Observer Based Fault Estimation and Fault-Tolerant Sliding Mode Control for Nonlinear Systems

In this article, a observer based sliding mode fault tolerant control (SMFTC) problem is discessed to guarantee the reliability for a class of nonlinear systems. To handle the unpredictable and simultaneous sensor fault and actuator fault, a unknown input sliding mode observer (UISMO) is developed with the integral switching function by taking the unmeasured states into account. By utilizing the UISMO, an augmented error system is established and guaranteed to be stable. Ulteriorly, both the unmeasured states and faults can be estimated and reconstructed simultaneously, which relaxes the strict constraints of actuator fault must being a step signal. Based on the fully estimation information, a SMFTC strategy is established under adaptive reaching law (ARL) to reduce the system chattering and restore the system performance when the unpredictable actuator and sensor faults occur. Then, a flexible joint robot arm system is provided to illuminate the reasonableness of the proposed UISMO and SMFTC scheme.