Collision-Free Rendezvous Maneuvers for Formations of Unmanned Aerial Vehicles

Abstract

This article discusses the rendezvous maneuver for a fleet of small fixed-wing Unmanned Aerial Vehicles (UAVs). Trajectories have to be generated on-line while avoiding collision with static and dynamic obstacles and minimizing rendezvous time. An approach based on Model Predictive Control (MPC) is investigated which assures that the dynamic constraints of the UAVs are satisfied at every time step. By introducing binary variables, a Mixed Integer Linear Programming (MILP) problem is formulated. Computation time is limited by incorporating the receding horizon technique. A shorter planning horizon strongly reduces computation time, but delays detection of obstacles which can lead to an infeasible path. The result is a robust path planning algorithm that satisfies the imposed constraints. However, further relaxation of the constraints and fine-tuning is necessary to limit complexity.

Pilóta nélküli légi járművek kooperatív irányítása

“…Napjaink egyik legfontosabb kihívása az UAV-k kötelék-vagy együttes precíziós repülésének összeütközés-mentes automatizálása. E témakörrel a [3], [5], [6], [7], [10], [11], [14], [15], [18], [20] irodalmak foglalkoznak behatóan.…”

“…• az UAV kezdeti térbeli helyzetének definiálása a 𝑡 𝑜 = 0 időpillanatra: (10) • az UAV kezdeti repülési sebessége a 𝑡 𝑜 = 0 időpillanatra: (11) • a P pont kezdeti térbeli helyzetének definiálása a 𝑡 𝑜 = 0 időpillanatra:…”

UAV Formation Control under Communication Constraints Based on Distributed Model Predictive Control

“…Moreover, DMPC is used to avoid collision [31,32]. In [33], state keeping, obstacles avoidance and precision formation control are performed with the use of DMPC based on introducing binary variables. However, a fundamental distinction of the above methods, this paper adopts different nonlinear error models and control methods to realize formation control which can reflect the relationship between the input and the error more intuitively.…”

Trajectory Generation for Multiagent Point-To-Point Transitions via Distributed Model Predictive Control

“…In contrast, inspired by the incremental inclusion of all collision constraints over an infinite horizon proposed in [6], we introduce on-demand collision avoidance in a DMPC framework, where we detect and resolve only the first collision in the finite prediction horizon, reducing computation time and increasing the success rate for transition tasks. Our method is further enhanced by the use of soft collision constraints, as in [19].…”