USING ARTIFICIAL NEURAL NETWORKS AND SELF-ORGANIZING MAPS FOR DETECTION OF AIRFRAME ICING

A method of using Artificial Neural Networks (ANNs) and Kohonen Self-Organizing Maps (SOMs) to detect ice on a horizontal tail is proposed and investigated. It is hypothesized that ANN systems trained on the aircraft dynamics would converge to different connection weights for iced and clean aircraft. Kohonen SOMs are used to detect these differences automatically and therefore identify the configuration as being clean or being iced. The system is shown to be capable of acting in an advisory role for the flight crew. The fidelity of the approach is shown to depend on the level of atmospheric turbulence, as well as the magnitude of the elevator input NOMENCLATURE Subscripts denote derivatives, unless stated otherwise. [A] state matrix [B] control matrix

[1]  Ajoy Kanti Ghosh,et al.  Estimation of Aircraft Lateral-Directional Parameters Using Neural Networks , 1998 .

[2]  Geoffrey E. Hinton,et al.  Learning internal representations by error propagation , 1986 .

[3]  Geoffrey E. Hinton,et al.  Schemata and Sequential Thought Processes in PDP Models , 1986 .

[4]  Michael Papadakis,et al.  Experimental study of simulated ice shapes on a NACA 0011 airfoil , 1999 .

[5]  Frederic M. Hoblit,et al.  Gust Loads on Aircraft: Concepts and Applications , 1988 .

[6]  G. Gregorek,et al.  Airfoil aerodynamics in icing conditions , 1986 .

[7]  Sam Lee,et al.  EFFECTS OF SIMULATED-SPAN-WISE-ICE SHAPES ON AIRFOILS: EXPERIMENTAL INVESTIGATION , 1999 .

[8]  Judith F. Van Zante,et al.  Investigation of Dynamic Flight Maneuvers With an Iced Tailplane , 1999 .

[9]  T. Teichmann,et al.  Dynamics of Flight: Stability and Control , 1959 .

[10]  James L. McClelland,et al.  Parallel distributed processing: explorations in the microstructure of cognition, vol. 1: foundations , 1986 .

[11]  Gregory J. Hancock An introduction to the flight dynamics of rigid aeroplanes , 1995 .

[12]  R. J. Ranaudo,et al.  Icing effects on aircraft stability and control determined from flight data: Preliminary results , 1993 .

[13]  Jaiwon Shin,et al.  Experimental and computational ice shapes and resulting drag increase for a NACA 0012 airfoil , 1992 .

[14]  J. G. Batterson,et al.  Determination of longitudinal aerodynamic derivatives using flight data from an icing research aircraft , 1989 .

[15]  James E. Steck,et al.  Application of an artificial neural network as a flight test data estimator , 1995 .

[16]  W. C. Schinstock,et al.  The Measurement of Aircraft Performance and Stability and Control After Flight Through Natural Icing Conditions , 1986 .

[17]  Judith Foss VanZante,et al.  In-Flight Aerodynamic Measurements of an Iced Horizontal Tailplane , 1999 .

[18]  Robert J. Shaw,et al.  Predictions of airfoil aerodynamic performance degradation due to icing , 1990 .

[19]  Laurene V. Fausett,et al.  Fundamentals Of Neural Networks , 1993 .

[20]  Kumpati S. Narendra,et al.  Identification and control of dynamical systems using neural networks , 1990, IEEE Trans. Neural Networks.

[21]  James G. Batterson,et al.  Estimation of longitudinal stability and control derivatives for an icing research aircraft from flight data , 1989 .

[22]  P. Smolensky,et al.  Neural and conceptual interpretation of PDP models , 1986 .