Anti-windup synthesis via sampled-data piecewise affine optimal control (Q1433055)

The authors propose a receding horizon optimal control approach to the anti-windup design problem for linear sampled data systems subject to input saturations. The goal of the proposed anti-windup scheme is to ensure that (1) the linear control input is not modified if the state is outside the saturation region and (2) the performance output under saturation converges to the unconstrained one in an \(L_2\) sense. For the actual computation of the controller the authors design a suitable discrete-time finite-horizon optimal control problem which is applied in a receding horizon fashion. The desired properties of the resulting piecewise affine feedback law are proved rigorously and illustrated by a numerical example.