441 research outputs found
An Environment for the Integrated Modelling of Systems with Complex Continuous and Discrete Dynamics
No full textThe modelling and simulation of sophisticated technical systems is a demanding task. On the one hand, the physical part consists of a large number of subsystems which exhibit predominantly continuous dynamics, sometimes with (infrequent) discontinuities. On the other hand, the distributed computerised control systems constitute complex discrete-time and discrete-event systems that require completely different modelling and simulation methods. For an evaluation of the behaviour and the performance of the overall systems, both types of models have to be combined and simulated efficiently. This contribution presents the requirements for a modelling environment for such systems and discusses an approach that consists of object-oriented modelling and efficient simulation of the physical part using the physical systems modelling language MODELICA, a software environment for the definition of discrete-event models using various formalisms, and the integration of both parts of the system via model translation. The coordination of both parts is performed by the MODELICA simulator. The modelling environment called DES/M (discrete-event systems for Modelica) supports the interoperation of different domain specific discrete-event formalisms. To illustrate the usage of the environment, a laboratory batch plant model is presented. A more elaborate example is described in another contribution in this volume (Mosterman et al., 2002)
Simulation for Analysis of Aircraft Elevator Feedback and Redundancy Control
No full textSafety critical systems such as aircraft require functional and hardware redundancy to achieve prescribed safety levels. Discrete event control is applied to ensure that a safe system configuration is available at all times. Since, at present, formal verification techniques are restricted to models with few continuous states, in this paper, simulation is used to verify that the overall system operates according to the requirements when an actuator failure occurs. The feasibility study to modelling and simulation of complex controlled systems presented here is characterised by (i) a complex object-oriented model of aircraft dynamics, including gravity, aerodynamics, etc., (ii) the specification of the discrete event redundancy control by a domain specific formalism that includes statecharts, (iii) the usage of energy based hybrid bond graphs to model the dynamics of the hydraulic actuators, (iv) model integration on the model level as well as on the data level, (v) support of DAEs with dynamically changing index and (vi) illustrative simulation results
Iterative Fahrweisenoptimierung der annularen Elektro-Chromatographie
No full textZusammenfassung
Die kontinuierliche annulare Elektrochromatographie (Continuous Annular Electro-chromatography, CAEC) ist ein neuartiges kontinuierliches Trennverfahren für die Gewinnung hochwertiger Substanzen aus Produktgemischen. Eine entsprechende Anlage wird derzeit im Rahmen eines EU-Projekts1 entwickelt und getestet. In diesem Beitrag beschreiben wir eine iterative Fahrweisenoptimierung dieses Prozesses auf Basis eines rigorosen 2D-Modells des Trennprozesses und der sog. Gradientenmodifikationsmethode, die für die Optimierung von Batch-Chromatographie-Prozessen bereits von Gao und Engell [7] angewendet wurde.</jats:p
Formale Methoden für die Entwicklung von eingebetteter Software in kleinen und mittleren Unternehmen
No full textThe number of control units within upper class vehicles has steadily increased over the last 15 years and is currently around 100 units. The majority of all innovations in the vehicle are generated by electronics and software. This makes software, on the one hand, one of the most important drivers of innovation for companies in the automotive industry. On the other hand, it involves a high risk potential: programming errors. With international standards and stricter requirements for software quality the industry is trying to counteract this risk. However, the complexity of the systems is growing due to the increasing diversity of variants in order to be able to fulfil all the wishes of the end customer individually. The testing of such software systems can only show the presence of errors, but not their absence. A guarantee that a system fulfils the requirements can be provided by formal methods. The approaches presented in this thesis help to improve and support software development in small and medium-sized enterprises by means of formal verification methods. In doing so, the advantages of small over large companies should be utilised and enhanced. This includes the close proximity to the customer as well as a high degree of flexibility and profitability. The system complexity of most projects as well as the process structures can positively contribute to the adoption of formal methods in the respective development process. The first approach deals with the analysis of timing requirements for embedded systems based on the formal method of model checking. In this case, a task system is modelled using Uppaal for an existing variant system for control units, and a schedulability analysis based on timed automata is presented. To manage the variants, a framework based on pure::variants was designed and an existing software platform was transformed into a product line. This allows companies to focus more on individual customer requirements and to reuse existing components efficiently and with high quality. The second approach to improve the quality of software is to verify the program code of embedded systems through the model checker Arcade. Specifically, the formalization of formal requirements and the applicability in the industrial environment were analysed. Errors in the program code could be localized as well as compliance with requirements were shown. The use of binary code verification can reduce the test effort, but will not replace it. The advantage for companies is, however, that this method can prove the absence of errors, which is not possible by conventional testing. Overall, an approach to the integration of formal methods into the development process of a small and medium-sized enterprise was presented, successfully implemented and evaluated with appropriate tool support. With the methods shown, it is possible to reduce the test effort and to increase the quality of automotive control systems at a nearly stage in the important phases of development
MILP Optimization Models for Short-term Scheduling of Batch Processes
No full textAs there has been a large number of promising developments to the short-term scheduling of batch plants in the last 20 years, the main goal of this work is to provide a general classification of batch scheduling problems and an up-to-date review of the state-of-the-art of this important area. Main features, strengths and limitations of the existing mixed integer linear programming (MILP) optimization techniques will be examined through this paper. We first present a general road-map for scheduling problems of batch plants as well as for the available optimization models. Subsequently, a discussion of modeling aspects of representative MILP models is introduced for both discrete and continuous time models. A comparison of effectiveness and efficiency is presented for discrete- and continuous-time models using a benchmark example taken from the literature. Finally, we draw some general conclusions and point out directions for future research.Fil: Mendez, Carlos Alberto. Consejo Nacional de Investigaciones Científicas y Técnicas. Centro Científico Tecnológico Conicet - Santa Fe. Instituto de Desarrollo Tecnológico para la Industria Química. Universidad Nacional del Litoral. Instituto de Desarrollo Tecnológico para la Industria Química; ArgentinaFil: Grossmann, Ignacio E.. University of Carnegie Mellon; Estados UnidosFil: Harjunkoski, Iiro. No especifíca;Fil: Fahl, Marco. No especifíca
An improved iterative real-time optimization scheme for slow processes
No full textIterative Real-Time Optimization (RTO) has gained increasing attention in
the context of model-based optimization of the operating points of chemical
plants in the presence of plant-model mismatch. In all these schemes, it is
necessary to wait for the plant having reached a steady-state to obtain the
required information on plant performance and constraint satisfaction, which
leads to slow convergence in the case of processes with slow dynamics. This
works addresses this issue by considering both parametric, and structural
plant-model mismatch. First, a simple approach to determine the type of
plant-model mismatch with the use of transient data is discussed. An approach
for dealing with parametric mismatch based on a sensitivity analysis of the
nominal dynamic model is presented, and its performance is evaluated with
the case-study of a Continuously Stirred Tank Reactor (CSTR), where fast
convergence to the optimum can be obtained, even with noisy measurements.
For the case of structural mismatch, nonlinear system identification is integrated
with iterative RTO. The identified models are used to predict the
steady-state of the system, thus reducing the total optimization time. The
performance of the strategy is illustrated by simulation studies of a CSTR and
a hydroformylation process. It is shown that a mixed scheme, where both a
linear and nonlinear model are used for steady-state prediction, results in fast
convergence to a neighborhood of the true optimum, even in the presence of
measurement noise. The use of taylored nonlinear models for dynamic system
identification is shown to be a promising approach for reducing the time
necessary to reach the optimum of a process
Resource effiency indicator-based decision support for the operation of batch and mixed batch-continuous processing plants
No full textSteigende Konzentrationen von Treibhausgasen in der Atmosphäre sind der Grund für
den globalen Klimawandel. Da die chemische Industrie wesentlich zu den Treibhausgasemissionen
beiträgt, schaffen politische Entscheidungsträger Anreize und Gesetze, um die
Industrie zu einer nachhaltigeren Produktion zu bewegen. In dieser Arbeit wird ein Rahmen
zur Definition und Nutzung von Echtzeit-Ressourcene zienzindikatoren (REI) entwickelt,
um die Ressourceneffizienz industrieller Produktionsprozesse kontinuierlich zu
überwachen und zu optimieren. Die Ressourceneffizienz ist eine mehrdimensionale Größe,
die in Relation zur Wirtschaftlichkeit bewertet werden kann. Der Fokus der Arbeit liegt
dabei auf Batch-Prozessen und Prozessen, die diskontinuierliche und kontinuierliche Teilprozesse
kombinieren. Diese stellen eine Herausforderung für die korrekte Erfassung relevanter
Prozessgrößen und die anschlie ende Analyse dar. Das vorgeschlagene Propagationskonzept
ermöglicht es, den Gesamtwirkungsgrad der Anlage auf Basis der Leistung
ihrer Komponenten zu berechnen. Die daraus resultierenden REIs spiegeln die technische
Leistung der Anlage wieder und werden zur Optimierung der gesamten Ressourceneffizienz eines Anwendungsbeispiels verwendet. Die Optimierung der Ressourceneeffizienz
stellt ein mehrdimensionales Optimierungsproblem dar, bei dem die Pareto-optimalen Betriebspunkte
die möglichen Kompromisse zwischen konkurrierenden Interessen angeben.
Die Auswahl eines gewünschten Betriebspunktes aus der Paretomenge ist nicht trivial und
kann sich ändernden Präferenzen folgen. Daher befasst sich der zweite Teil der Arbeit mit
der Synthese eines effizienten und effektiven Entscheidungsunterstützungssystems (Decision
Support System, DSS) zur Auswahl eines Betriebspunktes mit dem gewünschten
Leistungsprofil. Die Methodik wird auf ein Beispiel angewendet und durch eine experimentelle
Usability-Studie validiert. Damit leistet diese Arbeit einen Beitrag zur Optimierung
der Ressourceneffizienz in der Prozessindustrie durch die Identifikation von
ressourcenoptimalen Betriebszuständen. Die ganzheitliche Betrachtung der Ressourceneffizienz in Batchprozessen stellt eine wichtige Erweiterung der industriellen Praxis dar, die
sich derzeit in der Regel auf eine Energieeffizienzanalyse nach ISO50001 beschränkt.Increasing concentrations of greenhouse gases (GHG) in the atmosphere are the reason
for global climate change. Since the chemical industry is a signficant contributor to the
GHG emissions, policy makers are creating incentives and legislation to steer the industry
towards a more sustainable production. This thesis proposes a framework to defie and
utilize real-time resource effiency indicators (REI) to constantly monitor and optimize
the resource effiency of industrial production processes. Resource effiency is a multidimensional
entity that can be evaluated in relation to the economic performance. The
focus of the thesis is on batch- and hybrid - coupled batch and continuously operated -{ processes that introduce further challenges for the correct recording of relevant process
variables and the subsequent analysis. The proposed propagation concept makes it
possible to calculate the overall effiency of the plant based on the performance of its
components. The resulting REIs reflect the technical performance of the plant and are
used to optimize the overall resource effiency of an application case. Optimizing the resource
effiency of a process poses a multi-dimensional optimization problem, where the
Pareto optimal operating points reflect the potential trade-offs between competing interests.
The selection of a desired operational point among the optimal set is not trivial and
may be subject to changing preferences. Thus, the second part of the thesis addresses the
synthesis of an effcient and effective decision support system (DSS) to select an operating
point with the desired performance profile. The methodology is applied and validated by
an experimental usability-study. In summary, the thesis contributes to the optimization
of resource effiency in the process industry by identifying resource-optimal operating
conditions. The holistic consideration of resource effiency in batch processes represents
an important extension of industrial practice, which is up to now usually limited to an
energy effiency analysis according to ISO50001
EKF based State Estimation in a CFI Copolymerization Reactor including Polymer Quality Information
No full textState estimation is an integral part of modern control techniques, as it allows to characterize the state information of complex plants based on a limited number of measurements and the knowledge of the process model. The benefit is twofold: on one hand it has the potential to rationalize the number of measurements required to monitor the plant, thus reducing costs, on the other hand it enables to extract information about variables that have an effect on the system but would otherwise be inaccessible to direct measurement.
The scope of this thesis is to design a state estimator for a tubular copolymerization reactor, with the aim to provide the full state information of the plant and to characterize the quality of the product.
Due to the fact that, with the existing set of measurements, only a small number of state variables can be observed, a new differential pressure sensor is installed in the plant to provide the missing information, and a model for the pressure measurement is developed.
Following, the state estimation problem is approached rigorously and a comprehensive method for analyzing, tuning and implementing the state estimator is assembled from scientific literature, using a variety of tools from graph theory, linear observability theory and matrix algebra. Data reduction and visualization techniques are also employed to make sense of high dimensional information.
The proposed method is then tested in simulations to assess the effect of the tuning parameters and measured set on the estimator performance during initialization and in case of estimation with plant-model mismatch.
Finally, the state estimator is tested with plant data
Process optimization under uncertainty
No full textThe ability of a production plant to be flexible by adjusting the operating conditions
to changing demands, prices of the products and the raw materials is crucial to
maintain a profitable operation. In this respect, the application of mathematical
optimization techniques is unanimously recognized to be successful to improve the
decision-making process. Typical examples are production planning, scheduling,
real-time optimization and advanced process control. The more information are available
to the optimization approach, the more "optimal" are the resulting decisions: the
"optimal" production strategy cannot reduce the inventory costs if no supply-chain
model is integrated into the production planning optimization. This thesis lies in the
context of Enterprise-wide optimization with the goal of integrating decision layers
and functions while accounting for uncertain information. A stochastic programming
approach is adopted to integrate production scheduling with energy management
and production planning with predictive maintenance. The approaches are analysed
from a formulation perspective and from a computational point of view, which is
necessary to deal with one of the challenges of the presented methods consisting in
the size of the resulting optimization problems.
To reduce the electricity cost that is generated by the uncertain peaks of the dayahead
price, a two-stage risk-averse optimization is proposed to simultaneously
define the optimal bidding curves for the day-ahead market and the optimal production
schedule. The large-scale MILP problem is solved with a scenario-based
decomposition technique, the progressive hedging algorithm. Heuristic procedures
are applied to speed up the solution phase and to avoid the oscillatory behaviour due
to the integer variables. Since large electricity consumers rely on Time-Of-Use power
contracts to handle the volatility of the day-ahead price, the two-stage formulation
is expanded into a multi-stage optimization to optimally purchase electricity from
different sources and to generate electric power with a power plant. The unpractical
size of the resulting problem is handled by approximating the multi-stage tree with a
series of two-stage scenario-trees within a rolling horizon procedure. A mixed time
grid handles the multi-scale nature of the problem by making short-term decisions
with a detailed model and catching their effect on the long-term future with an aggregated
model.
While the electricity prices introduce exogenous uncertain information into the optimization
problem, the predictive maintenance optimization carries endogenous
uncertain sources into the production planning problem. Endogenous uncertainties,
contrary to the exogenous ones, are uncertain information that can be modified (in the
probability or in the timing of the realization) by the decision maker. The prognosis
technique of the Cox model is embedded into a multi-stage stochastic program to
consider an uncertain Remaining Useful Life of the equipment when the optimal
operating conditions of the plant are defined. Two modelling approaches (based on
superstructure-scenario trees and on conditional non-anticipativity constraints) are
proposed to formulate the optimization problem with endogenous uncertainties. Two
Benders-like decomposition techniques and several branching priority schemes are
applied to handle the high complexity of the resulting optimization problems
Clusterverfahren zur datenbasierten Generierung interpretierbarer Regeln unter Verwendung lokaler Entscheidungskriterien
No full textBetrachtet wird die Aufgabe der datenbasierten Modellierung von Prozessen für die Datenpunkte erhoben wurden. Jeder Datenpunkt besteht aus den Werten einer oder mehrerer Eingangsgrößen und einer dazugehörenden Ausgangsgröße. Das Modellierungsziel besteht darin, ausgehend von den Datenpunkten ein Modul zu lernen, das die Ausgangsgrößenwerte für zukünftige Datenpunkte, für die nur die Eingangsgrößenwerte bekannt sind, vorhersagt. Clusterverfahren können in diesem Kontext zur datenbasierten Regelgenerierung eingesetzt werden, indem jedes Cluster direkt als eine WENN-DANN Regel interpretiert wird. Dazu sind im Rahmen dieser Arbeit die zwei Clusterverfahren PNC2 und SMBC entwickelt worden. Der PNC 2-Algorithmus basiert auf dem Konzept der hierarchischen agglomerativen Clusterverfahren, bei denen - ausgehend von einer Partitionierung, in der jeder Datenpunkt ein eigenes Cluster darstellt - schrittweise, bis zum Erreichen eines Abbruchkriteriums, immer zwei einander anhand eines Ähnlichkeitskriteriums nahe liegende Cluster miteinander vereinigt werden. Die Grundidee ist es nun, eine Vereinigung nur dann zuzulassen, wenn das dann entstehende generalisierte Cluster einen Regeltest besteht.Der SMBC Algorithmus erweitert unüberwacht arbeitende modell-basierte Clusterverfahren in Richtung eines überwachten Generierens von Clustern. Cluster werden individuell nach Relevanzgesichtspunkten bewertet. Dies erhöht ihre Interpretierbarkeit. Durch Mechanismen zum Löschen, Hinzufügen und Vereinigen von Clustern wird automatisch eine passende Clusteranzahl ermittelt. Beim experimentellen Vergleich verschiedener Lernalgorithmen miteinander ist es notwendig, freie Parameter der jeweiligen Algorithmen systematisch einzustellen. Basierend auf einer Arbeit von Salzberg wird die sogenannte harte Validierung eingeführt, bei der alle freien Parameter mittels Kreuz-Validierung oder ähnlicher Ansätze innerhalb der jeweiligen Lernstichprobe eingestellt werden
- …
