Optimisation of Kinetic Parameters for an After-treatment Catalyst

Mathematical modelling has become an essential tool in the design of modern catalytic systems. Emissions legislation is becoming increasingly stringent, and so mathematical models of aftertreatment systems must become more accurate in order to provide confidence that a catalyst will convert pollutants over the required range of conditions. 
Automotive catalytic converter models contain several sub-models that represent processes such as mass and heat transfer, and the rates at which the reactions proceed on the surface of the precious metal. Of these sub-models, the prediction of the surface reaction rates is by far the most challenging due to the complexity of the reaction system and the large number of gas species involved. The reaction rate sub-model uses global reaction kinetics to describe the surface reaction rate of the gas species and is based on the Langmuir Hinshelwood equation further developed by Voltz et al. [1] The reactions can be modelled using the pre-exponential and activation energies of the Arrhenius equations and the inhibition terms. 
The reaction kinetic parameters of aftertreatment models are found from experimental data, where a measured light-off curve is compared against a predicted curve produced by a mathematical model. The kinetic parameters are usually manually tuned to minimize the error between the measured and predicted data. This process is most commonly long, laborious and prone to misinterpretation due to the large number of parameters and the risk of multiple sets of parameters giving acceptable fits. Moreover, the number of coefficients increases greatly with the number of reactions. Therefore, with the growing number of reactions, the task of manually tuning the coefficients is becoming increasingly challenging. 
In the presented work, the authors have developed and implemented a multi-objective genetic algorithm to automatically optimize reaction parameters in AxiSuite®, [2] a commercial aftertreatment model. The genetic algorithm was developed and expanded from the code presented by Michalewicz et al. [3] and was linked to AxiSuite using the Simulink add-on for Matlab. 
The default kinetic values stored within the AxiSuite model were used to generate a series of light-off curves under rich conditions for a number of gas species, including CO, NO, C_{3H8 and C3H6. These light-off curves were used to generate an objective function.} 
This objective function was used to generate a measure of fit for the kinetic parameters. The multi-objective genetic algorithm was subsequently used to search between specified limits to attempt to match the objective function. In total the pre-exponential factors and activation energies of ten reactions were simultaneously optimized. 
The results reported here demonstrate that, given accurate experimental data, the optimization algorithm is successful and robust in defining the correct kinetic parameters of a global kinetic model describing aftertreatment processes.

Geotechnical failure risk reviewed

Slope instabilities – commonly triggered by rainfall – pose a geotechnical risk causing disruption to transport routes and incur significant financial resources. This article details laboratory, ground and remote sensing investigations carried out by Queen’s University Belfast and Transport Northern Ireland (TNI) to characterise and monitor slope instability on two higher risk infrastructure slopes in Northern Ireland. The research is used to update a noninvasive risk assessment model of slopes across the country’s road network to direct resources for future investigation.

Consistency of arsenic speciation in global tobacco products with implications for health and regulation

Background: Tobacco smoke is a major risk to the health of its users and arsenic is among the components of smoke present at concentrations of toxicological concern. There are significant variations in human toxicity between inorganic and organic arsenic species and the aim of this study was to determine whether there are predictable relationships among major arsenic species in tobacco that could be useful for risk assessment.

Methods: 14 samples of tobacco were studied spanning a wide range of concentrations in samples from different geographical regions, including certified reference materials and cigarette products. Inorganic and major organic arsenic species were extracted from powdered tobacco samples by nitric acid using microwave digestion. Concentrations of arsenic species in these extracts were determined using HPLC-ICPMS.

Results: The concentrations of total inorganic arsenic species range from 144 to 3914 mu g kg(-1), while organic species dimethylarsinic acid (DMA) ranges from 21 to 176 mu g As kg(-1), and monomethylarsonic acid (MA) ranges from 30 to 116 mu g kg(-1). The percentage of species eluted compared to the total arsenic extracted ranges from 11.1 to 36.8% suggesting that some As species (possibly macro-molecules, strongly complexed or in organic forms) do not elute from the column. This low percentage of column-speciated arsenic is indicative that more complex forms of arsenic exist in the tobacco. All the analysed species correlate positively with total arsenic concentration over the whole compositional range and regression analysis indicates a consistent ratio of about 4:1 in favour of inorganic arsenic compared with MA + DMA.

Conclusions: The dominance of inorganic arsenic species among those components analysed is a marked feature of the diverse range of tobaccos selected for study. Such consistency is important in the context of a WHO expert panel recommendation to regulate tobacco crops and products using total arsenic concentration. If implemented more research would be required to develop models that accurately predict the smoker's exposure to reduced inorganic arsenic species on the basis of leaf or product concentration and product design features.

Quality Assurance for GCxGC-FID, Accreditation for TPH CWG Protocol

A new technological approach in the analysis and forensic interpretation of Total Hydrocarbons in soils and waters using 2D Gas Chromatography method (GC-GC) was developed alongside environmental forensic and the assessment models to provide better customer products for the environmental industry.
The objective was to develop an analytical methodology for TPH CWG. Raw data from this method is then to be evaluated for forensic interpretation and risk assessment modelling. Access will be made available to the expertise in methods of forensic tracing contaminant sources, transport modelling, human health risk modelling and detailed quantitative risk assessment.
The quantification of internal standards was key to the development of this method. As the laboratory does not test for TPH in 1D, it was requested during INAB ISO 17025 audit to individually map out where each compound falls chromatographically in the 2D. This was done through comparing carbon equivalent numbers to the n-alkane carbons. This proved e.g. 2-methylnaphthalene has 11 carbons in its structure; its carbon equivalent is 12.84 , the result of which falls within the band of Aromatic eC12-eC16 as opposed to expected eC10-eC12. This was carried out for all 16 PAH (polyaromatic hydrocarbons) and BTEX (benzene, toluene, ethylbenzene and o, m and p-xylenes). The n-alkanes were also assigned to their corresponding aliphatic bands e.g. nC8 would be expected to be in nC8-nC10.
The method was validated through a designated systematic experimental protocol and was challenged with spikes of known concentration of hydrocarbon parameters such as recoveries, precision, bias and linearity. The method was verified by testing a certified reference material which was used as a proficiency round of testing for numerous laboratories.
It is hoped that the method will be used in conjunction with the analysis through Bonn Agreement with their OSINet group. This is a panel of experts and laboratories (including CLS) who forensically identify oil spill contamination from a water source.
This method can prove itself to be a robust method and benefit the industry for contaminated land and water but the method needs to be seen as separate from the regular 1D chromatography. It will help identify contaminants and assist consultants, regulators, clients and scientists valuable information not seen in 1D

A primer for the Environmental Impact Assessment of mining at seafloor massive sulfide deposits

Seafloor massive sulfides (SMS) contain commercially viable quantities of high grade ores, making them attractive prospect sites for marine mining. SMS deposits may also contain hydrothermal vent ecosystems populated by high conservation value vent-endemic species. Responsible environmental management of these resources is best achieved by the adoption of a precautionary approach. Part of this precautionary approach involves the Environmental Impact Assessment (EIA) of exploration and exploitative activities at SMS deposits. The VentBase 2012 workshop provided a forum for stakeholders and scientists to discuss issues surrounding SMS exploration and exploitation. This forum recognised the requirement for a primer which would relate concepts underpinning EIA at SMS deposits. The purpose of this primer is to inform policy makers about EIA at SMS deposits in order to aid management decisions. The primer offers a basic introduction to SMS deposits and their associated ecology, and the basic requirements for EIA at SMS deposits; including initial data and information scoping, environmental survey, and ecological risk assessment. © 2013 Elsevier Ltd.

Clopidogrel use and cancer-specific mortality: a population-based cohort study of colorectal, breast and prostate cancer patients

PURPOSE: Concerns were raised about the safety of antiplatelet thienopyridine derivatives after a randomized control trial reported increased risks of cancer and cancer deaths in prasugrel users. We investigate whether clopidogrel, a widely used thienopyridine derivative, was associated with increased risk of cancer-specific or all-cause mortality in cancer patients.

METHODS: Colorectal, breast and prostate cancer patients, newly diagnosed from 1998 to 2009, were identified from the National Cancer Data Repository. Cohorts were linked to the UK Clinical Practice Research Datalink, providing prescription records, and to the Office of National Statistics mortality data (up to 2012). Unadjusted and adjusted hazard ratios (HRs) for cancer-specific and all-cause mortality in post-diagnostic clopidogrel users were calculated using time-dependent Cox regression models.

RESULTS: The analysis included 10 359 colorectal, 17 889 breast and 13 155 prostate cancer patients. There was no evidence of an increase in cancer-specific mortality in clopidogrel users with colorectal (HR = 0.98 95% confidence interval (CI) 0.77, 1.24) or prostate cancer (HR = 1.03 95%CI 0.82, 1.28). There was limited evidence of an increase in breast cancer patients (HR = 1.22 95%CI 0.90, 1.65); however, this was attenuated when removing prescriptions in the year prior to death.

CONCLUSIONS: This novel study of large population-based cohorts of colorectal, breast and prostate cancer patients found no evidence of an increased risk of cancer-specific mortality among colorectal, breast and prostate cancer patients using clopidogrel.

Optimal control and numerical optimization applied to epidemiological models

A relação entre a epidemiologia, a modelação matemática e as ferramentas computacionais permite construir e testar teorias sobre o desenvolvimento e combate de uma doença. Esta tese tem como motivação o estudo de modelos epidemiológicos aplicados a doenças infeciosas numa perspetiva de Controlo Ótimo, dando particular relevância ao Dengue. Sendo uma doença tropical e subtropical transmitida por mosquitos, afecta cerca de 100 milhões de pessoas por ano, e é considerada pela Organização Mundial de Saúde como uma grande preocupação para a saúde pública. Os modelos matemáticos desenvolvidos e testados neste trabalho, baseiam-se em equações diferenciais ordinárias que descrevem a dinâmica subjacente à doença nomeadamente a interação entre humanos e mosquitos. É feito um estudo analítico dos mesmos relativamente aos pontos de equilíbrio, sua estabilidade e número básico de reprodução. A propagação do Dengue pode ser atenuada através de medidas de controlo do vetor transmissor, tais como o uso de inseticidas específicos e campanhas educacionais. Como o desenvolvimento de uma potencial vacina tem sido uma aposta mundial recente, são propostos modelos baseados na simulação de um hipotético processo de vacinação numa população. Tendo por base a teoria de Controlo Ótimo, são analisadas as estratégias ótimas para o uso destes controlos e respetivas repercussões na redução/erradicação da doença aquando de um surto na população, considerando uma abordagem bioeconómica. Os problemas formulados são resolvidos numericamente usando métodos diretos e indiretos. Os primeiros discretizam o problema reformulando-o num problema de optimização não linear. Os métodos indiretos usam o Princípio do Máximo de Pontryagin como condição necessária para encontrar a curva ótima para o respetivo controlo. Nestas duas estratégias utilizam-se vários pacotes de software numérico. Ao longo deste trabalho, houve sempre um compromisso entre o realismo dos modelos epidemiológicos e a sua tratabilidade em termos matemáticos.

Genotoxic risk of herbicides to Anguilla anguilla L.

No contexto dos contaminantes aquáticos, os herbicidas são considerados como um dos grupos mais perigosos. Uma vez aplicados, estes são facilmente transportados para cursos de água, quer devido a uma pulverização pouco cuidada ou devido a fenómenos de escorrência superficial e/ou subterrânea. A presença destes agroquímicos no ambiente tem vindo a ser associada a efeitos nefastos em organismos não-alvo, como é o caso dos peixes. Contudo, existe ainda uma grande lacuna no que diz respeito à informação científica relacionada com o seu impacto genotóxico. Deste modo, a presente tese foi delineada com o intuito de avaliar o risco genotóxico em peixes de duas formulações de herbicidas: o Roundup®, que tem como princípio activo o glifosato, e o Garlon®, que apresenta o triclopir na base da sua constituição, produtos estes largamente utilizados na limpeza de campos agrícolas, assim como em florestas. Foi ainda planeado desenvolver uma base de conhecimento no que diz respeito aos mecanismos de dano do ADN. Como último objectivo, pretendeu-se contribuir para a mitigação dos efeitos dos agroquímicos no biota aquático, nomeadamente em peixes, fornecendo dados científicos no sentido de melhorar as práticas agrícolas e florestais. Este estudo foi realizado adoptando a enguia europeia (Anguilla anguilla L.) como organismo-teste, e submetendo-a a exposições de curta duração (1 e 3 dias) dos produtos comerciais mencionados, em concentrações consideradas ambientalmente realistas. Para a avaliação da genotoxicidade foram aplicadas duas metodologias: o ensaio do cometa e o teste das anomalias nucleares eritrocíticas (ANE). Enquanto o ensaio do cometa detecta quebras na cadeia do ADN, um dano passível de ser reparado, o aparecimento das ANE revela lesões cromossomais, sinalizando um tipo de dano de difícil reparação. O ensaio do cometa foi ainda melhorado com uma nova etapa que incluiu a incubação com enzimas de reparação (FPG e EndoIII), permitindo perceber a ocorrência de dano oxidativo no ADN. No que diz respeito ao Roundup®, o envolvimento do sistema antioxidante como indicador de um estado próoxidante foi também alvo de estudo. Uma vez que as referidas formulações se apresentam sob a forma de misturas, o potencial genotóxico dos seus princípios activos foi também avaliado individualmente. No caso particular do Roundup®, também foram estudados o seu surfactante (amina polietoxilada; POEA) e o principal metabolito ambiental (ácido aminometilfosfórico; AMPA). Os resultados obtidos mostraram a capacidade do Roundup® em induzir tanto dano no ADN (em células de sangue, guelras e fígado) como dano cromossómico (em células de sangue). A investigação sobre o possível envolvimento do stresse oxidativo demonstrou que o tipo de dano no ADN varia com as concentrações testadas e com a duração da exposição. Deste modo, com o aumento do tempo de exposição, os processos relacionados com o envolvimento de espécies reactivas de oxigénio (ERO) ganharam preponderância como mecanismo de dano no ADN, facto que é corroborado pela activação do sistema antioxidante observado nas guelras, assim como pelo aumento dos sítios sensíveis a FPG em hepatócitos. O glifosato e o POEA foram também considerados genotóxicos. O POEA mostrou induzir uma maior extensão de dano no ADN, tanto comparado com o glifosato como com a mistura comercial. Apesar de ambos os componentes contribuirem para a genotoxicidade da formulação, a soma dos seus efeitos individuais nunca foi observada, apontando para um antagonismo entre eles e indicando que o POEA não aumenta o risco associado ao princípio activo. Deste modo, realça-se a necessidade de regulamentar limiares de segurança para todos os componentes da formulação, recomendando, em particular, a revisão da classificação do risco do POEA (actualmente classificado com “inerte”). Uma vez confirmada a capacidade do principal metabolito do glifosato – AMPA – em exercer dano no ADN assim como dano cromossómico, os produtos da degradação ambiental dos princípios activos assumem-se como um problema silencioso, realçando assim a importância de incluir o AMPA na avaliação do risco relacionado com herbicidas com base no glifosato. A formulação Garlon® e o seu princípio activo triclopir mostraram um claro potencial genotóxico. Adicionalmente, o Garlon® mostrou possuir um potencial genotóxico mais elevado do que o seu princípio activo. No entanto, a capacidade de infligir dano oxidativo no ADN não foi demonstrada para nenhum dos agentes. No que concerne à avaliação da progressão do dano após a remoção da fonte de contaminação, nem os peixes expostos a Roundup® nem os expostos a Garlon® conseguiram restaurar completamente a integridade do seu ADN ao fim de 14 dias. No que concerne ao Roundup®, o uso de enzimas de reparação de lesões específicas do ADN associado ao teste do cometa permitiu detectar um aparecimento tardio de dano oxidativo, indicando deste modo um decaimento progressivo da protecção antioxidante e ainda uma incapacidade de reparar este tipo de dano. O período de pós-exposição correspondente ao Garlon® revelou uma tendência de diminuição dos níveis de dano, apesar de nunca se observar uma completa recuperação. Ainda assim, foi evidente uma intervenção eficiente das enzimas de reparação do ADN, mais concretamente as direccionadas às purinas oxidadas. A avaliação das metodologias adoptadas tornou evidente que o procedimento base do ensaio do cometa, que detecta apenas o dano nãoespecífico no ADN, possui algumas limitações quando comparado com a metodologia que incluiu a incubação com as enzimas de reparação, uma vez que a última mostrou reduzir a possibilidade de ocorrência de resultados falsos negativos. Os dois parâmetros adoptados (ensaio do cometa e teste das ANE) demonstraram possuir aptidões complementares, sendo assim recomendado a sua utilização conjunta com vista a efectuar uma avaliação mais adequada do risco genotóxico. Globalmente, os resultados obtidos forneceram indicações de grande utilidade para as entidades reguladoras, contribuindo ainda para a (re)formulação de medidas de conservação do ambiente aquático. Neste sentido, os dados obtidos apontam para a importância da avaliação de risco dos herbicidas incluir testes de genotoxicidade. A magnitude de risco detectada para ambas as formulações adverte para a necessidade de adopção de medidas restritivas em relação à sua aplicação na proximidade de cursos de água. Como medidas mitigadoras de impactos ambientais, aponta-se o desenvolvimento de formulações que incorporem adjuvantes selecionados com base na sua baixa toxicidade.

Toxicidade de químicos em mistura: o caso da albufeira do Alqueva

In the environment humans and biota are generally exposed to chemical mixtures rather than individual chemicals. Therefore, when assessing the environmental risk of chemicals, it is important to consider chemical mixtures and their possible interactions. The main objective of this work focused on the environmental risk assessment of pesticides found in the water of the Alqueva reservoir and their binary combinations. In this aquatic ecosystem several pesticides were above of the environmental quality standards. But in addition, there were several sampling points of the reservoir where ecotoxicity was observed despite the presence of these contaminants at low concentrations. Here, a component-based approach was used to assess the effects of the pesticide mixtures. The effects of the binary combinations of four herbicides, atrazine (ATR), terbuthylazine (TER), simazine (SIM) and metolachlor (MET), on the growth rate of the microalgae Pseudokirchneriella subcapitata and the effects of the binary combinations of the s-triazine herbicides ATR and TER and the insecticide chlorpyrifos (CPF) on the swimming behaviour and acetylcholinesterase (AChE) activity of the zebrafish Danio rerio were assessed using the two reference models of concentration addition (CA) and independent action (IA). Moreover, the combined effects of the herbicides (ATR, TER and MET) and the insecticide CPF were also tested on the swimming behaviour and AChE activity of the aquatic midge Chironomus riparius after the cholinesterases characterization. In this risk characterization, the calculated risk quotients for the herbicides ATR, TER, SIM and MET were higher than 1, meaning that these herbicides present a high risk for the Alqueva ecosystem. As expected, the microalgae P. subcapitata was the most sensitive species to the herbicides. However, despite these herbicides pose no or low risk to other aquatic organisms tested in this study, with EC50 values much higher than the concentrations found in this aquatic ecosystem, they are able to increase the toxic effects of CPF when they are tested in binary mixtures. Moreover, the risk quotients of mixtures of these herbicides present simultaneously in three different locations of the reservoir were also higher than 1, so this confirms the fact that these herbicides when present in mixtures, present a greater risk for this ecosystem than the expected considering each single chemical by its own.

Effect assessment of nanoparticles toxicity in the terrestrial compartment

Over 11 million tons of nanomaterials (NMs) have been produced in 2012 and predictions point the increase in production. Despite predictions and extended usage via consumer products and industry, the understanding of the potential impact of these materials on the environment is virtually absent. The main aim of this thesis is to understand how a selected group of nanomaterials (metal based particles) may impact soil invertebrates, with special focus on the mechanisms of response. Since a case-by-case Environmental Risk Assessment (ERA) of all the emerging contaminants (particularly NMs) is impossible, among others due to time and cost reasons, to gain understanding on the mechanism of action and response is very important to reach a common paradigm. Understanding the modes of action provides predictive characters in cross particle extrapolation. Besides, it also provides insight for the production of new and sustainable materials. Overall, the effects of the selected NMs (Copper and Silver, Titanium and Zirconium oxides) and the respective salt forms, were investigated at the gene expression (using high-throughput tools, microarray and qPCR technology), biochemical (using enzymatic assays for analysis of oxidative stress markers) and organism (survival and reproduction as in OECD test guidelines) levels, this using standard soil species (Enchytraeus albidus, Enchytraeus crypticus, Eisenia fetida). Gene expression analysis provided valuable information on the mechanisms affected by each of the NMs. The gene expression profile highlighted a (nano)material signature and the effect of the duration of exposure. The functional analyses integrated with the biochemical and organism data, revealed a good understanding power. The biochemical parameters (oxidative stress related) were distinct across the materials and also influenced by duration of exposure and concentration. The standardized organismal responses differed the least between the various materials. The overall outcome is that, in this context of NMs effect assessment, gene expression and enzymatic assays introduced a very important knowledge gap, which could not had been achieved by the standard organismal effects alone. A reoccurring issue with some metal based NMs is the possible dissolution and subsequent release of ions that then causes toxicity e.g. Cu-NPs or Ag-NPs release Cu2+ or Ag+. The oxidation state of the particles was investigated, although this was not the focus of the thesis. The study of fate, e.g. dissolution of NPs, is also only in its beginning and the appropriate techniques are currently being developed. The results showed a specific nanoparticle effect. The UV exposure with titanium dioxide nanoparticles increased its effect.

Seismic vulnerability assessment of slender masonry structures

Slender masonry structures are distributed all over the world and constitute a relevant part of the architectural and cultural heritage of humanity. Their protection against earthquakes is a topic of great concern among the scientific community. This concern mainly arises from the strong damage or complete loss suffered by this group of structures due to catastrophic events and the need and interest to preserve them. Although the great progress in technology, and in the knowledge of seismology and earthquake engineering, the preservation of these brittle and massive structures still represents a major challenge. Based on the research developed in this work it is proposed a methodology for the seismic risk assessment of slender masonry structures. The proposed methodology was applied for the vulnerability assessment of Nepalese Pagoda temples which follow very simple construction procedure and construction detailing in relation to seismic resistance requirements. The work is divided in three main parts. Firstly, particular structural fragilities and building characteristics of the important UNESCO classified Nepalese Pagoda temples which affect their seismic performance and dynamic properties are discussed. In the second part the simplified method proposed for seismic vulnerability assessment of slender masonry structures is presented. Finally, the methodology proposed in this work is applied to study Nepalese Pagoda temples, as well as in the efficiency assessment of seismic performance improvement solution compatible with original cultural and technological value.

Organic contaminants in urban soils: major inputs and potential risks

Urban soil quality may be severely affected by hydrophobic organic contaminants (HOCs), impairing environmental quality and human health. A comprehensive study was conducted in two contrasting Portuguese urban areas (Lisbon and Viseu) in order to assess the levels and potential risks of these contaminants, to identify sources and study their behaviour in soils. The concentrations of HOCs were related to the size of the city, with much higher contamination levels observed in Lisbon urban area. Source apportionment was performed by studying the HOCs profiles, their relationship with potentially toxic elements and general characteristics of soil using multivariate statistical methods. Lisbon seems to be affected by nearby sources (traffic, industry and incineration processes) whereas in Viseu the atmospheric transport may be playing an important role. In a first tier of risk assessment (RA) it was possible to identify polycyclic aromatic hydrocarbons (PAHs) in Lisbon soils as a potential hazard. The levels of PAHs in street dusts were further studied and allowed to clarify that traffic, tire and pavement debris can be an important source of PAHs to urban soils. Street dusts were also identified as being a potential concern regarding human and environmental health, especially if reaching the nearby aquatic bodies. Geostatistical tools were also used and their usefulness in a RA analysis and urban planning was discussed. In order to obtain a more realistic assessment of risks of HOCs to environment and human health it is important to evaluate their available fraction, which is also the most accessible for organisms. Therefore, a review of the processes involved on the availability of PAHs was performed and the outputs produced by the different chemical methods were evaluated. The suitability of chemical methods to predict bioavailability of PAHs in dissimilar naturally contaminated soils has not been demonstrated, being especially difficult for high molecular weight compounds. No clear relationship between chemical and biological availability was found in this work. Yet, in spite of the very high total concentrations found in some Lisbon soils, both the water soluble fraction and the body residues resulting from bioaccumulation assays were generally very low, which may be due to aging phenomena. It was observed that the percentage of soluble fraction of PAHs in soils was found to be different among compounds and mostly regulated by soil properties. Regarding bioaccumulation assays, although no significant relationship was found between soil properties and bioavailability, it was verified that biota-to-soil bioaccumulation factors were sample dependent rather than compound dependent. In conclusion, once the compounds of potential concern are targeted, then performing a chemical screening as a first tier can be a simple and effective approach to start a RA. However, reliable data is still required to improve the existing models for risk characterization.

Cape Verde coastal ecosystem: a study of community structure, trophic interactions and exploitation pattern

Tese de dout., Ciências e Tecnologia das Pescas, Faculdade de Ciências do Mar e do Ambiente, Universidade do Algarve, 2005

Design of neuro-fuzzy models by evolutionary and gradient-based algorithms

All systems found in nature exhibit, with different degrees, a nonlinear behavior. To emulate this behavior, classical systems identification techniques use, typically, linear models, for mathematical simplicity. Models inspired by biological principles (artificial neural networks) and linguistically motivated (fuzzy systems), due to their universal approximation property, are becoming alternatives to classical mathematical models. In systems identification, the design of this type of models is an iterative process, requiring, among other steps, the need to identify the model structure, as well as the estimation of the model parameters. This thesis addresses the applicability of gradient-basis algorithms for the parameter estimation phase, and the use of evolutionary algorithms for model structure selection, for the design of neuro-fuzzy systems, i.e., models that offer the transparency property found in fuzzy systems, but use, for their design, algorithms introduced in the context of neural networks. A new methodology, based on the minimization of the integral of the error, and exploiting the parameter separability property typically found in neuro-fuzzy systems, is proposed for parameter estimation. A recent evolutionary technique (bacterial algorithms), based on the natural phenomenon of microbial evolution, is combined with genetic programming, and the resulting algorithm, bacterial programming, advocated for structure determination. Different versions of this evolutionary technique are combined with gradient-based algorithms, solving problems found in fuzzy and neuro-fuzzy design, namely incorporation of a-priori knowledge, gradient algorithms initialization and model complexity reduction.

Mathematical modelling of supercritical CO2 extraction of volatile oils from aromatic plants

The modelling of the experimental data of the extraction of the volatile oil from six aromatic plants (coriander, fennel, savoury, winter savoury, cotton lavender and thyme) was performed using five mathematical models, based on differential mass balances. In all cases the extraction was internal diffusion controlled and the internal mass transfer coefficienty (k(s)) have been found to change with pressure, temperature and particle size. For fennel, savoury and cotton lavender, the external mass transfer and the equilibrium phase also influenced the second extraction period, since k(s) changed with the tested flow rates. In general, the axial dispersion coefficient could be neglected for the conditions studied, since Peclet numbers were high. On the other hand, the solute-matrix interaction had to be considered in order to ensure a satisfactory description of the experimental data.