Sensor Fault Diagnosis of Locomotive Electro-Pneumatic Brake Using an Adaptive Unscented Kalman Filter

Algorithm 1

The procedure of sensor fault diagnosis.

1: Determination of the system state equation and observation equation: the system state equation and the observation equation ;

2: Discretization of the system equation and observation equation: discretized system equation and discretized observation equation ;

3: Initialization: for set: initial estimation error covariance , initial process noise covariance , measurement noise covariance ;

4: Time update and measurement update for;

5: residual generation by making a difference between the pressure sensor measurement and the UKF estimation ;

6: residual evaluation through Sequential Probability Ratio Test;

7: gotoTime update and measurement update