Marten Völker

A METHODOLOGY FOR CONTROL STRUCTURE SELECTION BASED ON RIGOROUS PROCESS MODELS

Abstract In this contribution, a systematic procedure for control structure selection based on rigorous models is presented. The basic idea is that a feedback controller that regulates certain measurable quantities to their set-points should steer the process towards its economic optimum in the presence of disturbances and model uncertainties. This part of the analysis is performed for the stationary behaviour of the regulated process, dynamic aspects are considered in a second step where the dynamic controllability of the economically superior structures is assessed. The stationary analysis is performed for rigorous nonlinear plant models. The methodology is applied to a reactive distillation process.

Computation of Tight Uncertainty Bounds from Time-domain Data with Application to a Reactive Distillation Column

When deriving data-driven uncertainty descriptions that are compatible with robust controller design methods, it is still a challenge to quantify potentially nonlinear and time-varying model errors. One way to deal with such errors is to choose norm bounds on the error that are not contradicted by measured data which is here called model unfalsification. This methodology enables particularly reliable statements about the model error, if no a priori assumptions about measurement noise are made. Usually, in order to reduce the impact on the control performance, linear filters must be chosen. In this paper, in order to systematise the choice of linear filters, a methodology that minimises the conservatism in the frequency domain while maintaining the unfalsification property in the time domain is proposed, leading to a mixed frequency-and time-domain optimisation problem. The advantages of the approach are that tightness issues can be addressed in the frequency domain while time-domain unfalsification enables the consideration of nonlinearities and time-varying errors. The method is applied to experimental data from a reactive distillation process that exhibits nonlinear and time-varying behaviour.

Multiobjective controllability assessment by finite dimensional approximation

This paper deals with the calculation of the optimal control performance for linear time invariant plant models under various types of constraints on performance channels in the time as well as in the frequency domain. The approach presented here is based on finite dimensional Q-parametrisation of all stable closed-loop systems. The objective function and the constraints are evaluated on a grid in the time domain and in the frequency domain which facilitates the specification of problem specific goals and constraints. The focus of this paper is on the integration of H/sub /spl infin//-constraints in this design concept. To render the computation numerically tractable, a sequence of quadratic optimisations instead of a linear matrix inequality (LMI) formulation is used. The method is benchmarked against the standard LMI formulation for multiobjective performance calculation which introduces conservatism due to requiring common Lyapunov matrices for each criterion. As expected, the degree of conservatism is reduced showing the efficiency of the new approach.

CONTROL STRUCTURE SELECTION FOR A REACTIVE DISTILLATION COLUMN

Abstract In recent years integrated processes have attracted considerable attention in both academic research and industrial applications. The integration, for instance of reaction and separation, reduces the degrees of freedom and requires tight control to operate the processes at optimal conditions. From the available number of measurements and input variables, a control structure has to be identified which guarantees the best performance. In this paper, we discuss the application of a systematic control structure selection consisting of pre-screening of structures and calculation of the attainable performance for the heterogenous catalyzed esterification of methyl acetate in a batch reactive distillation column. The approach is applied to a linearized model of 85 th order of the nonlinear system. To verify the predictions of the control structure selection method, linear controllers are designed for the favorable structures by using a frequency response approximation method. The resulting linear controller is applied to the nonlinear model and shows good performance.

Quantitative I/O-controllability assessment for uncertain plants

This paper deals with the calculation of the optimal control performance for linear time invariant plant models in the context of the screening of possible control structures. Compared to other approaches, e.g. the use of the relative gain array or other controllability indices, the proposed computation of the attainable control performance gives a quantitative assessment of a control structure without reference to a particular controller structure other than the assumption of linear control. In the time domain, control performance is measured in a least squares sense while constraints on manipulated and controlled variables can be included. The extension of the method presented here integrates the frequency domain condition for robust stability. Uncertainties are represented by the well known MDelta structure which renders the method generically applicable

DESIGN OF ROBUST LOW-ORDER CONTROLLERS FOR COMPLEX PROCESSES: A CASE STUDY ON REACTIVE DISTILLATION IN A MEDIUM-SCALE PILOT PLANT

Abstract This paper deals with robust low-order controller design for a medium-scale pilot reactive distillation column. In the first step, a linear model of the column which is identified from experiments is used to compute the attainable control performance. In this step, actuator limitations and model uncertainty resulting from a priori knowledge as well as confidence intervals provided by the identification procedure are considered. In the second step, the result of the optimal control performance computation is employed in a frequency response approximation scheme to generate a low-order controller. Finally, the synthesised controller is validated at the reactive distillation column.

LEARN2CONTROL: a project-oriented approach to teaching control engineering

LEARN2CONTROL is a new computer-based learning system, that can be used to deepen basic knowledge in control engineering by self-study in a project-oriented environment. The didactic concept aims at teaching the dependencies and interactions between various methods for modelling, analysis and control system design

A Barrier Certificate Approach to the Verification of the Safe Operation of a Chemical Reactor

In industrial practice, extensive simulations are performed in order to analyse the safety and the correct operation of controlled chemical processes. One aspect of verifying the safe operation is to prove that the states of the system stay within a safe region for a certain set of inputs or disturbances which is the main theme of this paper. Recently, a rigorous method for this type of verification problem has been proposed which makes use of Barrier Certificates for verifying whether an undesired set of states can be reached. If the system dynamics can be described in polynomial form, the safety of the system can be proven algorithmically. The determination of a barrier certificate is a sum-of-squares (SOS) problem which can further be transformed into a non-convex Bilinear Matrix Inequality (BMI) problem. This paper deals with proving the safety of a Continuously Stirred Tank Reactor (CSTR), a non-linear system, using barrier certificates. Uncertainties are represented by a bounded disturbance acting on the system. Safety is explicitly proven for a convex set of initial conditions and a non-convex unbounded unsafe set. Two situations are considered, the uncontrolled plant and the closed-loop system with a state-feedback controller. For the solution of the BMI problem, three different numerical approaches are compared. It turned out that solving the non-convex BMI problem directly is more efficient than solving it using the convex iterative approach.

Einstellregeln für Füllstandsregelungen in der industriellen Praxis

Zusammenfassung Die Füllstandsregelung in chemischen Anlagen ist ein wichtiges praktisches Problem. Die Füllstände müssen entweder in bestimmten Grenzen gehalten oder auf einem festen Niveau geregelt werden. Diese Regelung ist eine besondere Herausforderung, da eine integrierende Strecke geregelt wird. Es werden im Allgemeinen PID-Regler des vorhandenen Leitsystems eingesetzt – oft ohne Nutzung des D-Anteils. In diesem Aufsatz werden die für Standregler in Chemieanlagen relevanten Fälle betrachtet. Der für die Fahrweise wichtige Fall der Störungsdämpfung durch Puffer wird besonders behandelt. Ziel des Beitrages ist es, auf Basis der analytischen Lösung des Problems und des H∞-Ansatzes sehr einfache Formeln zu entwickeln, mit denen der Ingenieur vor Ort schnell und sicher Standregler einstellen kann.

LEARN2CONTROL – Projektorientiertes Lernen für ein besseres Gesamtverständnis in der regelungstechnischen Ausbildung (LEARN2CONTROL – Project-oriented Learning for a Better Understanding in Control Engineering Education)

Abstract LEARN2CONTROL ist eine computergestützte Lernumgebung, die es ermöglicht, vorhandenes Grundlagenwissen im Bereich der Regelungstechnik durch selbstständiges Lernen in einem projektorientierten Umfeld zu vertiefen. Der Lernschwerpunkt liegt dabei besonders auf der Vermittlung der inneren Abhängigkeiten und Wechselwirkungen der unterschiedlichen Verfahren und Methoden der Regelungstechnik.