University of Twente Student Theses

Login

Using machine learning algorithms to improve traffic state estimation : a study on the usability of machine learning techniques in traffic state and speed estimation

Dommerholt, A.J. (2019) Using machine learning algorithms to improve traffic state estimation : a study on the usability of machine learning techniques in traffic state and speed estimation.

[img] PDF
2MB
Abstract:Traffic state is defined by the traffic variables intensity, speed and density. When two of the three defining variables are known a traffic state can be determined. When only one of the variables is known, additional information is needed for traffic state estimation. INWEVA is an overview of intensities on the Dutch national roads. For the parts of the road network that are not covered by detection loops, intensities are estimated by a model. Since only an estimation of intensities is known for these road sections, traffic state cannot be determined directly. Additional procedures have to be taken to estimate speed, and thus defining traffic state. In this research the relation between intensities and speeds is studied. This research aims to give a good estimation of speed, based on intensity data only. When speeds are known, it can be determined whether or not congestion occurs, and traffic state is defined. The estimations of traffic state are made by inputting intensities into machine learning models. The main question for this research is given below and is answered by researching the appropriate machine learning technique for traffic estimation and researching if there are additional attributes that may improve estimation. The last subquestion tries to find an answer to whether or not characteristics of road section are transferable to other road sections, in order to train models on other road sections than they will be tested on. How can machine learning techniques estimate traffic state based on intensity data?
Item Type:Essay (Master)
Faculty:ET: Engineering Technology
Subject:56 civil engineering
Programme:Civil Engineering and Management MSc (60026)
Link to this item:http://purl.utwente.nl/essays/79686
Export this item as:BibTeX
EndNote
HTML Citation
Reference Manager

 

Repository Staff Only: item control page