Citation

Chai M, Li, Sun, Guo, Huang. Transp. Res. D Trans. Environ. 2019; 66: 95-103.

Copyright

(Copyright © 2019, Elsevier Publishing)

DOI

10.1016/j.trd.2018.07.007

PMID

unavailable

Abstract

Drowsy driving is one of the main causes of road traffic accidents. It is of great significance to study the use of steering wheel status to detect the drowsiness of the driver. In the studies of the steering wheel state, there is a general problem of the parameter selection being not comprehensive and individual differences in the way of the controlling of the steering wheel not being considered. A driving simulator was used to collect eleven parameters related to the steering wheel, where four parameters having significant correlations with driver status were selected using variance analysis. A multilevel ordered logit (MOL) model, support vector machine (SVM) model and BP neural network (BP) model were built based on the selection of the parameters. Under the same conditions of classification, the recognition accuracy of the MOL model was shown to be much higher than that of the two other models. It was concluded that the MOL model using the steering wheel parameters and considering differences among individuals outperforms the others in terms of driver's state recognition.

Language: en

Keywords

Drowsiness detection; multilevel ordered logit (MOL) model; Non-intrusive; Steering wheel parameter; Transportation safety