Obstacle Avoidance-Driven Formation Scaling Maneuver Based on Distributed Optimization

This article investigates a novel distributed optimization-based method of formation scaling maneuver control with obstacle avoidance. The agents track the formation center in a geometric configuration and perform collective scaling maneuvers to avoid convex obstacles. To handle such a problem, we propose a method consisting of a bearing controller and a distributed optimizer. Combined with the bearing controller, not only does the optimizer under consensus constraint controls the generation of the expected geometric configuration, but it also maneuvers the scale of the configuration to avoid convex obstacles by optimizing a specialized cost function. In particular, the optimizer adopts an adaptive dynamic average consensus observer to estimate the summation of the local gradients, partial derivatives, and Hessian matrices. A simulation is conducted to distinguish the proposed method from existing obstacle-avoiding formation controllers. An experiment on unmanned ground vehicles is performed to verify the practicability of the proposed method.