基于鲁棒观测器的电液伺服系统传感器故障检测与隔离

Sensor Fault Detection and Isolation for an Electro-Hydraulic Servo System Based on Robust Observer

在电液伺服系统的故障检测中,经常会引入额外的传感器,这些传感器一旦发生故障就会对系统的运行状态产生误判;而且,电液伺服系统本身是一个典型的非线性系统,其运行过程中各种未知和时变的干扰力也会对传感器故障的检测与隔离造成影响.针对这些问题,文中提出了一个基于非线性鲁棒观测器的传感器故障检测与隔离方案,采用非线性鲁棒观测器来对电液伺服系统的非线性及未知干扰进行处理,并采用线性矩阵不等式协助设计观测器以方便求解;对于传感器故障的隔离,采用观测器组结合逻辑判断的方式.文中还通过仿真和实验对所提出的故障检测与隔离方案进行了验证,并根据实验数据的特征设计了自适应阈值对传感器故障进行了故障决策,结果证明所提出的故障检测与隔离方案是有效的.

In the electro-hydraulic servo system,extra sensors are often added for fault detection and isolation（ FDI）,but the sensor fault can cause a false alarm. The electro-hydraulic servo system itself is a typical nonlinear system and is often subjected to time-varying and unknown disturbances,which brings about great challenges to the sensor FDI. In order to solve these problems,a FDI scheme based on a nonlinear robust observer is proposed. In this scheme,the robust observer is used to handle the system nonlinearity as well as unknown disturbances and a linear matrix inequality method is adopted to facilitate the observer design. For the sensor fault isolation,a batch of robust observers is designed and some logic rules are made. Then,the proposed FDI scheme is verified by simulations and experiments,and an adaptive threshold is designed to make decision according to the characteristics of experimental data. Both simulation and experimental results show that the proposed FDI scheme is effective.