Adaptive model-free consensus control for a network of nonlinear agents under the presence of measurement noise

Abstract

In this paper, a decentralized model-free consensus control is proposed for a network of nonlinear agents with unknown nonlinear dynamics, unknown process disturbances and white noise measurement disturbances. Here, the main purpose of control protocol is to first synchronize the states of all follower agents in the network to a leader and then track a reference trajectory in the systems state-space. Here, the leader has at least one information connection with one of the follower agents in the network. The design procedure includes adaptive laws for estimating the unknown linear and nonlinear terms of each agent's dynamics. The estimation for nonlinear terms is in a model-free format. Moreover, a robust observer is designed to approximate the values of leader's control parameters at the agents without connection to the leader. Another observer is incorporated to filter the noisy measurements of the system states. The entire design procedure is analysed successfully for the stability using Lyapunov stability theorem. Finally, the simulation results for application of the algorithm on a network of Duffing-Holmes chaotic systems are presented.