As autonomous ground robots fulfill greater roles within the military, there is a requirement for an operator to be able to quickly give minimal route-planning guidance in support of an autonomous mission. The objective of this thesis is to develop a route-planning algorithm that uses open-source satellite imagery to allow a user to plot a start point, a goal point, and identify large-scale obstacles within the robot’s operating area. In this thesis, we build on previous work that developed a potential field obstacle avoidance algorithm. We advance the development of the autonomous mission capability by creating a global path-planning algorithm. The algorithm uses the visibility graph and A* search method to produce the optimal path from the given start point to the goal. The navigation algorithm developed allows users to generate imagery-based obstacle maps in Google Earth Pro and successfully produces an optimal path in the form of global positioning satellite coordinates via extensive MATLAB code development. The method was evaluated on a ground robot navigating in an outdoor environment using the waypoints generated. The path-planning algorithm was successfully implemented, but due to difficulties encountered with the navigation node of the mobile robot, a complete verification was not possible. Improvements to the robot’s ability to traverse over rugged terrain will make this solution more viable for a wider range of outdoor environments.
Yun, Xiaoping Calusdian, James
Naval Postgraduate School
Master of Science in Electrical Engineering
Electrical and Computer Engineering (ECE)
Approved for public release; distribution is unlimited.