Driving Simulator Validation Based On Characteristics Of Visual Cognition

With the acceleration of urbanization, more and more urban transportation problems appear. Good traffic guidance system can effectively guide the path of traffic, increase safety, reduce vehicle bypass, and then slow down traffic congestion. Currently, the traffic guidance system is the focus of attention at home and abroad, mainly based on theoretical analysis of system design from a macro perspective, induced by a single field experimental study design of the facility. Consider the safety of the experiment, study efficiency and the conclusion validity, how to make research and evaluation about the traffic guidance system more effective is a difficult problem. Based on virtual reality technology in the driving simulator set up traffic guidance system model, can solve the relevant problems.Virtual reality space is a real scene model simulation. According to previous studies showed that results obtained in the virtual and field experimental are different. Virtual scenes do not have the authenticity of the scene model. In order to ensure the validity of the experiment in virtual scene, the validation of driving simulator need to be taken. This paper contains the following work:First of all, validation of driving simulator were reviewed, the development of virtual reality technology and its applications, and driving simulation system. Driving simulator is divided into physical validation and behavior validation. This study limited to behavior validation of fixed driving simulator.Secondly, methods of legibility distance measurement virtual and real scene were proposed. Legibility distance in this paper was defined. Time points of recognition were recorded by time recorder software. Using GPS measured dynamic data of vehicles in real scene. Legibility distance was calculated after the system synchronization.Thirdly, driving simulator validation was conducted, guide sign as study object and legibility distance as measure of effectiveness. Analyzing data of legibility in virtual scene, speed and text size, models of the relationship between the three parameters were established. Results show that legibility distance and speed of negative correlation, positive correlation with the text size. Models were set up between legibility distance and speed in real scenes. Results show that legibility distance and speed of negative correlation. Conclusion is that behavior validation of driving simulator is relative validity.Fourth, In order to improve the validation of driving simulator, parameters of guiding sign were revised. Based on relationship models between legibility distance, speed and text size, formula of signs'coefficient calculation were derived. Guide signs in the real scene text height 30cm, considering the speed of the legibility distance of the virtual scene, text magnification, respectively 4.1 (20km / h), 3.7 (30km / h), 3.4 (40km / h), 2.9 (50km / h), text height was 124cm, 111cm, 101cm, 87cm.Fifth, effectiveness test was conducted in virtual space after parameter modification. Process of parameter calibration for driving simulator validation was established. Conclusion is that the validity of guide sign in driving simulator is effectively increased by modifying the parameters of it, legibility distance as measure of effectiveness. After parameter modification, the behavior validation of driving simulator is absolute validity.Main findings were summarized at the end of the paper. Directions for future research in this topic were proposed.