【作者】 王杰；

【导师】 李道飞；

【作者基本信息】 浙江大学 ， 车辆工程， 2019， 硕士

【摘要】 在车辆智能化发展趋势下,由于取消了方向盘与转向轮之间的机械连接,线控转向系统在转向性能、结构布置与整车舒适性等方面具有较大的优势,在智能驾驶车辆领域逐渐受到关注。但是,由于没有直接的机械连接,系统失效情况下车辆可控性降低,使得目前的线控转向系统尚不能完全满足功能安全等级要求,这是线控转向系统研究的重点与难点。本文结合实际进行的科研项目需求,从智能车辆线控转向系统功能安全角度出发,针对线控转向系统传感器故障诊断与容错控制展开研究。首先,本文根据ISO26262-3标准对线控转向系统的使用场景与危险事件进行综合分析,完成了系统的ASIL等级评估,确定了系统危险事件对应的安全目标。通过主要安全机制与措施的确定,指出目前线控转向技术功能安全研究重点集中在传感器、电机、ECU的故障检测与容错补偿方面,并结合课题需要选取转角传感器的故障诊断与容错补偿算法作为研究重点。根据电动助力转向系统与线控转向系统之间的结构相似性,确定线控转向系统原型结构,并以此为基础进行线控转向系统动力学建模。考虑车辆动力学模型的非线性、轮胎模型的非线性,搭建完整的转向系统-轮胎-车辆动力学模型。其次,以动力学模型为基础,选择对非线性系统滤波效果较好的无迹卡尔曼滤波算法设计车辆前轮转角估计器,并验证该算法对车辆前轮转角估计的准确性。同时针对估计器本身设计自诊断策略,保证车辆前轮转角估计的有效性。对于传感器故障诊断,本文通过对转角传感器典型失效形式进行建模分析,提出了基于传感器自身测量序列统计学特征的传感器卡死故障诊断算法与基于传感器残差序列统计学特征的传感器噪声与漂移故障诊断算法,并在第三章论述该故障诊断算法的正确性。在故障容错补偿方面,针对在可修正域范围内的噪声故障,设计自适应卡尔曼滤波器进行实时降噪处理,得到的补偿输出序列比估计序列精度高。针对其余故障类型,故障容错补偿最终都是输出估计序列,为了避免输出估计序列时的转角值突变,设计切换权重函数,对输出估计序列过程进行平滑处理,避免对控制系统产生较大干扰。最后,选取智能车辆路径跟踪控制作为验证该算法容错控制性能的上层控制功能。根据线性二次型调节器最优控制理论设计路径跟踪控制算法的状态反馈控制,并针对系统的稳态跟踪误差添加前馈控制,实现系统理想控制状态下的零位置误差控制。使用CarSim-MATLAB/Simulink联合仿真平台搭建基于故障诊断与容错补偿算法的路径跟踪控制系统,对传感器容错控制效果进行分析。基于NI PXIe8135实时控制器,搭建转角传感器硬件在环实验平台,对转角传感器进行故障注入,验证故障诊断与容错补偿算法运行的实时性与有效性。在实车试验方面,通过设计并改装实车线控转向系统,使用NI CompactRIO作为控制器原型完成试验车辆主动转角控制。运用dSPACE MicroAutoBox II作为算法实时运算平台进行实车试验,验证算法在实车运用中的可行性与有效性。更多还原

【Abstract】 Without any mechanical connections between the steering wheel and the front wheel,steerby-wire(SBW)system gives the steering system great advantages in steering performance,structural arrangement and ride comfort,and gradually attracts attention in the field of intelligent vehicles.However,the current SBW system cannot fully meet the functional safety requirements due to the reduction of vehicle controllability in the case of system failure,which is the focus and difficulty of the research of SBW systems.In this paper,according to the actual needs of the scientific research project,the sensor’s fault diagnosis and fault-tolerant control of SBW system for intelligent vehicles are studied from the perspective of functional safety.Firstly,according to ISO26262-3,the scenarios and risk events of the SBW system are comprehensively analyzed.The ASIL evaluation of the system is completed,and the corresponding safety objectives of the system risk events are determined.Through the determination of the main safety mechanisms and measures,it is pointed out that the current functional safety research of SBW technology focuses on the fault detection and fault-tolerant compensation of sensors,motors and ECUs.Combined with the scientific research project,the fault diagnosis and fault-tolerant compensation algorithm of the angle sensor is selected as the research focus.According to the structural similarity between the electric power steering system and the SBW system,the prototype structure of the SBW system is determined,and a dynamic model is built.Considering the nonlinearity of vehicle dynamics model and tire model,a complete steeringtire-vehicle dynamics model is built.Secondly,based on the dynamic model,the Unscented Kalman Filter(UKF)algorithm with better filtering effect for the non-linear system is selected to design the vehicle front wheel angle estimator,and the accuracy of the algorithm for the vehicle front wheel angle estimation is validate.At the same time,the self-diagnosis strategy of the estimator is designed to ensure the validity of the front wheel angle estimation.For sensor fault diagnosis,by modeling and analysis of typical failure modes of sensors,the sensor stuck fault diagnosis algorithm based on the statistical characteristics of the sensor’s own measurement sequence and the sensor noise and offset fault diagnosis algorithm based on the statistical characteristics of the sensor residual sequence are proposed.In the third chapter,the correctness of the fault diagnosis algorithm is discussed.For fault-tolerant compensation,an Adaptive Kalman Filter(AKF)is designed for noise faults in the correctable domain,and the compensation output sequence is more accurate than the estimation sequence.For the other fault types,fault tolerant compensation output is the estimation sequence.To avoid a sudden change of the angle value,the switching weight function is designed to smooth the process of outputting the estimation sequence and avoid the large disturbance to the control system.Finally,the intelligent vehicle path tracking control is selected as the upper control function to validate the fault-tolerant control performance of proposed algorithms.According to the optimal control theory of linear quadratic regulator,the state feedback control of path tracking control is designed.Feedforward control is added to decrease the steady-state tracking error of the system and to achieve zero position error control under the ideal control of the system.A path tracking control system based on fault diagnosis and fault-tolerant compensation algorithm is built by CarSim-MATLAB/Simulink simulation platform,and the effect of sensor fault-tolerant control is analyzed.Based on the NI PXIe8135 real-time controller,the hardware-in-the-loop(HIL)experimental platform of the angle sensor is built,and the fault injection of the angle sensor is carried out to validate the real-time performance and effectiveness of the fault diagnosis and fault-tolerant compensation algorithm.In the field tests,by designing and modifying the real vehicle SBW system,the active angle control of the test vehicle is completed by using NI Compact RIO as the prototype of controller.Vehicle test is carried out by using dSPACE MicroAutoBox II as the realtime computation platform of the algorithm,and the feasibility and effectiveness of the algorithm in vehicle application are validated.更多还原