Bi-Objective Study for the Assignment of
Unmanned Aerial Vehicles to Targets
Authors:
- Dr. James Crowder, CAES APD
-
Ryan D. Friese, Pacific NW National Lab
-
Howard Jay Siegel, Elect. & Computer Engineering and Computer Science Depts, Colorado State University
-
John N. Carbone, Elect. & Computer Engineering and Computer Science Depts, Southern Methodist University
Abstract:
Historically, research shows that multi-vehicle, multiconstraint,
surveillance problems require a combinatorial optimization
solution. In many of these surveillance missions, the overall objective
is to provide plans for surveillance tasks for UAVs (unmanned aerial
vehicles) visiting, or “surveilling,” targets across geographically
distributed areas. Surveillance plans are created with the goal of
maximizing the number of targets the fleet of UAVs can surveil in a
given period of time (in this paper, 24 hours) under a given set of
constraints related to total energy usage (the energy available to each
UAV).. Here, we present a bi-objective task planning genetic algorithm
(GA) that provides a Pareto set of near optimal surveillance plans, given
the above conflicting energy and surveillance objectives. Future work
will expand these algorithms to support multi-objective mission
planning, including speed, distance, weather conditions, and other
factors that would affect the overall surveillance opportunities.
Full access to this whitepaper requires completing and submitting the following form: