Publicação
Modelação e estimativa de incertezas no ensaio quase-dinâmico de coletores solares térmicos
| Resumo: | The present work discusses and evaluates the quasi-dynamic test procedure for solar thermal collectors according to EN12975, in terms of the determination of the collector parameters and their uncertainties. The collector parameters are identified by multiple linear regression, applied using a Python 3.3 environment. Two approaches were pursuit, one using a weighted multiple linear regression (Weighted Least Squares) and another using a standard multiple linear regression (Ordinary Least Squares). In the weighted multiple linear regression, the uncertainty of the input and output variables is estimated using the GUM procedure, while for the standard least squares approach, the uncertainty of those variables is accounted using the Monte Carlo method. The two approaches were tested for the characterization of a glazed flat plate collector and an evacuated tube collector with satisfying results. Comparisons are made with the results already obtained at LES. For the evacuated tube collector, two approaches for the modeling of the incidence angle modifier were taken, one using the factorized transversal incident angle for the direct beam radiation (dummy variables method), and another using a predefined polynomial equation cited on the literature. It is observable that the last gives a slight more accurate result than the first. |
|---|---|
| Autores principais: | Galvão, Carolina Isabel da Silva |
| Assunto: | Mínimos quadrados simples Mínimos quadrado pesados GUM Método de Monte Carlo Coletores planos Coletores de tubos de vácuo Estimatiza de incertezas Teses de mestrado - 2014 |
| Ano: | 2014 |
| País: | Portugal |
| Tipo de documento: | dissertação de mestrado |
| Tipo de acesso: | acesso aberto |
| Instituição associada: | Universidade de Lisboa |
| Idioma: | português |
| Origem: | Repositório da Universidade de Lisboa |
| Resumo: | The present work discusses and evaluates the quasi-dynamic test procedure for solar thermal collectors according to EN12975, in terms of the determination of the collector parameters and their uncertainties. The collector parameters are identified by multiple linear regression, applied using a Python 3.3 environment. Two approaches were pursuit, one using a weighted multiple linear regression (Weighted Least Squares) and another using a standard multiple linear regression (Ordinary Least Squares). In the weighted multiple linear regression, the uncertainty of the input and output variables is estimated using the GUM procedure, while for the standard least squares approach, the uncertainty of those variables is accounted using the Monte Carlo method. The two approaches were tested for the characterization of a glazed flat plate collector and an evacuated tube collector with satisfying results. Comparisons are made with the results already obtained at LES. For the evacuated tube collector, two approaches for the modeling of the incidence angle modifier were taken, one using the factorized transversal incident angle for the direct beam radiation (dummy variables method), and another using a predefined polynomial equation cited on the literature. It is observable that the last gives a slight more accurate result than the first. |
|---|