Search

Graphical abstract Principle of SARS-CoV RNA synthesis detection by a fluorescence-based High Throughput Screening assay. During the enzymatic step, the RNA template was incubated with the SARS-CoV RTC, the nucleotides triphosphate and the compound to be tested (putative inhibitor). The polymerase activity released a double stranded RNA which was detected by adding a fluorescent intercalant agent (Picogreen®, ex: 480 nM, em: 530 nm). In presence of an inhibitor, the fluorescence signal decreased and is quantified. NTP: Nucleotides Tri-Phosphate., Highlights • A new non-radioactive SARS-CoV RNA polymerase assay was developed. • The robotized assay is suitable to identify RdRp inhibitors based on HTS. • A total of 1520 FDA-approved compounds were tested using the SARS-CoV-RTC assay., The Severe Acute Respiratory Syndrome Coronavirus (SARS-CoV) emergence in 2003 introduced the first serious human coronavirus pathogen to an unprepared world. To control emerging viruses, existing successful anti(retro)viral therapies can inspire antiviral strategies, as conserved viral enzymes (eg., viral proteases and RNA-dependent RNA polymerases) represent targets of choice. Since 2003, much effort has been expended in the characterization of the SARS-CoV replication/transcription machinery. Until recently, a pure and highly active preparation of SARS-CoV recombinant RNA synthesis machinery was not available, impeding target-based high throughput screening of drug candidates against this viral family. The current Severe Acute Respiratory Syndrome Coronavirus-2 (SARS-CoV-2) pandemic revealed a new pathogen whose RNA synthesis machinery is highly (>96 % aa identity) homologous to SARS-CoV. This phylogenetic relatedness highlights the potential use of conserved replication enzymes to discover inhibitors against this significant pathogen, which in turn, contributes to scientific preparedness against emerging viruses. Here, we report the use of a purified and highly active SARS-CoV replication/transcription complex (RTC) to set-up a high-throughput screening of Coronavirus RNA synthesis inhibitors. The screening of a small (1520 compounds) chemical library of FDA-approved drugs demonstrates the robustness of our assay and will allow to speed-up drug discovery against the SARS-CoV-2.

In this paper, the robust control of a class of uncertain process systems with stable and unstable outputs that considers input constraints and control zones is addressed. In the proposed approach, a robust Linear Quadratic Regulator (LQR) controls the system outputs by manipulating the systems inputs and a robust Model Predictive Control (MPC) manipulates the output set-points to maintain the system outputs inside their zones and to enforce the input constraints and targets. The performance of the proposed controller is tested through the simulation of a low order industrial process.

This paper focuses on the application of the Linear Quadratic Regulator (LQR) in industrial process systems where one has zone control of the outputs and input economic targets as well as constraints on the inputs and input moves. In the approach proposed here, the LQR is combined with the Model Predictive Control (MPC) in a framework where the system outputs are controlled through the LQR state feedback control law and the output set points are manipulated by the MPC to enforce the input constraints and input targets. The performance of the proposed controller is tested through the simulation of the control of an industrial deisobutanizer column.

Background: Substantial heterogeneity in the epidemiology and management of Staphylococcus aureus bacteraemia (SAB) occurs in Latin America. We conducted a prospective cohort study in 24 hospitals from nine Latin American countries. Objectives: To assess the clinical impact of SAB in Latin America. Patients and methods: We evaluated differences in the 30 day attributable mortality among patients with SAB due to MRSA compared with MSSA involving 84 days of follow-up. Adjusted relative risks were calculated using a generalized linear model. Results: A total of 1030 patients were included. MRSA accounted for 44.7% of cases with a heterogeneous geographical distribution. MRSA infection was associated with higher 30 day attributable mortality [25% (78 of 312) versus 13.2% (48 of 363), adjusted RR: 1.94, 95% CI: 1.38–2.73, P

In this paper, the robust Model Predictive Control (MPC) of systems with model uncertainty is addressed. The robust approach usually involves the inclusion of nonlinear constraints to the optimization problem upon which the controller is based. At each time step the sequence of control actions is then calculated through the resolution of a NonLinear Programming problem, which can be too computer demanding for high dimension systems. Here, the conventional Multi-model Predictive Control (MMPC) problem is re-casted as an LMI-based problem that can be solved with a lower computational cost. The conventional and LMI-based robust controllers’ performances and computational costs are compared through simulations of the control of an industrial C3/C4 splitter.

Nesta tese, com a ambição de desenvolver um pacote de controle que poderia ser implementado a baixo custo nas indústrias brasileiras, é estudado o controle preditivo (MPC) de sistemas estáveis e instáveis com modelos incertos, baseado em um modelo de realinhamento e o uso de técnicas de desigualdades matriciais lineares (LMI), para resolução de problemas de controle robusto. Na primeira parte da tese, a aplicabilidade de um controlador baseado em um modelo de realinhamento é considerada. Como o controlador preditivo baseado em um modelo de realinhamento não requer o uso de um observador de estados, espera-se que seja mais eficiente e mais robusto a distúrbios não medidos que um controlador que precise de um observador de estados. Essa hipótese é testada comparando-se o desempenho e a robustez a distúrbios não medidos deste controlador com um controlador que requer o uso de um observador de estados, através da simulação do controle de um separador polipropileno/propano (PP) industrial. Por outro lado, a desvantagem de um controlador baseado em um modelo de realinhamento é a sensibilidade da construção do modelo e das matrizes do controlador a erros numéricos que aumentam com o tamanho do sistema. As etapas do algoritmo de controle mais sensíveis a erros numéricos são então destacadas com o objetivo de discutir a aplicabilidade de um controlador baseado em um modelo de realinhamento a sistemas de grande porte. Além disso, sempre que possível, os métodos utilizados para reduzir a sensibilidade das etapas problemáticas a erros numéricos são apresentados. A segunda parte da tese trata da solução do problema de controle robusto preditivo baseado em um modelo de sistema incerto. Este problema é usualmente abordado através da inclusão de restrições não lineares sobre os custos associados aos modelos da planta no problema de controle, de tal modo que a ação de controle é obtida a cada instante de tempo através da resolução de um problema de otimização não linear, cujo custo computacional pode ser proibitivo para sistemas de grande porte. Nesta parte da tese, o problema de MPC robusto é então reformulado como um problema baseado em LMIs, que pode ser resolvido com uma fração do esforço computacional. A abordagem proposta é comparada com o MPC robusto convencional e testada através da simulação do controle de um reator e de um separador C3/C4 industriais. Finalmente, na terceira parte da tese, o MPC de sistemas estáveis e instáveis com incertezas no modelo que permite a saturação das entradas manipuladas é abordado. Um subconjunto das entradas manipuladas é alocado ao controle das saídas instáveis através de uma lei de realimentação de estados, enquanto as outras entradas são deixadas livres para controlar as saídas estáveis restantes. Assume-se que as saídas são controladas dentro de faixas e que os setpoints das saídas são tratados como entradas manipuladas adicionais. Os controladores desenvolvidos nesta parte da tese permitem a saturação das entradas associadas às saídas instáveis através da manipulação das entradas livres e dos setpoints das saídas instáveis. A viabilidade recursiva dos controladores desenvolvidos é garantida, permitindo-se que os setpoints das saídas instáveis deixem temporariamente suas faixas. O desempenho dos controladores desenvolvidos é testado através da simulação do controle de dois processos integradores da indústria química. In this thesis, with the ambition of developing a homemade control package that could be implemented at a low cost in Brazilian industries, the model predictive control (MPC) of uncertain stable and unstable systems based on a realigned state space representation and on the use of the Linear Matrix Inequality (LMI) techniques is addressed. In the first part of this thesis, the practical implementability of a controller based on a realigned model is considered. Since a Model Predictive Controller based on a realigned model does not require the use of a state observer, it is expected to be more efficient and more robust to unmeasured disturbances than a controller requiring the use of a state observer. This assumption is tested by comparing the performance and the robustness to unmeasured disturbances of a controller based on a realigned model with a controller requiring the use of a state observer through the simulation of the control of a nonlinear industrial propylene/propane (PP) splitter. On the other hand, a disadvantage of a controller based on a realigned model is that the construction of the model and controller matrices is very sensitive to numerical errors. The steps of the control algorithm that are more sensitive to numerical errors are then highlighted in order to discuss the practical applicability of an MPC based on the realigned model to large-scale systems. Also, wherever possible, some ways of decreasing the sensitiveness of the problematic steps to numerical errors are presented. The second part of this thesis addresses the solution to the problem of robust MPC of systems with model uncertainty. The usual approach of dealing with this kind of problem is through the inclusion of nonlinear cost constraints in the control problem so that the control action is obtained at each sampling time as the solution to a nonlinear programming (NLP) problem that, for high order systems, can be computationally expensive. In this part of the thesis, the robust MPC problem is then recast as an LMI problem that can be solved in real time with a fraction of the computer effort. The proposed approach is compared with the conventional robust MPC and tested through the simulation of the control of a reactor system and a C3/C4 splitter of the process industry. Finally, in the third part of the thesis, the model predictive control of uncertain process systems with stable and unstable outputs that allows input saturation is addressed. A subset of the manipulated inputs is allocated to the control of the unstable outputs through a state feedback control law while the other inputs are left free to control the remaining stable outputs. It is assumed that the outputs are controlled inside zones and that the output setpoints are treated as additional free manipulated inputs. The proposed controllers allow the saturation of the inputs related to the unstable outputs by manipulating the free inputs and the setpoints of the unstable outputs. The recursive feasibility of the controllers is guaranteed by allowing the setpoints of the unstable outputs to temporarily leave their bounds. The performance of the proposed approach is tested through the simulation of the control of two integrating chemical process systems.

We investigated the suitability of surface plasmon resonance (SPR) for providing quantitative binding information from direct screening of a chemical library on protein tyrosine phosphatase 1b (PTP1B). The experimental design was established from simulations to detect binding with K(D)10⁻⁴ M. The 1120 compounds (cpds) were injected sequentially at concentrations [C(cpd)] of 0.5 or 10 μM over various target surfaces. An optimized evaluation procedure was applied. More than 90% of cpds showed no detectable signal in four screens. The 30 highest responders at C(cpd)=10 μM, of which 25 were selected in at least one of three screens at C(cpd)=0.5 μM, contained 22 promiscuous binders and 8 potential PTP1B-specific binders with K(D) ~10⁻⁵ M. Inhibition of PTP1B activity was assayed and confirmed for 6 of these, including sanguinarine, a known PTP1B inhibitor. C(cpd) dependence studies fully confirmed screening conclusions. The quantitative consistency of SPR data led us to propose a structure-activity relationship (SAR) model for developing selective PTP1B inhibitors based on the ranking of 10 arylbutylpiperidine analogs.

International audience; The chemokine CXCL12 and the receptor CXCR4 play pivotal roles in normal vascular and neuronal development, in inflammatory responses, and in infectious diseases and cancer. For instance, CXCL12 has been shown to mediate human immunodeficiency virus-induced neurotoxicity, proliferative retinopathy and chronic inflammation, whereas its receptor CXCR4 is involved in human immunodeficiency virus infection, cancer metastasis and in the rare disease known as the warts, hypogammaglobulinemia, immunodeficiency, and myelokathexis (WHIM) syndrome. As we screened chemical libraries to find inhibitors of the interaction between CXCL12 and the receptor CXCR4, we identified synthetic compounds from the family of chalcones that reduce binding of CXCL12 to CXCR4, inhibit calcium responses mediated by the receptor, and prevent CXCR4 internalization in response to CXCL12. We found that the chemical compounds display an original mechanism of action as they bind to the chemokine but not to CXCR4. The highest affinity molecule blocked chemotaxis of human peripheral blood lymphocytes ex vivo. It was also active in vivo in a mouse model of allergic eosinophilic airway inflammation in which we detected inhibition of the inflammatory infiltrate. The compound showed selectivity for CXCL12 and not for CCL5 and CXCL8 chemokines and blocked CXCL12 binding to its second receptor, CXCR7. By analogy to the effect of neutralizing antibodies, this molecule behaves as a small organic neutralizing compound that may prove to have valuable pharmacological and therapeutic potential.

The complete sequence of the human genome has been deciphered at the dawn of the new century. This historic event immediately challenged researchers with new needs both in terms of concepts and of working methods. Each scientific community considered how it could tackle these new challenges and it quickly became clear that using small chemical molecules would help discovering and characterizing the function of new proteins. The importance of the genes that the encode new proteins could thus be established in cells, organs and whole organisms. At the initiative of a handful of researchers, French chemists have organized the collection of their molecules and provided them to biologists. By doing so they killed two birds with one stone: on the one hand they created a unique opportunity to add value to their molecules by creating the first academic chemical library, and on the other hand they stimulated the launch of biologically active molecules discovery programs by biologists from the academic sector. It was necessary, however, to raise many compounds and ensure consistent quality control, which quickly became a priority for the chemical libraries to become reliable tools.

International audience; des années 2000, la séquence complète du génome humain a été déchiffrée. Cet événement historique a immédiatement lancé d'énormes défis aux chercheurs, tant sur le plan des concepts que sur celui des méthodes de travail. Chaque com-munauté a considéré la manière dont elle pouvait apporter une aide. Il est rapidement apparu que l'utilisation des petites molécules était un moyen de découvrir et de caractériser la fonction d'une protéine et, par inférence, l'importance du gène qui l'encode, dans la construction des cellules, des organes et des organismes. À l'initiative d'une poignée de chercheurs, les chimistes français se sont organisés pour collecter leurs molécules et les mettre à disposition de biologistes. Ils faisaient ainsi d'une pierre deux coups : ils donnaient une nouvelle vie à leurs molécules en créant la première chimiothèque académique d'une part, et ils per-mettaient pour la première fois à des biologistes du domaine public de se lancer dans des programmes de découverte de molécules biologiquement actives d'autre part. Il fallait toutefois réunir des composés en grand nombre et assurer un contrôle qualité uni-forme, ce qui est rapidement devenu une priorité pour que les chimiothèques deviennent des outils fiables. > que ces deux secteurs poursuivent des buts différents, leurs démarches expérimentales comportent des points communs. Chaque opération de criblage de collections de molécules conduit à l'identification de composés appelés touches, ou hits en anglais, qui sont des substances ayant exercé un effet biologique dans l'essai mis en oeuvre. Les essais de criblage peuvent reposer sur des mesures d'interac-tions (liaison d'une molécule à une protéine nécessairement identifiée) ou sur des mesures d'une réponse biologique (stimulation ou blocage d'une activité cellulaire, ou animale, sur une cible généralement non identi-fiée). L'activité de la touche doit être confirmée, c'est-à-dire répétée, et caractérisée par l'établissement, par exemple, d'une relation dose-effet. Nous avons participé à plus de 40 campagnes de criblage qui ont porté sur un total de 250 000 molécules testées et plusieurs centaines de touches identifiées. Il apparaît que le taux de confirmation d'une touche, ainsi que le taux d'identification de la même touche par deux centres de criblage, peuvent varier de manière importante. Ces taux se situent en moyenne aux alentours de 30 % (S. Gioria, P. Villa, communication personnelle) [1]. Les raisons de ces différences sont analysées dans cet article ; elles pro-viennent de quatre sources dont la qualité des collections de molécules fait partie. Une collection de molécules contient entre mille et plusieurs millions de composés. La caractérisation des molécules qui la composent est un point important, mais il est évident que les coûts de gestion de grosses collections peuvent exploser. Entrons dans le coeur du problème. Pourquoi faire un contrôle qualité : les sources de bruit dans un criblage L'identification d'une touche dans un essai de criblage est la résultante de plusieurs opérations indépendantes les unes des autres, mais dont la combinaison crée des conditions potentiellement très fluctuantes.

International audience; Cyclin-dependent kinases (CDKs) control many cellular processes and are considered important therapeutic targets. Large collections of inhibitors targeting CDK active sites have been discovered, but their use in chemical biology or drug development has been often hampered by their general lack of specificity. An alternative approach to develop more specific inhibitors is targeting protein interactions involving CDKs. CKS proteins interact with some CDKs and play important roles in cell division. We discovered two small-molecule inhibitors of CDK-CKS interactions. They bind to CDK2, do not inhibit its enzymatic activity, inhibit the proliferation of tumor cell lines, induce an increase in G1 and/or S-phase cell populations, and cause a decrease in CDK2, cyclin A, and p27(Kip1) levels. These molecules should help decipher the complex contributions of CDK-CKS complexes in the regulation of cell division, and they might present an interesting therapeutic potential

With the use of [ 3 H]gamma-hydroxybutyric acid, binding experiments allowed the screening of new compounds as ligands of gammahydroxybutyric acid receptors. Starting from the acid-alcohol gamma-hydroxybutyric acid structure, structure‐activity relation analysis and lead optimization highlighted gamma-hydroxybutyric acid derivatives with significantly increased affinities, when compared with the affinity of gamma-hydroxybutyric acid. Further pharmacological studies with the use of gamma-hydroxybutyric acid derivatives allowed the characterization of the first competitive antagonist acting at gamma-hydroxybutyric acid receptors (NCS 382). © 2000 Elsevier Science Inc. All rights reserved.

We have developed a surface plasmon resonance (SPR)-based inhibition in solution assay (ISA) to search for inhibitors of the medium affinity (KD = 0.8 μM) interaction between an E6-derived peptide (E6peptide) immobilized on the sensor and a PDZ domain (MAGI-1 PDZ1) in the mobile phase. DZ domains are widespread protein-protein interaction modules that recognize the C-terminus of various partners. Simulations indicated that relatively low compound concentrations (10 μM) and limited peptide densities (Rmax200 resonance units) should allow the detection of inhibitors with a target affinity close to 100 μM, which was then demonstrated experimentally. ISA screening, carried out on the Prestwick Chemical Library® (1120 compounds), identified 36 compounds that inhibited the interaction by more than 5%. Concentration-dependent ISA, carried out on a subset of 19 potential inhibitors, indicated that 13 of these indeed affected the interaction between MAGI-1 PDZ1 and the E6peptide. No effect was observed for 84 compounds randomly chosen among noninhibitors. One of the four best inhibitors was a peptide binder, and three were PDZ binders with KD in the 10-50 μM range. We propose that a medium (μM) affinity between the target and surface-bound partner is optimal for SPR-based ISA screening.

The α,β-unsaturated lactam 3 was submitted to 1,4-addition of organocuprate reagents R2CuLiI2 (R = Me, Bu or Ph) to provide the 4-substituted compounds 5a–c, 6a–c as a mixture of diastereomers; these intermediates were then transformed into the corresponding α-amino-γ-substituted adipic acids 7a–c, 8a–c.

A program combining ethnopharmacology and bioinformatic approaches has successfully been applied on anti-inflammatory activity. (i) An ethnobotanical study allowed the identification of several plants associated with putative anti-inflammatory properties as potential leads. (ii) On the other hand, it is well known that phospholipase A(2) is a target implicated in the pro-inflammatory process. Thus, (iii) some selected plant extracts were experimentally tested on phospholipase A(2). Finally, (iv) these experimental results combined with bioinformatic tools, such as database exploitation and molecular modeling, allowed to suggest that one compound, betulin and its oxidative form betulinic acid, might be responsible of the anti-PLA(2) activity. This suggestion was confirmed experimentally.

The tripeptide sequence Arg-Gly-Asp (RGD) has been shown to inhibit the adhesive and aggregatory functions of platelets by binding to platelet receptor GP IIb/IIIa. Fibrinogen binding to GP IIb/IIIa represents the final common event that leads to platelet aggregation regardless of platelet activation and, in certain circumstances, is the primary cause of a variety of human cerebral and cardiovascular diseases[1]. The knowledge of the preferred conformation of the RGD sequence together with structure-activity relationship (SAR) analyses should lead to a better understanding of the characteristics important for high affinity binding to GP IIb/IIIa.

A conformational analysis, performed on muscarinic M1 agonists, identified four structural features characteristic of the muscarinic M1 pharmacophore: (i) a protonable basic or quaternary nitrogen acting as a cationic head; (ii) an electronegative dipole usually part of a planar mesomeric ester, amide, or amidine function which can be replaced by an ether (muscarine) or a dioxolane (AF 30); (iii) an intercharge distance of 5 +/- 0.5 A between the cationic head and the electronegative atom of the dipole; (iv) an elevation of 0.5 +/- 0.03 A of the cationic head over the plane containing the electronegative dipole. During a reinvestigation of the conformational behavior of published structures of 5-HT3 antagonists, similar features were observed for the 5-HT3 pharmacophore. However many 5-HT3 antagonists possess additional aromatic planes not present in the muscarinic M1 agonists. These observations brought us to predict the chemical modifications that would change muscarinic M1 agonists into 5-HT3 antagonists. Four of the predicted aminopyridazines were actually synthesized and submitted to testing. The observed IC50 values for 5-HT3 receptor binding ([3H] BRL 43694) ranged from 10 to 425 nM, whereas the affinities for the muscarinic receptor preparations ([3H] pirenzepine) layed over 10,000 nM. In electrophysiological studies the two most active compounds 10 and 13 produced antagonist-like effects on the 5-HT receptor channel complexes responsible for the generation of the rapidly desensitizing ionic currents, and agonist-like effects on those responsible for the slowly desensitizing components.

Depuis 1994, le centre APTE accueille des toxicomanes et des alcooliques dans le meme lieu et releve le defi de les soigner ensemble dans les memes groupes de therapie et avec le meme programme de traite~ment. Inspiree par le modele du Minnesota, cette prise en charge de deux dependances, reputees parfois incompatibles, merite que l’on s’y attarde. La comprehension de cette approche necessite un bref historique du mo~dele de traitement et de la genese du centre Apte. Cette therapie, innovante en France, nous en apprend davantage sur les points communs de ces deux addictions. L’article met en avant l’importance des groupes d’en~traide dans ce type de traitement et partage reflexions et remarques clini~ques autour de celui-ci.

– La generalisation recente du terme resilience dans le champ psychosocial ouvre des perspectives encore peu exploitees. L’article cherche quel nouvel eclairage le concept de resilience pourrait donner aux parcours de toxicomanes abstinents depuis plusieurs annees. A travers des entretiens, les temoignages de retablissement ont mis en avant certains traits communs. Les recits individuels montrent encore que l’entree dans la toxicomanie avait aussi consiste en une sorte de resilience mal adaptee. Notre enquete recense certains etayages indispensables a leur retablisse~ment et explique pourquoi, par necessite, le choix de l’abstinence s’est impose. Cette etude questionne aussi la place des professionnels des soins aux personnes toxicomanes du point de vue de la resilience.

The α,β-unsaturated lactam 3 was submitted to 1,4-addition of organocuprate reagents R2CuLiI2 (R = Me, Bu or Ph) to provide the 4-substituted compounds 5a–c, 6a–c as a mixture of diastereomers; these intermediates were then transformed into the corresponding α-amino-γ-substituted adipic acids 7a–c, 8a–c.

The prominent role of Ca 2 + in cell physiology is mediated by a whole set of proteins involved in Ca 2 + -signal generation, deciphering and arrest. Among these intracellular proteins, calmodulin (CaM) known as a prototypical calcium sensor, serves as a ubiquitous carrier of the intracellular calcium signal in all eukaryotic cell types. CaM is assumed to be involved in many diseases including Parkinson, Alzheimer, and rheumatoid arthritis. Defects in some of many reaction partners of CaM might be responsible for disease symptoms. Several classes of drugs bind to CaM with unwanted side effects rather than specific therapeutic use. Thus, it may be more promising to concentrate at searching for pharmacological interferences with the CaM target proteins, in order to find tools for dissecting and investigating CaM-regulatory and modulatory functions in cells. In the present study, we have established a screening assay based on fluorescence polarization (FP) to identify a diverse set of small molecules that disrupt the regulatory function of CaM. The FP-based CaM assay consists in the competition of two fluorescent probes and a library of chemical compounds for binding to CaM. Screening of about 5300 compounds (Strasbourg Academic Library) by displacement of the probe yielded 39 compounds in a first step, from which 6 were selected. Those 6 compounds were characterized by means of calorimetry studies and by competitive displacement of two fluorescent probes interacting with CaM. Moreover, those small molecules were tested for their capability to displace 8 different CaM binding domains from CaM. Our results show that these CaM/small molecules interactions are not functionally equivalent. The strategy that has been set up for CaM is a general model for the development and validation of other CaM interactors, to decipher their mode of action, or rationally design more specific CaM antagonists. Moreover, this strategy may be used for other protein binding assays intended to screen for molecules with preferred binding activity. This article is part of a Special Issue entitled: 12th European Symposium on Calcium.

PONTIFÍCIA UNIVERSIDADE CATÓLICA DO RIO DE JANEIRO COORDENAÇÃO DE APERFEIÇOAMENTO DO PESSOAL DE ENSINO SUPERIOR PROGRAMA DE EXCELENCIA ACADEMICA A indústria 4.0 impulsionou o desenvolvimento de novas tecnologias para atender as demandas atuais do mercado. Uma dessas novas tecnologias foi a incorporação de técnicas de inteligência computacional no cotidiano da indústria química. Neste âmbito, este trabalho avaliou o desempenho de controladores baseados em aprendizado por reforço em processos químicos industriais. A estratégia de controle interfere diretamente na segurança e no custo do processo. Quanto melhor for o desempenho dessa estrategia, menor será a produção de efluentes e o consumo de insumos e energia. Os algoritmos de aprendizado por reforço apresentaram excelentes resultados para o primeiro estudo de caso, o reator CSTR com a cinética de Van de Vusse. Entretanto, para implementação destes algoritmos na planta química do Tennessee Eastman Process mostrou-se que mais estudos são necessários. A fraca ou inexistente propriedade Markov, a alta dimensionalidade e as peculiaridades da planta foram fatores dificultadores para os controladores desenvolvidos obterem resultados satisfatórios. Foram avaliados para o estudo de caso 1, os algoritmos Q-Learning, Actor Critic TD, DQL, DDPG, SAC e TD3, e para o estudo de caso 2 foram avaliados os algoritmos CMA-ES, TRPO, PPO, DDPG, SAC e TD3. Industry 4.0 boosted the development of new technologies to meet current market demands. One of these new technologies was the incorporation of computational intelligence techniques into the daily life of the chemical industry. In this context, this present work evaluated the performance of controllers based on reinforcement learning in industrial chemical processes. The control strategy directly affects the safety and cost of the process. The better the performance of this strategy, the lower will be the production of effluents and the consumption of input and energy. The reinforcement learning algorithms showed excellent results for the first case study, the Van de Vusse s reactor. However, to implement these algorithms in the Tennessee Eastman Process chemical plant it was shown that more studies are needed. The weak Markov property, the high dimensionality and peculiarities of the plant were factors that made it difficult for the developed controllers to obtain satisfactory results. For case study 1, the algorithms Q-Learning, Actor Critic TD, DQL, DDPG, SAC and TD3 were evaluated, and for case study 2 the algorithms CMA-ES, TRPO, PPO, DDPG, SAC and TD3 were evaluated.

Submitted by Natasha Valladão (natashasilvaa4@gmail.com) on 2021-01-29T14:42:53Z No. of bitstreams: 1 LucasFerreiraBernardino-min.pdf: 6666505 bytes, checksum: e7769d1ac5c3a10538630eb7ab6499c2 (MD5) Made available in DSpace on 2021-01-29T14:42:53Z (GMT). No. of bitstreams: 1 LucasFerreiraBernardino-min.pdf: 6666505 bytes, checksum: e7769d1ac5c3a10538630eb7ab6499c2 (MD5) Previous issue date: 2019-08 A exploração de campos com alto conteúdo de CO2 é desafiadora, devido ao baixo valor agregado e ao impacto ambiental associado a este componente. Nesse contexto, um uso importante para o CO2 gerado é sua reinjeção no reservatório, e a separação submarina de CO2 resulta em um processamento mais eficiente. O prognóstico é uma atividade chave nesse processo, devido à necessidade de minimizar intervenções nos equipamentos. O objetivo principal deste trabalho é investigar aspectos do controle consciente da saúde de um sistema de separação submarina de CO2. O modelo do sistema considerado, que consiste em um sistema de equações algébricodiferenciais, foi adaptado da literatura. Todas as simulações foram realizadas utilizando bibliotecas desenvolvidas pelo autor, baseadas nos métodos das bibliotecas Scipy e Assimulo. Um estado estacionário de referência foi obtido para as condições de projeto, e as funções de transferência em tempo contínuo foram identificadas a partir dos dados de resposta ao degrau. Os modelos identificados foram utilizados em um controlador preditivo, e simulações em malha fechada foram realizadas para avaliar a sintonia do controlador. No sentido de prognóstico, um modelo estocástico de degradação de bomba multifásica, sensível às condições de operação do processo, foi proposto, e um filtro de partículas foi implementado para estimação online do estado de degradação e predição do tempo de vida útil. Por fim, um controlador consciente da saúde foi projetado, e foram investigadas algumas dificuldades na combinação dos objetivos de extensão de vida útil e de controle. Os resultados obtidos indicam que o tratamento do problema de controle consciente da saúde sob a teoria de otimização multiobjetivo pode trazer resultados mais satisfatórios, assim como abordar a extensão de vida útil em uma camada de otimização. The exploitation of fields with high CO2 content is challenging, due to the low economic value and the environmental impact associated with this component. In this context, a significant use for the generated CO2 is its reinjection into the reservoir, and subsea CO2 separation allows for more efficient processing. Prognostics is a key activity in this process, due to the necessity of minimizing intervention in equipment. The main objective of this work is to investigate the health-aware control of a subsea CO2 separation system. The considered system model, which consists of a differential-algebraic equation system, was adapted from the literature. All simulations were performed using libraries developed by the author, based on methods from the Scipy and Assimulo libraries. A reference steady state was obtained for the design conditions, and continuous-time transfer functions were identified from step response data. The identified models were used in a predictive controller, and closed-loop simulations were performed to evaluate the controller tuning. In the sense of prognostics, a stochastic model of multiphase pump degradation, sensitive to the process operating condition, was proposed, and a particle filter was implemented to perform online degradation state estimation and remaining useful lifetime prediction. At last, a health-aware controller was designed, and some difficulties in combining reference tracking and lifetime extension objectives were investigated. The obtained results indicate that dealing with the health-aware control problem through the multiobjective optimization theory and addressing the useful lifetime extension in an optimization layer may result in more satisfactory results.

Submitted by Natasha Valladão (natashasilvaa4@gmail.com) on 2021-02-04T23:05:32Z No. of bitstreams: 1 LeonardoDorigoRibeiro-min.pdf: 2308256 bytes, checksum: 6ba8b3ab3f9efe986769f73976fb06cd (MD5) Made available in DSpace on 2021-02-04T23:05:32Z (GMT). No. of bitstreams: 1 LeonardoDorigoRibeiro-min.pdf: 2308256 bytes, checksum: 6ba8b3ab3f9efe986769f73976fb06cd (MD5) Previous issue date: 2019-06 A indústria de óleo e gás é impactada por ciclos de baixo crescimento econômico. Nesse contexto, técnicas de NMPC e de otimização que auxiliam a operabilidade da planta são de fundamental importância para o negócio. Contudo, o uso em escala industrial requer a superação de desafios, como por exemplo: determinação dos modelos de processo, tempo computacional adequado para obtenção da solução e integração entre as camadas de controle e otimização. Nesse trabalho é proposta uma metodologia para obter modelos analíticos de predição baseado na estrutura de Hammerstein, reduzindo o esforço computacional. As abordagens tradicionais transformam o modelo do NMPC, descrito por equações algebrico-diferenciais (DAE), em uma aproximação numérica baseada em equações algébricas não lineares (NLA), como por exemplo, colocação ortogonal. Na abordagem proposta, o NLA obtido é uma descrição eficaz do sistema original de DAE’s, sendo idependente de aproximações numéricas. O algoritmo proposto foi aplicado em um CSTR não-isotérmico integrado com camada de otimização. Os resultados mostram que a proposta apresenta redução no tempo computacional sem perda de desempenho, quando comparado com o NMPC usando modelo rigoroso. Além disso, o NMPC proposto conduziu a planta ao ponto ótimo vindo da otimização, sem apresentar problemas devido à diferença de modelos entre as camadas. Uma importante contribuição do trabalho é a aplicação do NMPC proposto integrado com camada de otimização em um processo de produção de óleo e gás. Os resultados também mostram bom desempenho do NMPC proposto em estabilizar o processo durante regimes de fluxo golfante, mantendo os requisitos de qualidade dentro das restrições e conduzindo o processo até o ponto ótimo obtido na otimização. The oil and gas industry is impacted by low growth economic cycles. For this reason, techniques of NMPC and optimization that help the operability of the plant are very important for the business. However, the use of these tools in industry requires overcoming some challenges, for instance: find accurate models of the process, computational time suitable for obtaining the solution and integrate the layers of control and optimization. In this work a methodology is proposed to obtain analytical models based on hammerstein structure, reducing the computational time. Unlike most common approaches that transform NMPC internal model, described by differential-algebraic equations (DAE), into an approximate system of nonlinear algebraic equations (NLA) using, for instance, orthogonal collocation. In the proposed approach, the obtained NLA is an exact description of the original DAEs system. The proposed algorithm was applied to a non-isothermal CSTR integrated with an optimization layer. The results show that the proposed structure presents significant reduction in computational time without performance loss, when compared with the NMPC using rigorous model. Moreover, the proposed strategy demonstrated good performance in tracking the targets sent by the optimization layer, without model mismatches between layers. An important contribution of the work is the application of the proposed NMPC integrated with an optimization layer in an oil and gas production process. The results show good performance of the proposed algorithm to stabilize the process during slug flow conditions, keeping the quality requirement inside the constraints and driving the process to the optimal point obtained in the optimization layer.