Cooperative Avoidance Control With Velocity-Based Detection Regions

Abstract

Guaranteed collision avoidance control laws for multi-agent systems typically rely on constant detection regions. This constraint tends to generate conservative and slower agents’ trajectories. To reduce the conservatism of avoidance control laws, this letter presents two decentralized, cooperative strategies for arbitrarily large groups of agents that decrease the vehicles’ effective detection regions by using velocity information. The vehicles are modeled as a class of nonlinear Lagrangian systems which full state vector represents absolute position. The control laws are proven to guarantee collision avoidance at all times and are shown to be more energy-efficient and to generate faster and smoother trajectories than traditional methods. Moreover, by decreasing the detection regions, the agents are able to converge to destinations closer to each other’s avoidance regions, a feature not possible with traditional avoidance control.