engleski

Comparison of simulation results obtained by calibrated and non-calibrated VISSIM

The micro-simulation model VISSIM are unquestionably a very useful tool, but it is questionable whether they can be expected to give realistic modeling results that can be applied within the methodology, analysis and design of traffic infrastructure in local conditions. I'd like to introduce a new method of calibration that we have developed and which is based on the application of neural network. The comparison of the simulated and measured traffic indicators, in real traffic conditions, provides the best insight into the success of the model calibration process. The traffic indicator, used for the calibration of the model, is the travel time between the measuring points at a chosen urban intersection. The calibrated model has predicted the travel time for new sets of measured data at the same intersection and second intersection with the prediction error smaller than 5%.A higher standard of validation is reached, if it can be confirmed that the model calibrated with one type of data can also give good estimates of other types of traffic data, such as queue parameters. VISSIM can provide realistic modeling results when the calibration of the model is done.

traffic micro-simulation model; VISSIM; calibration; validation

nije evidentirano