[pt] Foi desenvolvido, neste trabalho, um sistema automatizado para o prétratamento de inibidor à base de ácido fosfinopolicarboxílico (PPCA), usado na exploração de petróleo no mar com a finalidade de minimizar a ocorrência de incrustações inorgânicas. Os componentes que constituem o sistema são: válvulas solenóides com corpo de teflon de três, quatro (ou seis) vias; circuito elaborado para acionamento das válvulas; bomba peristáltica; um microcomputador; aplicativo desenvolvido em linguagem Delphi (PRETRAT). O circuito eletrônico, conectado via porta paralela do computador, tem a função de reduzir a voltagem de alimentação das válvulas de 12 para 4,5 V, permitindo que as válvulas permaneçam ligadas por tempo ilimitado. A fonte do computador é usada para alimentação das válvulas, enquanto que a bomba peristáltica é comandada pela porta serial. O programa permite optar distintamente pelas etapas do prétratamento, definindo volumes e vazões das soluções, calculando os tempos de acionamento das válvulas e executando o processo automaticamente. O prétratamento se faz necessário devido à alta salinidade e elevado teor de fósforo inorgânico presentes na matriz estudada (água produzida) sendo a separação do fósforo orgânico foi realizada utilizando mini-colunas de sílica-C18 (SEP-PAK clássica, 360 mg). A otimização do pré- tratamento resultou nos seguintes parâmetros de trabalho: (1) condicionamento com metanol 80% v/v; (2) taxas de percolação de 5 mL/min para condicionamento, passagem e eluição da amostra; (3) eluição do fósforo orgânico com 3,5 mL de solução tampão de H3BO3 0,025 M (ajustada a pH 9 com NaOH 0,2 M). A quantificação do inibidor foi feita de forma indireta através da determinação de fósforo, a qual foi realizada pelas técnicas de ICP-OES e ICP-MS, sendo que nesta última foi possível a determinação em linha de fósforo através do acoplamento do sistema ao espectrômetro ELAN 6000. Os limites de detecção (3 ômega ) na amostra (matriz de tampão de borato 0,025 M) foram 0,20 (Mi)g.L(elevado a -1) e 92 (Mi)g.L(elevado a -1), para as técnicas de ICP-MS e ICP-OES, respectivamente (ambos utilizando nebulizador Cross-flow e câmara de nebulização Ryton ). A recuperação de fósforo orgânico ficou tipicamente entre 90 e 95%. Para verificar a repetitividade do método, oito alíquotas contendo 0,40 mg.L- 1 de P, na forma de inibidor, foram processadas nas condições mencionadas, sendo obtida uma média de 0,42 +/- 0,02 mg.L(elevado a -1) de P, o que corresponde a um desvio padrão relativo de 4,8 %. A reprodutibilidade foi comparada com a da metodologia de pré-tratamento manual (Rocha, 1997), e o desvio padrão relativo obtido para amostras de campo foi inferior a 25%. Uma freqüência analítica de 6 h(elevado a -1}) pode ser obtida para uma pré-concentração de 10 vezes (volume de amostra igual a 35 mL). O pré-tratamento automatizado foi aplicado na comparação de produtos comerciais e utilizado como ferramenta complementar nos ensaios de adsorção/desorção do inibidor PPCA em testemunhos de rochas (tratamento de squeeze). Através da especiação/fracionamento de fósforo orgânico e inorgânico, o método aqui desenvolvido permitiu a identificação de diferentes teores de fósforo orgânico nos produtos comerciais, nomeados pelos fornecedores como sendo à base de PPCA. Amostras de campo provenientes de quatro poços que passaram por tratamento de squeeze foram submetidas ao pré- tratamento automatizado, o que possibilitou um acompanhamento do tratamento realizado nestes poços. A técnica mostrou-se eficaz no controle de qualidade dos inibidores a serem aplicados nos tratamentos de squeeze. / [en] In the present study, a system was developed for the automatized pretreatment of the inhibitor phosphinocarboxilic acid (PPCA) used during off-shore oil exploration to minimize the occurrence of inorganic scale formation. The components of the system used are: 3, 4 or 6 way solenoid valves with Teflon body, a home-made electronic circuit for activation of the valves, a peristaltic pump, a microcomputer and software (PRETRAT) written in Delphi. The electronic circuit, connected via the parallel port of the PC, has the function to reduce the supply voltage for the solenoid valves form 12 V to 4.5 V, thus permitting their uninterrupted use for longer periods of time. The power module of the microcomputer is used to obtain the supply voltage of the solenoid valves, while its serial port commands the peristaltic pump. The developed software program PRETRAT permits automatic control over all relevant parameters used in the pre-treatment procedure, including solution volumes and flow rates. This pre-treatment, mandatory due to the high salinity of the produced water and its elevated concentrations of inorganic phosphorus, was achieved by separating the organic phosphorus (contained in the PPCA molecule) by using silica-C18 columns (SEP-PAK classic, 360 mg). Optimization of the pre-treatment procedure resulted in the following experimental parameters: (1) conditioning of the column with methanol 80% v/v; (2) percolation rates of 5 mL.min-1 for conditioning, sample throughput and elution; (3) elution of organic phosphorus with 3.5 mL of a buffer solution of H3BO3 0.025 M (adjusted at pH 9 with NaOH 0.2 M). The quantification of the PPCA inhibitor was performed indirectly by measuring its phosphorus content by on-line inductively coupled plasma mass spectrometry (ICP-MS) and/or by optical emission spectrometry (ICP-OES). Detection limits (3 ) in the borate matrix (0.025 M) were 0.20 g.L-1 e 92 g.L-1 for ICP-MS and ICP-OES, respectively, using in both methods a cross-flow nebulizer and a Ryton spray chamber. Recoveries of organic phosphorus were, typically, in the range of 90% to 95%. For assessing the repeatability of the procedure, eight sample replicates containing each 0.40 mg.L-1 of phosphorus as inhibitor were processed under the conditions already mentioned, resulting in a mean recovery of de 0.42 +/- 0.02 mg.L-1 and a relative standard deviation of 4.8 %. This value compares advantageously with the mean reproducibility of about 25% obtained by a manual pre- treatment (Rocha, 1997). Frequency of analyses is about six samples per hour using a pre-concentration factor of 10 and a sample volume of 35 mL. The automatized pre- treatment method was applied for comparing the behavior of commercial inhibitors during field use and in laboratory studies on the adsorption/desorption characteristics of PPCA on rock samples, simulating the squeeze process. Produced waters from four wells, in which the squeeze process was applied, were analyzed by the here-developed method, permitting the follow-up of PPCA release and the observation that some of the used inhibitor lots did not perform adequately during the squeeze process. Speciation analysis of inorganic and organic phosphorus in those lots revealed significant differences in composition when compared to well performing ones, also confirmed by ion chromatography. This methodology has demonstrated efficient for quality assurance of inhibitors prior to their application.

City-wide travel time prediction in real-time is an important enabler for efficient use of the road network. It can be used in traveler information to enable more efficient routing of individual vehicles as well as decision support for traffic management applications such as directed information campaigns or incident management. 3D speed maps have been shown to be a promising methodology for revealing day-to-day regularities of city-level travel times and possibly also for short-term prediction. In this paper, we aim to further evaluate and benchmark the use of 3D speed maps for short-term travel time prediction and to enable scenario-based evaluation of traffic management actions we also evaluate the framework for traffic flow prediction. The 3D speed map methodology is adapted to short-term prediction and benchmarked against historical mean as well as against Probabilistic Principal Component Analysis (PPCA). The benchmarking and analysis are made using one year of travel time and traffic flow data for the city of Stockholm, Sweden. The result of the case study shows very promising results of the 3D speed map methodology for short-term prediction of both travel times and traffic flows. The modified version of the 3D speed map prediction outperforms the historical mean prediction as well as the PPCA method. Further work includes an extended evaluation of the method for different conditions in terms of underlying sensor infrastructure, preprocessing and spatio-temporal aggregation as well as benchmarking against other prediction methods. / <p>QC 20190531</p>

Le thème principal de cette thèse est d’étudier la détection d’anomalies dans des flux de données de grande dimension avec une application spécifique au Health Monitoring des moteurs d’avion. Dans ce travail, on considère que le problème de la détection d’anomalies est un problème d’apprentissage non supervisée. Les données modernes, notamment celles issues de la surveillance des systèmes industriels sont souvent des flux d’observations de grande dimension, puisque plusieurs mesures sont prises à de hautes fréquences et à un horizon de temps qui peut être infini. De plus, les données peuvent contenir des anomalies (pannes) du système surveillé. La plupart des algorithmes existants ne peuvent pas traiter des données qui ont ces caractéristiques. Nous introduisons d’abord un algorithme de clustering probabiliste offline dans des sous-espaces pour des données de grande dimension qui repose sur l’algorithme d’espérance-maximisation (EM) et qui est, en plus, robuste aux anomalies grâce à la technique du trimming. Ensuite, nous nous intéressons à la question du clustering probabiliste online de flux de données de grande dimension en développant l’inférence online du modèle de mélange d’analyse en composantes principales probabiliste. Pour les deux méthodes proposées, nous montrons leur efficacité sur des données simulées et réelles, issues par exemple des moteurs d’avion. Enfin, nous développons une application intégrée pour le Health Monitoring des moteurs d’avion dans le but de détecter des anomalies de façon dynamique. Le système proposé introduit des techniques originales de détection et de visualisation d’anomalies reposant sur les cartes auto-organisatrices. Des résultats de détection sont présentés et la question de l’identification des anomalies est aussi discutée. / The subject of this Thesis is to study anomaly detection in high-dimensional data streams with a specific application to aircraft engine Health Monitoring. In this work, we consider the problem of anomaly detection as an unsupervised learning problem. Modern data, especially those is-sued from industrial systems, are often streams of high-dimensional data samples, since multiple measurements can be taken at a high frequency and at a possibly infinite time horizon. More-over, data can contain anomalies (malfunctions, failures) of the system being monitored. Most existing unsupervised learning methods cannot handle data which possess these features. We first introduce an offline subspace clustering algorithm for high-dimensional data based on the expectation-maximization (EM) algorithm, which is also robust to anomalies through the use of the trimming technique. We then address the problem of online clustering of high-dimensional data streams by developing an online inference algorithm for the popular mixture of probabilistic principal component analyzers (MPPCA) model. We show the efficiency of both methods on synthetic and real datasets, including aircraft engine data with anomalies. Finally, we develop a comprehensive application for the aircraft engine Health Monitoring domain, which aims at detecting anomalies in aircraft engine data in a dynamic manner and introduces novel anomaly detection visualization techniques based on Self-Organizing Maps. Detection results are presented and anomaly identification is also discussed.

Le thème principal de cette thèse est d'étudier la détection d'anomalies dans des flux de données de grande dimension avec une application spécifique au \emph{Health Monitoring} des moteurs d'avion. Dans ce travail, on considère que le problème de la détection d'anomalies est un problème d'apprentissage non supervisée. Les données modernes, notamment celles issues de la surveillance des systèmes industriels sont souvent des flux d'observations de grande dimension, puisque plusieurs mesures sont prises à de hautes fréquences et à un horizon de temps qui peut être infini. De plus, les données peuvent contenir des anomalies (pannes) du système surveillé. La plupart des algorithmes existants ne peuvent pas traiter des données qui ont ces caractéristiques. Nous introduisons d'abord un algorithme de clustering probabiliste offline dans des sous-espaces pour des données de grande dimension qui repose sur l'algorithme d'espérance-maximisation (EM) et qui est, en plus, robuste aux anomalies grâce à la technique du trimming. Ensuite, nous nous intéressons à la question du clustering probabiliste online de flux de données de grande dimension en développant l'inférence online du modèle de mélange d'analyse en composantes principales probabiliste. Pour les deux méthodes proposées, nous montrons leur efficacité sur des données simulées et réelles, issues par exemple des moteurs d'avion. Enfin, nous développons une application intégrée pour le Health Monitoring des moteurs d'avion dans le but de détecter des anomalies de façon dynamique. Le système proposé introduit des techniques originales de détection et de visualisation d'anomalies reposant sur les cartes auto-organisatrices. Des résultats de détection sont présentés et la question de l'identification des anomalies est aussi discutée.

No description available.