Community Research and Development Information Service - CORDIS


MONSTER Report Summary

Project ID: 14367
Funded under: FP5-GROWTH
Country: Italy

Airplane footprint

In civil aviation, the impact of aircraft noise is a prime issue receiving more regulatory and technological attention than any other aviation environmental problem. It is currently one of main bottlenecks for further growth, that constraints aircraft noise emissions during specific manoeuvres around airports.

The result consists in the implementation of a passive (only listening) system, designed, unlike electromagnetic radar or ultrasonic sonar, to allow the identification of airplane types and manoeuvres by processing only the sound emissions. The system is composed by an algorithm for the acoustic signature identification and a dedicated neural network classifier, trained with a set of experimental aircraft noise data collected @ Naples airport of Capodichino.

The applied method for aircraft acoustic signature identification employs a wavelet multiresolution analysis of noise signals and a statistical analysis of the noise events of each aircraft class. This investigation plays a crucial role to learn the system classifier, a dedicated neural network, with features parameters of reasonable size and condense, at the same time, all the peculiar characteristics of each aircraft noise. The developed system processes noise time histories of airplanes, giving as output the identified airplane and manoeuvre, together with an index of the percentage of successful identification. It s an on-line system for noise events processing and requires a collection of aircraft noise events to learn the system classifier. The algorithm input is the noise time history of the airplane to be identified; identification task is possible only for airplane types and manoeuvres for which the classifier has been trained.

The algorithm has been developed under MATLAB programming environment with a friendly Graphical User Interface. It is a stand alone executable application that requires only Matlab runtime installation on the target machine; it needs MATLAB license with Neural Network toolbox only for new classifier implementation (i.e. trained with a new or extended set of experimental noise data).

The algorithm test and validation have been performed for different airplanes during take-off and landing manoeuvres. For this purpose both numerical simulation and experimental measurement of aircraft noise emissions have been analysed. A preliminary evaluation of the developed algorithm for acoustic signature recognition has been numerically performed by simulating different airplane noise sources.

Finally an experimental activity of ground noise measurements has been carried out at Naples airport of Capodichino. More than 200 aircraft noise events of five aircraft types (Airbus A320, Boeing B737, Mc Donnel Douglas MD80, Fokker F100, Aerospatiale/Alenia ATR72), during both take off and landing manoeuvre, have been measured. It can be applied as support for the radar monitoring of the airports and for the verification of compliance with limitations of annoyance in the area around them. It can allow the surveillance of isolated or dangerous areas and the verification of compliance with peace agreements ("no-fly zone").

Reported by

81043 CAPUA
Follow us on: RSS Facebook Twitter YouTube Managed by the EU Publications Office Top