Developing Data Mining-Based Prognostic Models for CF-18 Aircraft

Zaluski, Marvin; ́tourneau, Sylvain Le; Bird, Jeff; Yang, Chunsheng

doi:10.1115/gt2010-22944

Cited by 9 publications

(13 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, in some applications, there may be the possibility to have a large amount of data, as in Ref. [25] where the Weibull distribution may be obtained directly. At this stage of the procedure, the methodology results calibrated on the available data trends.…”

Section: Introductionmentioning

confidence: 99%

“…In the same area, Lipowsky et al [24] present a novel statistical method called Bayesian Forecasting as an instrument to forecast gas turbine performance, in order to fit historical data. Zaluski et al [25] develop a data mining methodology to build prognostic models from operational and maintenance data to predict the failures of a bearing and main fuel control of CF-18. Li and Nilkitsaranont [15] develop both a linear and a quadratic regression technique to predict the remaining useful life of gas turbine engines.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Prediction Reliability of a Statistical Methodology for Gas Turbine Prognostics

Venturini

Puggina²

2012

Journal of Engineering for Gas Turbines and Power

View full text Add to dashboard Cite

The performance of gas turbines degrades over time and, as a consequence, a decrease in gas turbine performance parameters also occurs, so that they may fall below a given threshold value. Therefore, corrective maintenance actions are required to bring the system back to an acceptable operating condition. In today’s competitive market, the prognosis of the time evolution of system performance is also recommended, in such a manner as to take appropriate action before any serious malfunctioning has occurred and, as a consequence, to improve system reliability and availability. Successful prognostics should be as accurate as possible, because false alarms cause unnecessary maintenance and nonprofitable stops. For these reasons, a prognostic methodology, developed by the authors, is applied in this paper to assess its prediction reliability for several degradation scenarios typical of gas turbine performance deterioration. The methodology makes use of the Monte Carlo statistical method to provide, on the basis of the recordings of past behavior, a prediction of future availability, i.e., the probability that the considered machine or component can be found in the operational state at a given time in the future. The analyses carried out in this paper aim to assess the influence of the degradation scenario on methodology prediction reliability, as a function of a user-defined threshold and minimum value allowed for the parameter under consideration. A technique is also presented and discussed, in order to improve methodology prediction reliability by means a correction factor applied to the time points used for methodology calibration. The results presented in this paper show that, for all the considered degradation scenarios, the prediction error is lower than 4% (in most cases, it is even lower than 2%), if the availability is estimated for the next trend, while it is not higher than 12%, if the availability is estimated five trends ahead. The application of a proper correction factor allows the prediction errors after five trends to be reduced to approximately 5%

show abstract

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Prediction Reliability of a Statistical Methodology for Gas Turbine Prognostics

Venturini

Puggina²

2012

Journal of Engineering for Gas Turbines and Power

View full text Add to dashboard Cite

show abstract

“…However, in some applications, it is possible to have a large amount of data, as in Ref. [31] where the Weibull distribution may be obtained directly. At this stage of the procedure, the methodology has been calibrated on the available data trends.…”

Section: Prognostic Methodologymentioning

confidence: 99%

“…In the same area, Lipowsky et al [30] present a novel statistical method called Bayesian forecasting as an instrument to forecast gas turbine performance, in order to fit historical data. Zaluski et al [31] develop a data mining methodology to build prognostic models from operational and maintenance data to predict the failures of a bearing and of the main fuel control of CF-18. Li and Nilkitsaranont [14] develop both a linear and a quadratic regression technique to predict the remaining useful life of gas turbine engines.…”

Section: Introductionmentioning

confidence: 99%

Application of a Statistical Methodology for Gas Turbine Degradation Prognostics to Alstom Field Data

Venturini

Therkorn

2013

Journal of Engineering for Gas Turbines and Power

View full text Add to dashboard Cite

In this paper, a prognostic methodology is applied to gas turbine field data to assess its capability as a predictive tool for degradation effects. On the basis of the recordings of past behavior, the methodology provides a prediction of future performance, i.e., the probability that degradation effects are at an acceptable level in future operations. The analyses carried out in this paper consider two different parameters (power output and compressor efficiency) of three different Alstom gas turbine power plants (gas turbine type GT13E2, GT24, and GT26). To apply the prognostic methodology, site specific degradation threshold values were defined, to identify the time periods with acceptable degradation (i.e., higher-than-threshold operation) and the time periods where maintenance activities are recommended (i.e., lower-than-threshold operation). This paper compares the actual distribution of the time points until the degradation limit is reached (discrete by nature) to the continuously varying distribution of the time points simulated by the probability density functions obtained through the prognostic methodology. Moreover, the reliability of the methodology prediction is assessed for all the available field data of the three gas turbines and for two values of the threshold. For this analysis, the prognostic methodology is applied by considering different numbers of degradation periods for methodology calibration and the accuracy of the next forecasted trends is compared to the real data. Finally, this paper compares the prognostic methodology prediction to a “purely deterministic” prediction chosen to be the average of the past time points of higher-than-threshold operations. The results show that, in almost all cases, the prognostic methodology allows a better prediction than the “purely deterministic” approach for both power and compressor efficiency degradation. Therefore, the prognostic methodology seems to be a robust and reliable tool to predict gas turbine power plant “probabilistic” degradation

show abstract

“…We use the training dataset of each component to build models following the steps described above. This involved a large-scale computational effort that we have automated using an in-house tool named EBM3 1 [19]. In Table 3, the models The second experiment is to perform fault identification in order to evaluate the usefulness of the validated FMEA for model ranking.…”

mentioning

confidence: 99%

APU FMEA Validation and Its Application to Fault Identification

Yang

Létourneau

Zaluski

et al. 2010

Volume 3: 30th Computers and Information in Engineering Conference, Parts a and B

Self Cite

View full text Add to dashboard Cite

FMEA (Failure Mode and Effects Analysis) is a systematic method to characterize product and process problems. As a standard document, an FMEA is produced during the design of a product or system. However, once a system is deployed, the corresponding FMEA is rarely validated and updated. This is mainly due to the lack of method to validate and update FMEA. This paper argues that historical maintenance and operational data could be used to help address this problem. Building on data mining and database techniques, the paper introduces a FMEA validation and update technique. The proposed technique derives statistics from real world historical operation and maintenance data and uses these statistics to update key FMEA parameters such as Failure Rate and Failure Mode Probability. The paper then shows how the validated FMEA can be used with data mining for fault detection and identification of root contributing component for a given failure mode or failure effect. The paper presents the developed methodology for FMEA validation and experimental results for fault identification. The results show that the updated FMEA can provide more reliable and accurate information that could benefit the decision-making process and improve maintenance efficiency.

show abstract

Developing Data Mining-Based Prognostic Models for CF-18 Aircraft

Cited by 9 publications

References 9 publications

Prediction Reliability of a Statistical Methodology for Gas Turbine Prognostics

Prediction Reliability of a Statistical Methodology for Gas Turbine Prognostics

Application of a Statistical Methodology for Gas Turbine Degradation Prognostics to Alstom Field Data

APU FMEA Validation and Its Application to Fault Identification

Contact Info

Product

Resources

About