Evolution Of Control Programs For A Swarm Of Autonomous Unmanned Aerial Vehicles

Unmanned aerial vehicles (UAVs)are rapidly becoming a critical military asset. In the future, advances in miniaturization are going to drive the development of insect sizeU AVs. New approaches to controlling these swarms are required. The goal of thi... Read more

Unmanned aerial vehicles (UAVs)are rapidly becoming a critical military asset. In the future, advances in miniaturization are going to drive the development of insect sizeU AVs. New approaches to controlling these swarms are required. The goal of this research is to develop a controller to direct a swarm of UAVs in accomplishing a given mission. While previous efforts have largely been limited to a two dimensional model, a three dimensional model has been developed for this project. Models of UAV capabilities including sensors, actuators and communications are presented. Genetic programming uses the principles of Darwinian evolution to generate computer programs to solve problems. A genetic programming approach is used to evolve control programs for UAV swarms. Evolved controllers are compared with a handcrafted solution using quantitative and qualitative methods. Visualization and statistical methods are used to analyze solutions. Results indicate thatgenetic programming is capable of producing effective solutions to multiobjective control problems. Less