A Navigation Function For Uncertain Environment
This addresses motion planning for robots in uncertain environments, representing an incremental advance by adapting existing navigation function concepts to stochastic scenarios.
The paper tackles motion planning in stochastic environments by extending navigation functions to incorporate Gaussian location probabilities and obstacle geometry, proving the resulting map is a navigation function and demonstrating it in various scenarios.
This paper introduces a novel motion planning algorithm for stochastic scenarios. We extend the concept of a navigation function to such scenarios. Our main idea is to consider both the Gaussian distribution probabilities of the players' locations and disc (or star sets) geometry of the objects operating in the work space. We do so by formulating a probability density function that encloses both. We use the PDF to define a metric between the robot, the obstacles and the configuration space boundary. In order to define the navigation function we formulate a safe probability value for collision. By analytically investigating the PDF we find a convenient approximation for a safe distance in the sense of that metric. We prove that the resulting map is a navigation function and demonstrate our algorithm for various scenarios.