Consensus seeking of multi-agent systems from an iterative learning perspective

Iterative learning control approach for a kind of heterogeneous multi-agent systems with distributed initial state learning

“…The result of the formation control with the iteration k = 100 is shown in Fig. 9, which shows that the desired formation is well achieved by using distributed learning control protocol (21) with the switching topology. …”

“…Let v(t) be a scalar function of t, the upper right-hand derivative D + v(t) is defined by (21) with initial state learning (22). If for all σ ∈ P, there exists the parameters γ 1 , γ 2 and β satisfying σ < 1 with…”

“…Suppose each switching communication graphḠ σ and sub-graphḠ σ,1 are connected for all σ ∈ P. Consider the heterogeneous multi-agent system (1)- (3) and apply the distributed iterative learning protocols (21) with initial state learning (22). If we utilize the matrix norm as || · || 2 and there exist parameters γ 1 , γ 2 , β satisfying [ −I σ T σ −I ] < 0 for all σ ∈ P, the results in Theorem 3 also hold.…”

“…Theorem 4. Suppose each switching communication graphḠ σ and sub-graphḠ σ,1 are connected for all σ ∈ P. Consider the heterogeneous multi-agent system (1)- (3) and employ the distributed iterative learning protocols (21) with initial state learning (22) and the consensus error (20). If there exist parameters γ 1 , γ 2 and β satisfying σ < 1 for all σ ∈ P, then all follower agents can form the desired formation and the first m follower agents achieve velocity consensus perfectly on [0, T] when the iteration number k approaches infinity.…”

Event‐Triggered Iterative Learning Control for Multi‐Agent Systems with Quantization

“…Remark Assumption implies that the directed graph contains a spanning tree where the leader is the root and has an influence directly or indirectly on all followers through a series of directed paths in the network. In addition, the effect of Assumption is to ensure that all the eigenvalues of H have positive real parts known by Lemma , thereby the inverse of matrix H exists, which plays a key role in the convergence analysis of systems and is a fundamental assumption as well as easy to realize for leader‐following MAS .…”