SySMo stands for Synthetic Sweden Mobility Model (1, 2), which creates a synthetic population, a simplified microscopic representation of an actual population in Sweden. Statistically representative of the population, it offers crucial data for simulation models, notably agent-based models, in fields like transportation, land use, economics, and epidemiology.
-
Tozluoğlu, Çağlar, Swapnil Dhamal, Sonia Yeh, Frances Sprei, Yuan Liao, Madhav Marathe, Christopher L. Barrett, and Devdatt Dubhashi. "A synthetic population of Sweden: datasets of agents, households, and activity-travel patterns." Data in Brief 48 (2023): 109209. Link
The project delivers a synthetic replica of Swedish car users (i.e., car agents), their household characteristics, activity-travel plans, and detailed traveling along the road network on an average weekday. This repository documents the methodology of generating these data which can be further used in mobility patterns analysis, agent-based simulations, and studying the impact of electrification of private car fleets in Sweden.
| Step | Script/Procedure | Objective |
|---|---|---|
| 1 | src\1-data-preparation.ipynb |
Prepare data for simulating car agents' movement in MATSim. This step uses functions in lib\workers.py. |
| MATSim simulation* | Each region's car agents are simulated using all road networks within the region and major roads for the rest of Sweden. | |
| 2 | src\2-activity-plans-merging.py |
Merge original SySMo's plans and MATSim simulations' experienced plans, organized for further processing. |
| 3 | src\3-network-events-processing.ipynb |
Combine all regions' network and save as one GIS file, network_o.shp and extract events from the simulations' results, using functions in lib\workers.py |
| 4 | src\4-data-description.ipynb |
Visualize individual plans and events, i.e., driving trajectories on the road network. |
| 5 | src\5-feasible-agents\ |
For each agent, we give it a label, feasible/infeasible, indicating his/her activity plan's feasibility. For example, overly busy plans maybe impossible to finish within 24 hours, considered infeasible. |
| 6 | src\6-events-filtering.ipynb |
Prepare agents' sociodemographic attributes with feasibility label and plans and events only for feasible agents. |
| 7 | src\7-data-product-description.ipynb |
A simple overview of the four main data products. |
*We input the agents' daily activity plans along with the road network into MATSim for iterative replanning, aiming for a convergence on optimal activity plans for all agents.
Subsequently, we retrieve the individual mobility trajectories of agents from the MATSim simulation.
The setup adheres to the MATSim 13.0 benchmark scenario (src\config_samm.xml), with slight adjustments.
The strategy for replanning integrates BestScore (60%), TimeAllocationMutator (30%), and ReRoute (10%)—
the percentages denote the proportion of agents utilizing these strategies.
| # | Dataset | File | Description |
|---|---|---|---|
| 1 | Synthetic car agents | syn_pop_all.parquet |
Contains all car agents and their attributes. |
| 2 | Activity plans | plans.parquet |
Contains the agents' daily activity plan for an average weekday. |
| 3 | Driving trajectories | events\*_events_batch*.parquet |
Cars' moving along the road network at the second level. |
| 4 | Road network | netowrk_o.shp |
Road network with the 'link_osm' corresponding to 'link' in Dataset 3. |