| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| 7.7 MB | Adobe PDF |
Authors
Advisor(s)
Abstract(s)
A iniciação análise de circuitos elétricos durante o primeiro ano de Engenharia Eletrotécnica pode revelar-se uma tarefa desafiante. Desde a correta identificação dos Ramos, Nós e Malhas, à identificação das Leis de Kirchhoff e ainda a aplicação de teoremas de simplificação de circuitos, envolvem alguma complexidade. Num curto espaço de tempo, os alunos familiarizam-se com os simuladores de circuitos elétricos, sendo que estes lhes oferecem inicialmente uma validação para exercícios teóricos e análises experimentais, e mais tarde revelam-se fundamentais para o projeto e aperfeiçoamento de circuitos elétricos. No entanto, os simuladores tradicionais não demonstram como chegar aos resultados nem sequer as leis que os fundamentam. Assim, a presente dissertação aborda o design e implementação de uma framework de análise de circuitos elétricos – U=RIsolve – que fornece uma solução mais pedagógica para o Método das Tensões nos Nós (MTN). Este método permite obter a tensão em todos os Nós do Circuito em relação a um Nó de referência, e consequentemente a corrente em todos os Ramos. A ferramenta U=RIsolve é uma aplicação web que utiliza um modelo descritivo de um determinado circuito (netlist), gerado pelo simulador QUCS, de forma a analisar e gerar uma solução por etapas, que descreve pormenorizadamente a aplicação do MTN. Para além da aplicação, este trabalho apresenta um algoritmo robusto e abrangente, baseado no paradigma dos Supernós, capaz de resolver circuitos de qualquer complexidade, ao contrário do MTN tradicional. O algoritmo proposto fornece ainda uma lógica de execução que facilita a sua implementação em sistemas computacionais e permite obter uma solução passo a passo, completa e intuitiva.
The first steps for rookie Electrical Engineering students to learn circuit analysis are quite challenging. The correct identification of Branches, Nodes and Loops, the expression of Kirchhoff Current and Voltage Laws and the application of analysis/simplification Theorems and Algorithms, all embed tricky and error-prone steps for beginners. Hopefully, students get more and more used to circuit simulators with a very quick learning curve, turning these tools extremely appealing for validating theoretical and experimental analysis and, later on, a paramount help for designing, validating and fine-tuning hardware projects. Nonetheless, traditional circuit simulators neither explain how to obtain those results nor the fundamental laws behind them. In this context, this Thesis addresses the design and implementation of a circuit analysis framework – U=RIsolve – for teaching/self-learning the Nodal voltage Method (NVM, for short). This method enables to determine the voltage in every Node in the circuit, related to a reference (Ground) Node, and consequently the Current in every Branch. U=RIsolve (read ”you resolve”) is a web-based application that uses a circuit description model (netlist) generated by the QUCS simulator, to analyse and output a symbolic step-based solution that describes how to apply the NVM and “solve” the circuit. Furthermore, this essay proposes a robust and comprehensive NVM algorithmic approach based on the Supernode paradigm – NVM-SA, which is able to tackle circuits with any complexity level, unlike the baseline NVM algorithm. The proposed NVM-SA algorithm provides a straightforward execution logic, making it suitable for computer implementation and enabling a step-by-step, very complete and intuitive output.
The first steps for rookie Electrical Engineering students to learn circuit analysis are quite challenging. The correct identification of Branches, Nodes and Loops, the expression of Kirchhoff Current and Voltage Laws and the application of analysis/simplification Theorems and Algorithms, all embed tricky and error-prone steps for beginners. Hopefully, students get more and more used to circuit simulators with a very quick learning curve, turning these tools extremely appealing for validating theoretical and experimental analysis and, later on, a paramount help for designing, validating and fine-tuning hardware projects. Nonetheless, traditional circuit simulators neither explain how to obtain those results nor the fundamental laws behind them. In this context, this Thesis addresses the design and implementation of a circuit analysis framework – U=RIsolve – for teaching/self-learning the Nodal voltage Method (NVM, for short). This method enables to determine the voltage in every Node in the circuit, related to a reference (Ground) Node, and consequently the Current in every Branch. U=RIsolve (read ”you resolve”) is a web-based application that uses a circuit description model (netlist) generated by the QUCS simulator, to analyse and output a symbolic step-based solution that describes how to apply the NVM and “solve” the circuit. Furthermore, this essay proposes a robust and comprehensive NVM algorithmic approach based on the Supernode paradigm – NVM-SA, which is able to tackle circuits with any complexity level, unlike the baseline NVM algorithm. The proposed NVM-SA algorithm provides a straightforward execution logic, making it suitable for computer implementation and enabling a step-by-step, very complete and intuitive output.
Description
Keywords
Método da Tensão nos Nós Análise de circuitos Abordagem dos Supernós QUCS Simulador Netlists Aplicação web Auto-aprendizagem Nodal Voltage Method Circuit analysis Supernode approach Circuit simulator Web-application Self-learning JavaScript