• JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
 
  Bookmark and Share
 
 
Master's Dissertation
DOI
https://doi.org/10.11606/D.3.2022.tde-20072023-134526
Document
Author
Full name
Renata Akemi Marçal Imai
E-mail
Institute/School/College
Knowledge Area
Date of Defense
Published
São Paulo, 2022
Supervisor
Committee
Cunha, Claudio Barbieri da (President)
Ribeiro, Glaydston Mattos
Rodrigues, Vinícius Picanço
Title in Portuguese
Análise do impacto do agrupamento da demanda na qualidade da solução de problemas de localização de instalações.
Keywords in Portuguese
Agrupamento da demanda
Modelagem
Problema de localização
Abstract in Portuguese
Este trabalho trata da análise do impacto do agrupamento de pontos de demanda na qualidade da solução de problemas de localização. Foram desenvolvidos dois conjuntos de dados com localizações reais de farmácias no Brasil, a partir dos quais 18 instâncias do problema de localização de instalações com custo fixo foram criadas. São propostos dois métodos de agrupamentos de pontos de demanda: um que considera os limites geográficos do município e outro que emprega o algoritmo K-Means. Constatou-se que as soluções dos problemas com demanda agregada por município apresentam diferenças de 0,43%, enquanto os problemas com demanda agregada por K-Means apresentam diferenças de no máximo 0,03% com relação ao problema desagregado. Para identificar os municípios nos quais a alocação da demanda pudesse ser diferente entre os modelos agregados e o desagregado, utilizou-se o algoritmo DBSCAN Density Based Spatial Clustering of Applications with Noise. Para analisar se o DBSCAN identifica adequadamente essas regiões, foram selecionadas quatro instâncias, sendo duas de cada conjunto. Em 3 das 4 instâncias analisadas, observou-se que o DBSCAN identificou as regiões do espaço nas quais a alocação de demanda é diferente entre modelos agregados e desagregados. Nestas mesmas instâncias, também se observou que as diferenças nos custos de transporte variam entre -84,99% e 662,90% no agrupamento por município e -23,25% e 110,31% no agrupamento por K-Means.
Title in English
Analysis of the impact of demand aggregation on the solution quality of facility location problems.
Keywords in English
Demand aggregation
Facility location
Modelling
Abstract in English
This thesis analyses the impact in the quality of solution for location problems with aggregated demand points. Two datasets containing the location of drugstores in Brazil were developed, and 18 instances of the fixed cost facility location problem were created. Two aggregation methods are proposed: aggregating demand points regarding the municipality boundary or using the K-Means algorithm. The models solutions with aggregated demand points by municipality showed a 0,43% difference in the objective function value at most, while the difference using K-Means algorithm was 0,03% at most. To identify the municipalities where differences in the allocation of unaggregated and aggregated models are different, DBSCAN (Density Based Spatial Clustering of Applications with Noise) algorithm was used. To identify whether DBSCAN properly identified these regions, four instances were selected, two from each dataset were selected. For 3 out of 4 instances, DBSCAN identified properly the location of demand points allocated to different facilities. For the same instances, the differences on transportation costs between aggregated and unaggregated problems range between -84,99% to 662,90% for demand points aggregated by municipality, and between -23,25% and 110,31% using K-Means clustering.
 
WARNING - Viewing this document is conditioned on your acceptance of the following terms of use:
This document is only for private use for research and teaching activities. Reproduction for commercial use is forbidden. This rights cover the whole data about this document as well as its contents. Any uses or copies of this document in whole or in part must include the author's name.
Publishing Date
2023-07-25
 
WARNING: Learn what derived works are clicking here.
All rights of the thesis/dissertation are from the authors
CeTI-SC/STI
Digital Library of Theses and Dissertations of USP. Copyright © 2001-2024. All rights reserved.