Diagnosis of aircraft engine performance deterioration based on support vector machines

In order to analyze the problem of aircraft engine performance deterioration, a diagnosis model based on SVM (Support Vector Machines)was built using state-related parameters including compressor outlet pressure, compressor outlet temperature, turbine outlet pressure, turbine outlet temperature, fuel flow, thrust and etc which were obtained from the simulation results of aircraft engine component-level simulation model under the condition that performance deterioration existed in compressor, turbine or combustor. The simulation results were used as learning samples and comparisons of the diagnosis results with the data sample used for test were given. It shows that SVM can be effectively applied to diagnose the aircraft engine performance deterioration and provide enough accuracy for failure location, which is of theoretical importance and application value for gas turbine health management.