避碰
碰撞
理论(学习稳定性)
纵向静稳定性
防撞系统
计算机科学
计算机安全
工程类
航空航天工程
机器学习
空气动力学
作者
Xiang Chen,Shuo Cheng,Shaohua Li,Wanzhong Zhao,Chunyan Wang
标识
DOI:10.1109/tte.2023.3323643
摘要
Collision avoidance control is a significant but challenging issue for safety-critical automated vehicles. Intractable challenges posed by dynamic and complex traffic situations have attracted many attempts from both the academia and industrial sectors. Prevalent methods include longitudinal emergency braking and lateral collision-free path planning. However, collision avoidance maneuvers could cause chassis handling instability due to vehicles' highly nonlinear dynamics, especially at high speeds, which remains a challenging open problem. Besides, the longitudinal and lateral coordinated control is not studied sufficiently in current studies. To address the above-mentioned issues, this paper proposes a lateral and longitudinal integrated emergency collision avoidance system, and its generated optimal trajectories can improve both chassis handling stability and geometrical collision-free. This work formulates emergency avoidance planning and control as a multiple-objective optimization problem, and this optimization formulation concerns trajectory smoothness, collision risk, and vehicle dynamics response. Furthermore, the proposed system employs rigorous constraints derived from the geometrical collision check and chassis dynamic stability analysis. Then, the optimal trajectories of collision avoidance are solved, which consists of both longitudinal speeds and lateral offsets. Various test scenarios with different initial speeds and distances are carried out. Compared with existing controllers, our proposed method performs better and avoids collisions while ensuring dynamic stability. Those results validate the effectiveness and feasibility of the proposed method.
科研通智能强力驱动
Strongly Powered by AbleSci AI