LAUSR

The primary challenge in designing a consensus algorithm under switching topologies is to deal with discontinuous interactive information. At present, most consensus control results are limited to either agents with mismatched uncertainties in a time-invariant interaction environment or integrator dynamics in a dynamically changing interaction environment. This work addresses consensus problems of heterogeneous nonlinear multiagent systems without these limitations. We propose a novel, simple, yet effective higher-order filter to yield differentiable control variables employing neighbor position information and discontinuous topology information. By applying this filter and defining a suitable auxiliary error, we develop distributed consensus controllers for second-order and uncertain higher-order systems. One of the salient attributes of the proposed methodology is to allow each agent to require minimal local feedback, i.e., only the position information of itself and its neighboring agents. Theoretical and simulation verifications of the scheme guaranteeing asymptotic convergence of consensus and the boundedness of each signal in the closed-loop system are carefully examined.