In this study, the performance of a slope Autonomous Emergency Braking (AEB) system that estimates the angle of inclination in a V2V-based vehicle was evaluated. Existing AEB systems do not take changes in the angle of inclination into account or they use a reference angle, which results in limited braking properties on a slope. To improve the braking properties of AEB systems on slopes, the angle of inclination must be estimated. We propose an AEB system that employs an extended Kalman filter, along with an in-vehicle sensor (ESC) combined with a bicycle model and linear tire model to estimate the angle of inclination. We conducted comparative analysis via simulations of the performance of the proposed AEB system with those of an AEB system that does not consider slope and another one that uses a reference angle of inclination at a variety of angles of inclination and velocities. The results obtained verify the utility of the proposed extended Kalman filter-based AEB system on slopes.

No abstract

Abstract. In this study, a brake intervention time algorithm of an ego vehicle is proposed to consider various road surface conditions when driving, through vehicle to vehicle (V2V) communication with a front vehicle in a curved road. Conventional autonomous emergency braking (AEB) systems have a consistent brake intervention time regardless of road surface conditions. To avoid a collision with a vehicle in front, a control method is introduced, in which the brake intervention time changes according to the road surface conditions. To verify the brake intervention time according to various road friction factors, the simulation scenario reflected the vehicle velocity and road curvature radius. The control method, which changed the brake intervention time according to the road surface conditions, had better collision-avoidance performance than the conventional AEB system control method.

This paper proposes a system to determine the minimum collision avoidance distance with vehicles in other lanes during lane change when vehicle-to-vehicle (V2V) communication is established between the host vehicle and the vehicle ahead. Existing collision avoidance systems have the limitations that blind spots exist because they employ sensor-based recognition and that the moving time in the lateral direction should be calculated in advance before the collision risk distance is calculated. To resolve these problems, V2V communication and minimum collision avoidance distance (CAD) method were adopted in this study. To verify the minimum CAD for lane change, the relative velocity and the time required for lateral direction movement were calculated and reflected in the simulation scenarios. The collision risk during a lane change was then calculated using the relative velocity and CAD.

This study proposes a way to generate a virtual lane by using a vehicle's driving trajectory during the operation of a V2V-based vehicle. The existing method of road shape detection using Lidar, Radar, and camera sensors has limitations in the range of detection with blind spots. To overcome such limitations, a method to generate virtual lanes was proposed to detect the road shape along with the driving trajectory of the leading vehicle based on V2V communications. A Clothoid road model was applied for the calculation of the driving trajectory of the leading vehicle. A driving scenario with various changes in curvature of road was also configured. Furthermore, the feasibility of the virtual lane generation method was verified by comparing the reference virtual lane and the virtual lane generated by the driving trajectory. The simulation results confirmed the usability of the virtual lane generated from Clothoid road model.