Detalhes bibliográficos
Ano de defesa: |
2015 |
Autor(a) principal: |
Romera, João Paulo Rodrigues [UNESP] |
Orientador(a): |
Não Informado pela instituição |
Banca de defesa: |
Não Informado pela instituição |
Tipo de documento: |
Dissertação
|
Tipo de acesso: |
Acesso aberto |
Idioma: |
por |
Instituição de defesa: |
Universidade Estadual Paulista (Unesp)
|
Programa de Pós-Graduação: |
Não Informado pela instituição
|
Departamento: |
Não Informado pela instituição
|
País: |
Não Informado pela instituição
|
Palavras-chave em Português: |
|
Link de acesso: |
http://hdl.handle.net/11449/143004
|
Resumo: |
The determination of the fiber content is an important parameter to sugar factory because it is directed linked to the yield of process for juice extraction, that is raw material in sugar and ethanol production. Several chemometrics strategies were used in this study such as design of experiments and multivariate calibration models with regression by Partial Least Squares. In the design of experiments, was possible to evaluate the best experimental conditions in Laser Induced Breakdown Spectroscopy to perform the measurements in natura in bagasse sugarcane samples. A total of 5327 LIBS spectra were acquired and composed the data matrix, as independent variables, to investigation of the multivariate calibration models by PLS. The reference values corresponding to the fiber content, the parameter that was predicted by PLS models, were obtained using the method recommended by the sugarcane, sugar and alcohol producers of the São Paulo states (CONSECANA). A better PLS model was achieved through the 119 variables selection with Root Mean Square Error (RMSE) in % to training set (RMSE of calibration) and validation (RMSE of validation) of 2.1 and 2.2, respectively. The developed method in this master's degree study show the inedited possibility in the determination of the fiber content with characteristics such as: no toxic waste generation, less use of electricity and increasing in the analytical frequency. |