Not logged in.
Quick Search - Contribution
Contribution Details
Type | Bachelor's Thesis |
Scope | Discipline-based scholarship |
Title | Applying Genetic Algorithms to Perform a Global Sensitivity Analysis on a Stock and Flow Model |
Organization Unit | |
Authors |
|
Supervisors |
|
Language |
|
Institution | University of Zurich |
Faculty | Faculty of Business, Economics and Informatics |
Date | 2022 |
Abstract Text | Genetic Algorithms employ the underlying mechanisms that give rise to evolution in order to approach computational challenges. In this context, a Genetic Algorithm is implemented and used to perform a global Sensitivity Analysis on a simulated Swiss city as part of the ongoing Post Fossil Cities project. Spanning three scenarios, the Genetic Algorithm is able to identify the ensemble of Population, Energy Intensity, and the collective of Typesplit and Splitshare resources as the main factors that can reduce the Carbon Dioxide (CO2) emissions and energy consumption within this city. Further investigation into the Genetic Algorithm’s behaviour using a multiple linear regression shows that both the population size and the number of generations have a significant impact on the fitness that can be reached. Additionally, as a side-product of the genetic mechanisms, an instability of the diversity emerges at small population sizes that closely resembles genetic drift. When comparing the performance of the Genetic Algorithm to a Monte-Carlo approach, there is a highly significant advantage in favour of the Genetic Algorithm. |
PDF File | Download |
Export | BibTeX |