基于等价空间的自动驾驶车辆故障检测方法

为了提高自动驾驶车辆的稳定性,保证驾驶人和乘客的安全,本文对车辆动力系统的故障检测进行了深入的研究。在准确地描述实际系统动态特性的前提下,对车辆动力系统进行化简,得到车辆的单轨模型。根据车辆单轨模型的特性,给出了包含检测残差、评价函数以及诊断逻辑设计在内的适用于该模型的故障检测方法。通过仿真实例可以看出,该故障检测体系可以快速准确地检测出车辆横摆率和转向角传感器故障,有效地提高了自动驾驶车辆的可靠性。

To improve the performance stability of autonomous vehicles and ensure the safety of drivers and passengers, the fault detection of vehicle power system was studied in this paper. On the premise of accurately describing the actual dynamic characteristics of the system, the vehicle power system was simplifi ed to obtain the monorail model of the vehicle. Based on the characteristics of the monorail model, a fault detection method which includes detection residual, evaluation function and diagnosis logic design was proposed. It can be seen from the simulation example that the proposed fault detection method can quickly and accurately detect the vehicle yaw rate and the steering angle sensor fault, which effetely improves the reliability of the autonomous vehicles.