Advances and Applications in Statistics
Volume 46, Issue 3, Pages 225 - 236
(September 2015) http://dx.doi.org/10.17654/ADASSep2015_225_236 |
|
KALMAN FILTERING APPLIED TO ENGINE FAULT DETECTION
Sutawanir Darwis, Nusar Hajarisman, Teti Sofia Yanti, Mohammad Sobri and Irfan Muharam
|
Abstract: Fault detection concerns with monitoring a system, identifying when a fault has occurred and pinpointing the type of faults. Two approaches can be distinguished: a direct recognition of sensor reading that indicates a fault and an analysis of the discrepancy between the readings and the expected values derived from some model. Fault detection typically waits until a sufficiently large fault signature is present before an alarm is raised. However, faults that affect safety must be detected quickly. Otherwise, undesired consequences arise such as engine power loss, severe vibration and no response to command. The idea of Kalman filter is to view these parameters as constant state and estimate them using a Kalman filter. If there is no fault, then the estimated state variable should not change rapidly. This survey paper explores the Kalman filter approach for engine fault detection. Kalman filter is used to estimate the state and the engine parameter. In diesel engine fault detection, the model of the engine in the fault free is described by four state and five state equations that described the fault model. Kalman filter based approaches seem to be the most commonly used method. |
Keywords and phrases: engine modeling, fault detection, Kalman filter, statistical process control. |
|
Number of Downloads: 440 | Number of Views: 1257 |
|