There is often a desire to use microsimulation models to evaluate road improvements or new traffic management strategies under different weather conditions. However, conventional simulation models do not provide the ability to directly specify weather conditions as an input. Instead, it is necessary to determine ( a) the impact that a specific weather condition has on traffic operations, quantified in terms of the traffic stream parameters (i.e., free speed, speed at capacity, capacity, and jam density); and ( b) appropriate values for the microsimulation model input parameters to generate a simulated traffic stream that has the desired characteristics. This paper addresses this second challenge for the Vissim microsimulation model. On the basis of existing literature, an initial list of 21 input parameters was identified. A sensitivity analysis was performed, and nine key input parameters were selected. A set of simulation runs was conducted; these runs used various combinations of values for the nine input parameters. A neural network model was calibrated and validated with these data. The model requires, as inputs, the free speed, the speed at capacity, the capacity, and the jam density of the traffic stream. The model provides as outputs the values for the nine Vissim input parameters. These values, combined with the default values for the remaining parameters, provide a traffic stream with the desired characteristics. The neural network model has been coded into a software tool, which is available online.
Identifying Parameters for Microsimulation Modeling of Traffic in Inclement Weather
Transportation Research Record: Journal of the Transportation Research Board
Transportation Research Record: Journal of the Transportation Research Board ; 2613 , 1 ; 52-60
2017-01-01
Article (Journal)
Electronic Resource
English