Powered by OpenAIRE graph
Found an issue? Give us feedback
image/svg+xml art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos Open Access logo, converted into svg, designed by PLoS. This version with transparent background. http://commons.wikimedia.org/wiki/File:Open_Access_logo_PLoS_white.svg art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos http://www.plos.org/ Estudo Geralarrow_drop_down
image/svg+xml art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos Open Access logo, converted into svg, designed by PLoS. This version with transparent background. http://commons.wikimedia.org/wiki/File:Open_Access_logo_PLoS_white.svg art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos http://www.plos.org/
Estudo Geral
Master thesis . 2020
Data sources: Estudo Geral
addClaim

Modelos de previsão de quebras de folha para a industria da pasta e papel

Authors: Dias, Márcia Vanessa Pereira;

Modelos de previsão de quebras de folha para a industria da pasta e papel

Abstract

As quebras de folha apresentam-se como um dos maiores problemas na indústria da pasta e do papel, influenciando não só o desempenho do processo mas também a qualidade e valor do produto final. Assim, conseguir prever e evitar uma quebra de folha pode significar ganhos significativos de produtividade e eficiência. Com a colaboração da The Navigator Company, o presente trabalho visa construir um modelo de previsão do risco de ocorrer uma quebra de folha na máquina de papel tissue. Numa primeira fase, procede-se a uma análise do sistema de forma a identificar variáveis importantes ao problema. Segue-se uma análise de dados que procura conhecer a relação entre as variáveis selecionadas e que, para além disso, procura preparar a amostra que será usada na fase de aprendizagem do modelo. Por último, recorre-se a árvores de decisão de forma a construir o modelo de previsão. Percebe-se que fazer uma previsão sólida é dificultada pela complexidade do fabrico de tissue e que cada condição não consegue ser explicada por um pequeno conjunto de variáveis. Alcança-se um modelo de previsão que apresenta uma exatidão superior a 80% numa fase de teste com dados novos e desconhecidos para o mesmo, sendo que se recorre apenas a 19 variáveis independentes na sua versão final. São identificadas sete variáveis com o maior poder decisivo e que devem ser o foco de atenção perante o crescimento de risco de quebra: pH das águas brancas, lâmina de limpeza, gramagem, velocidade, percentagem de broke no hood layer, vácuo de sucção e pH da pasta slush.

Web breaks are one of the biggest problems in the pulp and paper industry, not only affecting the performance of the process, but also the quality and value of the final product. Therefore, being able to predict and avoid web breaks means significant gains in productivity and efficiency. In collaboration with The Navigator Company, this dissertation aims at creating a prediction model for the risk of web breaks in a tissue machine. Firstly, a system analysis is made in order to identify the essential characteristics of the problem. This step is followed by a data analysis that seeks to understand the relationships between selected variables. Besides that, it seeks to prepare the dataset for the learning phase of the model. Finally, decision trees are used in order to build the prediction model. It is noticed that making a solid web break prediction is hampered by the complexity of the process and that each condition cannot be explained by a small set of variables. A prediction model is achieved and it is tested with new and unknown data. It has accuracy greater than eighty percent and it only uses nineteen independent variables in its final version. A total of seven variables with the highest importance are identified and it should be the focus of attention in case of increased risk: pH value of white water, utilisation of the cleaning blade, grammage of the paper web, speed of the tissue machine, percentage of broke in the hood layer, vacuum of the suction roll and pH value of slush pulp.

Dissertação de Mestrado em Engenharia e Gestão Industrial apresentada à Faculdade de Ciências e Tecnologia

Country
Portugal
Related Organizations
Keywords

Pulp and paper industry, Tissue machine, Web breaks, Decision trees, Quebras de folha, Modelos de previsão, Árvores de decisão, Máquina de papel, Indústria da pasta e papel, Prediction models

  • BIP!
    Impact byBIP!
    selected citations
    These citations are derived from selected sources.
    This is an alternative to the "Influence" indicator, which also reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
    0
    popularity
    This indicator reflects the "current" impact/attention (the "hype") of an article in the research community at large, based on the underlying citation network.
    Average
    influence
    This indicator reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
    Average
    impulse
    This indicator reflects the initial momentum of an article directly after its publication, based on the underlying citation network.
    Average
Powered by OpenAIRE graph
Found an issue? Give us feedback
selected citations
These citations are derived from selected sources.
This is an alternative to the "Influence" indicator, which also reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
BIP!Citations provided by BIP!
popularity
This indicator reflects the "current" impact/attention (the "hype") of an article in the research community at large, based on the underlying citation network.
BIP!Popularity provided by BIP!
influence
This indicator reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
BIP!Influence provided by BIP!
impulse
This indicator reflects the initial momentum of an article directly after its publication, based on the underlying citation network.
BIP!Impulse provided by BIP!
0
Average
Average
Average
Green