Optimal Mission Path for the Uninhabited Loyal Wingman
Clay J. Humphreys, Richard G. Cobb, David R. Jacques, Jonah A. Reeger · 2015
The Department of Defense has established a need for manned unmanned teaming (MUM-T), which includes the concept of the uninhabited loyal wingman. Given the vast number of technology pieces which must be advanced, this work first establishes operational objectives for the manned unmanned teaming of aerial vehicles used to guide this research and establishes a candidate scenario useful for demonstrating a multipletarget, low communication bandwidth, online optimal mission plan, which includes breaking from formation to complete an assigned task, and rejoining formation once the task is complete. A literature review establishes the use of direct orthogonal collocation as a viable solution technique for solving the optimal mission planning problem. A two-dimensional small-scale optimal control problem is formulated and results are provided. These results reveal the sensitivity of locally optimal solutions to the initial guess provided to the nonlinear programming solver. A discussion ensues suggesting the use of a particle swarm optimization algorithm as a technique for providing an initial guess to the nonlinear programming solver which will ameliorate concerns of converging on locally optimal solutions that would otherwise not be chosen in practical (real-world) applications. This paper suggests future research which will include a conversion to a full-scale three-dimensional model in which the loyal wingman must successfully estimate and avoid dynamic nondeterministic threats as well as be able to dynamically replan the mission in the presence of pop-up threats.