High Performance Control of Dynamical Formation by Estimation of Leader Information

To enhance the robustness and tracking performances of distributed systems with multiple agents in the presence of topological uncertainties and heterogeneous dynamics simultaneously, a distributed holographic state estimator of leader information is designed by sliding mode control theory and analyzed theoretically. The convergence performance of the designed information estimator does not depend on the dynamical model of agent, and only requires that the minimum eigenvalue of the topological matrix is greater than zero. With the estimated information of leader, a model predictive controller is designed for each one to ensure running safety and formation. In order to verify the adaptability of the designed system to uncertain communication topology and agent heterogeneity, both simulations and experiments have been carried out.