Hydraulic Servo System Fault Diagnosis Based on SVM

Hydraulic servo system plays an important role in aircraft power supply system. It is of great significance to study the fault diagnosis method for aircraft health management. This paper introduces a Support Vector Machine (SVM) approach to build an observer on Hydraulic servo system and obtain the fault diagnosis results by identifying of the residuals. A support vector observer estimates the output of the system. By comparing with system output, the residual is generated and then clustered by a support vector classifier for fault detection. Finally we use MATLAB/Simulink to establish fault diagnosis model and verify the correctness of this method.