Abstract:Model-based fault diagnosis serves as a powerful technique for addressing fault detection and isolation issues in control systems. However, diagnosing faults in closed-loop control systems is more challenging due to their inherent robustness. This paper aims to detect and isolate actuator and sensor faults in the cascade electro-hydraulic control system of a turbofan engine. Based on the fault characteristics, we design a robust unknown perturbation decoupling residual generator and an optimal fault observer specifically for the inner and outer control loops to detect potential faults. To locate the faults, we analyze the steady-state propagation laws of actuator and sensor faults within the loops using the final value theorem. Based on this, we establish the minimal-dimensional fault influence distribution matrix specific to the cascade turbofan engine control system. Subsequently, we construct the normalized residual vectors and monitor its vector angles against each row of the fault influence distribution matrix to isolate faults. Experiments conducted on an electro-hydraulic test bench demonstrate that our proposed method can accurately locate four typical faults of actuators and sensors within the cascade electro-hydraulic control system. This study enriches the existing fault isolation methods for complex dynamic systems and lays the foundation for guiding component repair and maintenance.

Fault Symptoms Generation via Reachable Set Estimation

Iterative Interval Estimation-Based Fault Detection for Discrete Time T-S Fuzzy Systems.

Fault Detectability Analysis for Linear Systems Within a Set-Based Framework

Active fault detection based on set‐membership approach for uncertain discrete‐time systems

Robust Fault Diagnosis for Uncertain Linear Systems: A Set-Based Geometric Approach

Fault tolerant estimation for a class of networked systems with sensor faults

Networked Fault Detection for Fuzzy MJSs

Robust Fault Detection and Diagnosis for Uncertain Nonlinear Systems

Residual evaluation of fault detection in discrete‐time systems

Fault Detection and Isolation for Uncertain Systems Based on Unknown Input Set-Membership Observer

ZKF-based Optimal Robust Fault Estimation of Descriptor LPV Systems with Measurement Error-Affected Scheduling Variables.

Steady-State Fault Propagation Characteristics and Fault Isolation in Cascade Electro-Hydraulic Control System

Active Fault Diagnosis under Hybrid Bounded and Gaussian Uncertainties

Set-Membership Estimation for Fault Diagnosis of Nonlinear Systems

State Estimation based Fault Analysis and Diagnosis in a Receiving-end Transmission System

A New Comprehensive Fault Diagnosis Approach Based on Fuzzy Theory

Quantized and event-triggered modeling and fault detection for networked fuzzy systems

Fault estimation for discrete time‐variant systems subject to actuator and sensor saturations

Fault Detection Using an Adaptive Fuzzy System

A Smoothing Approach for Active Fault Diagnosis with a Unified Set Representation.

An Analytic Model for Fault Diagnosis in Power Systems Utilizing Redundancy and Temporal Information of Alarm Messages