Adaptive predictive control of a high purity distillation column using irregularly sampled multi-rate data

This work aims at the development of multi-rate adaptive model predictive control (MR-AMPC) based on the fast rate model, which is identified from irregularly sampled multi-rate data. The model is assumed to have output error structure and is parameterized using generalized orthonormal basis filters. The identified model is used to generate inter-sample estimates of the irregularly sampled outputs and for performing future trajectory predictions in the proposed MRMPC formulation. The effectiveness of the proposed adaptive multi-rate control scheme is demonstrated by conducting simulation studies on a benchmark binary distillation column system. The results from the simulation reveals that the proposed adaptive multi-rate model predictive control successfully manages transition of the distillation column from moderate purity region to the high purity region where the system exhibits highly nonlinear dynamics.