Research on robustness of intelligent vehicle longitudinal active collision avoidance system based on equivalent external interference
Chaochun Yuan,Yang Xiao,Jie Shen,Long Chen,Yingfeng Cai,Youguo He,Shuofeng Weng,Yuqi Yuan,Yuxuan Gong,Qiuye Yu
DOI: https://doi.org/10.1177/09544070241228640
2024-02-20
Proceedings of the Institution of Mechanical Engineers Part D Journal of Automobile Engineering
Abstract:Proceedings of the Institution of Mechanical Engineers, Part D: Journal of Automobile Engineering, Ahead of Print. The rate of fatal traffic accidents caused by sudden risks such as pedestrians crossing the road in violation of regulations and goods falling in front of the vehicles has remained high under the complex dynamic environment. The appearance of sudden obstacles/pedestrians will indirectly trigger sudden changes in sensor signals and control signals, and cause the drivers' improper driving behaviors. At the same time, the parameter perturbation of the vehicle itself and the sensor noise and other direct disturbances will also put forward high requirements for the robust performance of the vehicle braking system controller. Traditional control methods are difficult to ensure the robust stability required by the vehicle braking system in the actual driving process under sudden emergency conditions. In order to improve the braking control performance and robust performance of the active collision avoidance system under sudden emergency conditions under the influence of its own parameter perturbation, sensor perturbation and other perturbations, this manuscript analyzes the accident risk characteristics of sudden emergency conditions, and establishes an equivalent virtual external force interference model [math]. Considering the virtual external force interference model, self-driving parameter perturbation and sensor perturbation, we adopt the [math] control theory-design brake controller. Under different perturbation conditions the [math] controller, PID controller, and [math] controller are simulated and compared. The results show that the [math] controller can effectively suppress sudden emergency conditions under parameter uncertainty and external sensor noise interference compared with the PID controller and the [math] controller, realizing emergency braking and collision avoidance, and having better closed-loop robustness.
engineering, mechanical,transportation science & technology