PTV Visum SBA Video Series – Dynamic Traffic Assignment in a tiered system

There are three types of private transport assignment available for PTV Visum users:

  1. Static assignment
  2. Analytical dynamic assignment
  3. Pseudo-dynamic assignment with ICA

They provide good results for long-term and strategic planning but there are situations where users need a more detailed representation of the dynamics of congestion and traffic flow.

That is why PTV has developed a fourth assignment method – Simulation Based Dynamic Assignment (SBA). In SBA, much like in microsimulation, individual vehicles and their interactions are simulated which means that congestion effects and delays over time can be more realistically represented. This makes SBA is the ideal stepping stone between strategic modelling in PTV Visum and mesoscopic simulation in PTV Vissim, in turn significantly improving the workflow between transport planners and traffic engineers.

To find out more, watch Michael Oliver’s six-episode video series explaining and demonstrating the fundamentals of SBA.

Created and narrated by Michael Oliver, PTV Visum Solution Director

An overview of SBA and a description of each episode has been provided below. For a more detailed SBA description please see the PTV Visum 16 Highlights.


PTV Visum SBA Video Series – Episode Overview: 

01 – Introduction – SBA fundamentals 

Episode 1. begins by addressing the theory behind analytical DTA and simulation based DTA, followed by an overview of when to use which method and for what project typologies.

02 – Network Coding 

Episode 2. discusses the typical workflow for coding a network for SBA using a typical static demand model. To demonstrate this, a real PTV Visum model of South Oregon is used (*the data has been disturbed for confidentiality purposes). The video concludes by explaining how this workflow can be further accelerated by working with Synchro Imports, OSM, PTV Vistro.

The workflow demonstrated includes:

  1. Cutting out territories
  2. Adding network detail
  3. Using presets for visualising the network
  4. Viewing data in lists
  5. Calibrating the reaction time and effective vehicle length: for a more realistic representation of the real world location being modelled.
  6. Coding intersection and nodes

03 – Subnetwork generation and demand profiling 

With the extra detail coded in the model, episode 3. demonstrates how to cut out a sub-area network to make the final preparations for SBA. Next, a demand profile is created to distribute the static demand over the DTA zones and time period.

04 – Runtime troubleshooting 

Episode 4. covers how to:

  1. Get a first successful run of SBA
  2. Troubleshoot potential runtime errors and warnings messages: using the built-in features.
  3. Setup SBA parameters

05 – Calibration 

Episode 5. looks at the outputs produced in episode 4. and discusses how to improve the fit of the model to reality. The first step being, identifying hot spots in the network (areas where the assignment is not performing realistically) by checking: volumes, speeds, SBA density and flow changes throughout the assignment period.

The next step is to calibrate the model to better fit the empirical data (derived from traffic counts, travel time data and GPS). Calibration can be achieved by:

  1. Improving network representation: e.g. coordinating signal timings
  2. Adjusting simulation parameters: reaction time factor, effective vehicle time factor, gap values
  3. Adjusting matrices and demand profiles

06 – Creating and comparing scenarios 

The final episode assesses the calibrated model and selects an underperforming node. Next, an attempt is then made to develop a scheme to improve its performance by developing and testing a new scenario with re-optimised signals timings and the addition of a traffic lane. The results of the scenario are then compared with that of the base model.