Fault Diagnosis of EHA Based on Improved Unknown Input Observer

The objective of this paper is to propose a fault diagnosis method based on an improved unknown input observer (UIO) for electro-hydraulic actuator (EHA) systems. The method is designed to enhance the accuracy and reliability of fault diagnosis in EHA systems, which are known to be complex and nonlinear. The improved UIO is capable of effectively filtering out unknown interference by optimising the observer gain and state estimation algorithm. This paper presents a validation of the proposed method through the use of the MATLAB/Simulink simulation platform, which allows for the testing of the system performance under normal and faulty working conditions. The results demonstrate that the enhanced UIO outperforms the conventional adaptive observer in terms of estimation accuracy, fault detection rate, isolation rate, detection time, and false alarm rate. This is particularly evident in the cases of sensor fault, actuator fault, and hydraulic leakage fault, where the detection rate reaches 95%, 90%, and 88%, respectively, and the isolation rate reaches 90%, 88%, and 85%, respectively. These enhancements significantly enhance the fault diagnosis capability of the EHA system, thereby providing an effective guarantee for its safety and reliability and having important practical application value.