Redes elétricas inteligentes: desenvolvimento de modelos computacionais e aplicações para a gestão do lado da procura

Esta tese incide sobre o desenvolvimento de modelos computacionais e de aplicações para a gestão do lado da procura, no âmbito das redes elétricas inteligentes. É estudado o desempenho dos intervenientes da rede elétrica inteligente, sendo apresentado um modelo do produtor-consumidor doméstico. O problema de despacho económico considerando previsão de produção e consumo de energia obtidos a partir de redes neuronais artificiais é apresentado. São estudados os modelos existentes no âmbito dos programas de resposta à procura e é desenvolvida uma ferramenta computacional baseada no algoritmo de fuzzy-clustering subtrativo. São analisados perfis de consumo e modos de operação, incluindo uma breve análise da introdução do veículo elétrico e de contingências na rede de energia elétrica. São apresentadas aplicações para a gestão de energia dos consumidores no âmbito do projeto piloto InovGrid. São desenvolvidos sistemas de automação para, aquisição monitorização, controlo e supervisão do consumo a partir de dados fornecidos pelos contadores inteligente que permitem a incorporação das ações dos consumidores na gestão do consumo de energia elétrica; SMART GRIDS - COMPUTATIONAL MODELS DEVELOPMENT AND DEMAND SIDE MANAGMENT APPLICATIONS Abstract: This thesis focuses on the development of computational models and its applications on the demand side management within the smart grid scope. The performance of the electrical network players is studied and a domestic prosumer model is presented. The economic dispatch problem considering the production forecast and the energy consumption obtained from artificial neural networks is also presented. The existing demand response models are studied and a computational tool based on the fuzzy subtractive clustering algorithm is developed. Energy consumption profiles and operational modes are analyzed, including a brief analysis of the electrical vehicle and contingencies on the electrical network. Consumer energy management applications within the scope of InovGrid pilot project are presented. Computational systems are developed for the acquisition, monitoring, control and supervision of consumption data provided by smart meters allowing to incorporate consumer actions on their electrical energy management.

The role of context in image annotation and recommendation

With the rise of smart phones, lifelogging devices (e.g. Google Glass) and popularity of image sharing websites (e.g. Flickr), users are capturing and sharing every aspect of their life online producing a wealth of visual content. Of these uploaded images, the majority are poorly annotated or exist in complete semantic isolation making the process of building retrieval systems difficult as one must firstly understand the meaning of an image in order to retrieve it. To alleviate this problem, many image sharing websites offer manual annotation tools which allow the user to “tag” their photos, however, these techniques are laborious and as a result have been poorly adopted; Sigurbjörnsson and van Zwol (2008) showed that 64% of images uploaded to Flickr are annotated with < 4 tags. Due to this, an entire body of research has focused on the automatic annotation of images (Hanbury, 2008; Smeulders et al., 2000; Zhang et al., 2012a) where one attempts to bridge the semantic gap between an image’s appearance and meaning e.g. the objects present. Despite two decades of research the semantic gap still largely exists and as a result automatic annotation models often offer unsatisfactory performance for industrial implementation. Further, these techniques can only annotate what they see, thus ignoring the “bigger picture” surrounding an image (e.g. its location, the event, the people present etc). Much work has therefore focused on building photo tag recommendation (PTR) methods which aid the user in the annotation process by suggesting tags related to those already present. These works have mainly focused on computing relationships between tags based on historical images e.g. that NY and timessquare co-exist in many images and are therefore highly correlated. However, tags are inherently noisy, sparse and ill-defined often resulting in poor PTR accuracy e.g. does NY refer to New York or New Year? This thesis proposes the exploitation of an image’s context which, unlike textual evidences, is always present, in order to alleviate this ambiguity in the tag recommendation process. Specifically we exploit the “what, who, where, when and how” of the image capture process in order to complement textual evidences in various photo tag recommendation and retrieval scenarios. In part II, we combine text, content-based (e.g. # of faces present) and contextual (e.g. day-of-the-week taken) signals for tag recommendation purposes, achieving up to a 75% improvement to precision@5 in comparison to a text-only TF-IDF baseline. We then consider external knowledge sources (i.e. Wikipedia & Twitter) as an alternative to (slower moving) Flickr in order to build recommendation models on, showing that similar accuracy could be achieved on these faster moving, yet entirely textual, datasets. In part II, we also highlight the merits of diversifying tag recommendation lists before discussing at length various problems with existing automatic image annotation and photo tag recommendation evaluation collections. In part III, we propose three new image retrieval scenarios, namely “visual event summarisation”, “image popularity prediction” and “lifelog summarisation”. In the first scenario, we attempt to produce a rank of relevant and diverse images for various news events by (i) removing irrelevant images such memes and visual duplicates (ii) before semantically clustering images based on the tweets in which they were originally posted. Using this approach, we were able to achieve over 50% precision for images in the top 5 ranks. In the second retrieval scenario, we show that by combining contextual and content-based features from images, we are able to predict if it will become “popular” (or not) with 74% accuracy, using an SVM classifier. Finally, in chapter 9 we employ blur detection and perceptual-hash clustering in order to remove noisy images from lifelogs, before combining visual and geo-temporal signals in order to capture a user’s “key moments” within their day. We believe that the results of this thesis show an important step towards building effective image retrieval models when there lacks sufficient textual content (i.e. a cold start).

Development of procedures for the triazole fungicides determination in fruits and liquid samples using microextraction techniques and chromatographic separation

Agricultural crops can be damaged by funguses, insects, worms and other organisms that cause diseases and decrease the yield of production. The effect of these damaging agents can be reduced using pesticides. Among them, triazole compounds are effective substances against fungus; for example, Oidium. Nevertheless, it has been detected that the residues of these fungicides in foods as well as in derivate products can affect the health of the consumers. Therefore, the European Union has established several regulations fixing the maximum residue of pesticide levels in a wide range of foods trying to assure the consumer safety. Hence, it is very important to develop adequate methods to determine these pesticide compounds. In most cases, gas or liquid chromatographic (GC, LC) separations are used in the analysis of the samples. But firstly, it is necessary to use proper sample treatments in order to preconcentrate and isolate the target analytes. To reach this aim, microextraction techniques are very effective tools; because allow to do both preconcentration and extraction of the analytes in one simple step that considerably reduces the source of errors. With these objectives, two remarkable techniques have been widely used during the last years: solid phase microextraction (SPME) and liquid phase microextraction (LPME) with its different options. Both techniques that avoid the use or reduce the amount of toxic solvents are convenient coupled to chromatographic equipments providing good quantitative results in a wide number of matrices and compounds. In this work simple and reliable methods have been developed using SPME and ultrasound assisted emulsification microextraction (USAEME) coupled to GC or LC for triazole fungicides determination. The proposed methods allow confidently determine triazole concentrations of μg L‐1 order in different fruit samples. Chemometric tools have been used to accomplish successful determinations. Firstly, in the selection and optimization of the variables involved in the microextraction processes; and secondly, to overcome the problems related to the overlapping peaks. Different fractional factorial designs have been used for the screening of the experimental variables; and central composite designs have been carried out to get the best experimental conditions. Trying to solve the overlapping peak problems multivariate calibration methods have been used. Parallel Factor Analysis 2 (PARAFAC2), Multivariate Curve Resolution (MCR) and Parallel Factor Analysis with Linear Dependencies (PARALIND) have been proposed, the adequate algorithms have been used according to data characteristics, and the results have been compared. Because its occurrence in Basque Country and its relevance in the production of cider and txakoli regional wines the grape and apple samples were selected. These crops are often treated with triazole compounds trying to solve the problems caused by the funguses. The peel and pulp from grape and apple, their juices and some commercial products such as musts, juice and cider have been analysed showing the adequacy of the developed methods for the triazole determination in this kind of fruit samples.

Processamento Inteligente de Sinais de Pressão e Temperatura Adquiridos Através de Sensores Permanentes em Poços de Petróleo

Originally aimed at operational objectives, the continuous measurement of well bottomhole pressure and temperature, recorded by permanent downhole gauges (PDG), finds vast applicability in reservoir management. It contributes for the monitoring of well performance and makes it possible to estimate reservoir parameters on the long term. However, notwithstanding its unquestionable value, data from PDG is characterized by a large noise content. Moreover, the presence of outliers within valid signal measurements seems to be a major problem as well. In this work, the initial treatment of PDG signals is addressed, based on curve smoothing, self-organizing maps and the discrete wavelet transform. Additionally, a system based on the coupling of fuzzy clustering with feed-forward neural networks is proposed for transient detection. The obtained results were considered quite satisfactory for offshore wells and matched real requisites for utilization

Association of kinase insert domain-containing receptor (KDR) gene polymorphism/ haplotypes with recurrent spontaneous abortion and genetic structure

Background: Recurrent spontaneous abortion is one of the diseases that can lead to physical, psychological, and, economical problems for both individuals and society. Recently a few numbers of genetic polymorphisms in kinase insert domain-containing receptor (KDR) gene are examined that can endanger the life of the fetus in pregnant women. Objective: The risk of KDR gene polymorphisms was investigated in Iranian women with idiopathic recurrent spontaneous abortion (RSA). Materials and Methods: A case controlled study was performed. One hundred idiopathic recurrent spontaneous abortion patients with at least two consecutive pregnancy losses before 20 weeks of gestational age with normal karyotypes were included in the study. Also, 100 healthy women with at least one natural pregnancy were studied as control group. Two functional SNPs located in KDR gene; rs1870377 (Q472H), and rs2305948 (V297I) as well as one tag SNP in the intron region (rs6838752) were genotyped by using PCR based restriction fragment length polymorphism (PCR-RFLP) technique. Haplotype frequency was determined for these three SNPs’ genotypes. Analysis of genetic STRUCTURE and K means clustering were performed to study genetic variation. Results: Functional SNP (rs1870377) was highly linked to tag SNP (rs6838752) (D´ value=0. 214; χ2 = 16.44, p<0. 001). K means clustering showed that k = 8 as the best fit for the optimal number of genetic subgroups in our studied materials. This result was in agreement with Neighbor Joining cluster analysis. Conclusion: In our study, the allele and genotype frequencies were not associated with RSA between patient and control individuals. Inconsistent results in different populations with different allele frequencies among RSA patients and controls may be due to ethnic variation and used sample size.

Aplicaciones de los modelos de predicción bio-inspirados en la administración

Las organizaciones y sus entornos son sistemas complejos. Tales sistemas son difíciles de comprender y predecir. Pese a ello, la predicción es una tarea fundamental para la gestión empresarial y para la toma de decisiones que implica siempre un riesgo. Los métodos clásicos de predicción (entre los cuales están: la regresión lineal, la Autoregresive Moving Average y el exponential smoothing) establecen supuestos como la linealidad, la estabilidad para ser matemática y computacionalmente tratables. Por diferentes medios, sin embargo, se han demostrado las limitaciones de tales métodos. Pues bien, en las últimas décadas nuevos métodos de predicción han surgido con el fin de abarcar la complejidad de los sistemas organizacionales y sus entornos, antes que evitarla. Entre ellos, los más promisorios son los métodos de predicción bio-inspirados (ej. redes neuronales, algoritmos genéticos /evolutivos y sistemas inmunes artificiales). Este artículo pretende establecer un estado situacional de las aplicaciones actuales y potenciales de los métodos bio-inspirados de predicción en la administración.

Applying fuzzy logic to comparative distribution modelling: a case study with two sympatric amphibians.

We modelled the distributions of two toads (Bufo bufo and Epidalea calamita) in the Iberian Peninsula using the favourability function, which makes predictions directly comparable for different species and allows fuzzy logic operations to relate different models. The fuzzy intersection between individual models, representing favourability for the presence of both species simultaneously, was compared with another favourability model built on the presences shared by both species. The fuzzy union between individual models, representing favourability for the presence of any of the two species, was compared with another favourabilitymodel based on the presences of either or both of them. The fuzzy intersections between favourability for each species and the complementary of favourability for the other (corresponding to the logical operation “A and not B”) were compared with models of exclusive presence of one species versus the exclusive presence of the other. The results of modelling combined species data were highly similar to those of fuzzy logic operations between individual models, proving fuzzy logic and the favourability function valuable for comparative distribution modelling. We highlight several advantages of fuzzy logic over other forms of combining distribution models, including the possibility to combine multiple species models for management and conservation planning.

Summary and Conclusion

This monograph outlines the process and results of development of a common educational programme at masters level in health and social care management, which was supported by the Erasmus Life Long Education project CareMan (Cul- ture and Care Management). The CareMan project brought together university partners actively involved in educating social and health care professionals in leadership and management at master’s level in Europe. The five partners of the consortium were Lahti University of Applied Sciences – Lahti UAS (administra- tive and academic coordinator, Finland), Charles University – CU (the Czech Republic), Edinburgh Napier University – ENU (Scotland), Hammeline University of Applied Sciences – HAMK (Finland), and University of Évora – UoE (Portugal). The objectives of the project were to achieve lower -level educational goals that included the development through education cultural and value -driven leadership, quality of care and quality management to effectively manage an integrated health and social care service. Through mapping the situation in the field and comparing curricula of all participating universities the overall aim was to develop a joint masters programme in social and healthcare management. After the detailed understanding of national and institutional specifics of each of the individual academic entities it was recognised that, due to a number of regulation issues, the original aim was not achievable. Following subsequent analytical work, it was decided to develop a set of three master’s level modules. At the end of the project it was intended that all created modules would be available virtually to the participating programmes and would contribute some added value to existing curricula. In the future these ready -to -use modules are intended to be taught in cooperation with the participating universities or as a separate module in each university. The chosen theoretical framework of the project that underpinned the devel- opment, management and evaluation of the inter -cultural educational provision relied on the combination of two learning theories – ‘cooperative collaborative and social learning’ and ’transformational’ (Mezirow, 2009). This theoretical framework helped to align with European collaborative policy and its application on all levels of implementation of the project.

Supporting Scientific Research Through Machine and Deep Learning: Fluorescence Microscopy and Operational Intelligence Use Cases

Although the debate of what data science is has a long history and has not reached a complete consensus yet, Data Science can be summarized as the process of learning from data. Guided by the above vision, this thesis presents two independent data science projects developed in the scope of multidisciplinary applied research. The first part analyzes fluorescence microscopy images typically produced in life science experiments, where the objective is to count how many marked neuronal cells are present in each image. Aiming to automate the task for supporting research in the area, we propose a neural network architecture tuned specifically for this use case, cell ResUnet (c-ResUnet), and discuss the impact of alternative training strategies in overcoming particular challenges of our data. The approach provides good results in terms of both detection and counting, showing performance comparable to the interpretation of human operators. As a meaningful addition, we release the pre-trained model and the Fluorescent Neuronal Cells dataset collecting pixel-level annotations of where neuronal cells are located. In this way, we hope to help future research in the area and foster innovative methodologies for tackling similar problems. The second part deals with the problem of distributed data management in the context of LHC experiments, with a focus on supporting ATLAS operations concerning data transfer failures. In particular, we analyze error messages produced by failed transfers and propose a Machine Learning pipeline that leverages the word2vec language model and K-means clustering. This provides groups of similar errors that are presented to human operators as suggestions of potential issues to investigate. The approach is demonstrated on one full day of data, showing promising ability in understanding the message content and providing meaningful groupings, in line with previously reported incidents by human operators.

Advanced statistical models for the segregation, identification and measurement of coexisting sound sources

Long-term monitoring of acoustical environments is gaining popularity thanks to the relevant amount of scientific and engineering insights that it provides. The increasing interest is due to the constant growth of storage capacity and computational power to process large amounts of data. In this perspective, machine learning (ML) provides a broad family of data-driven statistical techniques to deal with large databases. Nowadays, the conventional praxis of sound level meter measurements limits the global description of a sound scene to an energetic point of view. The equivalent continuous level Leq represents the main metric to define an acoustic environment, indeed. Finer analyses involve the use of statistical levels. However, acoustic percentiles are based on temporal assumptions, which are not always reliable. A statistical approach, based on the study of the occurrences of sound pressure levels, would bring a different perspective to the analysis of long-term monitoring. Depicting a sound scene through the most probable sound pressure level, rather than portions of energy, brought more specific information about the activity carried out during the measurements. The statistical mode of the occurrences can capture typical behaviors of specific kinds of sound sources. The present work aims to propose an ML-based method to identify, separate and measure coexisting sound sources in real-world scenarios. It is based on long-term monitoring and is addressed to acousticians focused on the analysis of environmental noise in manifold contexts. The presented method is based on clustering analysis. Two algorithms, Gaussian Mixture Model and K-means clustering, represent the main core of a process to investigate different active spaces monitored through sound level meters. The procedure has been applied in two different contexts: university lecture halls and offices. The proposed method shows robust and reliable results in describing the acoustic scenario and it could represent an important analytical tool for acousticians.

Efeito da adição de cinza da casca de arroz em misturas cimento-casca de arroz

The rice husk and its ash are abundant and renewable and can be used to obtain alternative building materials. An increase in the consumption of such waste could help minimize the environmental problems from their improper disposal. This study aimed to evaluate the use of ashes as a cargo mineral (filler). However, the rice husk chemically interferes in the conduct of the based cement mixtures. Thus, different mixes cement-rice husk with and without the addition of ash were evaluated in order to highlight the influence of its components (husk; ash), which could otherwise be excluded or be underestimated. Cylindrical samples (test of simple compression and traction by diametrical compression) and samples extracted from manufactured pressed board (test of bending and parallel compression to the surface), were used to evaluate the behavior of different mixtures of components (rice hush; RHA - rice husk ahs). The results of the mechanical tests showed, in general, there is not a statistical difference between the mixtures, which are associated with the chemical suppressive effect of the rice husk ash. The mixture of rice husk of 10 mm, with an addition of 35% of the rice husk ash, is notable for allowing the highest consumption of rice husk and rice husk ash, to reduce 25% the consumption of cement and to allow the storage (without emissions to the atmosphere), around 1.9 ton of CO2 per ton of cement consumed, thus contributing to the reduction of CO2 emissions, which can stimulate rural constructions under an ecological point of view.

Ontogenesis of stomata in Velloziaceae: paracytic versus tetracytic?

In Velloziaceae, the number of subsidiary cells has been used to characterize species and support groups. Nevertheless, the homology of the stomatal types have not been scrutinized. Stomatal ontogenesis of Vellozia epidendroides and V. plicata, assigned to have tetracytic stomata, and of V. glauca and Barbacenia riparia, assigned to have paracytic stomata, were investigated. In the four species studied, stomata followed perigenic development. Subsidiary cells arise from oblique divisions of neighbouring cells of the guard mother cell (GMC). These cells are elongated and parallel to the longer axis of the stoma. Polar cells show wide variation, following the shape and size of the epidermal cells in the vicinity. Hence, these cells cannot be called subsidiary cells. This wide variation is due to a much higher density of stomata in some regions of the leaf blade. This distribution of stomata forces the development of short polar cells, leading to an apparently tetracytic stomata. In regions of low concentration of stomata, higher spatial availability between the GMCs allows the elongation of polar cells, leading to evident paracytic stomata. Therefore, the four studied species are considered braquiparacytic, questioning the classification of stomata into tetracytic and paracytic in Velloziaceae.

Causas de variação nos preços de bovinos nelore elite no Brasil

Este trabalho teve por objetivo estudar as causas de variação nos preços de bovinos da raça nelore pertencentes a rebanhos de seleção, os quais foram comercializados em leilões, para verificar as influências das avaliações genéticas e dos julgamentos de exterior sobre esses preços. Para tanto, foram computados os preços de venda de 426 bovinos da referida raça em 12 leilões ocorridos em diversas localidades brasileiras (regiões Centro-Oeste, Norte e Sudeste), entre os anos de 2002 e 2005. O valor médio foi de R$ 3.325,49, sendo o mínimo de R$ 1.400,00 e o máximo de R$ 10.500,00. Esses dados foram digitados juntamente com outras informações que eram apresentadas nos catálogos dos leilões. As informações registradas incluíram o sexo de cada animal, o nome do leilão e as DEPs informadas nos catálogos. Além da avaliação da influência das informações dos catálogos, também foi avaliada a influência das informações dos reprodutores, pais dos animais vendidos nos leilões, envolvendo suas DEPs publicadas em um sumário de reprodutores da raça e as pontuações de suas progênies em julgamentos. Os métodos estatísticos aplicados foram análises de variâncias e análises de agrupamento (método K-médias). Como resultado, foi observado que animais com superioridade genética em características relacionadas a desempenho ponderal, considerando-se os efeitos diretos e maternos, foram valorizados ao serem comercializados nos leilões. Em contra-partida, a pontuação dos reprodutores nos julgamentos não teve influência significativa sobre os preços médios de venda de suas progênies nos leilões.

Isolamento de Salmonella enterica em gambás (Didelphis aurita e Didelphis albiventris) do Estado de São Paulo, Brasil

No Brasil, não há relato de estudos de Salmonella em gambás, sendo assim, este trabalho tem por objetivo determinar a frequência de isolamento de Salmonella enterica em gambás (D. aurita e D. albiventris) no Estado de São Paulo. No período de janeiro de 2005 a dezembro de 2006, foram necropsiados 106 D. aurita e 40 D. albiventris e colhidos fragmentos de intestinos delgado, grosso e suabe da cloaca. As amostras foram plaqueadas diretamente em ágar Mac Conkey, paralelamente suspendidas nos caldos Rappaport-Vassiliadis e Tetrationato e posteriormente plaqueados em ágar XLT4. As colônias sugestivas de Salmonella foram confirmadas através de provas bioquímicas e sorotipagem. Encontrou-se Salmonella enterica em 17,0% (18/106) dos D. aurita. Destes, 50% apresentaram positividade no intestino delgado (ID), 88,9% no intestino grosso (IG) e 66,7% na cloaca. Da espécie S. enterica, as subespécies encontradas foram: diarizonae (11,1%) houtenae e enterica (5,5% cada um); enquanto da subespécie S. enterica enterica os sorotipos foram Newport (83,3%), Typhimurium e Cerro (5,5% cada um). Nos D. albiventris, 17,5% (7/40) eram positivos, sendo que se encontraram 42,8% no ID, 85,7% no IG e 71,4% na cloaca. O sorotipo mais prevalente também foi Newport (71,4%), seguido por Typhimurium, Bareilly e Thompson (14,3% cada um). Através dos resultados obtidos neste estudo pode-se comprovar a presença de Salmonella enterica no trato intestinal de gambás no Brasil.

Chicken skeletal muscle-associated macroarray for gene discovery

Macro- and microarrays are well-established technologies to determine gene functions through repeated measurements of transcript abundance. We constructed a chicken skeletal muscle-associated array based on a muscle-specific EST database, which was used to generate a tissue expression dataset of similar to 4500 chicken genes across 5 adult tissues (skeletal muscle, heart, liver, brain, and skin). Only a small number of ESTs were sufficiently well characterized by BLAST searches to determine their probable cellular functions. Evidence of a particular tissue-characteristic expression can be considered an indication that the transcript is likely to be functionally significant. The skeletal muscle macroarray platform was first used to search for evidence of tissue-specific expression, focusing on the biological function of genes/transcripts, since gene expression profiles generated across tissues were found to be reliable and consistent. Hierarchical clustering analysis revealed consistent clustering among genes assigned to 'developmental growth', such as the ontology genes and germ layers. Accuracy of the expression data was supported by comparing information from known transcripts and tissue from which the transcript was derived with macroarray data. Hybridization assays resulted in consistent tissue expression profile, which will be useful to dissect tissue-regulatory networks and to predict functions of novel genes identified after extensive sequencing of the genomes of model organisms. Screening our skeletal-muscle platform using 5 chicken adult tissues allowed us identifying 43 'tissue-specific' transcripts, and 112 co-expressed uncharacterized transcripts with 62 putative motifs. This platform also represents an important tool for functional investigation of novel genes; to determine expression pattern according to developmental stages; to evaluate differences in muscular growth potential between chicken lines, and to identify tissue-specific genes.