Regulation of Drivers' Distance Cognition Difference between Daytime and Night-Time

In order to study the impact of drivers' distance cognition difference on traffic safety in dynamic environment of daytime and night-time, a real road tests was carried out by asking 19 drivers randomly selected to percept the distances of obstacles with different distances and velocities on daytime and nighttime. The values of cognition are obtained by statistical methods. The distance cognition difference is predicted in a simulation using BP neural network and the variation laws and impact on traffic safety are analyzed. Results demonstrate that BP neural network model can efficiently simulate the cognition difference which augments with distance increasing. With the same distance, the absolute distance cognitive difference first increases and then reduces as the speed increases; while the relative distance cognition difference first increases and then decreases only when the distance is small; and it first decreases followed by an increase , and then decreases again in case of a medium and long-distance.