This project elaborates on the work developed for MSSI - Modelação e Simulação de Sistemas @FEUP.
Even though public transportation is getting cheaper the general tendency remains that private transportation is the majority. This implies cost is not the main driver for the decision, thus there are other variables that need to be factored in. The goal of the project is to factor the most variables possible and do a sensitivity analysis of the system. Ultimately it is possible to infer the main drivers on the decision ofthe transportation. Policy makers are able to substantiate their positions and take better decisions.
During the development of this study, it was possible to provide commuters with a cognitive model that allowed them to make the most correct decision when choosing the means of transport to use. For this, Reinforcement Learning based on the Exploration–Exploitation trade-off dilemma was used. After simulating several scenarios, it was possible to perceive, with a high degree of confidence, the factors that most influence the decision-making of commuters. Although the analysis presented here elaborates on a large number of variables, it is important to point out that all analyzes were carried out individually, that is, varying only one at a time. For a better analysis and applicability to reality, it would be necessary to incorporate in the simulation the variability of multiple variables simultaneously. This study presents the initial foundations and concept in order to facilitate the expansion of the work and scenarios under analysis.
- Date : 4th Year, 2nd Semester, 2020/2021
- Couse : Modelação e Simulação de Sistemas | Systems Modelling and Simulation (MSSI)
- Contributors : Cláudia Martins, Joaquim Rodrigues, José Silva, Vítor Gonçalves
To find out how to run the simulation, check this file.