Evaluating day type simulations
At the end of each simulation performed by TRE, the TRE Viewer opens. Here you can check the simulation results.
What to check
The results of each daily simulation must be stable and realistic:
- There must be no gridlock situations, which means that all queues must extinguish before the end of the simulation.
- Traffic flow, speeds, travel times, and queues predicted by the model for each day type must be coherent with the average detected data available for the same day type.
- All demand has to reach destinations within the typical time.
- Results of dynamic simulation for each day type must be realistic in terms of flows and queue generation, propagation and dispersion. Compare simulation results with available measurement data to validate the quality of the calibration.
- The maximum relative deviations in the observables have to be less than 0.001 (where applicable).
What to do if the results are not satisfactory
If TRE results are not satisfactory for one or more day types, you need to repeat the simulations changing either:
- The demand model
- The supply model
- TRE parameters (→ Traffic Real-time Equilibrium (TRE))