Development of Collision Avoidance System Integrated With Real-Time Tire-Road Friction Coefficient Estimator

Collision avoidance systems that utilize both steering and braking have been widely studied. However, information about the tire-road friction coefficient, which plays a critical role in collision avoidance, is assumed to be a known constant value in many studies. Because of this, it is difficult to properly respond to changes in road surface conditions that commonly occur in practice. In this paper, a collision avoidance system that estimates and applies the tire-road friction coefficient in real-time is proposed to overcome these limitations. First, a path planner and tracking controller are designed. It applies friction coefficient changes to the algorithm and provides prompt feedback of the updated vehicle state variables. Next, a friction coefficient estimator is proposed that enables rapid and highly accurate estimation in a short period of time, even in collision avoidance situations where sufficient excitation required for estimation may not be continuously achieved for a long time. Finally, by integrating the collision avoidance system and the friction coefficient estimator, the estimated tire-road friction coefficient is applied to the collision avoidance system in real-time, and the reactions are used to estimate the tire-road friction coefficient again. This demonstrates the excellent collision avoidance performance despite changes in road friction and also proves the virtuous cycle interaction of the integrated system.