Publications of Pierre Dewallef
Bookmark and Share    
Full Text
See detailA Way to Deal with Model-Plant Mismatch for a Reliable Diagnosis in Transient Operation
Borguet, Sébastien ULiege; Dewallef, Pierre ULiege; Léonard, Olivier ULiege

in Proceedings of the ASME Turbo Expo 2006 (2006, May)

Least-squares health parameter identification techniques such as the Kalman filter have been massively used to solve the problem of turbine engine diagnosis. Indeed, such methods give a good estimate ... [more ▼]

Least-squares health parameter identification techniques such as the Kalman filter have been massively used to solve the problem of turbine engine diagnosis. Indeed, such methods give a good estimate provided that the discrepancies between the model prediction and the measurements are zero-mean, white random variables. In turbine engine diagnosis, however, this assumption does not always hold due to the presence of biases in the model. This is especially true for transient operation. As a result, the estimated parameters tend to diverge from their actual values which strongly deteriorates the diagnosis. The purpose of this contribution is to present a Kalman filter diagnosis tool where the model biases are treated as an additional random measurement error. The new methodology is tested on simulated transient data representative of a current turbofan layout. While relatively simple to implement, the newly developed diagnosis tool exhibits a much better accuracy than the original Kalman filter in the presence of model biases. [less ▲]

Detailed reference viewed: 33 (2 ULiège)
Full Text
See detailCombining classification techniques with Kalman filters for aircraft engine diagnostics
Dewallef, Pierre ULiege; Romessis, C.; Léonard, Olivier ULiege et al

in Journal of Engineering for Gas Turbines & Power (2006), 128(2), 281-287

A diagnostic method consisting of a combination of Kalman filters and Bayesian Belief Network (BBN) is presented. A soft-constrained Kalman filter uses a priori information derived by a BBN at each time ... [more ▼]

A diagnostic method consisting of a combination of Kalman filters and Bayesian Belief Network (BBN) is presented. A soft-constrained Kalman filter uses a priori information derived by a BBN at each time step, to derive estimations of the unknown health parameters. The resulting algorithm hers improved identification capability in comparison to the stand-alone Kalman filter. The paper focuses on a way of combining the information produced by the BBN with the Kalman filter. An extensive set of fault cases is used to test the method on a typical civil turbofan layout. The effectiveness of the method is thus demonstrated, and its advantages over individual constituent methods are presented. [less ▲]

Detailed reference viewed: 119 (7 ULiège)
Full Text
See detailOn-Line Transient Engine Diagnostics in a Kalman Filtering Framework
Borguet, Sébastien ULiege; Dewallef, Pierre ULiege; Léonard, Olivier ULiege

in Proceedings of the ASME Turbo Expo 2005 (2005, June)

A common assumption made in the performance assessment of a turbine engine for aircraft propulsion consists in restricting the data processing to steady-state data. This especially holds for on-board ... [more ▼]

A common assumption made in the performance assessment of a turbine engine for aircraft propulsion consists in restricting the data processing to steady-state data. This especially holds for on-board performance monitoring of a commercial aircraft which spends up to 90% of the time in cruise flight where such conditions are satisfied. The present contribution is intended to investigate the ability of a diagnosis method to process unsteady data rather than steady-state data. The aim of this unsteady approach is to strongly reduce the time and the efforts spent to obtain a reliable diagnosis. In order to assess the improvements in terms of diagnosis efficiency and engine operability, the resulting diagnostic method is tested for different degradations that can be expected on commercial turbofans. The results are also compared to those obtained from cruise flight steady-state data in order to balance the two approaches. [less ▲]

Detailed reference viewed: 44 (8 ULiège)
Full Text
See detailOn-Line Aircraft Engine Diagnostic Using a Soft-Constrained Kalman Filter
Dewallef, Pierre ULiege; Léonard, Olivier ULiege; Mathioudakis, Kostas

in Proceedings of the ASME Turbo Expo 2004 (2004, June)

Detailed reference viewed: 186 (1 ULiège)
Full Text
See detailCombining classification techniques with Kalman filters for aircraft engine diagnostics
Dewallef, Pierre ULiege; Léonard, Olivier ULiege; Mathioudakis, Kostas et al

in Proceedings of the ASME Turbo Expo 2004 (2004, June)

Detailed reference viewed: 23 (3 ULiège)
Full Text
See detailOn-Line Performance Monitoring and Engine Diagnostic Using Robust Kalman Filtering Techniques
Dewallef, Pierre ULiege; Léonard, Olivier ULiege

in Proceedings of the ASME Turbo Expo 2003 (2003, June)

In this contribution, an on-line engine performance monitoring is carried out through an engine health parameter estimation based on several gas path measurements. This health parameter estimation makes ... [more ▼]

In this contribution, an on-line engine performance monitoring is carried out through an engine health parameter estimation based on several gas path measurements. This health parameter estimation makes use of the analytical redundancy of an engine model and therefore implies the knowledge of the engine state. As the latter is a priori not known the second task is therefore an engine state variable estimation. State variables here designate working conditions such as inlet temperature, pressure, Mach number, rotational speeds, . . . Estimation of the state variables constitutes a general application of the Extended Kalman Filter theory, while the health parameter estimation is a classical recurrent regression problem. Recent advances in stochastic methods [1] show that both problems can be solved by two Kalman filters working jointly. Such filters are usually named Dual Kalman Filters. The present contribution aims at using a dual Kalman filter modified to provide robustness. This procedure should be able to cope with as much as 20 to 30% of faulty data. The resulting online method is applied to a turbofan model developed in the frame of the OBIDICOTE 1 project. Several tests are carried out to check the performance monitoring capability and the robustness that can be achieved. [less ▲]

Detailed reference viewed: 66 (0 ULiège)
Full Text
See detailOn-Line Measurement Validation and Performance Monitoring Using Robust Kalman Filtering Techniques
Dewallef, Pierre ULiege; Léonard, Olivier ULiege

in 5th European Conference on Turbomachinery - Fluid Dynamics and Thermodynamics - Conference Proceedings (2003, March)

In an on-line engine performance monitoring, two tasks must be simultaneously carried out: an on-line engine health parameter estimation and an on-line engine state variables estimation. Estimation of the ... [more ▼]

In an on-line engine performance monitoring, two tasks must be simultaneously carried out: an on-line engine health parameter estimation and an on-line engine state variables estimation. Estimation of the state variables constitutes a general application of the Extended Kalman Filter theory, while the health parameter estimation is a classical recurrent regression problem. Recent advances show that both problems can be solved by two Kalman filters working jointly. Such filters are usually named Dual Kalman Filters. The present contribution aims at using a modified dual Kalman filter to provide robustness. This procedure should be able to cope with as much as 20 to 30% of faulty data. The resulting on-line validation method has been applied to a turbofan model developed in the frame of the OBIDICOTE 1 project. Several tests have been carried out to check the performance monitoring capability and the robustness that can be achieved [less ▲]

Detailed reference viewed: 40 (2 ULiège)
Full Text
See detailOn-Line Validation of Measurements on Jet Engines Using Automatic Learning Methods
Dewallef, Pierre ULiege; Léonard, Olivier ULiege

in Proceedings od the 15th ISABE Conference (2001)

Nowadays, turbine engine tests are processed using an open loop, i.e. the measurements are verified and treated a posteriori, sometimes weeks or months after the end of the test. The purpose of the ... [more ▼]

Nowadays, turbine engine tests are processed using an open loop, i.e. the measurements are verified and treated a posteriori, sometimes weeks or months after the end of the test. The purpose of the present project is to develop a new methodology which enables real time detection of faulty measurements and the suppression of the source of these faults during the test. The validation of the measurements is achieved by a “robust” parameter identification [1]. Such a method is called robust in the sense that it can cope with 20 to 30% of faulty measurements. The robustness is insured by a distribution of the measurement noise, as introduced by Huber [7, 8], that takes into account the possibility of faults. The purpose of a parameter identification is to find the set of parameters which has most likely generated the measurements observed on the process. This leads to an optimisation problem that has to be solved for the parameters. The measurements are linked to the parameters through a non-linear model, leading to a large system of equations for modern jet engines. If no physical model of the process can be made available or if this model is too complex to allow real time validation, automatic learning methods may provide a solution: • either a mathematical representation is generated, directly based on the measurements (online learning), • or a database is first generated, based on the existing (but expensive) physical model, the database being subsequently used to build a statistical model (off-line learning). Neural networks seem to be very suitable for modeling the behavior of turbojets, avoiding the resolution of a time-consuming non-linear system. In this paper neural networks are tested to generate a mathematical representation of a single flow, single spool and variable geometry nozzle turbojet, from a data base of “measurements” generated by a physical model of the engine. Only the off-line learning approach is considered. [less ▲]

Detailed reference viewed: 56 (2 ULiège)
Full Text
See detailRobust Validation of Measurements on jet Engines
Dewallef, Pierre ULiege; Léonard, Olivier ULiege

in European Journal of Mechanical and Environmental Engineering (2001), 46(4), 237-242

Nowadays, most of turbine engine tests are processed using an open loop, i.e. the measurements are verified and treated a posteriori, sometimes weeks or months after the end of the test. The scope of the ... [more ▼]

Nowadays, most of turbine engine tests are processed using an open loop, i.e. the measurements are verified and treated a posteriori, sometimes weeks or months after the end of the test. The scope of the present project is to develop a new methodology which enables real time detection of faulty measurements and the suppression of the source of these faults during the test. This implies the development of a method which is sufficiently robust to cope with a lot of faulty data (up to 20 to 30 % of the measurements). Most of the existing methods make use of a least square approach so that a set of parameters is sought which minimizes the RMS error between the measurements and the values provided by a numerical model of the installation. This type of approach is efficient in finding whether a set of measurements is valid but cannot locate the fault(s). In this paper an alternative is proposed based on a distribution of the measurement noise introduced by Huber. This so-called robust validation method has been tested on a single spool, single flow and variable geometry nozzle turbojet. Both the least square and Huber's function approaches have been tested and compared in terms of efficiency. The robust estimator based on Huber's error function has shown to be much more powerful for both finding an invalid set of measurements and locating the faulty measurements. [less ▲]

Detailed reference viewed: 37 (4 ULiège)