Virtual Structure Formation Flight Control Based on Nonlinear MPC

A formation flight controller for unmanned aerial vehicles (UAV) capable of obstacle avoidance is designed in the framework of nonlinear model predictive control (MPC). The controller considers both input and state constraints. The formation flight controller is designed in a distributed way; each aircraft holds one. The formation architecture is based on the virtual structure approach. The objective function for each UAV is based on the tracking error and the relative distance to obstacles and other vehicles of the formation. The costs related to obstacle and inter-vehicle collision avoidance are added as penalties. Inter-vehicle collision avoidance is ensured using a priority strategy. The optimization problem is solved using sequential quadratic programming (SQP), which has better convergence. A simulation in a 3-D environment provided the results needed to evaluate the performance of the algorithm.