The Study Of Road Lighting Efficiency Based On Simulated Driving Scene

The road lighting is to provide a good visual environment for people driving vehicles on the road at night,so that they can find pedestrians or objects appearing on the road in time.Good road lighting design can effectively reduce the night accident rate and the degree of injuries and deaths in accident,providing an important guarantee for the safety of life and property.The luminance under road lighting condition generally belongs to the mesopic vision range.The widely used models to describe mesopic vision under road lighting condition are visibility level(VL)model and relative visual performance(RVP)model,but they have some limitations.First,VL model and RVP model were obtained by the static experiment which did not include the effect of the real driving load.Second,they cannot describe the interaction between various influencing factors.Based on the driving simulating platform,relevant experiments were designed to study the effect of background luminance,target size,target contrast,target initial appearance distance and target position on target visibility with real driving load.Results showed that,all factors have significant effects on the target visibility.Overall,the greater the background luminance,the larger the target size,the greater the target contrast,the closer position to the center lead to the greater the target visibility.Target initial appearance distance has no significant effect on the detection rate,as target initial appearance distance increases,the reaction time becomes longer.The VL model was used to verify the influence of background luminance,target size and other factors on the visual performance under real driving load condition.The results proved that the VL model is not suitable for directly calculating the target visibility in the situation which with actual driving operation.In terms of the limitations of the VL model,a Simulator based Visual Performance(SVP)model was proposed which included the influencing factors such as background luminance,target size and interaction of various factors.The results showed that the model can reflect the experimental data well.The R~2 of the SVP model for positive and negative contrast can reach 0.97and above.With reference to the behavioral validity verification method of the driving simulator,experiments with the same detection tasks are designed in the actual road scene and the laboratory driving simulation platform scene.The detection distances for the specified targets in the two experimental scenarios were compared to evaluate the validity of driving simulation platform.The results showed that detection distance of driving simulation platform was consistent with the actual road scene while ensuring that the environment information and target contrast in the driving simulation platform scene are consistent with the actual road scene.The virtual driving platform has certain behavior validity.