Aiming at the problems of imperfect index system for judging driver road test examination, high subjectivity and insufficient objective scientificity for judging driver road test examination hinders the prevention and control of road traffic accidents centering on safe driving. In this paper, based on natural driving data, 2926 cases of low-speed lane keeping segments with speeds lower than 30km/h are screened to extract the four indicators of lateral maneuvering evaluation, and then the thresholds of each indicator are determined by comprehensive consideration of the above indicators in combination with the Youden index, Boxplot method, and statistical measure values. In order to overcome the limitations of indicator weight distribution caused by a single assignment method, this paper proposes the subjective and objective weight combination calculation method based on the coefficient of variation to realize the scientific and reasonable allocation of evaluation index weights. According to the determined indicator thresholds and indicator weights, using the normalization method, the full score and passing score of each indicator are obtained, and created the road test lane keeping ability evaluation technique and grading standards. Finally, the effectiveness of the road test lane keeping ability grading evaluation method proposed in this paper was verified. The research results show that the road test lane keeping ability evaluation method based on natural driving data proposed in this paper can truly and objectively reflect the reality of the driver's road test, and help prevent and control road traffic accidents with safe driving as the core.
Study on the Evaluation of Lateral Lane Keeping Ability in Road Test Based on Natural Driving Data
28.06.2024
1185763 byte
Aufsatz (Konferenz)
Elektronische Ressource
Englisch
Lane keeping behavior ability evaluation method based on automatic driving technology
Europäisches Patentamt | 2024
|Driver's lane keeping ability with eyes off road: Insights from a naturalistic study
Online Contents | 2013
|Transportation Research Record | 2017
|