Showing papers by "Gabriele Pannocchia published in 2019"

Toward a Unifying Framework Blending Real-Time Optimization and Economic Model Predictive Control

Abstract: Nowadays, real-time optimization (RTO) and nonlinear as well as linear model predictive control (MPC) are standard methods in operation and process control systems. Hence there exists a good understanding of how to combine RTO and set point tracking MPC schemes. However, recently, there has been substantial progress in analyzing the properties of so-called economic MPC schemes. This paper proposes a conceptual framework to blend ideas from (output) modifier adaptation and offset-free economic MPC with recent results on economic MPC without terminal constraints. Specifically, we leverage recent insights into economic MPC based on turnpike and dissipativity properties of the underlying optimal control problem. Interestingly, the proposed scheme alleviates the need for a dedicated computation of steady-state targets by exploiting the turnpike property in the open-loop predictions. Two detailed simulation examples show that the proposed schemes deliver excellent performance, while being conceptually much simpler.

Biomethane Production: Mass and Energy Balances of Alternative Supply Chains

Abstract: Two supply chains for biomethane are here analyzed and modeled: gasification of short rotation forestry (SRF) poplar wood chips and anaerobic digestion of two species of microalgae, Chlorella vulga...

Numerical Investigation of the Mixing of Highly Viscous Liquids with Cowles Impellers

Abstract: This work is aimed at investigating the mixing process of highly viscous paints, used to colour leathers in the tanning industry, through Computational Fluid Dynamics (CFD). In particular, a mixing tank is fed with a master liquid and different liquid pigments and then stirred by Cowles impellers in order to obtain a paint of a uniform colour. The typical dynamic viscosity of the liquids in this process is µ ~ O(0.1-10) Pa·s, while the Cowles rotational speed is usually very high, i.e. 3000-5000 rpm.The numerical model is based on the solution of the unsteady Reynolds-Averaged Navier–Stokes (RANS) equations for continuity, momentum and species mass fractions, the latter being used to describe the different components. The impeller motion is modelled through the Sliding Deforming Mesh (SDM) approach, using rotating (unstructured) meshes in the impeller region and a static (structured) mesh in the remainder of the tank. The master liquid and coloured pigments are assumed to stratify within the tank at initial time and the steady rotational speed is then imposed abruptly to the impellers.The level of homogeneity in the stirred tank is evaluated through the analysis of component concentration fields over time. In particular, such local concentrations can be used to determine the mixture colour in different regions of the tank, and hence predict the degree of homogeneity at different times; this is accomplished by defining a proper homogeneity indicator based on the spatial variance of the estimated colour. The proposed numerical model provides an efficient method to investigate the colour of the mixture and to evaluate an appropriate mixing time. The methodology gives also important indications for the tank design, especially useful in the case of non-conventional impellers, high rotation rates and viscous fluids.

A performance monitoring algorithm for sustained optimal operation with economic MPC

Abstract: This paper addresses the problem of performance monitoring for Economic Model Predictive Control (EMPC) in the presence of plant parameter changes. In order to cope with plant-model mismatch, we adopt a recently developed offset-free EMPC algorithm which requires the gradient of the plant input-output steady-state map. A subspace identification method is used in order to estimate this plant gradient from transient measurements. However, when the plant parameters change, this method may fail unless re-identification is performed. Hence, to start a new data collection for the identification an event-triggered mechanism is proposed, based on a suitable performance monitoring strategy. In this case this mechanism investigates a possible, more profitable, steady-state equilibrium and, if convenient, it re-identifies the plant gradient. The proposed monitoring technique is then successfully tested over an illustrative example of a chemical reactor.