University of Twente Student Theses


Driver Pattern Clustering and Similarity Analysis Using Driving Simulation Data

Lakomski, Victoria (2023) Driver Pattern Clustering and Similarity Analysis Using Driving Simulation Data.

[img] PDF
Abstract:This study aims to explore driving behaviour patterns, whether they represent previously identified driver types and the similarities between them which could indicate a common driver type across multiple parameters. The driving behaviour patterns emerge from using the K-Means clustering algorithm on a dataset containing the parameters speed, acceleration, brake law, steering angle and heart rate, whereby two parameters are clustered together one at a time. The dataset used in this study was previously created by researchers using a driving simulator and contains 59 participants. To assess the similarity between the patterns, the clusters of parameter pairs are compared to each other using the Rand Index. The resulting similarity measure indicates whether the driving behaviour patterns of multiple parameters combined have commonalties that could imply an underlying driver type. Reason for this study is the fact that the most common reason for road accidents is human error. It is important to understand driver behaviour to increase driving performance. While driving profiles have been studied before, their underlying behavioural components have not been analysed regarding their commonalities towards each other. This could further provide insights for driving assistance systems. The results of this study show that the previously identified driver types are represented in this study. Furthermore, there are high similarity scores for a multitude of driving behaviour patterns which can indicate one common driver type.
Item Type:Essay (Bachelor)
Faculty:BMS: Behavioural, Management and Social Sciences
Subject:77 psychology
Programme:Psychology BSc (56604)
Link to this item:
Export this item as:BibTeX
HTML Citation
Reference Manager


Repository Staff Only: item control page