Fault diagnosis

Details Fault diagnosis Fault diagnosis

2000-06-12

2003-08-12

Hoff, Marc S. (Department: 2857)

Data processing: measuring, calibrating, or testing

Measurement system

Performance or efficiency evaluation

C702S182000, C702S183000, C702S186000, C702S188000

Reexamination Certificate

active

06606580

ABSTRACT:

FIELD OF THE INVENTION
This invention relates to methods and systems for fault diagnosis, in particular, although not necessarily exclusively, for the detection of faults in multi-component systems, typically power plant, whose performance is characterised by a series of indirectly measurable performance parameters. One specifically envisaged application of the invention is the detection and quantification of faults in a gas turbine.
BACKGROUND
Analysis of the performance of gas turbines is important for both testing of development engines and condition monitoring of operating engines. In particular, the ability to accurately and reliably identify the component or components responsible for loss of performance would be particularly beneficial. However, the nature of a gas turbine means there are inherent problems to be addressed if this is to be achieved in practice.
In common with other power plant, a gas turbine's performance can be expressed in terms of a series of performance parameters for the various components of the system. Typically, those chosen to characterise the performance of a gas turbine would be the efficiency and flow function of compressors and turbines and the discharge coefficient of nozzles. Significantly, these performance parameters are not directly measurable. They are, however, related to measurable parameters, such as spool speeds, averaged pressures and temperatures, thrust and air flows, and any loss in performance of a component will be reflected by changes in these measurable parameters (referred to hereinafter as “measurement parameters”).
Given a particular operating point, typically defined by environmental conditions and a power setting (inlet temperature and pressure, and fuel flow for example), or other such operating parameters, the relationship between measurement parameters and performance parameters can be represented, for example, in a performance simulation model describing the aerothermodynamics of the gas turbine's components. In principle this enables the performance parameters, and more importantly changes in them from respective reference values, to be calculated. In turn, it should be possible, from these calculated values, to detect a drop in the performance of one or more components and hence deduce the cause of a loss in performance.
In practice, however, it has been found difficult to accurately and reliably relate the calculated changes in performance parameters to the presence of faults in specific components, particularly when only a small number of engine components are faulty. This is due at least in part to the non-linearity of the relationship between the measurement parameters and the performance parameters.
There are, however, a number of further obstacles to this diagnostic task. In particular, sensors used to collect the measurement parameters necessarily work in a harsh operating environment. This leads to large measurement noise and a high probability of systematic errors (i.e. biases) in the sensor suite. The same problems affect the measured operating parameters (e.g. environment and power setting parameters). To be effective a diagnostic method has to be able to account for these measurement uncertainty effects.
SUMMARY OF THE INVENTION
It is a general object of the present invention to provide methods and systems for processing data relating to the performance of systems having a number of distinct components, with the aim of detecting “faults” in the components of the system and/or the measurements taken to collect the data from the system.
The term “fault” is used here and in the following to refer to deviations in performance parameters associated with the components from expected, or reference values, often referred to as base line values, as well as errors in measured values arising from measurement biases. Aspects of the present invention are concerned with the detection of both of these forms of “fault”.
A further object is the isolation of a faulty component or faulty components in a system and quantification of the loss in performance. In this respect, the identification of faulty measurements (i.e. measurement biases) can be important because they may result in false indications of the values of associated performance parameters. The invention is particularly useful in the detection of faults in only a small number of components (eg. 1, 2 or 3), rather than faults in a large number of components as might be expected as an engine slowly deteriorates in service.
These and other objects and aspects of the invention will be apparent from the detailed description and examples below.


REFERENCES:
patent: 5633800 (1997-05-01), Bankert et al.
patent: 5897629 (1999-04-01), Shinagawa et al.
patent: 6038540 (2000-03-01), Krist et al.
patent: 6119075 (2000-09-01), Dean et al.
patent: 6163840 (2000-12-01), Chrysos et al.
patent: 6418398 (2002-07-01), Duek et al.
patent: WO 91/14990 (1991-10-01), None
A J Volponi, “Sensor error compensation in engine . . . ” ASME Paper 94-GT-58.
L A Urban et al. “Mathematical Methods of relative engine . . . ” SAE Paper 922048.
D L Doel “An assessment of weighted-least-squares gas path analysis” ASME Paper 93-GT-119.
D L Doel “Temper: Gas-path analysis tool for commerical jet engines” ASME Paper 92-GR-315.
K Deb et al. “Multiobjective Optimization using nondominated . . . ” Journal of Evolutionary Computation 1994, 2(3), MIT.
A Stamatis et al. “Jet engine fault detection with differential . . . ” ISABE 89-7133.
R Eustace “Neural Network fault diagnosis . . . ” ISABE 93-7091.
R Eustace et al. “Fault diagnosis of fleet engines . . . ” ISABE 95-7085.
M Zedda et al. “Fault Diagnosis of a turbofan engine . . . ” AIAA 98-3602.

Affiliated with

Also associated with

Rating

Fault diagnosis does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with Fault diagnosis, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Fault diagnosis will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFUS-PAI-O-3092513