Lane changing and keeping as mediating variables to investigate the impact of driving habits on efficiency: An EWM‐GRA and CB‐SEM approach with trajectory data

• Published in: IET intelligent transport systems 2024-02, Vol.18 (2), p.230-243
• Main Authors: Wang, Tianshi, Lu, Huapu, Sun, Zhiyuan, Wang, Jianyu
• Format: Article
• Language: English
• Subjects: CB‐SEM; driving habits; efficiency; EWM‐GRA; lane‐change; NGSIM
• ISSN: 1751-956X; 1751-9578
• DOI: 10.1049/itr2.12447

This paper uses the Entropy Weight Method‐Grey Relational Analysis (EWM‐GRA) and Covariance Base Structural Equations Model (CB‐SEM) to study the relationships between driving habits and efficiency. EWM‐GRA ranks 12 indicators in terms of their relevance of lane‐changing and driving efficiency. Based on this, a CB‐SEM‐based framework to describe the relevance between driving habits and lane‐changing is established, focusing on the effects of lane‐changing and car‐following behaviour. To validate the established framework, NGSIM trajectory data is used as measurement variables to describe latent variables. Several hypotheses about the relationships between the latent variables in this framework are proposed, and they are verified using trajectory data. The results show that driving habits have a direct impact on efficiency, and this impact becomes more significant when associated with lane‐change behaviour. This paper uses the Entropy Weight Method‐Grey Relational Analysis (EWM‐GRA) and Covariance Base Structural Equations Model (CB‐SEM) to study the relations between driving habits and efficiency. To validate the established framework, NGSIM trajectory data are used as measurement variables to describe latent variables. Our results show that driving habits impose a direct impact on efficiency, and this impact becomes more marked when associated with lane‐change behaviour.