Cooperative Formation Control of Connected Vehicles With Collision Avoidance and Unknown Input Reconstruction
Abstract
This paper addresses cooperative formation control problem of a group of connected vehicles, where unknown inputs, obstacles in the workspace, and limited inter-vehicle communication range are simultaneously taken into account. To handle the unknown inputs, a prescribed-time reconstruction scheme is developed by integrating an interval observer with a super-twisting differentiator, so that the reconstruction error converges to zero within a user-specified time. On this basis, a distributed observer is constructed to estimate the leader state at each follower within the same time. For collision avoidance and connectivity maintenance, an exponential artificial potential field is then designed, whose negative gradient is continuous at the activation boundary so as to prevent abrupt changes of the control input. A segmented scaling function synchronizes the above modules with the formation controller and forces the tracking errors to the origin within a user-specified time, regardless of initial conditions. Simulation experiment results confirm the effectiveness of the developed method. Note to Practitioners—This paper is motivated by the need for safe, smooth, and prescribed-time cooperative formation tracking of mobile robots or uncrewed ground vehicles (UGVs) in environments with obstacles. To avoid abrupt changes in the obstacle-avoidance force at the activation boundary, an exponential artificial potential field with a continuous negative gradient is constructed. Additionally, a segmented scaling function coordinates leader-state observation and unknown-input reconstruction with formation tracking over prescribed time intervals. The resulting design can inform controller development for autonomous vehicle fleets subject to bounded unknown inputs, obstacles, and communication constraints. Physical-platform validation remains outside the scope of this study.