By Tessa Hayman
Within Aimsun Next, traffic is added to the model using a traffic demand, which is a set of OD matrices or traffic states initiated over time. Aimsun Next provides a number of methods for profiling OD matrices:
The most basic way to profile your demand is to use traffic profiles.
Aimsun Next can store a traffic profile, which can be used to split a “total” matrix for the modelled period into a set of matrices where the volume changes over time. The traffic profile contains the percentages of traffic released for each interval during a time period.
You can create a new traffic profile in the demand data folder. Ensure that you label it clearly so that you can find it again later.
To apply this traffic profile to your OD matrix, open your matrix and go to the Cells tab. Choose the Apply Traffic Profiles tab. Drag over the cells you wish to apply it to, select the profile and click Apply. You can assign different profiles to each OD pair. Once you have assigned a traffic profile to each OD pair, click Execute and Aimsun Next will create a set of new matrices.
This method is good for small models where you are confident about your traffic profiles – this is because the split is permanent so it would need to be repeated if you wished to change the traffic profile.
If you wish to develop your matrices externally, then Aimsun Next can use externalized matrices imported from .omx, .csv or .txt files. You could use this to import matrix files that you have profiled externally.
Aimsun Next calls the matrices from these files at the start of a simulation and the link is maintained. This means that when the files are updated, the matrices are also updated in Aimsun Next.
This is a good method if you are making a larger model where detailed traffic profiles are required and you think that you may want to edit during the calibration phase. If you have a large number of matrices, then you can use a script to set up the external links.
Static OD departure adjustment adjusts the traffic demand to fit a real data set (RDS) with counts for short time intervals, hence it uses the profile of the counts at each location to profile the overall demand. The adjustment takes a static scenario, path assignment, real data set and maximum deviation matrix as inputs. The static scenario should be calibrated to a reasonable standard before carrying out this stage. We recommend that you remove sections where the observed flow is nearing or over capacity from the RDS.
The algorithm seeks to improve the correlation between modeled and observed for these time intervals by using R². So take care if you know that a centroid has a very different profile from the areas of high flow, such as a factory that operates night shifts. In this instance, ensure that you have enough iterations to allow the adjustment to fit to this centroid or another adjustment can be carried out with the highway flows removed.