Motion planning methods for stratified systems
István Harmati · 2006
This paper addresses the motion planning problem (MPP) of stratified kinematic systems (SKSs) without and with drift. Nonholonomic constraints and drift, two typical features of nonlinear systems, make the motion planning nontrivial while stratified nature demands additional attention for path planning. The paper highlights the issues of stratified motion planning from the aspect of these two properties and it provides algorithms to solve MPP. A procedure to drift neutralization within the stratified framework is also proposed. Due mainly to the Lie bracket averaging intensively used in the concept, the algorithm yields generally an approximation. The main motivation in all cases is to drive the robot along a reference path that assists resolving obstacle avoidance problem. Simulation results are provided on the example of a biped robot