L'industrie sidérurgique rencontre depuis des années un développement important. Les caractéristiques des produits destinés aux différents secteurs industriels deviennent de plus en plus exigeantes. Actuellement des régulateurs avancés de type PID sont utilisés avec une structure multi-boucles locales. Afin de traiter les cahiers des charges de plus en plus contraints, il devient intéressant de revoir cette stratégie de contrôle. Ce sujet de thèse CIFRE entre ArcelorMittal et le CRAN a pour objectif de proposer de nouvelles stratégies de contrôle avancé pour le laminage à froid des bandes d'acier de type tôle fine, notamment des stratégies centralisées de contrôle remplaçant des boucles de régulation locales et indépendantes, afin d'élargir le champ des matériaux traitables, d'augmenter et d'optimiser la productivité et les capabilités du procédé ainsi que de minimiser les coûts de production. Dans une première partie, une étude bibliographique sera effectuée afin de mieux cerner les différents modèles métallurgiques et dynamiques du système à piloter. Dans une seconde partie, le cahier des charges sera traduit en termes d'un problème d'optimisation dont la solution est un contrôleur de type retour d'état. Enfin dans une troisième et dernière partie, une synthèse de retour de sortie dynamique sera proposée pour résoudre le problème des mesures manquantes / The steel industry has been developing for years. The characteristics of the products intended for the various industrial sectors are becoming more and more demanded. Currently advanced PID controllers are used with a local multi-loop structure. In order to deal with the increasingly constrained specifications, it is interesting to review this control strategy. This CIFRE PhD thesis between ArcelorMittal and CRAN aims to propose new advanced control strategies for the cold rolling of thin sheet steel strips, including centralized control strategies replacing local and independent control loops, to broaden the scope of treatable materials, increase and optimize process productivity and capabilities, and minimize production costs. In a first part, a bibliographic study will be carried out in order to better define the various metallurgical and dynamic models of the system to be controlled. In a second part, the specifications will be translated in terms of an optimization problem whose solution is the state feedback controller. Finally, in a third and final part, a dynamic output feedback synthesis will be proposed to solve the problem of missing measurements

<p>Inom processindustrin finns en etablerad reglerhierarki där basreglering sker med PID-regulatorer och där avancerad, multivariabel styrning sköts av MPC-programvara. Steget mellan dessa två nivåer kan upplevas som stort. För mindre och snabba multivariabla processer undvikes helst en multivariabel ansats, med försämrad reglering som följd. På Preem AB har detta upplevts som ett problem. Syftet med examensarbetet har varit att utveckla en alternativ, multivariabel styrstrategi för en process med ett mindre antal reglerstorheter som interagerar. Detta har gjorts genom en utveckling av en LQG-regulator i styrsystemet DeltaV.</p><p>För att implementera en regulator i ett styrsystem måste hänsyn tas till en rad faktorer, såsom hantering av olika körlägen, bortfall av signaler, integratoruppvridning, kommunikation med slavregulatorer och inte minst operatörernas gränssnitt för hantering av regulatorn. Att sedan utveckla en regulator för en process kräver bland annat stegförsök, analys och anpassning av stegtestdata, modellidentifiering, framtagning av trimningskonstanter, testning av styrstrategi i simulerad miljö och idrifttagning. Den typen av frågeställningar addresseras i rapporten.</p><p>Examensarbetet visar att det finns en plats för LQG-regulatorn i processindustrin för en viss typ av problem. Den utvecklade regulatorn har implementerats på en avsvavlingsprocess på Preems oljeraffenaderi i Lysekil med lyckat resultat. Oscillationer i processen, som tidvis påverkat produktionen av propen, har kunnat reduceras.</p> / <p>Process control in process industry is done in a hierarchy in which PID controllers are used for basic control and MPC software is used for advanced, multivariable process control. The implementation of multivariable control using MPC software is a major undertaking and development of such controllers for small and fast multivariable processes is therefore avoided. To achieve better control for such processes, a simpler approach to multivariable control is often sought. The purpose of this masters thesis is to develop an alternative, multivariable control strategy for processes with a smaller number of interacting control variables. This is achieved by developing an LQG-controller in the DCS DeltaV at Preem AB.</p><p>Implementation of such a controller in a DCS requires that consideration is given to a number of factors, including handling of different modes, loss of signals, reset windup, communication with slave controllers and construction of operator interface. To develop a controller for a specific process also requires step testing, model identification, tuning of the controller parameters, simulation of the control strategy and commissioning. Solutions to such issues are addressed in this report.</p><p>The thesis shows that  LQG-controllers can be useful in process industry for some niche applications. The LQG-controller has successfully been applied to a desulphurisation process at Preem's oil refinery in Lysekil, where oscillations affecting the production of propylene have been reduced.</p>

Els objectius d'aquesta memòria són: a) El desenvolupament de procediments cinètics per l'anàlisi simultani de mescles de compostos químics, basats en el registre dels espectres UV-vis al llarg del temps de reacció i el seu tractament mitjançant tècniques de calibració multivaribles.b) Demostrar la seva aplicabilitat general en situacions d'elevada complexitat tant pel mecanisme cinètic com per la complexitat de la matriu de la mostra.Els quatre sistemes químics estudiats, es presenten de forma cronològica a la seva realització, que coincideix amb l'ordre de complexitat. En el primer, es quantifica l'acetat d'hidrocortisona, contingut en un preparat farmacèutic comercial presentat en forma de pomada. La reacció analítica, la condensació amb l'àcid isonicotínic hidrazina, segueix una mecanisme cinètic de pseudoprimer ordre. La pomada també conté un altre corticoesteroid. En el segon, es quantifiquen dues catecolamines d'estructura química similar (levodopa i benserazida, ambdues contingudes en comprimits) mitjançant la reacció d'oxidació amb excés de periodat sòdic. La levodopa evoluciona segons un mecanisme de primer ordre, mentre que la beserazida es desvia d'aquest comportament, seguint un mecanisme desconegut. A més, té lloc una reacció creuada, donant lloc a l'aparició de nous productes.En el tercer, es quantifiquen mescles ternàries de formaldehid, glioxal i glutaraldehid, contingudes en detergents comercials utilitzats en la desinfecció hospitalària. S'han utilitzat dos procediments experimentals diferents, degut a la interferència produïda per l'elevada concentració de tensioactiu a les mostres. Ambdós utilitzen la reacció de condensació amb 3-metil-2-benzotiazolona hidrazona (MBTH). En el primer mètode la mostra es dilueix prèviament per evitar aquest efecte, mentre que en el segon la mostra no diluïda s'addiciona directament a la mescla de reacció. En aquest últim cas, la calibració s'ha realitzat utilitzant un ampli conjunt mixt de mescles de calibració.En el quart, es quatifiquen mescles de fins a quatre sulfonamides (sulfatiazol, sulfadiazina, sulfamerazina, i sulfametazina) utilitzades conjuntament en diferents preparats farmacèutics per us veterinary, utilitzant la reacció de azo-copulació oxidativa amb MBTH en presència de Fe(III) com agent oxidant. La dificultat en la seva resolució no només és deguda a la gran similitud estructural entre els analits, sinó també al fet que el reactiu es descompon en processos laterals a una velocitat molt alta, originant que la reacció analítica es trobi lluny de les condicions de pseudoprimer ordre esperades quan inicialment s'addiciona un gran excés de reactiu al medi.La metòdica seguida al llarg del treball ha estat sempre la mateixa. Inicialment, es busquen les condiciones experimentals més adequades per tal d'obtenir la major discriminació cinètico-espectral entre els diferents analits. Aleshores, es prepara un conjunt de mescles dels diferents analits seguint un disseny factorial, i es construeixen els models de calibració. Els models de calibració s'obtenen utilitzant la regressió en Components Principals (PCR), la regressió Parcial per Mínims Quadrats bilinear (PLS), la regressió Parcial per Mínims Quadrats Multi Via (nPLS), i les Xarxes Neuronals Artificials (ANNs).La qualitat dels resultats obtinguts s'avalua per comparació amb els resultats trobats per cromatografia líquida d'alta resolució (HPLC), utilitzada com a mètode de referència. Es troba que tots els mètodes són comparables pel que fa a precisió i exactitud; excepte per la determinació de l'acetat d'hidrocortisona, en què el mètode cinètic proposat es mostra més precís atès que el mètode HPLC requereix d'una extracció prèvia de l'analit. Les conclusions més importants que s'extrauen d'aquesta memòria són:a) S'ha demostrat el gran potencial analític de la simbiosi entre un mètode cinètic i una calibració multivariable.b) El PLS bilinear és un mètode de calibració molt adequat, capaç de modelar un nombre molt elevat de sistemes no-lineals. El nPLS no millora la seva capacitat predictiva, però simplifica molt la interpretació qualitativa dels resultats.c) L'ús d'algoritmes tan complexes com les ANN, només es justifica en casos excepcionals. / The goals of this Thesis are: a) The development of kinetic procedures for the simultaneous analysis of mixtures of chemical compounds, based on the recording of UV-vis spectra along the reaction time and the use of multivariate calibration procedures.b) To prove its general applicability in very complex situations, both because of the kinetic mechanism and the complexity of the sample matrix.The four chemical studied systems are presented in the chronological sequence of their realisation, which agrees with the order of complexity. In the first one, hydrocortisone acetate, contained in an available pharmaceutical preparation presented as a pomade is quantified. The analytical reaction, condensation with isonicotinic acid hydrazine, conforms to pseudofirst order kinetics. The pomade also contains another corticosteroid. In the second one, two similar chemical structure catecholamines were quantified (levodopa and benserazide, both contained together in tablets) by using the oxidation reaction with excess of sodium periodate. Levodopa evolves according to a first order kinetics, while the benserazide deviates of this behaviour, following an unknown mechanism. Furthermore, a crossed reaction takes place, giving way to appearance of new products. In the third one, ternary mixtures of formaldehyde, glyoxal and glutaraldehyde, presents in commercially available detergents used for disinfecting purposes in hospitals are quantified. Two different experimental procedures has been used, because of the interference coming from the high concentration of surfactants in the samples. Both are based on the condensation reaction with 3-methyl-2-benzothiazolone hydrazone (MBTH). In the first procedure, the sample is previously diluted so that avoiding this effect, while in the second the undiluted sample is directly added to the reaction mixture. In the latter case, the calibration was done using an extended mixed calibration sample set.In the fourth one, mixtures up to four sulfonamides (sulfathiazole, sulfadiazine, sulfamerazine, and sulfamethazine) which are used together in different pharmaceutical preparations for veterinary uses, are quantified by using the oxidative azo coupling reaction with MBTH in the presence of Fe(III) as oxidizing agent. The difficulty in their resolution not only is due to the high chemical similarity between the analytes but also to the fact that the reagent decomposes in a lateral process at a very high rate, which keeps the analytical reaction far from the pseudofirst order conditions that could be expected when a large excess of reagent was initially added.The methodological procedure followed along the work, has been always the same. Initially, the more suitable experimental conditions are searched to obtain the highest kinetic-spectral discrimination between analytes. Then, a set of mixtures of the different analytes are prepared following a factorial design, and the calibration models constructed. Regression models were obtained using the Principal Components Regression (PCR), the bi-linear Partial Least Squares regression (PLS), the N-way Partial Least Squares regression (nPLS), and Artificial Neural Networks (ANNs).The quality of the results obtained was checked by comparison with those provided by high performance liquid chromatography (HPLC), used as a reference method. All the methods were found to be comparably precise and accurate; except for hydrocortisone acetate determination, where the proposed kinetic method proved to be more precise, as a result of the previous extraction of the analyte needed in the HPLC method. The most important conclusions of this work are:a) The high analytical potential of the combination of a kinetic and a multivariate calibration has been proved.b) The bilineal PLS is a very suitable calibration method, able to model a high number of non lineal systems. nPLS does not improve its forecasting capabilities, but it simplifies strongly the understanding of the results. c) The use of very complex algorithms as ANN can be justified only in rare cases.

Inom processindustrin finns en etablerad reglerhierarki där basreglering sker med PID-regulatorer och där avancerad, multivariabel styrning sköts av MPC-programvara. Steget mellan dessa två nivåer kan upplevas som stort. För mindre och snabba multivariabla processer undvikes helst en multivariabel ansats, med försämrad reglering som följd. På Preem AB har detta upplevts som ett problem. Syftet med examensarbetet har varit att utveckla en alternativ, multivariabel styrstrategi för en process med ett mindre antal reglerstorheter som interagerar. Detta har gjorts genom en utveckling av en LQG-regulator i styrsystemet DeltaV. För att implementera en regulator i ett styrsystem måste hänsyn tas till en rad faktorer, såsom hantering av olika körlägen, bortfall av signaler, integratoruppvridning, kommunikation med slavregulatorer och inte minst operatörernas gränssnitt för hantering av regulatorn. Att sedan utveckla en regulator för en process kräver bland annat stegförsök, analys och anpassning av stegtestdata, modellidentifiering, framtagning av trimningskonstanter, testning av styrstrategi i simulerad miljö och idrifttagning. Den typen av frågeställningar addresseras i rapporten. Examensarbetet visar att det finns en plats för LQG-regulatorn i processindustrin för en viss typ av problem. Den utvecklade regulatorn har implementerats på en avsvavlingsprocess på Preems oljeraffenaderi i Lysekil med lyckat resultat. Oscillationer i processen, som tidvis påverkat produktionen av propen, har kunnat reduceras. / Process control in process industry is done in a hierarchy in which PID controllers are used for basic control and MPC software is used for advanced, multivariable process control. The implementation of multivariable control using MPC software is a major undertaking and development of such controllers for small and fast multivariable processes is therefore avoided. To achieve better control for such processes, a simpler approach to multivariable control is often sought. The purpose of this masters thesis is to develop an alternative, multivariable control strategy for processes with a smaller number of interacting control variables. This is achieved by developing an LQG-controller in the DCS DeltaV at Preem AB. Implementation of such a controller in a DCS requires that consideration is given to a number of factors, including handling of different modes, loss of signals, reset windup, communication with slave controllers and construction of operator interface. To develop a controller for a specific process also requires step testing, model identification, tuning of the controller parameters, simulation of the control strategy and commissioning. Solutions to such issues are addressed in this report. The thesis shows that  LQG-controllers can be useful in process industry for some niche applications. The LQG-controller has successfully been applied to a desulphurisation process at Preem's oil refinery in Lysekil, where oscillations affecting the production of propylene have been reduced.

Diss. Luleå : Luleå tekniska univ., 2002.

With the increasing use of multimedia applications on the wireless network, the functionalities of the IEEE 802.11 WLAN was extended to allow traffic differentiation so that priority traffic gets quicker service time depending on their Quality of Service (QoS) requirements. The extended functionalities contained in the IEEE Medium Access Control (MAC) and Physical Layer (PHY) Specifications, i.e. the IEEE 802.11e specifications, are recommended values for channel access parameters along traffic lines and the channel access parameters are: the Minimum Contention Window CWmin, Maximum Contention Window CWmax, Arbitration inter-frame space number, (AIFSN) and the Transmission Opportunity (TXOP). These default Enhanced Distributed Channel Access (EDCA) contention values used by each traffic type in accessing the wireless medium are only recommended values which could be adjusted or changed based on the condition of number of associated nodes on the network. In particular, we focus on the Contention Window (CW) parameter and it has been shown that when the number of nodes on the network is small, a smaller value of CWmin should be used for channel access in order to avoid underutilization of channel time and when the number of associated nodes is large, a larger value of CWmin should be used in order to avoid large collisions and retransmissions on the network. Fortunately, allowance was made for these default values to be adjusted or changed but the challenge has been in designing an algorithm that constantly and automatically tunes the CWmin value so that the Access Point (AP) gives out the right CWmin value to be used on the WLAN and this value should be derived based on the level of activity experienced on the network or predefined QoS constraints while considering the dynamic nature of the WLAN. In this thesis, we propose the use of feedback based control and we design a controller for wireless medium access. The controller will give an output which will be the EDCA CWmin value to be used by contending stations/nodes in accessing the medium and this value will be based on current WLAN conditions. We propose the use of feedback control due to its established mathematical concepts particularly for single-input-single-output systems and multi-variable systems which are scenarios that apply to the WLAN.

O projeto de controladores baseado em dados tem como finalidade o ajuste dos parâ- metros de controle através de experimentos realizados no sistema, sem considerar mode- los que descrevem a dinâmica ou identificação do processo. Na literatura existem diferen- tes abordagens para este tipo de controle, e a análise e a aplicação de um destes métodos é o propósito desta dissertação. O objetivo deste trabalho é apresentar uma melhoria ao mé- todo de sintonia de controladores baseado em Referência Virtual (VRFT) para sistemas multivariáveis (MIMO), visto que a metodologia apresentada na literatura limita a apli- cação do método a uma classe restrita de processos. A metodologia de controle baseado em dados pode ser considerada consolidada para implementação em sistemas monovariá- veis, porém a extensão desta metodologia a sistemas MIMO é uma necessidade a fim de acrescentar a aplicabilidade em processos cujas diferentes variáveis a serem controladas interferem umas nas outras. Diferente do método apresentado na literatura, o método pro- posto neste trabalho permite a aplicação do VRFT em sistemas MIMO onde os requisitos de desempenho podem ser escolhidos de forma específica para cada variável. Além disso, a utilização de variável instrumental e/ou um filtro mostram-se adequadas para estimar os controladores quando os dados são afetados por ruído. Considerando que a eficácia da metodologia depende de fatores como as características da planta, o modelo de referên- cia e estrutura do controlador escolhidos, e a presença ou não de ruído, realiza-se uma comparação de resultados a fim de avaliar as melhorias obtidas. Finalmente, testes foram conduzidos em uma planta de nível do laboratório de controle, implementado o método proposto, o qual apresentou um desempenho satisfatório mostrando a aplicabilidade desta metodologia em sistemas reais. / Data-driven control methods are a variety of control designs that are developed to use batches of input-output data collected, from the process to be controlled without mak- ing explicitly use of parametric models of these processes. Although data driven control methods have been largely used in SISO systems, the extent of this methodology to mul- tivariable processes is yet a necessity, considering that the control design must take into account the multivariable nature of the process, that is, the interaction between different variables involved. The analysis and application of one of these methods is the main objective of this work, which is intended to present and improvement to the existing Virtual Reference Feedback Tuning Method (VRFT) for Multivariable systems, since the methodology disclosed in the literature restricts the application to a limited class of pro- cesses. Moreover, in this work it is proposed an extension of the VRFT method to the MIMO case, which does not present restrictions and closed loop performance of each variable can be determined accordingly. The method can be used to tune a centralized or decentralized controller. Furthermore, when the signals are corrupted by noise the use of an instrumental variable and/or an appropriate filter is proposed. All these contributions are demonstrated through simulations as well as tests conducted in a real plant, which showed satisfactory performance demonstrating the applicability of this method in real systems.

O projeto de controladores baseado em dados tem como finalidade o ajuste dos parâ- metros de controle através de experimentos realizados no sistema, sem considerar mode- los que descrevem a dinâmica ou identificação do processo. Na literatura existem diferen- tes abordagens para este tipo de controle, e a análise e a aplicação de um destes métodos é o propósito desta dissertação. O objetivo deste trabalho é apresentar uma melhoria ao mé- todo de sintonia de controladores baseado em Referência Virtual (VRFT) para sistemas multivariáveis (MIMO), visto que a metodologia apresentada na literatura limita a apli- cação do método a uma classe restrita de processos. A metodologia de controle baseado em dados pode ser considerada consolidada para implementação em sistemas monovariá- veis, porém a extensão desta metodologia a sistemas MIMO é uma necessidade a fim de acrescentar a aplicabilidade em processos cujas diferentes variáveis a serem controladas interferem umas nas outras. Diferente do método apresentado na literatura, o método pro- posto neste trabalho permite a aplicação do VRFT em sistemas MIMO onde os requisitos de desempenho podem ser escolhidos de forma específica para cada variável. Além disso, a utilização de variável instrumental e/ou um filtro mostram-se adequadas para estimar os controladores quando os dados são afetados por ruído. Considerando que a eficácia da metodologia depende de fatores como as características da planta, o modelo de referên- cia e estrutura do controlador escolhidos, e a presença ou não de ruído, realiza-se uma comparação de resultados a fim de avaliar as melhorias obtidas. Finalmente, testes foram conduzidos em uma planta de nível do laboratório de controle, implementado o método proposto, o qual apresentou um desempenho satisfatório mostrando a aplicabilidade desta metodologia em sistemas reais. / Data-driven control methods are a variety of control designs that are developed to use batches of input-output data collected, from the process to be controlled without mak- ing explicitly use of parametric models of these processes. Although data driven control methods have been largely used in SISO systems, the extent of this methodology to mul- tivariable processes is yet a necessity, considering that the control design must take into account the multivariable nature of the process, that is, the interaction between different variables involved. The analysis and application of one of these methods is the main objective of this work, which is intended to present and improvement to the existing Virtual Reference Feedback Tuning Method (VRFT) for Multivariable systems, since the methodology disclosed in the literature restricts the application to a limited class of pro- cesses. Moreover, in this work it is proposed an extension of the VRFT method to the MIMO case, which does not present restrictions and closed loop performance of each variable can be determined accordingly. The method can be used to tune a centralized or decentralized controller. Furthermore, when the signals are corrupted by noise the use of an instrumental variable and/or an appropriate filter is proposed. All these contributions are demonstrated through simulations as well as tests conducted in a real plant, which showed satisfactory performance demonstrating the applicability of this method in real systems.

O projeto de controladores baseado em dados tem como finalidade o ajuste dos parâ- metros de controle através de experimentos realizados no sistema, sem considerar mode- los que descrevem a dinâmica ou identificação do processo. Na literatura existem diferen- tes abordagens para este tipo de controle, e a análise e a aplicação de um destes métodos é o propósito desta dissertação. O objetivo deste trabalho é apresentar uma melhoria ao mé- todo de sintonia de controladores baseado em Referência Virtual (VRFT) para sistemas multivariáveis (MIMO), visto que a metodologia apresentada na literatura limita a apli- cação do método a uma classe restrita de processos. A metodologia de controle baseado em dados pode ser considerada consolidada para implementação em sistemas monovariá- veis, porém a extensão desta metodologia a sistemas MIMO é uma necessidade a fim de acrescentar a aplicabilidade em processos cujas diferentes variáveis a serem controladas interferem umas nas outras. Diferente do método apresentado na literatura, o método pro- posto neste trabalho permite a aplicação do VRFT em sistemas MIMO onde os requisitos de desempenho podem ser escolhidos de forma específica para cada variável. Além disso, a utilização de variável instrumental e/ou um filtro mostram-se adequadas para estimar os controladores quando os dados são afetados por ruído. Considerando que a eficácia da metodologia depende de fatores como as características da planta, o modelo de referên- cia e estrutura do controlador escolhidos, e a presença ou não de ruído, realiza-se uma comparação de resultados a fim de avaliar as melhorias obtidas. Finalmente, testes foram conduzidos em uma planta de nível do laboratório de controle, implementado o método proposto, o qual apresentou um desempenho satisfatório mostrando a aplicabilidade desta metodologia em sistemas reais. / Data-driven control methods are a variety of control designs that are developed to use batches of input-output data collected, from the process to be controlled without mak- ing explicitly use of parametric models of these processes. Although data driven control methods have been largely used in SISO systems, the extent of this methodology to mul- tivariable processes is yet a necessity, considering that the control design must take into account the multivariable nature of the process, that is, the interaction between different variables involved. The analysis and application of one of these methods is the main objective of this work, which is intended to present and improvement to the existing Virtual Reference Feedback Tuning Method (VRFT) for Multivariable systems, since the methodology disclosed in the literature restricts the application to a limited class of pro- cesses. Moreover, in this work it is proposed an extension of the VRFT method to the MIMO case, which does not present restrictions and closed loop performance of each variable can be determined accordingly. The method can be used to tune a centralized or decentralized controller. Furthermore, when the signals are corrupted by noise the use of an instrumental variable and/or an appropriate filter is proposed. All these contributions are demonstrated through simulations as well as tests conducted in a real plant, which showed satisfactory performance demonstrating the applicability of this method in real systems.

No description available.