Made available in DSpace on 2014-12-17T14:56:19Z (GMT). No. of bitstreams: 1 JanEMGP_DISSERT.pdf: 3028317 bytes, checksum: 5eeb8ec6954b59f2853f263ffa4c4d9c (MD5) Previous issue date: 2014-05-16 / Conselho Nacional de Desenvolvimento Cient?fico e Tecnol?gico / The area of research and development involving the PID tune of controllers is an active area in the academic and industrial sectors yet. All this due to the wide use of PID controllers in the industry (96% of all controllers in the industry is still PID). Controllers well tuned and tools to monitor their performance over time with the possibility of selftuning, become an item almost obligatory to maintain processes with high productivity and low cost. In a globalized world, it is essential for their self survival. Although there are several new tools and techniques that make PID tune, in this paper will explore the PID tune using the relay method, due its good acceptance in the industrial environment. In addition, we will discuss some techniques for evaluation of control loops, as IAE, ISE, Goodhart, the variation of the control signal and index Harris, which are necessary to propose new tuning for control loops that have a low performance. Will be proposed in this paper a tool for tuning and self tuning PID. Will be proposed in this paper a PID auto-tuning software using a relay method. In particular, will be highlighted the relay method with hysteresis. This method has shown tunings with satisfactory performance when applied to the didactic, simulated and real plants / O campo de pesquisa e desenvolvimento de softwares envolvendo a sintonia de controladores PID, ainda ? uma ?rea ativa dentro do meio acad?mico e industrial. Tudo isso devido ? larga utiliza??o de controladores PID na ind?stria (96% de todos os controladores na ind?stria ainda ? PID). Ter controladores bem sintonizados e com ferramentas que possam acompanhar seus desempenhos ao longo do tempo com a possibilidade de ressintoniz?-los, ou ainda autossintoniz?-los, passar a ser um item quase que obrigat?rio para manter processos com alta produtividade e baixo custo. J? que em um mundo globalizado, o n?vel mais acirrado de concorr?ncia entre as empresas, atualmente, est? no custeio e na produtividade. Apesar de existirem diversas novas t?cnicas e ferramentas que fazem sintonia de controladores PID, neste trabalho ser? explorada esta sintonia utilizando o m?todo do rel?, devido a sua boa aceita??o no ambiente industrial, simplicidade e robustez. Al?m disto, abordaremos algumas t?cnicas para avalia??o de desempenho de malhas de controle de processos, tais como IAE, ISE, Goodhart, Vari?ncia de sinais e ?ndice de Harris. Ser? proposta neste trabalho uma ferramenta de sintonia e autossintonia PID (usando o m?todo do rel?), em especial o m?todo do rel? com histerese. Este m?todo tem apresentado sintonias com desempenhos satisfat?rios quando aplicados em plantas simuladas e reais

The problem of designing control systems for entire plants is studied. A review of previous works, available techniques and current research challenges is presented, followed by the description of some theoretical tools to improve plantwide control, including the proposal of an augmented lexicographic multi-objective optimization procedure. With these, an augmented hierarchical plantwide control design technique and an optimal multi-objective technique for integrated control structure selection and controller tuning are proposed. The main contributions of these proposed techniques are the inclusion of system identification and optimal control tuning as part of the plantwide design procedure for improved results, support to multi-objective control specifications and support to any type of plant and controllers. Finally, the proposed techniques are applied to industrial benchmarks to demonstrate and validate its applicability. / O problema de projetar sistemas de controle para plantas inteiras é estudado. Uma revisão de trabalhos anteriores, técnicas disponíveis e atuais desafios de pesquisa é apresentada, seguida da descrição de algumas ferramentas teóricas para melhorar o controle plantwide, incluindo a proposta de um procedimento de otimização multi-objetivo lexicográfico aumentado. Com tais elementos, são propostas uma nova técnica hierárquica aumentada de projeto de sistemas de controle plantwide e uma técnica multi-objetivo para seleção de estrutura de controlador integrada à sintonia ótima do controlador. As principais contribuições das técnicas propostas são a inclusão de identificação de sistemas e sintonia ótima de controladores como parte do procedimento de projeto de controle plantwide para melhores resultados, suporte a especificações multi-objetivo e suporte a quaisquer tipos de plantas e controladores. Finalmente, as técnicas propostas são aplicadas a benchmarks industriais para demonstrar e validar sua aplicabilidade.

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / This works proposes parameters estimation algorithm to auto-tune the control laws of a speed sensorless servo. The identification process of the electrical and mechanical parameters is based on recursive least squares method (RLS) and utilizes just measured stator currents. The electrical parameters are obtained by the transfer function of the motor with locked rotor. Hence, this eliminates the need to realize classical tests for the electrical parameters obtention and impose robustness to the controller for parameters variations caused by aging of materials. Thus, the electrical parameters are used to project the current controllers and the rotor speed is estimated by a model reference adaptive systems (MRAS). The structure of MRAS is composed by two systems (reference model and adaptive system) based on instantaneous reactive power and using an adaptation mechanism. On the other hand, to obtain a robust control when the system is subject to disturbances or load s variations, mechanical parameters are identified using estimated rotor speed and mechanical model of the induction motor (IM). Finally, for the speed control is used a proportional plus integral controller (PI) self-tuned through the estimated inertia. / Este trabalho propõe um algoritmo de identificação paramétrica para auto-ajuste da lei de controle de um servo de velocidade sem sensor mecânico. O processo de identificação dos parâmetros elétricos e mecânicos do servo é baseado em algoritmos do tipo mínimos quadrados recursivo (RLS) e utiliza apenas as medidas das correntes estatóricas. Os parâmetros elétricos são identificados através da função de transferência do motor com rotor bloqueado. Isso elimina a necessidade de realizar ensaios clássicos para obtenção dos parâmetros elétricos e impõe robustez ao controlador frente a variações paramétricas que ocorrem devido aos desgastes e envelhecimento dos materiais que constituem o motor. Uma vez identificados, os parâmetros elétricos são utilizados no projeto dos controladores de corrente e a velocidade rotórica é estimada utilizando um sistema adaptativo por modelo de referência (MRAS). A estrutura do MRAS é composta por dois sistemas (modelo de referência e sistema adaptativo) baseados na potência reativa instantânea e utiliza um mecanismo de adaptação. Por outro lado, para impor robustez ao controlador frente a variações de carga, os parâmetros mecânicos são identificados a partir da velocidade rotórica estimada e do modelo mecânico do motor de indução (MI). Finalmente, para o controle de velocidade é usado um controlador proporcional e integral (PI) com seus ganhos sintonizados a partir da inércia estimada.

Ce mémoire de thèse s’inscrit dans le développement d’un microscope thermique à sonde locale.Ce système d’imagerie présente deux modes de fonctionnement permettant de déterminer soit unetempérature de surface soit des propriétés thermophysiques de matériaux. Un micro-thermocouplebifilaire a été utilisé comme capteur thermique. Il est peu invasif et permet d’accéder à destempératures de surface sur une large gamme de température. De plus, le microscope offrel’avantage d’être moins sensible à la nature optique des échantillons que les méthodes en champlointain. Dans le but de maitriser le contact entre la sonde et la surface, un résonateur à quartz(diapason) a été utilisé comme capteur de force. Un système d’excitation original basé sur l’effetphoto-thermo-élastique a été mis au point. Le microscope fonctionne donc comme un SThM puisqu’ilpermet d’extraire simultanément des images topographiques et thermiques (régime périodique 2 et3 oméga). En revanche, les résultats obtenus ont permis de mettre en évidence les avantages dumicro-thermocouple en termes de résolutions spatiales topographiques vis-à-vis des techniques àsondes résistives fonctionnant en mode 3 oméga. / This PhD thesis deals with the development of a thermalmicroscope using a local probe. This imagingsystem presents two functioning modes that allow determining either surface temperature or thermalproperties of materials. A micro-wire thermocouple is used as a thermal sensor. It is less invasiveand allows measuring the surface temperature with a large temperature range. Furthermore, themicroscope offers an advantage to be less sensitive to the optical nature of a sample surface thanoptical methods. To control the contact between the probe and the surface, a quartz tuning fork hasbeen used as a force sensor. An original excitation system has been developed based on the photothermaleffect. The microscope works also as a SThM since it permits to extract simultaneouslytopographical and thermal pictures (2 and 3 omega periodical modes). Results underlining themicro-thermocouple advantages, in terms of topographical compared to resistive probe techniquesfunctioning with the 3 omega method, have been obtained.

Databases are all around us, and ensuring their efficiency is of great importance. Database optimisation has many parts and many methods, two of these parts are database tuning and database optimisation. These can then further be split into methods such as indexing. These indexing techniques have been studied and compared between Database Management Systems (DBMSs) to see how much they can improve the execution time for queries. And many guides have been written on how to implement query optimisation and indexes. In this thesis, the question "How does indexing and query optimisation affect response time in PostgreSQL?" is posed, and was answered by investigating these previous studies and theory to find different optimisation techniques and compare them to each other. The purpose of this research was to provide more information about how optimisation techniques can be implemented and map out when what method should be used. This was partly done to provide learning material for students, but also people who are starting to learn PostgreSQL. This was done through a literature study, and an experiment performed on a database with different table sizes to see how the optimisation scales to larger systems. What was found was that there are many use cases to optimisation that mainly depend on the query performed and the type of data. From both the literature study and the experiment, the main take-away points are that indexes can vastly improve performance, but if used incorrectly can also slow it. The main use cases for indexes are for short queries and also for queries using spatio-temporal data - although spatio-temporal data should be researched more. Using the DBMS optimiser did not show any difference in execution time for queries, while correctly implemented query tuning techniques also vastly improved execution time. The main use cases for query tuning are for long queries and nested queries. Although, most systems benefit from some sort of query tuning, as it does not have to cost much in terms of memory or CPU cycles, in comparison to how indexes add additional overhead and need some memory. Implementing proper optimisation techniques could improve both costs, and help with environmental sustainability by more effectively utilising resources. / Databaser finns överallt omkring oss, och att ha effektiva databaser är mycket viktigt. Databasoptimering har många olika delar, varav två av dem är databasjustering och SQL optimering. Dessa två delar kan även delas upp i flera metoder, så som indexering. Indexeringsmetoder har studerats tidigare, och även jämförts mellan DBMS (Database Management System), för att se hur mycket ett index kan förbättra prestanda. Det har även skrivits många böcker om hur man kan implementera index och SQL optimering. I denna kandidatuppsats ställs frågan "Hur påverkar indexering och SQL optimering prestanda i PostgreSQL?". Detta besvaras genom att undersöka tidigare experiment och böcker, för att hitta olika optimeringstekniker och jämföra dem med varandra. Syftet med detta arbete var att implementera och kartlägga var och när dessa metoder kan användas, för att hjälpa studenter och folk som vill lära sig om PostgreSQL. Detta gjordes genom att utföra en litteraturstudie och ett experiment på en databas med olika tabell storlekar, för att kunna se hur dessa metoder skalas till större system. Resultatet visar att det finns många olika användingsområden för optimering, som beror på SQL-frågor och datatypen i databasen. Från både litteraturstudien och experimentet visade resultatet att indexering kan förbättra prestanda till olika grader, i vissa fall väldigt mycket. Men om de implementeras fel kan prestandan bli värre. De huvudsakliga användingsområdena för indexering är för korta SQL-frågor och för databaser som använder tid- och rum-data - dock bör tid- och rum-data undersökas mer. Att använda databassystemets optimerare visade ingen förbättring eller försämring, medan en korrekt omskrivning av en SQL fråga kunde förbättra prestandan mycket. The huvudsakliga användingsområdet för omskriving av SQL-frågor är för långa SQL-frågor och för nestlade SQL-frågor. Dock så kan många system ha nytta av att skriva om SQL-frågor för prestanda, eftersom att det kan kosta väldigt lite när det kommer till minne och CPU. Till skillnad från indexering som behöver mer minne och skapar så-kallad överhead". Att implementera optimeringstekniker kan förbättra både driftkostnad och hjälpa med hållbarhetsutveckling, genom att mer effektivt använda resuser.

In hierdie studie word die intonasie van fluitspel ondersoek met die doel om vas te stel watter faktore beïnvloed intonasie en watter metodes aangewend kan word om dit te verbeter. Intonasie is ’n besonder sensitiewe onderwerp vir enige musikant. Indien ’n musikant negatief gekritiseer word oor intonasie, word dit dikwels as ’n persoonlike aanslag ervaar. Deur die oorsprong van Westerse stemmingstelsels te ondersoek, word dit moontlik om die onderwerp van intonasie meer sensitief te benader. Hierdie bewusmaking kan moontlik nuwe insigte tot stand bring ten opsigte van die nodige aanpassings wat gemaak moet word om intonasie te verbeter tydens solo- en ensemble-spel. Verskeie faktore kan daartoe bydra dat intonasie as minder goed ervaar word. Sommige van hierdie faktore word nie noodwendig deur die uitvoerder beheer nie, maar hou verband met die fisiologie van gehoor – hoe klank waargeneem word. Die enigste manier waarop intonasie gemonitor kan word tydens ’n uitvoering, is met die menslike oor. Intonasie is direk gekoppel aan die frekwensie van ’n spesifieke toonhoogte. In die natuur bestaan daar faktore wat die waarneming van ’n frekwensie as ’n spesifieke toonhoogte kan beïnvloed, byvoorbeeld die tydsduur, intensiteit en toonkleur. Hierdie faktore word ook in hierdie studie ondersoek. Alvorens metodes ter verbetering van intonasie ondersoek kan word, moet die bou van die moderne fluit ondersoek word. Geen fluitskaal het perfekte ingeboude intonasie nie en deur bewus te wees van waar die tekortkominge voorkom, kan die uitvoerder die onderwerp van intonasie met groter omsigtigheid benader. Sodra die uitvoerder en onderwyser bewus is van watter faktore intonasie beïnvloed en wat nie deur die uitvoerder beheer word nie, kan metodes oorweeg word om faktore wat wel binne die uitvoerder se beheer is, toe te pas om die fluitspeler se intonasie te verbeter. Verskeie aspekte word bespreek, waaronder die plasing van die kurkstopper, die posisie van die kopstuk ten opsigte van die fluit, die manier waarop die fluit vasgehou word, opwarmingsmetodes en oefeninge, die metode van instemming en toonproduksie. Die toonhoogteneigings van note op die fluit word vergelyk soos dit voorkom in vyf verskillende bronne, waarna alternatiewe vingersettings voorgestel word vir pianissimo- en fortissimo-spel. Hierdie vingersettings kan handig te pas kom om die embouchure te assisteer en intonasie beter te beheer. Laastens word die verband tussen die afwykings in die fluit se intonasie, en klimaat en weerstoestande ondersoek, sodat die speler die uitwerking wat veral temperatuur op intonasie het, ook in ag kan neem. Alhoewel hierdie navorsing grootliks klem lê op fluitspelers, is die studie nie net beperk tot die fluit nie en ander instrumentaliste sal ook hierby kan baatvind. ENGLISH This study investigates the factors that influence intonation when playing the flute. Methods by which intonation may be improved are also examined. All musicians regard intonation as an extremely sensitive subject. Musicians often perceive negative criticism of intonation as a personal affront. By investigating the origin of the Western tuning systems, it becomes possible to approach the subject of intonation with more sensitivity. This awareness may establish new insights with regard to the necessary adjustments that must be made during both solo and ensemble playing. Different factors may contribute to intonation being perceived as unsatisfactory. Some of these factors are not necessarily controlled by the performer, but relate to the physiology of hearing – how sound is perceived. The only way in which intonation can be monitored during a performance, is with the human ear. Intonation is directly linked to the frequency of a specific pitch. Factors exist in nature which may influence the perception of a frequency as a specific pitch, for example duration, intensity and tone colour. These factors are also examined in this study. Before investigating methods to improve intonation, the way that the modern flute is built has to be examined. No flute scale has perfect, built-in intonation. By being aware of the shortcomings that may occur, the performer can approach the subject with greater circumspection. When the performer and the educator are aware of the factors which may influence intonation and which are not controlled by the performer, methods can be considered to apply factors that can be controlled by the performer to improve intonation. Different aspects are discussed, such as the placement of the cork stopper, the position of the head joint with regard to the flute, the way the flute is held, warm-up methods and exercises, how to tune and tone production. The pitch tendency of notes on the flute are compared as they occur in five different sources, after which alternative fingerings are suggested for pianissimo and fortissimo playing. These fingerings are useful in assisting the embouchure to maintain better control of intonation. Finally, the relation between the deviation in the flute’s intonation, and climate and weather conditions are investigated to assist the performer to take into consideration the effect that temperature has on intonation. Although this research focuses on flute players, this study is not restricted to the flute, and other instrumentalists can also benefit from it. / Dissertation (MMus)--University of Pretoria, 2007. / Music / unrestricted

De nombreuses applications en calcul scientifique traitent des matrices creuses de grande taille ayant des structures régulières ou irrégulières. Pour réduire à la fois la complexité spatiale et la complexité temporelle du traitement, ces matrices nécessitent l'utilisation d'une structure particulière de stockage (ou de compression) des données ainsi que l’utilisation d’architectures cibles parallèles/distribuées. Le choix du format de compression le plus adéquat dépend généralement de plusieurs facteurs dont, en particulier, la structure de la matrice creuse, l'architecture de la plateforme cible et la méthode numérique utilisée. Etant donné la diversité de ces facteurs, un choix optimisé pour un ensemble de données d'entrée peut induire de mauvaises performances pour un autre. D’où l’intérêt d’utiliser un système permettant la sélection automatique du meilleur format de compression (MFC) en prenant en compte ces différents facteurs. C’est dans ce cadre précis que s’inscrit cette thèse. Nous détaillons notre approche en présentant la modélisation d'un système automatique qui, étant donnée une matrice creuse, une méthode numérique, un modèle de programmation parallèle et une architecture, permet de sélectionner automatiquement le MFC. Dans une première étape, nous validons notre modélisation par une étude de cas impliquant (i) Horner, et par la suite le produit matrice-vecteur creux (PMVC), comme méthodes numériques, (ii) CSC, CSR, ELL, et COO comme formats de compression, (iii) le data parallèle comme modèle de programmation et (iv) une plateforme multicœurs comme architecture cible. Cette étude nous permet d’extraire un ensemble de métriques et paramètres qui influent sur la sélection du MFC. Nous démontrons que les métriques extraites de l'analyse du modèle data parallèle ne suffisent pas pour prendre une décision (sélection du MFC). Par conséquent, nous définissons de nouvelles métriques impliquant le nombre d'opérations effectuées par la méthode numérique et le nombre d’accès à la mémoire. Ainsi, nous proposons un processus de décision prenant en compte à la fois l'analyse du modèle data parallèle et l'analyse de l’algorithme. Dans une deuxième étape, et en se basant sur les données que nous avons extrait précédemment, nous utilisons les algorithmes du Machine Learning pour prédire le MFC d’une matrice creuse donnée. Une étude expérimentale ciblant une plateforme parallèle multicœurs et traitant des matrices creuses aléatoires et/ou provenant de problèmes réels permet de valider notre approche et d’évaluer ses performances. Comme travail futur, nous visons la validation de notre approche en utilisant d'autres plateformes parallèles telles que les GPUs. / Several applications in scientific computing deals with large sparse matrices having regular or irregular structures. In order to reduce required memory space and computing time, these matrices require the use of a particular data storage structure as well as the use of parallel/distributed target architectures. The choice of the most appropriate compression format generally depends on several factors, such as matrix structure, numerical method and target architecture. Given the diversity of these factors, an optimized choice for one input data set will likely have poor performances on another. Hence the interest of using a system allowing the automatic selection of the Optimal Compression Format (OCF) by taking into account these different factors. This thesis is written in this context. We detail our approach by presenting a design of an auto-tuner system for OCF selection. Given a sparse matrix, a numerical method, a parallel programming model and an architecture, our system can automatically select the OCF. In a first step, we validate our modeling by a case study that concerns (i) Horner scheme, and then the sparse matrix vector product (SMVP), as numerical methods, (ii) CSC, CSR, ELL, and COO as compression formats; (iii) data parallel as a programming model; and (iv) a multicore platform as target architecture. This study allows us to extract a set of metrics and parameters that affect the OCF selection. We note that data parallel metrics are not sufficient to accurately choose the most suitable format. Therefore, we define new metrics involving the number of operations and the number of indirect data access. Thus, we proposed a new decision process taking into account data parallel model analysis and algorithm analysis.In the second step, we propose to use machine learning algorithm to predict the OCF for a given sparse matrix. An experimental study using a multicore parallel platform and dealing with random and/or real-world random matrices validates our approach and evaluates its performances. As future work, we aim to validate our approach using other parallel platforms such as GPUs.

No description available.

Antenna is a key component in wireless devices such as cell phones, laptops and smartwatches. With the new generation of Wi-Fi technology: Wi-Fi 6E coming out, 1200MHz bandwidth from 5.925GHz to 7.125GHz is freed for use. Mobile devices are expected to meet the required bandwidth allocated for Wi-Fi 6E. However, electrically small antenna design for Wi-Fi 6 and 6E applications has been a challenge. Reducing the size of antennas while maintaining the performance is one major issue. The goal of this project is to evaluate different antennas and try to find the potential compact antenna that could cover Wi-Fi 6 and Wi-Fi 6E bands. In this project, we evaluated different types of electrically small antennas, explored parameters that affect antenna performance, characterized one nominated antenna in High Frequency Simulation Software(HFSS), prototyped the antenna in Altium Designer and measured the antenna performance using Network Analyzer. The measurement result shows that the proposed antenna is omni-directional, partially meet the fixed frequency requirement and has low efficiency at 2.4GHz but good efficiency at 6GHz bands. The conclusion is that the proposed antenna is good for the extended bandwidth of Wi-Fi 6 but for the whole Wi-Fi 6E the performance still needs to be improved in future work. / Antenn är en nyckelkomponent i trådlösa enheter som mobiltelefoner, bärbara datorer och smartklockor. Med den nya generationens Wi-Fi-teknik: Wi-Fi 6E kommer ut, 1200MHz bandbredd från 5,925GHz till 7,125GHz frigörs för användning. Mobila enheter förväntas uppfylla den erforderliga bandbredden som tilldelats för Wi-Fi 6E. Däremot har en elektrisk liten antenndesign för Wi-Fi 6 och 6E-applikationer varit en utmaning. Att minska storleken på antennerna samtidigt som prestandan bibehålls är en stor fråga. Målet med detta projekt är att utvärdera olika antenner och försöka hitta den potentiella kompakta antennen som kan täcka Wi-Fi 6- och Wi-Fi 6E-banden. I det här projektet utvärderade vi olika typer av elektriskt små antenner, utforskade parametrar som påverkar antennprestanda, karakteriserade en nominerad antenn i High Frequency Simulation Software (HFSS), prototypade antennen i Altium Designer och mätte antennens prestanda med hjälp av Network Analyzer. Mätresultatet visar att den föreslagna antennen är rundstrålande, delvis uppfyller kravet på fast frekvens och har låg verkningsgrad vid 2,4GHz men bra effektivitet vid 6GHz-band. Slutsatsen är att den föreslagna antennen är bra för den utökade bandbredden för Wi-Fi 6 men för hela Wi-Fi 6E behöver prestandan fortfarande förbättras i framtida arbete.

[pt] Neste trabalho, apresentamos um estudo sobre o fine-tuning de um LLM (Modelo de Linguagem Amplo ou Large Language Model) pré-treinado para a sumarização abstrativa de textos longos em português. Para isso, construímos um corpus contendo uma coleção de 7.450 artigos científicos na área de Ciências da Saúde em português. Utilizamos esse corpus para o fine-tuning do modelo BERT pré-treinado para o português brasileiro (BERTimbau). Em condições semelhantes, também treinamos um segundo modelo baseado em Memória de Longo Prazo e Recorrência (LSTM) do zero, para fins de comparação. Nossa avaliação mostrou que o modelo ajustado obteve pontuações ROUGE mais altas, superando o modelo baseado em LSTM em 30 pontos no F1-score. O fine-tuning do modelo pré-treinado também se destaca em uma avaliação qualitativa feita por avaliadores a ponto de gerar a percepção de que os resumos gerados poderiam ter sido criados por humanos em uma coleção de documentos específicos do domínio das Ciências da Saúde. / [en] In this work, we present a study on the fine-tuning of a pre-trained Large Language Model for abstractive summarization of long texts in Portuguese. To do so, we built a corpus gathering a collection of 7,450 public Health Sciences papers in Portuguese. We fine-tuned a pre-trained BERT model for Brazilian Portuguese (the BERTimbau) with this corpus. In a similar condition, we also trained a second model based on Long Short-Term Memory (LSTM) from scratch for comparison purposes. Our evaluation showed that the fine-tuned model achieved higher ROUGE scores, outperforming the LSTM based by 30 points for F1-score. The fine-tuning of the pre-trained model also stands out in a qualitative evaluation performed by assessors, to the point of generating the perception that the generated summaries could have been created by humans in a specific collection of documents in the Health Sciences domain.