Diagnosis for lipschitz nonlinear system with sliding mode observer and virtual sensor

摘要

This paper proposes a scheme for designing a robust sliding mode observer and a virtual sensor for uncertain Lipschitz nonlinear system with actuator and sensor faults are considered. First, a transformation of states and outputs is proposed to divide the original system into two subsystems in order to isolate the sensor faults, such that the first subsystem is affected by actuator faults and the second subsystem is subjected only to sensor faults. Then, we design a single SMO to estimate all stats and simultaneous actuator and sensor faults. Moreover, we use a technique called virtual sensor to replace a failed, missing or corrupted sensor to estimate and recover the sensor faults. Then, An illustrative example based on robotic manipulator model is conducted to verify the proposed approach effectiveness.