University of Twente Student Theses


Effects of junction delay functions on traffic modelling using large-scale strategic traffic models

Arem, M. van (2019) Effects of junction delay functions on traffic modelling using large-scale strategic traffic models.

[img] PDF
Abstract:This report will investigate the effects of different methods of junction modelling on the convergence of large-scale strategic traffic models. The main hypotheses would be a more simplistic approach, which would have less variables, could be able to increase the rate of convergence of strategic traffic models. A promising option could be Junction delay functions, as they only have the rough layout of the junction and traffic volumes as variables. To see what the effects of junction delay functions would be on the convergence. The main question is formulated to be: To what extent could junction delay functions increase the convergence of large-scale strategic traffic models, while reducing the amount of input and retaining a high accuracy in comparison to the old method? While answering this question, the first part will investigate different methods that could be used to describe the delays or time penalties that vehicles get when using the junction. As a result, three methods will be compared to one another (two focussed on junction delay functions and one that is compared to as the original method). These methods all have a different approach to junction modelling. The first method would be the method of Vasvári (Vasvári, 2015). This method works with a single equation and a set of parameters to describe each of the junctions. The delay curve of this equation set looks flat with low increase in delay until the capacity of the junction is reached. At this point the function gets an exponential growth. This gives a large delay when the junctions capacity is reached. The second method has a different approach. This method has different equations for signalized and unsignalized junctions (Aashtiani & Iravani, 1999). This method has a maximum delay value of 18 seconds for an unsignalized junction, which is a low value. This implies that the unsignalized junction always has a low delay time for these types of junctions. While for the signalized junctions, this method gives the roughly equal delays to the other two methods. This method gives all types of junctions a slowly increasing delay, where the delays would also increase if the traffic loads relatively low. The third method is already implemented in OmniTRANS (DAT.Mobility, 2016), which would thus be the one that is compared to. This method is different in the fact that the other two have volume factor as variable, whereas this method also has a capacity reduction due to the other traffic directions. This means that there are more complex calculations involved. After these three methods were described, they were tested to see which of the methods would have the best convergence. This was done in the modelling program OmniTRANS, where the tutorial network of Delft was used, as for this purpose a not too detailed network would satisfy. When the testing of the model was done, the results of 50 iterations of the network for each method were compared to each other. This showed that, unfortunately, the currently implemented version of junction modelling was the method that had the highest rate of convergence. Not all was lost, as the second method was not too far behind regarding the convergence and the eventual difference in load between iterations. This means that the expectations at the start very well give adjustments to these new methods, by calibration and validation, to give them the higher convergence rate that was expected in this report.
Item Type:Essay (Bachelor)
Faculty:ET: Engineering Technology
Subject:56 civil engineering
Programme:Civil Engineering BSc (56952)
Link to this item:
Export this item as:BibTeX
HTML Citation
Reference Manager


Repository Staff Only: item control page