Contributions on model-based controllers applied to dead-time systems

Abstract

This work addresses the analysis and design of model-based controllers applied to the control of single-input single-output (SISO) and multiple-input multiple-output (MIMO) stable, unstable and integrative dead-time systems. Dead-time, which appears in many industrial processes, is a rather challenging issue in the process control area since the transport delay can lead the closed-loop system to undesired oscillatory behaviour or even instability. The longer the delay, the more difficult it is to cope with it, and one solution to this problem consists of using dead-time compensator (DTC) structures. Thus, initially, this work proposes new guidelines for the tuning of a simplified DTC focusing on industrial processes. The new guidelines employ different poles in the robustness filter of the DTC in order to consider the trade-off between noise attenuation, disturbance rejection performance and overall closed-loop robustness. Furthermore, an extension of the simplified DTC structure for state-space systems is also proposed. The proposed structure has few adjustment parameters and can be applied to both continuous and discrete-time systems. Moreover, it allows improving rejection of both matched and unmatched unknown disturbances for linear time-invariant (LTI) systems with input delay. It is worth to highlight that finite spectrum assignment (FSA) based implementation is used in order to guarantee the internal stability of the proposed state-space controller, which is a novel strategy in the DTC literature for guaranteeing a safe implementation for non Hurwitz open-loop systems. Although the simplified DTC state-space structure is useful for dealing with MIMO systems, it is well known that model predictive controllers (MPCs) can yield some advantages when dealing with such class of processes, specially in the case of non-square and multiple-delay systems. Therefore, a generalized predictive control (GPC) based DTC structure that can deal with aforementioned issues is also developed. Finally, this work also presents a simplified control strategy based on GPC that is applied to two-inputs single-output (TISO) systems in an economic context when the two inputs are from different nature, thus yielding different operating costs. Through simulations and practical experiments, it is shown that the proposed approaches present better results in the control of dead-time processes than other recent works from the literature.