Felix Hanisch

Model-based optimization and control of chromatographic processes

Abstract This contribution presents an integrated approach to the optimal operation and automatic control of chromatographic separation processes in batch elution mode as well as in continuous SMB operation. The new approach is based on computationally efficient simulation models and combines techniques from mathematical optimization, parameter estimation and control theory. The resulting algorithms were implemented in an industrial standard control system and the capability of the proposed control approach is demonstrated on the separation of fructose and glucose, both in batch and SMB operation mode.

Model-based control of a simulated moving bed chromatographic process for the separation of fructose and glucose

Abstract Chromatographic separations are an expanding technology for the separation of high value products, particularly in the area of pharmaceutics, food, and fine chemicals. The simulated moving bed (SMB) process as a continuous chromatographic separation process is an interesting alternative to conventional batch chromatography, and gained more and more impact recently. The SMB process is realized by connecting several single chromatographic columns in series. A countercurrent movement of the bed is approximated by a cyclic switching of the inlet and outlet ports in the direction of the fluid stream. Because of its complex dynamics, the optimal operation and automatic control of SMB processes is a challenging task. This paper presents the design of a model-based optimization and control scheme for SMB chromatographic separation processes and its application to the separation of fructose and glucose. We propose a two-layer control architecture where the optimal operating trajectory is calculated off-line by dynamic optimization based on a rigorous process model. The parameters of the model are adapted based on online measurements. The low-level control task is to keep the process on the optimal trajectory despite disturbances and plant/model mismatch. Here identification models based on simulation data of the rigorous process model along the optimal trajectory are combined with a suitable local controller. The efficiency of the trajectory control algorithm is shown in a simulation study for the separation of fructose and glucose on an 8-column SMB plant.

Neural network-based identification of SMB chromatographic processes

Abstract In this contribution, the identification problem for the control of nonlinear SMB chromatographic processes is addressed. For process control the flow rates of extract. desorbent, and recycle of the SMB-process, and the switching time are the natural choices for the manipulated variables. However, these variables influence the process in a strongly coupled manner. Therefore, a new set of input variables is introduced by a nonlinear transformation of physical inputs, such that the couplings are reduced considerably. The front positions of the axial concentration profile are taken as model outputs. Multilayer neural networks are utilized as approximate models of the nonlinear input-output behaviour. The correlation functions between the input and output signals and the gradient distribution of the model outputs with respect to the inputs are used to determine their structural parameters. To illustrate the effectiveness of the identification method, a laboratory scale SMB process is taken as an example. The simulation results of the identified model confirm a very good approximation of the first principles models and have a satisfactory long range prediction performance.

Modellbasierte Regelung von Batch- Chromatographieprozessen

Im Bereich der chemischen Industrie gelten die sogenannten Life Science Produkte als der erfolgversprechendste Bereich der nächsten Jahre. Da jedoch pharmazeutische Präparate, Lebensmittel und Feinchemikalien immer komplexeren Anforderungen und strengeren gesetzlichen Auflagen gerecht werden müssen, werden hier effiziente Verfahren zur schonenden Produkttrennung benötigt. Chromatographische Trennverfahren, als ein Beispiel, stellen in der Regel den entscheidenden Kostenfaktor im gesamten Herstellungsprozeß dar. Eine vollständige Ausnutzung des Potentials der chromatographischen Trennung ist nur durch Einsatz einer zuverlässigen Prozeßüberwachung und Prozeßführung möglich. Diese muß sowohl die Produktqualität sicherstellen als auch den kostenoptimalen Betriebspunkt einhalten. Zur Prozeßführung der Batch-Chromatographie wird in diesem Beitrag eine Struktur bestehend aus einer Schätzung der Systemparameter aus Meßwerten und einer modellbasierten Echtzeitoptimierung der Betriebsparameter vorgeschlagen. Die dazu benötigten mathematischen Prozeßmodelle werden kurz vorgestellt, und die neu entwickelten Algorithmen zur Echtzeitoptimierung und Parameterschätzung beschrieben. Das Konzept wurde an einer Laboranlage mit industrieller Prozeßleittechnik der Siemens 57 Familie implementiert und getestet. Experimentelle Ergebnisse für ein Stoffsystem mit linearer Adsorptionsisotherme sowie eine Simulationsstudie für ein Stoffsystem mit nichtlinearer Adsorptionsisotherme schließen den Beitrag ab.

NEURAL NETWORK-BASED IDENTIFICATION AND MPC CONTROL OF SMB CHROMATOGRAPHY

Abstract In this contribution, the identification and control of nonlinear SMB-chromatographic processes are discussed. Instead of using the physical manipulated process variables, the flow rates of extract, desorbent, and recycle, and the switching time directly, a new set of input variables (β-factors) is employed as control inputs to reduce input/output couplings. A new measure of the front positions of the axial concentration profiles is used as outputs. Multi-layer neural network models are identified for this nonlinear MIMO system. The identified model is used in a model predictive control algorithm. In this algorithm a parameter varying linear model is employed which avoids the on-line computation of the nonlinear optimization problem. The simulation results show that the identified model gives a very good approximation of the process models and the LPVMPC scheme has a good control performance.

An Experimental Comparison of Nonlinear Controllers for a Neutralisation Process

Neutralisation processes are simple examples of chemical processes with inherently nonlinear dynamics. Linear controllers can only achieve good control performance for such processes in the vicinity of a fixed operation point but not over a range of pH-values. In this contribution, different approaches to the nonlinear control of such plants are described and compared in experiments at a real laboratory reactor. One control algorithm is gain-scheduled trajectory control which uses a combination of exact linearisation of the process model together with linear gain-scheduled control around the nominal trajectory, the other is a variant of nonlinear predictive control, the extended DMC-algorithm, which is used here both with an analytical process model and with a neural net model of the plant dynamics.

Neural network-based identification of the SMB chromatographic process

Abstract In this contribution, the identification problem for the control of nonlinear SMB chromatographic processes is addressed. For process control the flow rates of extract. desorbent, and recycle of the SMB-process, and the switching time are the natural choices for the manipulated variables. However, these variables influence the process in a strongly coupled manner. Therefore, a new set of input variables is introduced by a nonlinear transformation of physical inputs, such that the couplings are reduced considerably. The front positions of the axial concentration profile are taken as model outputs. Multilayer neural networks are utilized as approximate models of the nonlinear input-output behaviour. The correlation functions between the input and output signals and the gradient distribution of the model outputs with respect to the inputs are used to determine their structural parameters. To illustrate the effectiveness of the identification method, a laboratory scale SMB process is taken as an example. The simulation results of the identified model confirm a very good approximation of the first principles models and have a satisfactory long range prediction performance.