A Framework to Predict Failures for Ground Tests on Aircrafts

Ground testing is an essential process in the manufacturing cycle of an aircraft. Testing requires a large investment of resources and time on the part of the engineers of the company. It is common for failures to occur during the different tests during this process. Our group at the Electronic Technology Department of the University of Seville (Spain) worked with the Airbus Defense and Space Company (Airbus DS) to develop a framework for the prediction of failures (known in the terminology of the company as “incidences”) generated in the ground testing of the A400M aircraft. The prediction models are generated through a data mining process with the simulated data registered by the company on its testing process of the previous aircraft. In addition, the framework not only predicts incidences but also suggests changes in the test parameters so that these incidences will not occur again in future aircraft. A battery of tests on real cases was carried out to validate the different models. The success rate of the models was 87.5%, which is considered a good result by the company. In addition, the changes suggested by the framework to ensure that the incidences did not reoccur were studied and executed by Airbus DS in the tests. Currently, based on the framework developed with simulated data, Airbus is using the framework with the actual data of its aircraft.

[2]  N. A. Khovanova,et al.  Decision tree and random forest models for outcome prediction in antibody incompatible kidney transplantation , 2017, Biomed. Signal Process. Control..

[3]  Arjana Brezigar-Masten,et al.  CART-based selection of bankruptcy predictors for the logit model , 2012, Expert Syst. Appl..

[4]  Marc Böswald,et al.  Dynamic characterization of the A400M acoustics fuselage demonstrator , 2016 .

[5]  Ian H. Witten,et al.  Data mining: practical machine learning tools and techniques, 3rd Edition , 1999 .

[6]  Elcio Hideiti Shiguemori,et al.  Identification and Classification of Drop Zones and Helicopter Landing Zones in Images Obtained by Small Size Remotely Piloted Aircraft Systems , 2018, IGARSS 2018 - 2018 IEEE International Geoscience and Remote Sensing Symposium.

[7]  Ibrahim Ziedan,et al.  Machine Learning Methods for Spacecraft Telemetry Mining , 2019, IEEE Transactions on Aerospace and Electronic Systems.

[8]  A. Roudbari,et al.  EGT prediction of a micro gas turbine using statistical and artificial intelligence approach , 2018, IEEE Aerospace and Electronic Systems Magazine.

[9]  Eduardo S. Ogasawara,et al.  A Review on Flight Delay Prediction , 2017, ArXiv.

[10]  David Gil Méndez,et al.  Predicting seminal quality with artificial intelligence methods , 2012, Expert Syst. Appl..

[11]  Victor A. Skormin,et al.  Data mining technology for failure prognostic of avionics , 2002 .

[12]  Qi Xiao-lin.  Design of Avionic Fault Diagnosis Expert System Based on 1553B Bus , 2009 .