Cooperative extremum seeking for a one-dimensional non-linear map using a two-robot formation

This paper presents a cooperative extremum seeking control scheme that uses a linear formation of two robots to find a maximum of an unknown one-dimensional non-linear map. A consensus control algorithm is used to drive the centre of the formation and an integral feedback control loop is proposed to drive the gradient of an unknown field function to zero. Simulation results are provided to evaluate the performance of the proposed method.This research was supported by the Australian Research Council under grant DP16010112