No description available.

As Redes Neurais Artificiais (RNAs) vêm sendo empregadas com cada vez mais sucesso em diversas áreas de pesquisa, no campo da engenharia e em outros campos diversos. Neste trabalho foram modeladas séries fluviométricas relativas às afluências a quatro reservatórios, localizados em quatro bacias hidrográficas distintas que compõem a Bacia Metropolitana de Fortaleza, Ceará, Brasil. Tais afluências apresentam peculiaridades relativas à ocorrência de magnitudes nulas, que dificultam sua modelagem através dos convencionais modelos estatísticos da família Box-Jenkins. Neste estudo foram trabalhadas duas abordagens distintas, a primeira univariada, em que cada série era modelada de forma individual, e a segunda multivariada, em que as séries fluviométricas eram modeladas simultaneamente. Os resultados obtidos, segundo ambas as modelagens, demonstram que a técnica apresenta potencial para a aplicação pretendida. Estudos futuros merecem ser desenvolvidos ainda no sentido de verificar a melhor maneira de se enquadrar a componente aleatória nas séries sintéticas produzidas via RNAs. / Artificial Neural Networks (ANNs) are being used more and more in many different fields of research, in engineering applications or other applications. This research deals with modeling of inflows to four reservoirs, located in different watersheds that belong to the Metropolitan Watershed of Fortaleza city, Brazil. These discharge sequences have particular characteristics in that they have frequent occurence of null discharges which makes it difficult to use traditional statistical models such as those Box-Jenkis family. Two different modeling approaches were adopted in this study, the first univariate, in which each time series was modeled individually, and the second multivariate, in which the four time series were modeled simultaneously. The results from the both approaches show that the technique has potential for use in water resources planning and management. Future studies are required to propose better means of incorporing the random component in the generation of synthetic time series through ANNs.

As Redes Neurais Artificiais (RNAs) vêm sendo empregadas com cada vez mais sucesso em diversas áreas de pesquisa, no campo da engenharia e em outros campos diversos. Neste trabalho foram modeladas séries fluviométricas relativas às afluências a quatro reservatórios, localizados em quatro bacias hidrográficas distintas que compõem a Bacia Metropolitana de Fortaleza, Ceará, Brasil. Tais afluências apresentam peculiaridades relativas à ocorrência de magnitudes nulas, que dificultam sua modelagem através dos convencionais modelos estatísticos da família Box-Jenkins. Neste estudo foram trabalhadas duas abordagens distintas, a primeira univariada, em que cada série era modelada de forma individual, e a segunda multivariada, em que as séries fluviométricas eram modeladas simultaneamente. Os resultados obtidos, segundo ambas as modelagens, demonstram que a técnica apresenta potencial para a aplicação pretendida. Estudos futuros merecem ser desenvolvidos ainda no sentido de verificar a melhor maneira de se enquadrar a componente aleatória nas séries sintéticas produzidas via RNAs. / Artificial Neural Networks (ANNs) are being used more and more in many different fields of research, in engineering applications or other applications. This research deals with modeling of inflows to four reservoirs, located in different watersheds that belong to the Metropolitan Watershed of Fortaleza city, Brazil. These discharge sequences have particular characteristics in that they have frequent occurence of null discharges which makes it difficult to use traditional statistical models such as those Box-Jenkis family. Two different modeling approaches were adopted in this study, the first univariate, in which each time series was modeled individually, and the second multivariate, in which the four time series were modeled simultaneously. The results from the both approaches show that the technique has potential for use in water resources planning and management. Future studies are required to propose better means of incorporing the random component in the generation of synthetic time series through ANNs.

Changes in precipitation patterns, temperature, and other climatic variables have been shown to modify thehydrological cycle and hydrological systems, potentially resulting in a shift in river runoff behavior and an increasedrisk of floods. There have been several instances of devastating floods throughout Europe’s history, which haveresulted in devastation and enormous economic losses. As a result of the effects of climate change, floods areoccurring more frequently in Sweden as well as across Europe. Research on the subject of flood prediction has beengoing on for decades, where particularly data-driven models have advanced in recent years. This study examinedtwo different machine learning (data-driven) models for forecasting river discharge in the Nissan River: Linearregression and Random Forrest regression (RFR), with the use of ECMWF Reanalysis v5 ( ERA5 ) data and historicaldischarge data. The Linear regression model yielded a r2 score of 0.45 and could not be considered an acceptablemodel. The RFR model had a r2 score of 0.71. This implies, given ERA5 reanalysis data, that one might generatea moderately performing machine learning model for Nissan river. An additional investigation was carried out,to see if the trained model could be used with EC-EARTH CMIP6 future projection. The findings resulting fromapplying the EC-EARTH CMIP6 future data on the trained RFR indicated too many uncertainties, necessitatingmore investigation before any conclusions can be drawn.

By 2030, it is expected that 80% of all electric power will flow through power electronics systems. Wide bandgap power modules that can tolerate higher voltages and currents than silicon-based modules are the most promising solution to reducing the size and weight of power electronics systems. These wide-bandgap power modules constitute powerful building blocks for power electronics systems, and wide bandgap-based converter/power electronics building blocks are envisaged to be widely used in power grids in low- and medium-voltage applications and possibly in high-voltage applications for high-voltage direct current and flexible alternating current transmission systems. One of the merits of wide bandgap devices is that their slew rates and switching frequencies are much higher than silicon-based devices. However, from the insulation side, frequency and slew rate are two of the most critical factors of a voltage pulse, influencing the level of degradation of the insulation systems that are exposed to such voltage pulses. The shorter the rise time, the shorter the lifetime. Furthermore, lifetime dramatically decreases with increasing frequency. Thus, although wide bandgap devices are revolutionizing power electronics, electrical insulating systems are not prepared for such a revolution; without addressing insulation issues, the electronic power revolution will fail due to dramatically increased failure rates of electrification components. In this regard, internal partial discharges (PDs) have the most effect on insulation degradation. Internal PDs which occur in air-filled cavities or voids are localized electrical discharges that only partially bridge the insulation between conductors. Voids in solid or gel dielectrics are challenging to eliminate entirely and may result simply during manufacturing process. The objective of this study is to develop a Finite-Element Analysis (FEA) PD model under fast, repetitive voltage pulses, which has been done for the first time. The model is coded and implemented in COMSOL Multiphysics linked with MATLAB, and its simulation results are validated with experimental tests. Using the model, the influence of different parameters including void shape, void size, and void air pressure on PD parameters are studied. / M.S. / To decarbonize and reduce energy consumption for commercial aviation, the development of lightweight and ultra-efficient all-electric powertrain including electric motors, drives, and associated thermal management systems has been targeted. Using wide bandgap (WBG) power modules that can tolerate high voltages and currents can reduce the size and weight of the drive. However, the operation of WBG-based power converter can endanger the reliability of the electrified systems, most importantly, the insulation system. In this study, it is attempted to model the impact of such threats to the insulation system using numerical models.