考虑传感器与执行器故障的EPS主动容错控制

Active Fault-tolerant Control of EPS Considering Sensor and Actuator Fault

针对传感器及执行器故障对EPS助力性能的影响,提出一种EPS主动容错控制方法。建立含参数不确定性、传感器与执行器故障的EPS系统模型,将系统不确定性转化为故障估计误差系统的扰动,基于未知输入观测器及线性矩阵不等式推导故障估计误差系统稳定并对扰动具有鲁棒性的充分条件,采用LMI区域极点配置法提升故障估计性能;在此基础上,针对执行器故障设计控制律补偿容错控制算法,针对传感器故障设计信号重构容错控制算法。Matlab/Simulink环境下的仿真结果表明,当传感器与执行器单独或同时发生故障时,设计的故障估计算法均可较为准确地估计故障幅值,故障估计的误差较小;针对不同故障对助力性能的影响,提出的容错控制方法均可使故障EPS系统的助力性能有所恢复。基于LabVIEW PXI的硬件在环试验进一步验证容错控制应用于EPS系统的有效性,提升汽车转向行驶的安全性及可靠性。

Aiming at the effects on assist performance of EPS caused by the sensor and actuator fault, an active fault-tolerant control is proposed. EPS model with parameter uncertainty, sensor and actuator fault is built, and the parameter uncertainty is transformed into the disturbance of the fault estimation error system. Sufficient condition, that for error dynamics system which is stable and robust against disturbance, is derived based on unknown input observer and linear matrix inequalities. LMI region pole assignment is used to improve the accuracy of fault estimation;based on fault estimation, control law compensation is designed for the actuator fault tolerance;signal reconstruction is designed for the sensor fault tolerance. Simulation in Matlab/Simulink shows that, when single or simultaneous sensor and actuator fault happens, the designed algorithm can estimate the amplitude of the fault accurately while the estimation error is small;in view of the assist performance deteriorate caused by various faults, the proposed fault-tolerant method can restore the assist characteristic of failure EPS system to some extent. Hardware-in-loop test based on LabVIEW PXI further validates the effectiveness of fault-tolerant control in EPS which improves the safety of steering driving and reliability of the vehicle.