tube-based motion planning algorithm that guarantees safety and robustness for nonlinear systems under disturbances using a lattice-based approach.

Findings

Successfully applied to an Euler-Lagrange ship model with wind disturbance

System trajectories remain within a bounded tube around the nominal path

Ensures termination within a region around the final state

Abstract

This paper proposes a robust lattice-based motion-planning algorithm for nonlinear systems affected by a bounded disturbance. The proposed motion planner utilizes the nominal disturbance-free system model to generate motion primitives, which are associated with fixed-size tubes. These tubes are characterized through designing a feedback controller, that guarantees boundedness of the errors occurring due to mismatch between the disturbed nonlinear system and the nominal system. The motion planner then sequentially implements the tube-based motion primitives while solving an online graph-search problem. The objective of the graph-search problem is to connect the initial state to the final state, through sampled states in a suitably discretized state space, such that the tubes do not pass through any unsafe states (representing obstacles) appearing during runtime. The proposed strategy is implemented on an Euler-Lagrange based ship model which is affected by significant wind disturbance. It is shown that the uncertain system trajectories always stay within a suitably constructed tube around the nominal trajectory and terminate within a region around the final state, whose size is dictated by the size of the tube.