MPC Model Assessment of Highly Coupled Systems

Abstract

Systems with strong interactions among the variables are frequent in the chemical industry, and the use of model predictive control (MPC) is a standard tool in these scenarios. However, model assessment in this case is more complex when compared with fairly coupled systems, since the interactions make the system more sensitive to the model uncertainties. It means that, if the coupling is high, a small modeling error in a single variable could be spread to the entire system. As a result, all the controlled variables (CVs) of the MPC will have their performance deteriorated and the root of the model problem will not be evident. This paper presents a method of model assessment for highly coupled systems. This is an extension of the method proposed by Botelho et al. for model-plant mismatch evaluation in MPC applications, based on the use of the diagonal elements of the output sensitivity matrix. One of its advantages is that the method does not require previous knowledge regarding the systems coupling level. The effectiveness of the proposed method is illustrated by two case studies: a high-purity distillation column and the Shell heavy oil fractionator.

System Behavior and Predictive Controller Performance Near the Azeotropic Region

“…High RG values also result in large off‐diagonal values in the RGA matrix. This implies that the system is interactive, i.e., it exhibits strong coupling among the variables, which means that decentralized control cannot be used for effective control of the system , . The RG values of the models identified in this work are calculated using the following equation: …”

“…As a result, MPC is recommended for the control of such processes. However, the estimation of model parameters, especially time delay, in such systems is a challenge and the controller performance and stability are highly dependent on the magnitude of MPM .…”

Diagnosis of Poor Performance in Model Predictive Controllers: Unmeasured Disturbance versus Model–Plant Mismatch

“…Estimation of Nominal Outputs. The first step in our approach consists of the nominal output estimation following the method proposed by Botelho et al 12,13 We assume the control loops illustrated in Figure 1, with a MPC controller C and nominal model G 0 , which is used in the MPC to describe the real plant G. The model−plant mismatch (MPM) magnitude is ΔG. The theoretical system without mismatch is shown in Figure 1a, for which nominal closed-loop outputs are y 0 .…”

Optimization-Based Control Strategy with Wavelet Network Input–Output Linearizing Constraint for an Ill-Conditioned High-Purity Distillation Column

“…1,8,10−20 Model predictive control (MPC) is one of the successful implementations of needed control strategy. A model described as an ordinary differential equation, 10,11,20 a transfer function, or a state-space model 12−14 is typically used in the controller formulation. However, the physical-based models have some practical drawbacks.…”

“…In recent years, investigation into optimal control for improving the controllability and operability of the high-purity distillation column has been reported. ,,− Model predictive control (MPC) is one of the successful implementations of needed control strategy. A model described as an ordinary differential equation, ,, a transfer function, or a state-space model − is typically used in the controller formulation. However, the physical-based models have some practical drawbacks.…”