Reconfigurable Internal Model Control based on adaptive IIR lattice filters

In this article the design framework for a Reconfigurable Internal Model Controller (RIMC) with learning attributes is presented. The algorithm relies on lattice filtering and is suitable for linear discrete ARMA systems of unknown model (order-model). The adaptation rules depend on the identified dynamics through an adaptive lattice filter. The identification scheme is extended with a proposed algorithm for the model order selection. Within the employed Internal Model Control (IMC)-structure, a lattice controller is utilized in the forward path in cascade with a gain compensator and a lowpass detuning filter. As time progresses, the lattice filter estimates more accurately the system dynamics, and the learning scheme adjusts accordingly the attributes of the detuning filter. Simulation studies are used to investigate the efficacy of the suggested scheme.