A Data-Driven method of Engine Sensor on Line Fault Diagnosis and Recovery

Considering the requirements of convinced sensor measurements for engine control, a method of aircraft engine sensor on line fault diagnosis and recovery based on least squares support vector machine (LS-SVM) is proposed. First, sensor sets correlations are calculated and the sensor with high correlation is selected by correlation analysis. Then sensor LS-SVM prediction model is established with the sensor itself primary data series and used to sensor fault diagnosis. The sensor recovery module is obtained based on the LSSVM algorithm with the high correlated sensor set, and is activated as the sensor failure detected. Experimental results show that the engine sensor fault recognition rate is satisfied by the proposed method, and could be used to turbofan engine sensor fault diagnosis and data recovery.