Yi Ma, Graduate Student
Jana Kosecka, Postdoctoral Researcher
(Professor S. Shankar Sastry)
Funding Sources:
ARO-MURI
The goal of this project is to explore theories and methods of mobile robot motion planning with explicitly taking into account the nonholonomic constraints and the presence of obstacles. Two different classes of schemes will be studied: the open-loop schemes and feedback schemes.
In the open-loop schemes, the collision avoidance has been treated as a (optimal in some sense) path planning problem. The conditions of existence of the solution will be studied as the issue of system reachability. We wish to find techniques to constructively give a solution whenever exists. We will also try to explore algorithms for path planning in some optimal senses (such as minimum energy, or shortest path length). This approach assumes a priori knowledge of the environment.
The feedback schemes are more applicable in partially unknown environments where the information about the obstacles is acquired by available sensors on-line and consequently integrated with the lower level control. In the case of noholonomic mobile robot the integration of the sensory information, especially the visual sensor, within the control loop has been studied only in very simplified scenarios.
We will explore the use of " artificial potential field" based schemes in design of the feedback control laws for navigation of nonholonomic mobile robots in the presence of obstacles.