University of Twente Student Theses


A comparison between the traditional trip-based traffic demand model and the tour-based model : Analysis of the trip characteristics of the Trefcenter Venlo shopping mall in Limburg

Carvalho Bustamante, Hugo (2022) A comparison between the traditional trip-based traffic demand model and the tour-based model : Analysis of the trip characteristics of the Trefcenter Venlo shopping mall in Limburg.

[img] PDF
Abstract:Travel demand models have been used over the past decades to solve traffic problems such as congestion and high travel times by focusing on the increase in the capacity of the transport infrastructure. In this context, trip-based models were developed and used incessantly worldwide by traffic modellers and academics. However, the shift to demand management strategies to enhance the efficiency of the transport system has boosted the interest and understanding of advanced modelling, such as the tour-based model (TBM). TBM tries to overcome some limitations of the traditional trip-based model by temporally and spatially interlinking a sequence of trips in a tour. However, even after years of academic progress and knowledge share over advanced models, the trip-based model still widely dominates practice. The complexity and high costs of new models, as well as the scepticism of practitioners are among the reasons of this gap between theory and practice. Therefore, this research tries to fill in this gap by means of a comparison between the TBM and the trip-based model. Royal HaskoningDHV (RHDHV) has developed a TBM for the province of Limburg in the Netherlands. This model is used in this research aiming to evaluate and demonstrate the potential advantages of the tour-based approach in practice. For this purpose, a trip-based model was developed for the comparison to the TBM of Limburg. More specifically, the trips of visitors of the Trefcenter Venlo shopping mall are investigated. In addition, a travel survey was designed and conducted with the visitors of the mall. This survey is, together with traffic counts, used as the benchmark of the comparison. In other words, the best model is the one that replicates the Shop trip behaviour of the mall closer to the observations on the survey and the traffic counts. To perform a fair comparison, it is important to understand and specify the differences between the developed models. The aim is to make sure that the differences in the results of the models are mainly due to the intrinsic characteristics that distinguish them. That is, the link of trips in tours made by the TBM. Furthermore, four criteria were defined to compare the models. They are the modal split, the trip length, which consists of the travel time and distance frequency distributions as well as the average trip length, the home location of visitors, and trip assignment. The results indicate that the TBM better replicates the modal shares of both the pre-trips (i.e., Home-Shop trips) and post-trips (i.e., Shop-Home trips) of the visitors. However, this is a consequence of limitations in the development of the models. First, the TBM modal split considerably differs from the survey. The TBM underestimates car trips and overestimates the other modes. This is likely because of the Shop distribution functions which are averaged for various shopping facilities, not distinguishing whether it is in a major city centre or near a highway, where the behaviour of the visitors is different. Improving these functions is expected to enhance the TBM’s modal split accuracy. Second, the trip-based model results are hindered by the car distribution functions used in the model, which misrepresent people under 18 years old by enabling them to drive. However, improving it would likely reduce even more the car share of Shop trips in the model, worsening it in the comparison with the survey. Finally, although correcting the aforementioned drawbacks is expected to benefit the TBM, this could not be proved in this research. Thus, the modal split advantage is not exclusively due to the intrinsic characteristics of the TBM. Moreover, although the TBM estimates the trip length of the visitors of the mall better than the trip-based model, it is not possible to conclude whether this is due to the intrinsic distinction between the two approaches. This is because the trip-based model overestimates car trips by misrepresenting people under 18 years old. This increases the generalised costs of the model when taking the effects of the demand into account. In addition, the TBM cross-boundary trips are hindered more significantly by increasing the travel costs twice for trips to another country. Improving these issues would change the trip length of the models, but this could not be analysed in this research. Furthermore, the TBM simulates the home location of the visitors of the mall worse. The higher costs for traveling to a different country kept most of the TBM’s trips in the Netherlands. This restriction was not observed on the survey, as many visitors come from Germany. However, the disadvantage of the TBM is merely a drawback of this research and different conclusions could be drawn if the cross-border costs were reduced. Finally, the TBM performs better than the trip-based model on the trip assignment criterion. Yet, the results are not good since most of the T-values are not acceptable. Likewise the other criteria, it was not possible to deduce if the better replication of trip assignment is solely due to the intrinsic differences between the models. To conclude, the comparison of Shop trips performed in this research does not concretely demonstrate the benefits of the TBM over the trip-based model. However, improving the limitations of this research is expected to favour the TBM in the modal split comparison. If this expectation is correct, the TBM is more appropriate to dealing with emerging policies, such as management-oriented policies which aims to incentivise sustainable modes of transport. Nevertheless, this research provides relevant insights to traffic modelling, suggesting improvements to the models. In addition, it pinpoints aspects that deserve further interest, such as the implementation of deterrence functions comprising the different characteristics of the personas, which would increase the behavioural realism of the TBM, and the investigation of NHB trips, which are expected to be better modelled by the TBM.
Item Type:Essay (Master)
Faculty:ET: Engineering Technology
Subject:56 civil engineering
Programme:Civil Engineering and Management MSc (60026)
Link to this item:
Export this item as:BibTeX
HTML Citation
Reference Manager


Repository Staff Only: item control page