Abstract
Dear Editor, Aiming at the consensus tracking problem of a class of unknown heterogeneous nonlinear multiagent systems (MASs) with input constraints, a novel data-driven iterative learning consensus control (ILCC) protocol based on zeroing neural networks (ZNNs) is proposed. First, a dynamic linearization data model (DLDM) is acquired via dynamic linearization technology (DLT). Then, the pseudo partial derivative (PPD) estimation in DLDM is addressed as a time-varying zero-seeking task, tackled with noise-tolerant ZNNs (NT-ZNNs). Next, the consensus tracking of MASs with input constraints is transformed into time-varying quadratic programming (QP) with inequality constraints, resolved by NT-ZNNs. Moreover, theoretical analysis confirms the convergence of zero-seeking and QP problems solved by NT-ZNNs, ensuring no residual error and robust noise suppression. Finally, the simulation results exhibit that the proposed ILCC can effectively complete consensus tracking missions while actively handling input constraints in the presence of noise pollution.
Funding Information
  • National Nature Science Foundation of China (U21A20166)

This publication has 8 references indexed in Scilit: