Analyzing modelling problems of the frequency-bases methodology in Zenith

Akse, R. (2016) Analyzing modelling problems of the frequency-bases methodology in Zenith.

Abstract:Modelling Public Transport (PT) is done for many reasons, for example exploring possible new routes for train lines or analyzing commuter flows between two cities. The traditional 4-step model is a way to describe the behavior of a PT system. The last step of the 4-step model describes route-choice for PT travelers. The route choice algorithm of OmniTRANS, a computer software package to model travel behavior between origins and destinations, is called Zenith. It is based on the frequency-based methodology. Frequency is one of the main inputs for this methodoloy. The choice fractions to board a certain line of a certain path are based on the costs of that path and the frequency of that line. The total average travel time between an origin and destination is based on the average waiting time and the travel time. It appeared that for some path sets the removal of a line led to a lower total average travel time of all paths. This specific model output is counterintuitive and not realistic. In general, there are other disadvantages of the frequency-based methodology such as the average output relative to lines. The research goals are twofold: firstly there must be determined under which circumstances the specific modelling problem of the total average costs occurs and how this problem has an effect on the modelling output. Secondly, a recommendation should be written based on all advantages and disadvantages of the frequency-based methodology including the outcomes of the analysis of the specific modelling problem. The theoretical background includes a detailed analysis of the frequency-based algorithm in Zenith and an explanation of the cause of the modelling problem. The literature research discusses other assigning methodologies, such as the scheduled-based methodoloy. This methodology takes into account explicitly all travel times and waiting times. The methodology to analyze the specific modelling problem consists of two parts. The first part is a sensitivity analysis of how different combinations of input parameters (frequency, travel time) lead to the occurrence of the modelling problem. The second part consists of an analysis of the modelling problem in a large network in OmniTRANS, since it is not known to what extent the modelling problem occurs in practice. This enables to link the theoretical framework of the sensitivity analysis with the practical modelling world. Based on the literature search and the findings of the above-stated first parts, a recommendation is written concerning the frequency-based methodology. The sensitiviy analysis makes clear that the modelling problem only occurs for low travel-times. Moreover, only small differences in travel time lead to the modelling problem. In the large case network it appears that the modelling problem occurs quite often (6 % of the stop centroid combinations contains the modelling problem). In case of occurrence of the modelling problem, it appears that the problem severity is low since the difference in travel time (situation with or without a certain line) is small . In general, the theoretical framework matches the outcomes of the large case network. The recommendation concludes that fixing the specific modelling problem of the total average costs is not that difficult, since the severity is low. But since the frequency-based methodology shows fundamental modelling problems concerning average modelling output, it is recommended to exploit the possibilites of a new dynamic assignment algorithm in OmniTRANS such as the scheduled-based methodology. The transit modelling world increasingly needs dynamic models to tackle capacity problems. Secondly the development of dynamic car traffic models is way further in OmniTRANS, so not upgrading the frequency-based methodology leads to disintegral traffic models.
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