An Adaptive Collision Avoidance Decision Method for Various Driving Conditions

With the continuous advancement of autonomous driving technology, enabling the vehicles to avoid obstacles has gained increasing attention. However, traditional obstacle avoidance strategies have certain limitations as they often focus solely on nearby obstacles. In light of this, this paper proposes a more comprehensive collision avoidance approach. Suitable obstacle avoidance modes are firstly designed for various driving scenarios. The optimal collision avoidance mode is subsequently determined based on the actual distance to the obstacles. Effective obstacle avoidance for autonomous vehicles is finally achieved through the coordinated steering and braking control systems. To validate the feasibility of the proposed strategy, joint simulations using Carsim and MATLAB/Simulink have been conducted with consideration of two scenarios with different driving conditions. Experimental tests were further carried out using Quanser's latest autonomous vehicle-QCar platform. Both the simulation and the experimental results have verified the successful implementation of the obstacle avoidance approach under various driving conditions.