[pt] Meios porosos fraturados estão presentes em diferentes tipos de formações geológicas, como os maciços rochosos e os reservatórios de petróleo e gás. A modelagem adequada dos sistemas de fraturas presentes nesses meios é de grande relevância para o desenvolvimento de estratégias de exploração e produção dessas formações. Isso porque os processos de fluxo de fluido são fortemente influenciados pelas características dos sistemas de fraturas. Nesse contexto, diversas abordagens têm sido desenvolvidas para a modelagem desses problemas utilizando representações explícitas e implícitas para as fraturas. A representação explícita usando modelos de fraturas discretas fornece resultados precisos, mas possui um custo computacional elevado e apresenta dificuldades na construção de modelos mais complexos. Por outro lado, modelos de representação implícita, como o de dupla porosidade/dupla permeabilidade, são muito atrativos por incorporarem o efeito das fraturas nas simulações sem a necessidade de representá-las no modelo. No entanto, esses modelos são adequados para problemas envolvendo fraturas pequenas e conectadas, possuindo aplicabilidade limitada para representar fraturas principais de maior escala que podem dominar o fluxo. Assim, este trabalho apresenta algumas das abordagens disponíveis para a representação de formações porosas fraturadas. Diferentes cenários foram estudados para avaliar pontos fortes e limitações de cada método em diferentes aplicações. Além disso, uma nova formulação foi proposta para representar o efeito de fraturas isoladas, que se demonstrou eficiente em modelos com considerável número de fraturas. / [en] Fractured porous media are present in different types of geological formations as rock masses and oil and gas reservoirs. The proper modelling of the fractured systems present in these media is of high relevance to the development of production and exploitation strategies of these formations. This is because the fluid flow processes are strongly influenced by the fractured systems characteristics. In this context, several approaches have been developed to model these problems using explicit and implicit representations to fractures. The explicit representation using discrete fracture models provides accurate results, but has a high computational cost and exhibits difficulties to construct more complex models. On the other hand, implicit representation models, as the dual porosity/dual permeability, are very attractive because they incorporate the effect of fractures to simulations without the need to represent them explicitly in the models. However, these models are suitable to problems with small and connected fractures, and have limited capability to represent major fractures of larger scale that can dominate the flow. Therefore, this work shows some of the available approaches to represent fractured porous formations. Moreover, a new formulation was proposed to represent the effect of isolated fractures, which proved to be efficient in models with considerable number of fractures.

In the emerging field of audio generation using diffusion models, this project pioneers the adaptation of the AudioLDM model framework, initially designed for text-to-daily sounds generation, towards text-to-music audio generation. This shift addresses a gap in the current scope of audio diffusion models, predominantly focused on everyday sounds. The motivation for this thesis stems from AudioLDM’s remarkable generative capabilities in producing daily sounds from text descriptions. However, its application in music audio generation remains underexplored. The thesis aims to modify AudioLDM’s architecture and training objectives to cater to the unique nuances of musical audio. The re-engineering process involved two primary methods. First, a dataset was constructed by sourcing a variety of music audio samples from the A Dataset For Music Analysis (FMA) [1] and generating pseudo captions using a Large Language Model specified in music captioning. This dataset served as the foundation for training the adapted model. Second, the model’s diffusion backbone, a UNet architecture, was revised in its text conditioning approach by incorporating both the CLAP encoder and the T5 text encoder. This dualencoding method, coupled with a shift from the traditional noise prediction objective to the V-objective, aimed to enhance the model’s performance in generating coherent and musically relevant audio. The effectiveness of these adaptations was validated through both subjective and objective evaluations. Compared to the original AudioLDM model, the adapted version demonstrated superior quality in the audio output and a higher relevance between text prompts and generated music. This advancement not only proves the feasibility of transforming AudioLDM for music generation but also opens new avenues for research and application in text-to-music audio synthesis / Inom det framväxande området för ljudgenerering med användning av diffusionsmodeller, banar detta projekt för anpassningen av AudioLDMmodellramverket, som ursprungligen utformades för generering av text-tilldagliga ljud, mot ljudgenerering av text-till-musik. Denna förändring tar itu med en lucka i den nuvarande omfattningen av ljuddiffusionsmodeller, främst inriktade på vardagliga ljud. Motivationen för denna avhandling kommer från AudioLDM:s anmärkningsvärda generativa förmåga att producera dagliga ljud från textbeskrivningar. Dock är dess tillämpning i musikljudgenerering fortfarande underutforskad. Avhandlingen syftar till att modifiera AudioLDM:s arkitektur och utbildningsmål för att tillgodose de unika nyanserna av musikaliskt ljud. Omarbetningsprocessen involverade två primära metoder. Först konstruerades en datauppsättning genom att hämta en mängd olika musikljudprover från A Dataset For Music Analysis (FMA) [1] och generera pseudotexter med hjälp av en Large Language Model specificerad i musiktextning. Denna datauppsättning fungerade som grunden för att träna den anpassade modellen. För det andra reviderades modellens diffusionsryggrad, en UNet-arkitektur, i sin textkonditioneringsmetod genom att inkludera både CLAP-kodaren och T5-textkodaren. Denna dubbelkodningsmetod, i kombination med en övergång från det traditionella brusförutsägelsemålet till V-målet, syftade till att förbättra modellens prestanda för att generera sammanhängande och musikaliskt relevant ljud. Effektiviteten av dessa anpassningar validerades genom både subjektiva och objektiva utvärderingar. Jämfört med den ursprungliga AudioLDMmodellen visade den anpassade versionen överlägsen kvalitet i ljudutgången och en högre relevans mellan textmeddelanden och genererad musik. Detta framsteg bevisar inte bara möjligheten att transformera AudioLDM för musikgenerering utan öppnar också nya vägar för forskning och tillämpning inom text-till-musik ljudsyntes.