A multi-wavelength laser diode based optical sensor was designed, developed and evaluated for monitoring and control microalgae growth in real-time. The sensor measures optical density of microalgae suspension at three wavelengths: 650 nm, 685 nm and 780 nm, which are commonly used for estimating microalgae biomass concentration and chlorophyll content. The sensor showed capability of measuring cell concentration up to 1.05 g L⁻¹ without sample dilution or preparation. The performance of the sensor was evaluated using both indoor photobioreactors and outdoor paddle wheel reactors. It was shown that the sensor was capable of monitoring the dynamics of the microalgae culture in real-time with high accuracy and durability. Specific growth rate (μ) and ratios of optical densities (OD ratios) at different wavelengths were calculated and were used as good indicators of the health of microalgae culture. A series of experiments was conducted to evaluate the sensor's capability of detecting environmental disturbances in microalgae systems, for instance, induced by dust or Vampirovibrio chlorellavorus, a bacteria found to cause crash of microalgae culture. Optical densities measured from the sensor were insensitive to the amount of dust that consisted of 59.7% of dry weight of microalgae in the system. However, the sensor was able to detect multiple events of introduction of dust timely by μ and OD ratios. The sensor was also capable of detecting subtle changes of culture in color that leads to a total crash of the culture before it can be differentiated by naked eye. The sensor was further integrated into an existing outdoor raceway to demonstrate the sensor's potential of being a core component to control microalgae production system. A real-time monitoring and control program along with a graphical user interface (GUI) was developed for a central control station aiming at improving resource use efficiency for biomass production.

The burdensome capital cost of urban water distribution systems demands the use of efficient optimization methods capable of finding a relatively inexpensive design that guarantees a minimum functionality under all conditions of operation. The combinatorial and nonlinear nature of the optimization problem involved accepts no definitive method of solution. Adaptive search methods are well fitted for this type of problem (to which more formal methods cannot be applied), but their computational requirements demand the development and implementation of additional heuristics to find a satisfactory solution. This work seeks to employ adaptive search methods to enhance the search process used to find the optimal design of any water distribution system. A first study presented here introduces post-optimization heuristics that analyze the best design obtained by a genetic algorithm--arguably the most popular adaptive search method--and perform an ordered local search to maximize further cost savings. When used to analyze the best design found by a genetic algorithm, the proposed post-optimization heuristics method successfully achieved additional cost savings that the genetic algorithm failed to detect after an exhaustive search. The second study herein explores various ways to improve artificial neural networks employed as fast estimators of computationally intensive constraints. The study presents a new methodology for generating any large set of water supply networks to be used for the training of artificial neural networks. This dataset incorporates several distribution networks in the vicinity of the search space in which the genetic algorithm is expected to focus its search. The incorporation of these networks improved the accuracy of artificial neural networks trained with such a dataset. These neural networks consistently showed a lower margin of error than their counterparts trained with conventional training datasets populated by randomly generated distribution networks.

An evaluation for Lag time, defined as the time from the centroid of rainfall excess to the centroid of direct runoff, was performed for seven small watersheds in the Southwestern United States. The size of these watersheds range in size from 0.33 to 4.37 ha. The evaluation period goes from 2000 to 2010. This evaluation was compared versus 28 lag time equations.The USDA-ARS Southwest Watershed Research Center has developed an electronic data processing system where rainfall and runoff data is collected from sensors in the field and are transmitted to computers in the office.Prior to these data sets there were no measurable rainfall and runoff data. This digital data is generated by means of synchronized clocks in rain gages and flumes. As a result, the data from this electronic processing system offers a unique opportunity for hydrologic research. This infrastructure is a characteristic not often available in many other sites and a condition not enjoyed in previous investigations.There are several definitions of lag time depending on what particular time period is used to describe the occurrence of unit rainfall and runoff. But, time parameters currently lack commonly-accepted standard definitions.The various definitions for time parameters such as Lag Time, Time of Concentration, Time to Peak, Equilibrium Time and others, sometimes, are used interchangeably. Another aspect is that in Lag time determination, centers of mass play a critical role. This study found out that depending of the loss model chosen to estimate rainfall excess, it shall influence the determination of center of mass and hence the lag time.Some negative values were obtained for lag time when measured from peak of runoff to centroid of effective rainfall. However, in the lag time definition from centroid of effective rainfall to centroid of direct runoff, negative values were not obtained.One field in particular, time parameters currently lacks commonly-accepted standard definitions. This has become a source of profound confusion in this branch of science to the point where teams of scientists writing about and discussing hydrological Time Parameters can be compared to the aftermath at the Tower of Babel.

One of the motivating reasons for the development of hydroponics was avoidance of root pathogens. Hydroponics involves growing crops in relatively sterile media, isolated from the underlying soil which may have disease pressure. However, even when hydroponics is coupled with controlled environments such as high tunnels and climate-controlled greenhouses, soil-borne pathogens can enter the growing area and proliferate due to optimal environmental conditions for pathogen growth. Control of root pathogens is difficult and usually achieved through synthetic fungicides since few biocontrol options are available. Compost water extracts (CWE) have recently been gaining the attention of greenhouse growers because they may be a low-cost, environmentally friendly approach to control root disease. CWE are mixtures of compost and water incubated for a defined period of time, either with or without aeration, and with or without additives intended to increase microbial populations, which in turn suppress disease. Much anecdotal, but very little scientific, evidence exists describing CWE effect on suppressing soil-borne pathogens. The present study 1) examined the effect of an aerated CWE on disease suppression at the laboratory scale and in container studies using different soilless substrates, 2) investigated a phenotypic change at the root level caused by CWE that may be associated with disease suppression, and 3) isolated some factors in the production of CWE that affect the ability of a CWE to suppress disease. The common model pathogen-host system of Fusarium solani f.sp. pisi and pea was used to examine CWE-induced disease suppression, with information then being translatable to similar patho-systems involved in greenhouse crop production. In the first study, laboratory-based root growth and infection assays resulted in 100% suppression of F. solani when roots were drenched in CWE. These protected seedlings were then taken to a greenhouse and transplanted into fine coconut coir, watered with hydroponic nutrient solution, and grown for five weeks. At the end of the experiment, 23% of the shoots of the pathogen-inoculated, CWE-drenched seedlings remained healthy while only 2% of the inoculated seedlings without CWE drench remained healthy. All of the roots of the inoculated seedlings developed lesions, even those drenched in CWE. However, 29% of the CWE drenched roots were able to recover from disease, growing white healthy roots past the lesion, while only 2% recovered naturally. A shorter-term container study was conducted in the laboratory to determine the effects of CWE-induced suppression when peas were grown in different substrates and to determine if the hydroponic nutrient solution had an effect on the suppression. Peas were grown in sterilized fine and coarse coconut coir fiber and sand irrigated with water, with a second set of fine coir irrigated with hydroponic nutrient solution. Pea seeds with 20-25mm radicles were inoculated with pathogen and sown directly into CWE-drenched substrate and grown for three weeks. At the end of the experiment, 80%, 60%, 90%, and 50% of the shoots of the inoculated, CWE-drenched seedlings remained healthy when grown in fine coir, coarse coir, sand, and fine coir irrigated with hydroponic nutrient solution, respectively. Nearly 100% of the roots grown in coconut coir substrates again developed necrotic lesions but 83%, 87%, 100%, and 87% grew healthy roots beyond the disease region. The hydroponic nutrient solution had a negative effect on suppression, with a reduction of at least 30 percentage points. Sand demonstrated a natural ability to suppress F. solani. Only 23% of inoculated seedlings had dead or dying shoots by the end of the experiment (compared to 77-80% in coir substrates) and although all but one of the roots developed lesions, all were able to recover on their own with CWE. CWE further increased shoot health and also prevented 57% of the roots from developing lesions. In a second study, two different CWE were used to examine the effect on root border cell dispersion and dynamics in pea, maize, cotton, and cucumber and its relation to disease suppression. Dispersal of border cells after immersion of roots into water or CWE was measured by direct observation over time using a compound microscope and stereoscope. Pictures were taken and the number of border cells released into suspension were enumerated by counting the total number of cells in aliquots taken from the suspension. Border cells formed a mass surrounding root tips within seconds after exposure to water, and most cells dispersed into suspension spontaneously. In CWE, >90% of the border cell population instead remained appressed to the root surface, even after vigorous agitation. This altered border cell phenomena was consistent for pea, maize, and cotton and for both CWE tested. For most cucumber roots (n=86/95), inhibition of border cell dispersal in both CWE was similar to that observed in pea, maize, and cotton. However, some individual cucumber roots (8±5%) exhibited a distinct phenotype. For example, border cells of one root immersed into CWE remained tightly adhered to the root tip even after 30 minutes while border cells of another root immersed at the same time in the same sample of CWE expanded significantly within 5 minutes and continued to expand over time. In a previous study, sheath development over time in growth pouches also was distinct in cucumber compared with pea, with detachment of the sheaths over time, and root infection was reduced by only 38% in cucumber compared with 100% protection in pea (Curlango-Rivera et al. 2013). Further research is needed to evaluate whether this difference in retention of border cell sheaths plays a role in the observed difference in inhibition of root infection. In the third study, a series of investigations were conducted to isolate different factors that contribute to the suppression ability of a CWE by changing incrementally changing some aspect of the CWE production process. The basic aerated CWE recipe (with molasses, kelp, humic acid, rock phosphate, and silica) provided 100% protection of pea from root disease while the non-aerated basic recipe CWE provided 72% protection. Aerated CWE made of only compost and water resulted in 58% protection. It was found that molasses did not contribute to the suppression ability of the ACWE, while kelp contributed strongly. When soluble kelp was added by itself to the compost and water, the CWE provided 80% suppression. However, when all additives were included except molasses and kelp, suppression remained high (93%) indicating that humic acids, rock phosphate, and/or silica were also major contributors toward the suppression effect. Optimal fermentation time for ACWE was 24 hr to achieve 100% suppression, with increased time resulting in inconsistent suppression results. Optimal fermentation time for NCWE was 3 days or 8 days. These studies are important contributions to understanding the differences that might be expected in CWE suppression when growing in different substrates, some of the factors in the production of CWE that affects the ability of a CWE to suppress disease, and the phenotypic effect CWE has on the root zone of plants and the possible relationship between that effect and disease suppression.

Submitted by Eliana Katia Pupim (katiapupim@tupa.unesp.br) on 2018-06-28T13:30:07Z No. of bitstreams: 1 pupim_ek_do_mar.pdf: 5107031 bytes, checksum: 8847c7bb4b7bf86bf7add53a1e945d29 (MD5) / Approved for entry into archive by Satie Tagara (satie@marilia.unesp.br) on 2018-06-29T14:31:46Z (GMT) No. of bitstreams: 1 pupim_ek_dr_mar.pdf: 5107031 bytes, checksum: 8847c7bb4b7bf86bf7add53a1e945d29 (MD5) / Made available in DSpace on 2018-06-29T14:31:46Z (GMT). No. of bitstreams: 1 pupim_ek_dr_mar.pdf: 5107031 bytes, checksum: 8847c7bb4b7bf86bf7add53a1e945d29 (MD5) Previous issue date: 2018-05-30 / Não recebi financiamento / A profissão do Engenheiro de Biossistemas ainda é recente no Brasil, sendo que a implantação do primeiro curso superior ocorreu em 2009. Em função da sua afinidade com a Engenharia Agrícola, os órgãos reguladores da profissão de Engenharia ainda não reconhecem seus egressos como Engenheiros de Biossistemas, mas sim os intitulam de Engenheiros Agrícolas com especificidade em Biossistemas. Diante desta problemática, o objetivo desta pesquisa foi traçar paralelos entre os campos científicos das Engenharia Agrícola e Engenharia de Biossistemas, de forma a permitir a elaboração de subsídios que levem à compreensão dos pontos de proximidade e de distanciamento dos domínios estudados. O método utilizado para alcançar o objetivo proposto foi a Análise de Domínio, optando pelas abordagens com Estudos Históricos e Estudos Bibliométricos, com a adoção da análise de conteúdo de Bardin e a elaboração de nuvens de palavras para as analogias. Os resultados permitiram compreender que sim, há vários momentos em que a Engenharia de Biossistemas e a Engenharia Agrícola são análogas, como na organização e na duração e integralização dos cursos, porém quanto aos conteúdos disseminados oferecem uma variação quanto aos temas relacionados à Zootecnia e Ciências Biológicas, sendo que enquanto a Engenharia de Biossistemas os trata como temas centrais, a Engenharia Agrícola os tem como temáticas periféricas. A produção científica também tem suas características peculiares, havendo periódicos que publicam apenas o conteúdo de uma ou de outra Engenharia, contudo há em número maior os periódicos que publicam as duas Engenharias concomitantemente, demonstrando que apesar de suas especificidades, há muito em comum na Engenharia de Biossistemas e na Engenharia Agrícola. / The profession of the Biosystems Engineer is very recent in Brazil. The first undergraduate major was implemented in 2009. Due to its affinity with Agricultural Engineering, the regulatory entities of the Engineering majors in Brazil do not recognize its graduates as Biosystems Engineers, but rather call them Agricultural Engineers with specificity in Biosystems. In view of this problem, the objective of this research was to draw parallels between the scientific fields of Agricultural Engineering and Biosystems Engineering, in order to find information that lead to understanding the points of proximity and distancing of the studied domains. The method used to reach the proposed goal was Domain Analysis, opting for approaches with Historical Studies and Bibliometric Studies, with the adoption of the Bardin content analysis and the elaboration of word clouds to allow for comparisons. The results allowed us to understand that there are several moments in which Biosystems Engineering and Agricultural Engineering are similar, as in the organization and duration and completion of the majors. However, the disseminated contents offer a variation regarding the subjects related to Zootechnics and Sciences Biological Sciences. While Engineering of Biosystems treats them as central themes, Agricultural Engineering has them as peripheral themes. Scientific production also has its peculiar characteristics, and there are periodicals that publish only the content of one or the other Engineering, however there are in greater number periodicals that publish the two Engineering concomitantly, demonstrating that despite their specificities, there is much in common in Engineering Biosystems and Agricultural Engineering.

La construction des biosystèmes d’oxydation passive du méthane (BOPM) est une option économique et durable pour réduire les émissions de méthane des sites d’enfouissement de déchets et des effets subséquents du réchauffement climatique. Les BOPM sont constitués de deux couches principales: la couche d'oxydation du méthane (MOL) et la couche de distribution du gaz (GDL). L'oxydation du méthane se produit dans la MOL par les réactions biochimiques des bactéries méthanotrophes, et la GDL est construite sous la MOL pour intercepter et distribuer les émissions fugitives de biogaz à la base de la MOL. Fondamentalement, l'efficacité d'un BOPM est définie en fonction de l'efficacité d'oxydation du méthane dans la MOL. Par conséquent, il est indispensable de fournir des conditions adéquates pour les activités bactériennes des méthanotrophes. En plus des paramètres environnementaux, l'intensité et la distribution du biogaz influencent l'efficacité des BOPM, et ils peuvent rendre le matériau de la MOL - avec une grande capacité d'accueillir les activités bactériennes - inutilisables en termes d'oxydation du méthane sur place. L'effet de barrière capillaire le long de l'interface entre la GDL et la MOL peut provoquer des émissions localisées de méthane, due à la restriction ou la distribution non uniforme de l’écoulement ascendant du biogaz à la base de la MOL. L'objectif principal de cette étude est d'incorporer le comportement hydraulique non saturé des BOPM dans la conception des BOPM, afin d’assurer la facilité et la distribution adéquates de l'écoulement du biogaz à la base de la MOL. Les fonctions de perméabilité à l'air des matériaux utilisés pour construire la MOL des BOPM expérimentaux au site d’enfouissement des déchets de St Nicéphore (Québec, Canada), ainsi que celles d'autres de la littérature technique, ont été étudiés pour évaluer le comportement d'écoulement non saturé du gaz dans les matériaux et pour identifier le seuil de migration sans restriction du gaz. Ce dernier seuil a été introduit en tant que un paramètre de conception avec lequel le critère de conception recommandé ici, c’est à dire la longueur de la migration sans restriction de gaz (LMSG), a été défini. La LMSG est considérée comme la longueur le long de l'interface entre la GDL et la MOL où le biogaz peut migrer à travers la MOL sans restriction. En réalisant des simulations numériques avec SEEP/W, les effets de la pente de l'interface, des paramètres définissant la courbe de rétention d'eau, de la fonction de la conductivité hydraulique du matériau de la MOL sur la valeur de la LMSG (représentant la facilité d'écoulement du biogaz à l'interface) et de la distribution de l'humidité (et par conséquent celle du biogaz) ont été évalués. Selon les résultats des simulations, la conductivité hydraulique saturée et la distribution des tailles de pores du matériau de la MOL sont les paramètres les plus importants sur la distribution de l'humidité le long de l'interface. Ce dernier paramètre influe également sur la valeur du degré de saturation et donc la facilité du biogaz à la base de la MOL. La densité sèche du matériau de MOL est un autre paramètre qui contrôle la facilité d'écoulement ascendant du biogaz. Les limitations principales de la présente étude sont associées au nombre de matériaux de MOL testés et à l'incapacité de SEEP/W de considérer l'évapotranspiration. Toutefois, compte tenu des hypothèses raisonnables dans les simulations et en utilisant les données de la littérature, on a essayé de réduire ces limitations. En utilisant les résultats des expériences et des simulations numériques, des étapes et des considérations de conception pour la sélection du matériau de MOL et de la pente d'interface ont été proposées. En effet,le comportement hydraulique non saturé des matériaux serait intégré dans les nécessités de conception pour un BOPM efficace, de sorte que la capacité maximale possible d'oxydation du méthane du matériau de la MOL soit exploitée. / Abstract : Implementation of passive methane oxidation biosystems (PMOB) is a cost effective and sustainable solution to eliminate the methane emissions of landfills to the atmosphere and ensuing global warming effects. PMOBs consist of two main layers: methane oxidation layer (MOL) and gas distribution layer (GDL). The oxidation of methane occurs in MOL through the biochemical reactions of methanotrophic bacteria, and GDL is constructed beneath the MOL to intercept and distribute the fugitive biogas emissions at the base of MOL. Basically, the efficiency of a PMOB is defined based on the methane oxidation efficiency in MOL. Therefore, it is indispensable to provide adequate conditions for the bacterial activities of methanotrophs. In addition to the environmental parameters, the intensity and the distribution of the biogas reaching the MOL material influence the efficiency of PMOBs, and they may cause the MOL material possessing great capacity to host the bacterial activities to be unserviceable in terms of in field methane oxidation. The capillary barrier effect along the GDL MOL interface may provoke localized surface methane emissions, resulted from the restricted and/or non uniform distribution of upward flow of biogas at the base of MOL. The main focus of present study is to incorporate the unsaturated hydraulic behavior of PMOBs into the design of PMOBs, providing adequate ease and distribution of upward flow of biogas at the base of MOL. The air permeability functions of the materials used to construct the MOL of experimental PMOBs at the St Nicephore landfill (Quebec, Canada), along with other materials from the technical literature, were studied to evaluate the unsaturated gas flow behavior of the materials and to identify the threshold of unrestricted gas migration. This latter threshold was introduced as a design parameter based on which the recommended design criterion herein, i.e. the length of unrestricted gas migration (LUGM), was defined. LUGM is considered as the length along the GDL MOL interface along which biogas can migrate upwards without restriction. Performing sets of numerical simulations in SEEP/W, the effect of slope of interface and the parameters defining the water retention curve and hydraulic conductivity function of MOL material on value of LUGM (representing the ease of upward flow of biogas at the interface) and distribution of moisture (and therefore biogas) along the GDL MOL interface were assessed. The saturated hydraulic conductivity and the pore size distribution of the MOL material were the most influencing parameters in distribution of moisture along the interface. The latter parameter influences also the value of degree of saturation and therefore, the ease of biogas at the base of MOL. Dry density of MOL material is another parameter that controls the ease of upward flow of biogas. The main limitations of the present study are associated with the number of tested MOL materials and the inability of SEEP/W in considering the evapotranspiration. However, considering reasonable assumptions in simulations and using the data from the literature, it was attempted to reduce the limitations. Based on the results of experiments and numerical simulations, some design steps and considerations for selection of the MOL material and the slope of interface were suggested that incorporate the unsaturated hydraulic behavior into the design necessities for an efficient PMOB so that the maximum possible methane oxidation capacity of MOL material is exploited.

La biologie synthétique est une science issue du rapprochement entre les biotechnologies et les sciences pour l’ingénieur. Elle consiste à créer de nouveaux systèmes biologiques par une combinaison rationnelle d’éléments biologiques standardisés, découplés de leur contexte naturel. L’environnement, l’agroalimentaire et la santé figurent parmi ses principaux domaines d’application. Cette thèse s’est focalisée sur les aspects liés à la conception ex-vivo de ces biosystèmes artificiels. A partir des analogies réalisées entre les processus biologiques et certaines fonctions électroniques, l’accent a été mis sur la réutilisation et l’adaptation des outils de conception numériques, supportant l’approche de conception « top-down ». Ainsi, une adaptation complète des méthodes de CAO de la microélectronique a été mise en place pour la biologie synthétique. Dans cette optique, les mécanismes biologiques élémentaires ont été modélisés sous plusieurs niveaux d’abstraction, allant de l’abstraction numérique à des modèles flux de signal et des modèles conservatifs. Des modèles en logique floue ont aussi été développés pour faire le lien entre ces niveaux d’abstraction. Ces différents modèles ont été implémentés avec deux langages de description matérielle et ont été validés sur la base de résultats expérimentaux de biosystèmes artificiels parmi les plus avancés. Parallèlement au travail de formalisation des modèles destinés au flot de conception, leur amélioration a aussi été étudiée : la modélisation des interactions entre plusieurs molécules a été rendue plus réaliste et le développement de modèles de bruits biologiques a également été intégré au processus. Cette thèse constitue donc une contribution importante dans la structuration et l’automatisation d’étapes de conception pour les biosystèmes synthétiques. Elle a permis de tracer les contours d’un flot de conception complet, adapté de la microélectronique, et d’en mettre en évidence les intérêts. / Synthetic biology is a science derived from the rapprochement between biotechnology and engineering science. It aims to create new biological systems through a rational combination between standardized biological elements which are disconnected from their natural context. Its main areas of application are the environment, the food-processing industry and the health sector. This thesis focuses on the ex vivo design aspects of these artificial biosystems. Thanks to analogies between biological processes and some electronic functions, the emphasis was put on reusing and adapting digital design tools that are fitting the top-down design approach. Thus, microelectronics CAD methods have been completely adapted to synthetic biology. In this regard, basic biological mechanisms have been modelled with various levels of abstraction, from digital abstraction to signal flow and conservative models. Fuzzy logic models have also been developed as a link between these levels of abstraction. These models have been implemented with two hardware description languages. They have been proven correct thanks to experimental results from state-of-the-art artificial biosystems. Concurrently to their formalization, improvements of design flow models have been studied: the modelling of interactions between several molecules have been made more realistic and the development of models for biological noise have been integrated to the process. This thesis is an important contribution to the structuring and the automation of some design steps for synthetic biosystems. It has made possible to highlight and to trace the outlines of a complete design flow, adapted from microelectronics.

Living cells and their components play a key role within biotechnology industry. Cell cultures and their products of interest are used for the design of vaccines as well as in the agro-alimentary field. In order to ensure optimal working of such bioprocesses, the understanding of the complex mechanisms which rule them is fundamental. Mathematical models may be helpful to grasp the biological phenomena which intervene in a bioprocess. Moreover, they allow prediction of system behaviour and are frequently used within engineering tools to ensure, for instance, product quality and reproducibility.<p> <p>Mathematical models of cell cultures may come in various shapes and be phrased with varying degrees of mathematical formalism. Typically, three main model classes are available to describe the nonlinear dynamic behaviour of such biological systems. They consist of macroscopic models which only describe the main phenomena appearing in a culture. Indeed, a high model complexity may lead to long numerical computation time incompatible with engineering tools like software sensors or controllers. The first model class is composed of the first principles or white box models. They consist of the system of mass balances for the main species (biomass, substrates, and products of interest) involved in a reaction scheme, i.e. a set of irreversible reactions which represent the main biological phenomena occurring in the considered culture. Whereas transport phenomena inside and outside the cell culture are often well known, the reaction scheme and associated kinetics are usually a priori unknown, and require special care for their modelling and identification. The second kind of commonly used models belongs to black box modelling. Black boxes consider the system to be modelled in terms of its input and output characteristics. They consist of mathematical function combinations which do not allow any physical interpretation. They are usually used when no a priori information about the system is available. Finally, hybrid or grey box modelling combines the principles of white and black box models. Typically, a hybrid model uses the available prior knowledge while the reaction scheme and/or the kinetics are replaced by a black box, an Artificial Neural Network for instance.<p><p>Among these numerous models, which one has to be used to obtain the best possible representation of a bioprocess? We attempt to answer this question in the first part of this work. On the basis of two simulated bioprocesses and a real experimental one, two model kinds are analysed. First principles models whose reaction scheme and kinetics can be determined thanks to systematic procedures are compared with hybrid model structures where neural networks are used to describe the kinetics or the whole reaction term (i.e. kinetics and reaction scheme). The most common artificial neural networks, the MultiLayer Perceptron and the Radial Basis Function network, are tested. In this work, pure black box modelling is however not considered. Indeed, numerous papers already compare different neural networks with hybrid models. The results of these previous studies converge to the same conclusion: hybrid models, which combine the available prior knowledge with the neural network nonlinear mapping capabilities, provide better results.<p><p>From this model comparison and the fact that a physical kinetic model structure may be viewed as a combination of basis functions such as a neural network, kinetic model structures allowing biological interpretation should be preferred. This is why the second part of this work is dedicated to the improvement of the general kinetic model structure used in the previous study. Indeed, in spite of its good performance (largely due to the associated systematic identification procedure), this kinetic model which represents activation and/or inhibition effects by every culture component suffers from some limitations: it does not explicitely address saturation by a culture component. The structure models this kind of behaviour by an inhibition which compensates a strong activation. Note that the generalization of this kinetic model is a challenging task as physical interpretation has to be improved while a systematic identification procedure has to be maintained.<p><p>The last part of this work is devoted to another kind of biological systems: proteins. Such macromolecules, which are essential parts of all living organisms and consist of combinations of only 20 different basis molecules called amino acids, are currently used in the industrial world. In order to allow their functioning in non-physiological conditions, industrials are open to modify protein amino acid sequence. However, substitutions of an amino acid by another involve thermodynamic stability changes which may lead to the loss of the biological protein functionality. Among several theoretical methods predicting stability changes caused by mutations, the PoPMuSiC (Prediction Of Proteins Mutations Stability Changes) program has been developed within the Genomic and Structural Bioinformatics Group of the Université Libre de Bruxelles. This software allows to predict, in silico, changes in thermodynamic stability of a given protein under all possible single-site mutations, either in the whole sequence or in a region specified by the user. However, PoPMuSiC suffers from limitations and should be improved thanks to recently developed techniques of protein stability evaluation like the statistical mean force potentials of Dehouck et al. (2006). Our work proposes to enhance the performances of PoPMuSiC by the combination of the new energy functions of Dehouck et al. (2006) and the well known artificial neural networks, MultiLayer Perceptron or Radial Basis Function network. This time, we attempt to obtain models physically interpretable thanks to an appropriate use of the neural networks.<p> / Doctorat en sciences appliquées / info:eu-repo/semantics/nonPublished

S tím, jak se neustále vyvíjejí nové technologie pro energeticky náročná průmyslová odvětví, stávající zařízení postupně zaostávají v efektivitě a produktivitě. Tvrdá konkurence na trhu a legislativa v oblasti životního prostředí nutí tato tradiční zařízení k ukončení provozu a k odstavení. Zlepšování procesu a projekty modernizace jsou zásadní v udržování provozních výkonů těchto zařízení. Současné přístupy pro zlepšování procesů jsou hlavně: integrace procesů, optimalizace procesů a intenzifikace procesů. Obecně se v těchto oblastech využívá matematické optimalizace, zkušeností řešitele a provozní heuristiky. Tyto přístupy slouží jako základ pro zlepšování procesů. Avšak, jejich výkon lze dále zlepšit pomocí moderní výpočtové inteligence. Účelem této práce je tudíž aplikace pokročilých technik umělé inteligence a strojového učení za účelem zlepšování procesů v energeticky náročných průmyslových procesech. V této práci je využit přístup, který řeší tento problém simulací průmyslových systémů a přispívá následujícím: (i)Aplikace techniky strojového učení, která zahrnuje jednorázové učení a neuro-evoluci pro modelování a optimalizaci jednotlivých jednotek na základě dat. (ii) Aplikace redukce dimenze (např. Analýza hlavních komponent, autoendkodér) pro vícekriteriální optimalizaci procesu s více jednotkami. (iii) Návrh nového nástroje pro analýzu problematických částí systému za účelem jejich odstranění (bottleneck tree analysis – BOTA). Bylo také navrženo rozšíření nástroje, které umožňuje řešit vícerozměrné problémy pomocí přístupu založeného na datech. (iv) Prokázání účinnosti simulací Monte-Carlo, neuronové sítě a rozhodovacích stromů pro rozhodování při integraci nové technologie procesu do stávajících procesů. (v) Porovnání techniky HTM (Hierarchical Temporal Memory) a duální optimalizace s několika prediktivními nástroji pro podporu managementu provozu v reálném čase. (vi) Implementace umělé neuronové sítě v rámci rozhraní pro konvenční procesní graf (P-graf). (vii) Zdůraznění budoucnosti umělé inteligence a procesního inženýrství v biosystémech prostřednictvím komerčně založeného paradigmatu multi-omics.