The strictness of antipollution constraints, the necessity of fuel economy or the development of comfort imply new control strategies for the automotive engine. With the increasing complexity of modern systems, diagnosis becomes a major theme for improving industrial process safety and reliability. This paper describes how the fuzzy logic can be used in diagnostic problems dedicated to the fuel injection automotive engine. The design of the diagnostic scheme is the following. Pertinent symptoms that respond to a given fault are generated. The residual computation detect whether a fault has occurs or not. Then, the analysis of these residuals with fuzzy logic operators perform the localisation procedure to determine the cause of the fault. Finally, the identification procedure is performed to determine the size of the fault. The decision is processed by fuzzy rules. In cooperation with Bosch company, over 2500 sensor data have been analysed for the purpose of diagnosing throttle sensor bias and manifold pressure bias. Results proved that the diagnostic tool is able to identify throttle sensor bias with a precision of 4 degrees and 130mbar for the manifold pressure. Then, an aided-computer diagnostic tool will be developed and future work will lead to the implementation of the method in an onboard diagnosis for autoadaptative injection control strategies.
[1]
John J. Moskwa,et al.
Transient Air Flow Rate Estimation in a Natural Gas Engine Using a Nonlinear Observer
,
1994
.
[2]
Paul M. Frank,et al.
Fault diagnosis in dynamic systems using analytical and knowledge-based redundancy: A survey and some new results
,
1990,
Autom..
[3]
A. Willsky,et al.
Analytical redundancy and the design of robust failure detection systems
,
1984
.
[4]
J. Chen,et al.
A Review of Parity Space Approaches to Fault Diagnosis
,
1991
.
[5]
R. Patton,et al.
A Review of Parity Space Approaches to Fault Diagnosis
,
1991
.
[6]
Alan S. Willsky,et al.
A survey of design methods for failure detection in dynamic systems
,
1976,
Autom..