Resumen
Reasonable traffic signs are important for ensuring operation safety and efficiency of expressways. The overloaded information of the traffic signs and improper distance ahead of the signs present serious problems, for they may raise the driver's visual cognition burden, reduce expressway operation efficiency and cause potential safety hazards. This paper aims at developing design criteria of traffic sign information volume and reasonable AGS positions with regard to driving behavior and traffic safety. The relationship among driving behavior, traffic sign information and AGS positions has been studied based upon driving simulation tests in this paper. The information of traffic signs is processed and quantified with the method of information entropy under the framework of information theory. According to traffic flow theory and a mathematical model of vehicle lane change, AGS positions are analyzed and reasonable positions are recommended in relation to the operation conditions on multi-lane expressways. During the simulation tests, drivers aged 20-40 with 2-10 year driving experience were invited for the tests and asked to drive along the designed expressway scenes. The driver's visual behavior was monitored with an eye tracking system and data of driving behavior and the vehicle running tracks were also collected during the simulation tests. The relationship model between traffic sign information volume, AGS position and driving behavior parameters was developed as a result of further data mining, analysis and synthesis. Eventually, a multidimensional indicator of expressway sign information volume threshold has been proposed and reasonable AGS positions on multi-lane expressways are determined.