Fault diagnosis for sensors of aero-engine based on improved least squares support vector regression

In this paper, an intelligent online sensor faults diagnosis system of aero-engine is designed based on improved least squares support vector machine. In order to distinguish sensor offset fault and drift fault, the method, called threshold discriminance, is adopted. For common sensor faults, including the single or multiple sensor offset faults and drift faults, the sensor faults diagnosis system is able to accomplish the fault detection, isolation and signal reconstruction well. And the simulation experiments of aero-engine sensor faults show the effectiveness and validity of the proposed system.