Analyzing the Effect of Incremental Traffic Assignment Iteration Number on the Accuracy of Results: The Case of Mashad City

One important issue in transportation planning is the estimation of network link andpath flows, as outputs of traffic assignment process. Different methods of traffic assignment are based on different assumptions, mainly the behavioral assumption.Incremental assignment is one of the heuristic methods to estimate the equilibriumflow pattern, which although unable to produce the exact results similar to observations,is quite efficient due to its simplicity and amount of computation. An important criterion in this method, specifying the amount of calculation is the number of iterations. In this paper, the effect of the number of iterations on the accuracy of results is analyzed and computed for different iterations (2 to 12) for Mashad City, using VISUM software package. Results based on the comparison of model results with observations, indicate that increasing the number of iterations up to a certain threshold causes model outputs to replicate observations more accurately.However, beyond that threshold, coefficient of determination shows an asymptoticbehavior, so that further increase in iteration number does not affect the resultsnoticeably. The threshold for the case of this research was shown to be eight iterations.