The Mid-Term Forecast (MTF) module is an online simulation engine which is based on a dynamic traffic assignment. It provides a forecast for:
- The next days (rest of the week)
- The next hours until midnight (rest of the day).
The forecast can manage any planned events, like roadworks or other events.
The forecast allows the traffic operator to evaluate the outcomes of these events and the associated traffic changes.
In the Optima architecture, MTF is mainly a scheduler which invokes Short Term Forecast (STF) to perform a simulation at a specific timestamp.
MTF and STF
The MTF module extends the capabilities of the Short-Term Forecast (STF), extending the forecast horizon, and therefore the scope of the forecast. The driver behavior aims to contrast the impact of an event. The route change behavior (compliance to new optimum or stick to usual habits) has to be calibrated accordingly.
MTF (optionally) allows to the daily adaptation of the drivers to the current traffic situation.
From a methodological point of view, MTF works on a dynamic traffic management that has not yet reached an equilibrium (users can make non-optimal choices).
STF does not include route changes, while MTF does. Also, route changes computed by MTF do not affect the route choices used by STF. MTF is a scheduled-based forecast, exactly like STF, but differently from this last one it cannot be triggered.
On large models, forecast computation can take hours, since it is methodologically comparable to a dynamic user equilibrium: a typical configuration of the forecast for the next hours is calculated hourly, while the forecast for the next days is calculated daily.
Only active events are managed by MTF in the considered forecast period.
MTF Jobs, tasks, and operations
MTF configuration lets you schedule items called jobs.
Each job has a specific time entry that defines the associated run time.
Every job is a container of sub-items, called tasks.
Each task defines an operation. When a job is run, its tasks are performed sequentially in their given order. Thus, a job is a container of a list of operations to be performed.
Special jobs
There are a few special jobs that create their tasks automatically. You do not need to create tasks manually in the configuration. These tasks do not appear in the configuration but are created runtime. The values of the attributes of these tasks are set according to the job configuration.
An example of these special jobs are:
next days forecast
day types forecast
Topics in this section