A method of engine component fault diagnosis based on fuzzy decision fusion was proposed to improve the accuracy of component fault diagnosis
and reduce the probability of false alarming and missing detection.By using Kalman filter algorithm and least square support vector machine(LSSVM)
the sensor outputs were both sent to two different modules —— self-tuning model and data-based diagnostic module —— to estimate the health parameters.Fuzzy algorithm was operated to adjust the weights of two modules
and then the fault fusion decision was made by D-S(Dempster-Shafer) theory.Simulation on a turbofan engine shows that
as compared to these two modules
this method has better accuracy of fault diagnosis.