M. Storz

Asymptotically exact I/O-linearization of an industrial distillation column

A distillation column on a pilot plant scale is input-output linearized via asymptotically exact I/O-linearization. A model consisting of differential and algebraic equations (DAE) is used to derive an observer and the feedback law. The outputs to be controlled are functions of algebraic state variables. Therefore the theory of I/O-linearization is extended here to explicit DAE-systems of index 1 with implicit algebraic equations. The practical application demonstrates the excellent performance of asymptotically exact I/O-linearization.

Asymptotically Exact Linearization of Chemical Processes

Abstract Nonlinear control of chemical processes using methods based on geometric theory has been proposed by a number of researchers during the last years. The most important approach for practical applications is exact linearization of the I/O- behavior of nonlinear systems. However, there are a number of problems associated with this approach. In this paper, an extension to exact I/O-linearization for SISO- systems is proposed to obtain output feedback instead of state feedback as well as improved robustness properties. The results of the approach are illustrated by simulation examples.

Some aspects of integrated process operation

Abstract Operation and control of integrated processes is considered. Thereby, the aspect of process integration is examined on two levels. On the lower level, different unit operations, like reaction and heat recovery or reaction and separation, are integrated into multifunctional processing units. On the flowsheet level, different processing units are integrated via mass and energy recycles, like the heat integration of distillation columns. On both levels, complex dynamics including periodic and aperiodic oscillations and multiple attractors can arise as a consequence of strong interaction. For the lower level, nonlinear phenomena are illustrated with two examples. The first example is the so-called circulation loop reactor, which is intended to work as an autonomous periodic system. Second, dynamic operation of a reactive distillation column with multiple steady states is considered. To handle dynamics on the flowsheet level a hierarchical process control concept for plant-wide control is presented. The concept is based on a formal definition of product quality as an intensive state function in the thermodynamic sense. This quality function is evaluated for each product stream in a chemical plant and is employed as a plant-wide controlled variable. Further, such a plant-wide objective function can be used for the integration of process design and control.

Process design and control for quality assurance

Abstract In this paper a hierarchical process control concept for plant-wide quality control is proposed. The concept is based on a formal definition of product quality as an intensive state function in the thermodynamic sense. This quality function is evaluated for each product stream in a chemical plant and can be used as a plant-wide uniform controlled variable. For the assigned extensive state function, called the qualifying potential, a balance equation can be formulated and thus, the production of qualifying potential can be determined for each processing unit of a plant. Therefore, the quality concept can be used for process design as well as for process control.

Asymptotisch exakte Linearisierung des Eingangs-/Ausgangs Verhaltens nichtlinearer Systeme

During the last years good progress has been achieved in the field of nonlinear control theory. The most important approach for technical applications is the exact linearization of the input/output behavior of nonlinear systems. However, the method of the exact linearization assumes the knowledge of the whole process state. In this paper, an extension to the present procedure is proposed with the intention to linearize asymptotically exact the input /output behavior of a nonlinear single-input-single-output process via output feedback. Here, the basis of the asymptotically exact linearization of the process is the exact linearization of an adequate accurate process model in real time

Exact I/O-linearization of chemical processes within a plant-wide quality control framework

In this paper a process control concept for plant-wide quality control is presented and illustrated with a simple example process consisting of a continuous stirred tank reactor (CSTR), two distillation columns (DCs) and a recycle stream from one separation unit back into the reactor. The concept is based on a formal definition of product quality as an intensive state function in the thermodynamic sense. This quality function is evaluated for each product stream in a chemical plant and is used as the plant-wide controlled variable. The concept involves two hierarchically ordered levels using exact I/O-linearization (EIOL) for local control systems and LQ methods on the superior coordination level.

Plant-Wide Control Concepts on the Basis of the Qualifying Potential

Abstract In this paper some plant-wide control concepts are discussed based on an extensive thermodynamic state function, the so-called qualifying potential. The dynamic plant behavior with respect to the production objective is represented by the balance equation of the qualifying potential. This concept enables an information compression and forms the basis for plant-wide quality control. The complex plant-wide control task is split into two parts using a hierarchical control concept with two levels. This facilitates the controller design and allows to achieve quality assurance of the end product. Connections between process control and process design are pointed out and the usefulness of the qualifying potential for both is shown.

ASYMPTOTICALLY EXACT LINEARIZATION OF CHEMICAL PROCESSES

Nonlinear control of chemical processes using methods based on geometric theory has been proposed by a number of researchers during the last years. The most important approach for practical applications is exact linearization of the I/O-behavior of nonlinear systems. However, there are a number of problems associated with this approach. In this paper, an extension to exact I/O-linearization for SISO-systems is proposed to obtain output feedback instead of state feedback as well as improved robustness properties. The results of the approach are illustrated by simulation examples.