A Novel Leader-Follower Framework for Control of Helicopter Formation

Abstract

A framework for formation control of a group of autonomous helicopters is presented. We introduced two control schemes named as I - alpha and I - I, which are tailored to control the relative positions of a helicopter constrained by either one or two neighboring leaders, respectively. To stabilize the internal formation parameters of these schemes, a nonlinear model predictive controller is developed. The controller finds the future control commands by optimizing a cost function, which includes formation parameter errors among other parameters such as control forces. The gradient descent method is considered as a suitable optimizer candidate for our approach. The design steps of the I - I controller is presented in this work. By designing both the two l - alpha and I - I control schemes, any user-defined three dimensional grid pattern could be achieved by a group of autonomous helicopters.