Speed Optimization Control for Wheeled Robot Navigation with Obstacle Avoidance Based on Viability Theory

Abstract

The navigation efficiency of wheeled robots needs to be further improved. Although related research has proposed various approaches, most of them describe the relationship between the robot and the obstacle roughly. Viability theory concerns the dynamic adaptation of evolutionary systems to the environment. Based on viability, we explore a method that involves robot dynamic model, environmental constraints and navigation control. The method can raise the efficiency of the navigation. We treat the environment as line segments to reduce the computational difficulty for building the viability condition constraints. Although there exists lots of control values which can be used to drive the robot safely to the goal, it is necessary to build an optimization model to select a more efficient control value for the navigation. Our simulation shows that viability theory can precisely describe the link between robotic dynamics and the obstacle, and thus can help the robot to achieve radical high speed navigation in an unknown environment.