Laserfiche WebLink
VISUM automatically adjusts the distribution to match the I-X and X-I origins and <br />destinations at the external zones. It also automatically distributes the proper number of <br />destinations to each zone based upon the values derived during the trip generation <br />process. However, the distribution of those trips within the model can be modified by <br />changing the apparent distance traveled either approaching or departing the model <br />area. These were adjusted to get an expected distribution from the externals. <br />From these results, analyses are performed and potential changes or upgrades to the <br />entered data are made for the following simulation run. Overall high or low trends can <br />suggest that information needs to be upgraded concerning dwelling units, employment, <br />trip generation rates, and/or gravity model-spatial behavior coefficients. Throughout <br />the calibration of overall trends, each segment of entered data is questioned and <br />necessary changes are made. In addition, ground count data is also scrutinized. <br />Final Calibration Values <br />Changes were made to the parameters in an iterative fashion based upon judgment. The <br />final values used in the calibration are the following: <br />• Car impedance is weighted 0.985 on time and 0.15 on distance. <br />• Truck impedance is weighted 0.980 on time and .020 on distance <br />• Iterative distribution and assignment with equilibrium assignment was <br />used. The skim matrix (a measure of travel impedance) was averaged for <br />each iteration for up to 3 iterations after the initial step. Specification of <br />this option meant that the gravity model distributions are based upon <br />recalculated travel impedances in subsequent assignments. <br />• Table 8 illustrates the gravity model exponents set at the following: <br />Kittitas County, Washington Transportation Model-May 2009 Page38