Consider an under-determined system of nonlinear equations F(x)=0, F:R^m→R^n, where F is continuously differentiable and m > n. This system appears in a variety of applications, including parameter-dependent systems, dynamical systems with periodic solutions, and nonlinear eigenvalue problems. Robust, efficient numerical methods are often required for the solution of this system. Newton's method is an iterative scheme for solving the nonlinear system of equations F(x)=0, F:R^n→R^n. Simple to implement and theoretically sound, it is not, however, often practical in its pure form. Inexact Newton methods and globalized inexact Newton methods are computationally efficient variations of Newton's method commonly used on large-scale problems. Frequently, these variations are more robust than Newton's method. Trust region methods, thought of here as globalized exact Newton methods, are not as computationally efficient in the large-scale case, yet notably more robust than Newton's method in practice. The normal flow method is a generalization of Newton's method for solving the system F:R^m→R^n, m > n. Easy to implement, this method has a simple and useful local convergence theory; however, in its pure form, it is not well suited for solving large-scale problems. This dissertation presents new methods that improve the efficiency and robustness of the normal flow method in the large-scale case. These are developed in direct analogy with inexact-Newton, globalized inexact-Newton, and trust-region methods, with particular consideration of the associated convergence theory. Included are selected problems of interest simulated in MATLAB.

In visually guided control of a robot, a large residual problem occurs when the robot configuration is not in the neighborhood of the target acquisition configuration. Most existing uncalibrated visual servoing algorithms use quasi-Gauss-Newton methods which are effective for small residual problems. The solution used in this study switches between a full quasi-Newton method for large residual case and the quasi-Gauss-Newton methods for the small case. Visual servoing to handle large residual problems for tracking a moving target has not previously appeared in the literature. For large residual problems various Hessian approximations are introduced including an approximation of the entire Hessian matrix, the dynamic BFGS (DBFGS) algorithm, and two distinct approximations of the residual term, the modified BFGS (MBFGS) algorithm and the dynamic full Newton method with BFGS (DFN-BFGS) algorithm. Due to the fact that the quasi-Gauss-Newton method has the advantage of fast convergence, the quasi-Gauss-Newton step is used as the iteration is sufficiently near the desired solution. A switching algorithm combines a full quasi-Newton method and a quasi-Gauss-Newton method. Switching occurs if the image error norm is less than the switching criterion, which is heuristically selected. An adaptive forgetting factor called the dynamic adaptive forgetting factor (DAFF) is presented. The DAFF method is a heuristic scheme to determine the forgetting factor value based on the image error norm. Compared to other existing adaptive forgetting factor schemes, the DAFF method yields the best performance for both convergence time and the RMS error. Simulation results verify validity of the proposed switching algorithms with the DAFF method for large residual problems. The switching MBFGS algorithm with the DAFF method significantly improves tracking performance in the presence of noise. This work is the first successfully developed model independent, vision-guided control for large residual with capability to stably track a moving target with a robot.

Mémoire (M.Sc. A.)-- Université du Québec à Chicoutimi, 1984. / "Mémoire présenté comme exigence partielle de la maîtrise en sciences appliquées en ressources et systèmes" CaQCU bibliographie: p. 56-58. Document électronique également accessible en format PDF. CaQCU

Thesis (Ph. D.)--Ohio State University, 2005. / Available online via OhioLINK's ETD Center; full text release delayed at author's request until 2006 Aug 15.

Hydrauliska högtryckskopplingar av typen FEM ½” studeras med avsikten att fastställa en effektiv beräkningsmetodik som kan användas till att prediktera kopplingarnas hållfasthet. Metodiken utgörs av finita element analyser (FEA), och valideras av experimentella trycktester utförda på kopplingstypen FEM ½”. Genom FEA kan kopplingarnas hållfasthetsbeteende och maximala belastningskapacitet studeras virtuellt, vilket minskar behovet av experimentella tester och medför potential för optimering av produkterna. Arbetet utförs på Parker Hannifin AB i Skövde. Experimentella tester utförs på 20 stycken kopplingspar av typen FEM ½” för att utöka förståelsen av kopplingarnas beteende under brottsförloppet och för att prediktera trycket som medför haveri. Testernas genomförande och struktur baseras på metodiken Design of Experiments (DOE). Kritiska komponenter identifieras utifrån experimentets resultat, vilka sedan studeras närmare via FEA. Analyserna valideras utifrån standarder som kopplingarna ska efterfölja, och mätdata insamlad under de experimentella testerna. Från de experimentella testerna är det komponenterna: kulhållaren, styrningen och nippelhuset som upptar belastning i störst utsträckning. Vid haveri framgår två brottmoder som vanliga, att kulhållaren slits isär samt att styrningen brister, båda fallen uppkommer vid approximativt samma tryck. FE-analyserna för styrningen och kulhållaren visar god överensstämmelse med experimentella resultat. Deformationerna skiljer sig dock mellan analyserna och de experimentella testerna, var nippelhusets analyser uppvisar störst avvikelser. FE-modellerna uppvisar god potential för att prediktera samt utvärdera kopplingarnas mekaniska beteende under tryckbelastning. Analyserna är dock helt beroende av ingående data, var saknaden av en verklig materialmodell medför avvikelser från experimentella resultat. Förhållandet framgår tydligt av nippelhuset, vars relaterade härdningsegenskaper saknas. / Hydraulic FEM ½" high pressure couplings are studied with the purpose of establishing an effective methodology that can be used to predict the strength of the couplings. The methodology consists of finite element analyzes (FEA) and is validated by experimental pressure tests, performed on the FEM ½” couplings pairs.  Using FEA, the couplings solid mechanical behavior and maximum load ability can be viewed virtual, reducing the need for experimental tests and gives the potential for optimized products. The work is performed at Parker Hannifin AB in Skovde. Experimental tests are performed on 20 FEM ½” couplings pairs, to understand the solid mechanical behavior of the couplings until failure occurs, and to predict the maximum pressure that can be applied. The experimental structure and performance is based on the method Design of Experiments (DOE). Critical components are identified based on the results from the experimental tests, which are then studied more closely through FEA. The analysis are validated based on the applied material model, and data collected during the experimental tests. From the experimental tests it is shown that the components: ball cage, guide and plug housing are the components in which failure occur. In case of failure, two failure modes appear as common, that the ball cage is worn apart and that the guide burst, both types of failure modes occur at a similar pressure. The analysis for the guide and ball cage corresponds with the experimental outcomes. Differences occurs however when looked at the deformations, in which the plug housing shows the largest deviation when compared to the experimental results. The usage of FE-models appears to be appropriate for predicting and evaluating the mechanical strengths of the couplings during pressure loads. The analysis are however entirely dependent on the input data, where an incorrect material model generates incorrect results. The relationship is shown for the plug housing, which lack the mechanical properties related to curing processes.

Orientador: Marco Lucio Bittencourt / Dissertação (mestrado) - Universidade Estadual de Campinas, Faculdade de Engenharia Mecanica / Made available in DSpace on 2018-08-05T06:50:24Z (GMT). No. of bitstreams: 1 Gerardin_RodrigoCeccatto_M.pdf: 1433853 bytes, checksum: d1832bbbe32aa8b2f8edb3756793f3f5 (MD5) Previous issue date: 2005 / Resumo: Atualmente, devido as exigências comerciais e técnicas, os motores de combustão interna operam com pressões de combustão cada vez mais altas. Ao mesmo tempo, deve-se otimizar a operação de todo sistema e a forma dos componentes levando-se em conta a melhor performance e redução de peso. Para um melhor entendimento do comportamento dinâmico de um motor de multi cilindros, é necessário verificar a cinemática e a dinâmica para apenas um cilindro como feito neste trabalho. O foco principal deste trabalho é desenvolver um modelo matemático para determinar as distribuições de pressão e as forças atuantes nos mancais hidrodinâmicos para um cilindro de um motor de combustão interna. Um modelo dinâmico do sistema de pistão-biela-manivela é apresentado e permite calcular as forças dinâmicas e folgas no mancal principal, olhais maior e menor da biela derivados da pressão de combustão. O modelo matemático do mancal hidrodinâmico radial é oriundo da equação de Reynolds e resolvido utilizando o Método de Elementos Finitos. O sistema dinâmico não-linear é resolvido utilizando o método iterativo de Newton-Raphson para cada passo de integração no tempo / Abstract: Due to the current commercial and technical requirements, the internal combustion engines must operate under higher pressures. It is also necessary to optimize the system operation and the shape of the components aiming at increasing the performance and weight reducing. For a better understanding of the dynamical behavior of a multi-cylinder engine, it is necessary to verify the kinematics and dynamics for just one cylinder, as considered in this work. The main focus of this work is the development of a mathematical model to determine the pressure distributions and the hydrodynamic bearing forces for one cylinder internal combustion engine. The dynamical model of the piston-conrod-crank system is presented and allows the calculation of the dynamic forces and clearances obtained from the combustion pressure for the main, big-end and small-end bearings. The mathematical model of the hydrodynamic bearing comes from the Reynolds equation and is solved by the Finite Element Method. The non-linear dynamic system is solved by the iterative Newton-Rhapson method for each time integration step / Mestrado / Mecanica dos Sólidos e Projeto Mecanico / Mestre em Engenharia Mecânica

Orientadores: Walmir de Freitas Filho, Luiz Carlos Pereira da Silva / Dissertação (mestrado) - Universidade Estadual de Campinas, Faculdade de Engenharia Eletrica e de Computação / Made available in DSpace on 2018-08-11T12:47:41Z (GMT). No. of bitstreams: 1 Anzai_AlexandreHaruiti_M.pdf: 1115532 bytes, checksum: 8d2da875fae1c7141a82e80bcad115b7 (MD5) Previous issue date: 2008 / Resumo: Esta dissertação de mestrado apresenta um modelo monofásico de fluxo de carga para sistemas de distribuição capaz de representar as tecnologias mais comumente utilizadas em geração distribuída. Tal ferramenta permite a realização de estudos técnicos e de diagnósticos em relação à conexão de um novo gerador nos alimentadores de sistemas de distribuição. Propõe-se um método de cálculo de fluxo de carga expandido, no qual os modelos dos geradores, turbinas e seus controles são levados em conta na resolução das equações do fluxo de carga. O método de fluxo de carga expandido considera que o sistema é modelado através de equações algébricas e diferenciais num ponto de equilíbrio, ao invés da modelagem convencional baseada apenas em equações algébricas de balanço de potência. As tecnologias de geração analisadas são compostas por máquinas síncronas, máquinas de indução e geradores conectados via conversores estáticos de potência. Diferentes fontes primárias de energia são consideradas, incluindo turbinas a vapor, hidráulicas e eólicas, além dos sistemas de controle associados às partes elétrica e mecânica de caqa tecnologia. O método de fluxo de carga proposto é testado usando uma rede de distribuição real do Estado de São Paulo / Abstract: This dissertation presents a monophasic load flow model for distribution systems capable of representing the most common technologies used for distributed generation. The developed computational tool allows the accomplishment of technical studies and diagnosis related to the impacts of connecting new generators on distribution feeders. It is proposed an expanded load flow method, including models for generators, turbines and associated controllers, for the solution of load flow equations. The expanded load flow method represents the system by using algebraic and differential equations in a equilibrium point, instead of the conventional modeling based only on the power balance algebraic equations. The technologies used for generation are composed by synchronous machines, induction machines and inverter based generators. Different prime mover technologies are considered, including steam turbines, hydraulic turbines, wind turbines, and control systems associated with the electric and mechanical operation of each technology. The proposed load flow method is tested by using a real distribution network of the State of São Paulo / Mestrado / Energia Eletrica / Mestre em Engenharia Elétrica

Submitted by Geandra Rodrigues (geandrar@gmail.com) on 2018-07-04T15:02:14Z No. of bitstreams: 1 rafaeldepaivapontes.pdf: 2283859 bytes, checksum: b3ed61e125a30bbe054bb3183741ad41 (MD5) / Approved for entry into archive by Adriana Oliveira (adriana.oliveira@ufjf.edu.br) on 2018-07-06T14:23:02Z (GMT) No. of bitstreams: 1 rafaeldepaivapontes.pdf: 2283859 bytes, checksum: b3ed61e125a30bbe054bb3183741ad41 (MD5) / Made available in DSpace on 2018-07-06T14:23:02Z (GMT). No. of bitstreams: 1 rafaeldepaivapontes.pdf: 2283859 bytes, checksum: b3ed61e125a30bbe054bb3183741ad41 (MD5) Previous issue date: 2018-03-21 / Este trabalho, realiza uma revisão da metodologia tradicional, de representação dos limites de geração de potência reativa, na solução do problema de fluxo de potência, em coordenadas polares, pelo método de Newton-Raphson. Apresenta uma nova modelagem para o tratamento destes limites, baseada em uma formulação full Newton do problema. Para tanto, utiliza-se um conjunto de chaves sigmoides, que incorporam novas equações à matriz Jacobiana, de acordo com a geração de potência reativa da barra. De forma a tornar o sistema possível e determinado, a geração de potência reativa, é tratada como uma nova variável de estado do problema, corrigida a cada iteração do método de Newton. A formulação proposta, utiliza duas chaves sigmoides para o tratamento dos limites de geração de potência reativa, em barras PV e, duas chaves, que consideram o procedimento de retomada do controle (conhecido como estratégia de backoff ), das barras que atingiram limites, caso necessário. Dessa forma, não há uma mudança explícita no tipo da barra, como na metodologia tradicional, e a dimensão da matriz Jacobiana é mantida constante. Neste trabalho, todo o equacionamento e formulação, foi desenvolvido através do programa Matlab, e os resultados foram validados, utilizando-se o programa de análise de redes – ANAREDE, versão acadêmica, disponibilizada pelo Centro de Pesquisas de Energia Elétrica (CEPEL). São avaliados alguns sistemas benchmark IEEE e os resultados obtidos, demonstram a eficácia das formulações propostas. / This work, presents a review of the traditional methodology, of representing the limits of reactive power generation, in the solution of the power flow problem, in polar coordinates, by the Newton-Raphson method. Presents a new modeling for the treatment of these limits, based on a formulation full Newton of the problem. For this purpose, a set of sigmoid switches is used, to incorporate new equations to the Jacobian matrix, according to the generation of reactive power of the bus. In order to produce a system possible and determined, reactive power generation, is treated as a new problem state variable, corrected at each iteration of the Newton’s method. The proposed formulation, use two sigmoid switches for the tratment of reactive power generation limits, in PV buses and also, two switches, that consider the procedure of resumption of control (procedure known as backoff strategy), of the buses that reached some limit, if necessary. Thus, there is no explicit change in the bus type, as in traditional methodology, and the Jacobian matrix dimension is kept constant. In this work, all equation and formulation, were developed through Matlab software, and the results were checked, using the network analysis program – ANAREDE, academic version provided by Electrical Energy Research Center (CEPEL). Some IEEE benchmark systems are evaluated, and the results obtained, demonstrates the efficacy of the proposed formulations.

Orientador: Antonio Carlos Bannwart / Dissertação (mestrado) - Universidade Estadual de Campinas, Faculdade de Engenharia Mecânica e Instituto de Geociências / Made available in DSpace on 2018-08-27T04:22:37Z (GMT). No. of bitstreams: 1 BarbosaNeto_AntonioMarinho_M.pdf: 2354581 bytes, checksum: dada46fe50ab798d217e21bcad15eeb1 (MD5) Previous issue date: 2015 / Resumo: Fluidos de petróleo, incluindo gás, condensado e óleo leve, são misturas complexas de hidrocarbonetos. O conhecimento sobre o comportamento de fases destas misturas é de interesse para a indústria de petróleo, desde o reservatório, produção, transporte, até o processamento e refino de petróleo. Além disso, pacotes termodinâmicos computacionais para predizer o comportamento de tais misturas são de grande interesse nesta indústria. Neste sentido, uma ferramenta computacional para prever com precisão as propriedades termodinâmicas utilizando cálculos de equilíbrio de fases pode desempenhar um papel importante para aplicações de projeto e operações de campo. Equações de estado cúbicas tem sido amplamente utilizada para predizer o comportamento de fases e propriedades volumétricas de fluidos de petróleo devido a sua simplicidade e alto desempenho computacional. Diante deste contexto, o presente trabalho teve como objetivo principal desenvolver um simulador PVT composicional, de fácil manuseio, porém completo, capaz de avaliar o comportamento volumétrico e de fases de fluidos de petróleo em diferentes condições operacionais (P, T) a partir da modelagem do equilíbrio líquido-vapor e do equilíbrio líquido-líquido-vapor. A ferramenta computacional foi desenvolvida na linguagem de programação Visual Basic Applications (VBA) em aplicativo Excel. Uma interface gráfica foi construída em aplicativo Excel. O programa, denominado PVTpetro, consiste basicamente em três módulos computacionais e suas funções auxiliares. O módulo de cálculo de flash (P, T) 2 fases isotérmico para misturas de hidrocarbonetos consiste no equilíbrio líquido-vapor (ELV) para sistemas óleo-gás multicomponentes. O módulo de cálculo de flash (P, T) 3 fases isotérmico avalia sistemas envolvendo óleo, água e gás (OWG). E, o módulo de cálculo de propriedades volumétricas de fluidos de petróleo que calcula o peso molecular, o volume e a massa específica das fases existentes no sistema. A ferramenta foi validada via comparação entre os resultados de pacotes termodinâmicos comerciais (Wimprop, Hysys, COMThermo) e os do programa desenvolvido. Os resultados gerados pelo PVTpetro foram avaliados através do cálculo do desvio relativo em relação às respostas dos softwares comerciais. Desta forma, verificou-se para o módulo de cálculo flash (P, T) 2 fases que os valores de desvio em relação ao Wimprop e Hysys foram inferiores a 1% para os estudos realizados. Enquanto que, para o módulo de cálculo flash (P, T) 3 fases os valores de erro relativo se concentraram na faixa de 0 ¿ 5% quando comparados com os pacotes termodinâmicos Hysys e COMThermo. De uma maneira geral, os resultados gerados pelo PVTpetro apresentaram-se satisfatórios em todos os cálculos propostos. Na validação do módulo de cálculo das propriedades volumétricas, feita de forma qualitativa, observou-se que o comportamento das propriedades em função da pressão e em diferentes temperaturas, gerado pelo PVTpetro, reproduziram as curvas fornecidas pelo Hysys, em todas as condições investigadas. Portanto, além de se apresentar confiável em seus cálculos o PVTpetro mostrou ser de fácil manuseio e intuitivo para o usuário, facilitando assim o trabalho em vários casos / Abstract: Petroleum fluids, which include gas, condensate and light oil, are complex hydrocarbons mixtures. There is a great deal of interest in the oil industry in computational thermodynamic packages able to predict the phase behavior of such mixtures, from the reservoir, production, transportation until the oil processing and refining. In this sense, a computational tool that accurately predicts thermodynamic properties using phase equilibrium calculations have an important role in designing applications and field operations. Cubic equations of state have been widely used to predict the phase behavior and bulk properties of petroleum fluids due to its simplicity and high computational performance. This study focused on developing a PVT compositional simulator, easy to use, able to assess the bulk properties and phase behavior of petroleum fluids at different operating conditions (P, T) from the modeling of the vapor-liquid equilibrium (VLE) and the vapor-liquid-liquid equilibrium (VLLE). The computational tool was developed in Visual Basic Applications (VBA) programming language, with a graphic interface built in Excel. The program, named PVTpetro, consists in three basic computational modules and their auxiliary functions. The isothermal two phase flash (P, T) module calculation to hydrocarbon mixtures consists of VLE for multi-component oil-gas systems. The isothermal three phase flash (P, T) module calculation evaluates oil, water and gas (OWG) systems. Lastly, the bulk properties of petroleum fluids calculation module calculates the molecular weight, the volume and density of the system phases. The tool has been validated through the comparison between the results of commercial thermodynamic packages (Wimprop, Hysys, COMThermo) with the developed program. The results were evaluated by the relative deviation calculation between these commercial software and PVTpetro results. For the isothermal two-phase flash (P, T) calculation module were observed deviations below 1% for Wimprop and Hysys results. For the isothermal three phase flash (P, T) calculation module the deviations were about 0 and 5% for Hysys and COMThermo. For the bulk properties calculation module the results by PVTpetro and Hysys showed a good agreement in all investigated conditions, ensuring the PVTpetro reliability / Mestrado / Explotação / Mestre em Ciências e Engenharia de Petróleo

In this thesis, we present the use of logistic regression method to develop a credit scoring modelusing the raw data of 4447 customers of a bank. The data of customers is collected under 14independent explanatory variables and 1 default indicator. The objective of this thesis is toidentify optimal coefficients. In order to clean data, the raw data set was put through variousdata calibration techniques such as Kurtosis, Skewness, Winsorization to eliminate outliers.On this winsorized dataset, LOGIT analysis is applied in two rounds with multiple statisticaltests. These tests aim to estimate the significance of each independent variable and modelfitness. The optimal coefficients can be used to obtain the credit scores for new customers witha new data set and rank them according to their credit risk.