Neural Network-Based Safety Analysis of Driving Behavior in Desert Areas

In order to explore the influence of sand accumulation road surface on traffic safety under wind and sand environment. This paper relies on the desert environment real-vehicle test to study the correlation between drivers’ EEG data and driving behavior under different road alignments and different sand thicknesses, constructs behavioral spectra, uses K-means to classify dangerous driving behaviors into four grades, and applies MATLAB software to construct the BP and RBF neural network recognition model. The study shows that 1) driver’s EEG <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\delta $ </tex-math></inline-formula> and <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\beta $ </tex-math></inline-formula> wave PSD integral values are positively correlated with the thickness of sand on the road and the complexity of the road alignment; 2) drivers tend to change the speed to reduce the driving risk when there is no sand accumulation to 0-2mm and the flat straight road to the flat curved road, and drivers tend to change the driving direction to avoid the dangerous driving environment and reduce the driving risk when there is a combination of road sand thickness of 4-6mm and curved and sloped road; 3) spectrum of risky driving behavior characteristics is mainly distributed in the (0, 0.2] interval, and the risk of serious driving increases significantly in the <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\alpha $ </tex-math></inline-formula>/<inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\beta =0.81$ </tex-math></inline-formula> and <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\alpha $ </tex-math></inline-formula>/<inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\beta =0.89$ </tex-math></inline-formula> states, which account for 2.16 times and 5.28 times more than the <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\alpha $ </tex-math></inline-formula>/<inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\beta =1.01$ </tex-math></inline-formula> state, respectively. The results of the study can provide a basis for driving safety and security in desert areas, and reduce the rate of traffic accidents in desert areas to a certain extent.