| Along with the development of highway traffic, our country has been the country which has the most complicated tunnels with the fastest developing speed in the world. The highway tunnel was regarded as the highway network's bottleneck road section, it's inclining to happen traffic accident, and the extent of the accident hazard is very huge, after-treatment is so difficult to treat with. Through the related accident materials' analysis, we knew that majority of highway tunnel traffic accident mainly related with driver. Effecting by the tunnel environment driver's information cognition would change. And in driver's sensation system, the vision occupies the dominant position obviously. Therefore, studying the driver's visual characteristic in the tunnel road section is very vital significance.This article first analysis the statistics of 262 typical domestic highway tunnel traffic accidents, obtains distributed rule of accident's time, spatial, vehicle, accident type and the tunnel type. Based on the drivers' dynamic visual characteristic and related research, confirmed that the driver's average fixation time, fixation number, fixation distribution, the average saccade speed and average saccade scope to evaluate driver's visual characteristic. Through a number of drivers whose chose at random to carry the experiment at highway long tunnel road, using Eyelink record eye move data. Analyze the change role of these indexes above at tunnel road, using BP neural network simulate the indexes'change role at approach and exit section, get the function that index invariable with distance. Consult with the experiment video, applying cluster analysis, get the law of fixation distribution at approach, inner tunnel and exit road. Finding that the fixation inner tunnel can distribute regionalism into 6 kinds, they are left rear mirror, left of main fixation area, the middle of main fixation area, right of main fixation area, the road before vehicle and the gauge board region.
|
PDF Full Text Request