H∞ consensus control for high-order multi-agent systems with disturbances

This paper is devoted to consensus problems in directed networks of high-order agents with disturbances. A new distributed protocol is proposed with the consideration of model uncertainty, which only depends on the agent's own information and it's neighbors' first state, an auxiliary variable is included to describe the effects of all-order derivatives' relative information. Based on Lyapunov theory, for three cases: (a) network with fixed topology and zero time-delay; (b) network with switching topology and zero time-delay; (c) network with fixed topology and non-zero time-delay, sufficient conditions are derived correspondingly to make all agents reach H. consensus. Especially, the approach used in this paper does not need any model transformation. Finally, numerical simulations are provided to show the effectiveness of the obtained results.