Development of an Open-source Multi-objective Optimization Toolbox
The industrial trend is currently to increase product customization, and at the same time decrease cost, manufacturing errors, and time to delivery. These are the main goal of the e-FACTORY project which is initiated at Linköping university to develop a digital framework that integrates digitization...
| Main Author: | |
|---|---|
| Format: | Others |
| Language: | English |
| Published: |
Linköpings universitet, Maskinkonstruktion
2019
|
| Subjects: | |
| Online Access: | http://urn.kb.se/resolve?urn=urn:nbn:se:liu:diva-159985 |
| Summary: | The industrial trend is currently to increase product customization, and at the same time decrease cost, manufacturing errors, and time to delivery. These are the main goal of the e-FACTORY project which is initiated at Linköping university to develop a digital framework that integrates digitization technologies to stay ahead of the competitors. e-FACTORY will help companies to obtain a more efficient and integrated product configuration and production planning process. This thesis is a part of the e-FACTORY project with Weland AB which main mission is the optimization of spiral staircase towards multiple disciplines such a scost and comfortability. Today this is done manually and the iteration times are usually long. Automating this process could save a lot of time and money. The thesis has two main goals, the first part is related to develop a generic multi-objective optimization toolbox which contains NSGA-II and it is able to solve different kinds of optimization problems and should be easy to use as much as possible. The MOO-toolbox is evaluated with different kinds of optimization problems and the results were compared with other toolboxes. The results seem confident and reliable for a generic toolbox. The second goal is to implement the optimization problem of the spiral staircase in the MOO-toolbox. The optimization results achieved in this thesis shows the benefits of optimization for this case and it can be extended by more variables to obtain impressive results. |
|---|