Loading...
Publication
Sistema de apoio à decisão fabril
| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| 6.91 MB | Adobe PDF |
Authors
Abstract(s)
Esta dissertação desenvolve um modelo de otimização orientado para eficiência energética em
contexto industrial (têxtil) com o objetivo de resolver um problema de FJSSP (do anglo-saxónico
Distributed Flexible Job shop Scheduling Problem). Pretende-se analisar o impacto da integração de
recursos energéticos no planeamento da produção e, deste modo, avaliar eventuais reduções de
custos energéticos em diferentes condições.
O modelo é criado para responder aos desafios específicos do FJSSP, pautado por crescentes
pressões energéticas, económicas e ambientais num mercado cada vez mais global. A procura por
novas perspetivas e vantagens é essencial para reduzir custos, mantendo a competitividade e tendo
em conta também a sustentabilidade. Formula-se um modelo que considera processos produtivos
e aspetos energéticos, permitindo considerar presença ou ausência de recursos energéticos como
PV (do anglo-saxónico Photovoltaic) e BESS (do anglo-saxónico Battery Energy Storage System),
horários de operação e tarifas de fornecimento. São assim os casos de estudo que permitem avaliar
vários contextos.
Os casos de estudo iniciam-se com a otimização com foco em custos que, em comparação com a
otimização do makespan resultou numa poupança de 2,49% de custos. A integração da produção
PV no sistema gerou reduções imediatas nos custos (até 38,85% face ao cenário base), confirmando
assim a sua pertinência atual. A utilização isolada da bateria revelou ganhos mais modestos (4,03%),
mas verifica-se flexibilidade no sistema. A combinação entre PV e BESS potenciou poupanças
significativas de 41,97% (comparada com 24,19% quando otimizada para makespan).
Relativamente aos horários laborais e à sua influência direta nos custos, verificou-se que os turnos
mais longos reduzem os custos em 9,15% (com algumas limitações de aplicabilidade, uma vez que
possuem uma maior dispersão das tarefas) enquanto os turnos mais restritivos implicam acréscimo
de 1,93%. Na análise dos tarifários, um em três revela poupanças positivas (2,00%) enquanto os
restantes com perdas (0,07% e 2,06%). Por fim, foi também testado um período laboral de dois dias
que revela poupanças adicionais entre 1,69% e 6,02% comparativa à soma de dois dias isolados.
De um modo geral, comprova-se que o modelo desenvolvido é capaz de resolver problemas de
FJSSP e reduzir custos energéticos para que deste modo possa servir de apoio à decisão fabril. Os
resultados obtidos variam entre poupanças marginais e significativas, demonstrando a aplicação
científica e prática do modelo, apesar deste não se limitar ao sector têxtil.
This dissertation develops an optimization model focused on energy efficiency in an industrial context (textile) to solve an FJSSP (Distributed Flexible Job shop Scheduling Problem). The aim is to analyze the impact of the integration of energy resources in production planning and, in this way, evaluate possible reductions in energy costs under different conditions. The model is created to respond to the specific challenges of the FJSSP, driven by growing energy, economic and environmental pressures in an increasingly global market. The search for new perspectives and advantages is essential to reduce costs, maintain competitiveness and take sustainability into account. In order to understand how this optimization process translates into this context of a company in the sector and integrate sustainability. A model is formulated that considers production processes and energy aspects, allowing the presence or absence of energy resources such as PV (Photovoltaic) and BESS (Storage Unit), operating hours and supply tariffs to be considered. These case studies made it possible to evaluate various contexts. The case studies begin with the cost oriented optimization, which, compared to makespan optimization revealed a cost reduction of 2.49%. The PV integration into the system generated immediate cost reductions (up to 38.85% compared to the base scenario), thus confirming its current relevance. The isolated use of the BESS revealed more modest gains (4.03%), but there is flexibility in the system. The combination of PV and BESS led to significant savings of 41.97% (versus 35.19% when optimized for makespan). Regarding working hours and their direct influence on costs, longer shifts reduce costs by 9.15% (there are some limitations in this application, as they lead to a greater dispersion of jobs), while even more restrictive shifts imply an increase of 1.93%. When analyzing the tariffs, one in three shows positive savings (2.00%) while the rest show losses (0.07% and 2.06%). Finally, a two-day working period was also tested, revealing additional savings of between 1.69% and 6.02% compared to the sum of two isolated days. In general, it is proven that the model developed is capable of solving FJSSP problems and reducing energy costs so that it can serve as support for manufacturing decisions. The results obtained vary between marginal and significant savings, demonstrating the scientific and practical application of the model, although it is not limited to the textile sector.
This dissertation develops an optimization model focused on energy efficiency in an industrial context (textile) to solve an FJSSP (Distributed Flexible Job shop Scheduling Problem). The aim is to analyze the impact of the integration of energy resources in production planning and, in this way, evaluate possible reductions in energy costs under different conditions. The model is created to respond to the specific challenges of the FJSSP, driven by growing energy, economic and environmental pressures in an increasingly global market. The search for new perspectives and advantages is essential to reduce costs, maintain competitiveness and take sustainability into account. In order to understand how this optimization process translates into this context of a company in the sector and integrate sustainability. A model is formulated that considers production processes and energy aspects, allowing the presence or absence of energy resources such as PV (Photovoltaic) and BESS (Storage Unit), operating hours and supply tariffs to be considered. These case studies made it possible to evaluate various contexts. The case studies begin with the cost oriented optimization, which, compared to makespan optimization revealed a cost reduction of 2.49%. The PV integration into the system generated immediate cost reductions (up to 38.85% compared to the base scenario), thus confirming its current relevance. The isolated use of the BESS revealed more modest gains (4.03%), but there is flexibility in the system. The combination of PV and BESS led to significant savings of 41.97% (versus 35.19% when optimized for makespan). Regarding working hours and their direct influence on costs, longer shifts reduce costs by 9.15% (there are some limitations in this application, as they lead to a greater dispersion of jobs), while even more restrictive shifts imply an increase of 1.93%. When analyzing the tariffs, one in three shows positive savings (2.00%) while the rest show losses (0.07% and 2.06%). Finally, a two-day working period was also tested, revealing additional savings of between 1.69% and 6.02% compared to the sum of two isolated days. In general, it is proven that the model developed is capable of solving FJSSP problems and reducing energy costs so that it can serve as support for manufacturing decisions. The results obtained vary between marginal and significant savings, demonstrating the scientific and practical application of the model, although it is not limited to the textile sector.
Description
Keywords
Decision Support Energy Costs Industry Optimization Scheduling Apoio à Decisão Custos Energéticos Escalonamento Indústria Otimização