Comparison of Model-based Vs. Data-driven Methods for Fault Detection and Isolation in Engine Idle Speed Control System

Ruochen Yang and Giorgio Rizzoni
Submission Type: 
Full Paper
AttachmentSizeTimestamp
phmc_16_006.pdf3.25 MBAugust 3, 2016 - 8:00am

This paper presents a comparison of model-based vs data-driven diagnostic algorithms for an industrial application. The objective of the paper is to evaluate the effectiveness of different methods in identifying sensor and actuator faults in an automotive engine during idle speed control operation. An internal combustion engine operating at idle is regulated by a feedback controller so that it runs at a preset idle speed without stalling when no acceleration is requested from the driver. At idle, engine speed regulation is needed due to the complexity and uncertainty of the operating condition and the ambient environment. In particular, ambient temperature affects air density, as well as the warm-up behaviour of the engine (and hence engine friction). Further, accessories, such as the air conditioner compressor, alternator or power steering actuator, can introduce significant load torque disturbances. Because of the regulating behavior from the controller, faults, especially actuator faults, may affect sensor measurements in a way very similar to disturbance, system uncertainty or noise. This poses a challenge to the fault detection and isolation (FDI) problem for this system. In this paper, two fundamentally different fault diagnoses approaches are used to detect and isolate faults. A model-based residual generation scheme as well as a data-driven linear discriminant analysis approach are developed to solve the FDI problem even when faults occur in the presence of system uncertainty, disturbance and noise. The performances of the two algorithms are compared side by side using data gathered from an experimentally validated simulator for an engine idle system that considers an actuator fault, a sensor fault, several system uncertainties and disturbance (operating conditions), and sensor noise. The results show that comparable performance can be achieved with both schemes and some comments are made about the merits and challenges for each approach as well.

Publication Year: 
2016
Publication Volume: 
7
Publication Control Number: 
006
Page Count: 
9
Submission Keywords: 
Fault Diagnostics; engine operating at idle speed; Model-based method; Data-based method
Submission Topic Areas: 
Data-driven methods for fault detection, diagnosis, and prognosis
Industrial applications
Model-based methods for fault detection, diagnostics, and prognosis
Submitted by: 
  
 
 
 

follow us

PHM Society on Facebook Follow PHM Society on Twitter PHM Society on LinkedIn PHM Society RSS News Feed