A Study on Scenario Construction for the Autonomous and Regular Vehicles Mixed Traffic Conditions

Abstract

To respond to hazardous situations that may arise due to the limitations of self-driving vehicle recognition and judgment capability, technology development is in progress for infrastructure to recognize this and provide operation guidance to autonomous vehicles. In order to evaluate the performance of the infrastructure guidance technology under development and to use it for model advancement, this study presented a scenario assuming various traffic environments and mixed situations of autonomous and general vehicles. The proportion of autonomous vehicles was composed of optimistic, average, and conservative three cases using the figures predicted in previous studies, and traffic-related factors were composed of road traffic internal factors (intersection type, number of lanes, etc.) and external factors (e.g., unexpected situation, weather, etc.). Finally, the compliance rate of the C-ITS service was used to variable whether general vehicles comply with the guidance, and the number of four compliance rate cases was derived by applying a percentage to the compliance rate distribution for each event. A total of 1,843,200 combinations of scenarios were derived, but there is a limitation to test all cases. Therefore, the fractional factor design method was used to reduce them to the number of scenarios that can be analyzed, and as a result, it was possible to reduce the number to 35 scenarios. This study is meaningful in that it adds an element called driver compliance to the creation of a scenario that can verify safety in a mixture of actual self-driving vehicles and general vehicles, and presents a methodology that can reduce the number of scenarios to a number that can be analyzed realistically while statistically significant. In future studies, simulations will be conducted using the scenarios derived in this study, and a advanced autonomous driving scenarios will be derived through performance evaluation, feedback, and supplementation.