CONTACT US: Contact info
|
Citation |
Qi W, Zhu S, Hu J. Sustainability (Basel) 2022; 14(17): e10944. |
|
Copyright |
(Copyright © 2022, MDPI: Multidisciplinary Digital Publishing Institute) |
|
DOI |
|
PMID |
unavailable |
|
Abstract |
Due to inertia, heavy trucks are often involved in serious losses in accidents. To prevent such accidents, since 2020, the transportation department has promoted the free installation of intelligent video surveillance systems on key vehicles of "two passengers, one danger, and one cargo". The system can provide real-time warnings to drivers for various risky driving behaviors. The data collected by the system are often managed by third-party platforms, and such platforms do not have authority beyond the information that the authority system can collect. Therefore, it is necessary to use the trajectory data and warning behavior records that the system can collect for behavior analysis and accident prevention. To analyze the correlation between different warning factors, 88,841 warning records and 1033 trip records of heavy trucks for construction in the second half of 2021 were collected from a third-party supervision platform. The research associated the warning records with the vehicle operation records according to the warning time and the license plate and established a multiple linear regression equation associated with operational attributes and warning factors. The factor selection results showed that only two warning factors, "too close distance" and "lane change across solid line", can be used as dependent variables to construct a regression model. The results showed that many distracted behaviors had a significant impact on aggressive driving behavior. Companies need to focus on behaviors that are prone to other warning behaviors. This paper provides a theoretical basis for the optimization of the warning function of the electrified supervision system and the continuing education of drivers by exploring the internal correlation between different warning factors. Language: en |
|
Keywords |
electrified supervision system; heavy truck; lasso regression; multiple linear regression; optimal subset |