Flocking with Obstacle Avoidance for Fixed-Wing Unmanned Aerial Vehicles via Nonlinear Model Predictive Control

Inspired by biological swarms, collective motions such as flocking of multiple agents can be generated by local interactions. Nevertheless, most of the current swarm models cannot ensure safe flight for fixed-wing unmanned aerial vehicles (UAVs) in obstacle-dense environments. Recent works suggest that predictive models have the potential to integrate reliable obstacle avoidance capabilities with flocking, and make drones predict and coordinate their behavior in real-time. Therefore, this paper investigates the flocking problem in densely cluttered environments for swarms of fixed-wing unmanned aerial vehicles (UAVs) using predictive control. More specifically, a 3D flocking model is developed to realize self-organized, cohesive and safe flight by solving an optimization problem iteratively. Simulation results demonstrate the effectiveness of the proposed model based on predictive control to achieve stable flocking and reliable obstacle avoidance.