Distributed Stochastic Model Predictive Control for Heterogeneous UAV Swarm
Mengting Lin, Bin Li, Bin Zhou, Carlo Cecati
Abstract
A distributed stochastic model predictive control (DSMPC) algorithm is proposed for the cooperative control of a heterogeneous unmanned aerial vehicle (UAV) swarm in the presence of external disturbances. Additionally, obstacle avoidance is considered. By permitting only one UAV to optimize at each time step, the cooperative control of the UAV swarm is decoupled into a sequence of local subproblems with chance constraints. Based on the idea of distributionally robust optimization, the chance-constrained subproblems are reformulated into convex optimization problems, which are computationally tractable and can be implemented online. Furthermore, convergence and recursive feasibility of the proposed algorithm are proven. Experiments have been conducted to verify the effectiveness of the proposed method.