Modeling driver distance recognition and speed perception at night for freeway speed limit selection in china

Abstract

Developing a proper speed limit for freeway is critical for roadway safety. Due to the difference in visibility between day and night, it is necessary to have different speed limits for the two time periods on freeways with changing geometric features. Aiming to reduce the number of crashes caused by speeding at night on freeways, an exploratory study was conducted on the maximum speed limit at night. In order to investigate the potential relationship between drivers' distance recognition and driving speed and between speed perception and driving speed under different geometric design features, an experiment was carried out on a 22-km-long freeway segment on Chang-song freeway in China. Based on round-trips made by 10 drivers during day and night on this segment, drivers' recognition distance (distance between a sign and the location where the sign was clearly recognized the first time) and estimated speed were recorded. The data analysis results show that driver recognition distance at night decreases by about 7% compared with recognition distance at daytime. The accuracy of driver speed perception at nighttime is only 29%, whereas it is 67% at daytime. With the collected data, several multivariate non-linear regression models were established to capture the relationship among the variables of recognition distance, estimated speed at night, driving speed, and highway alignment indexes. Then the modeling results were used to develop the speed limit model by physical equations. A case study is introduced at the end of the paper Keywords Freeway, traffic safety, nighttime, recognition distance, estimated speed, maximum speed limit. Language: en