Entropy Statistics and Cluster Analysis Applied to Jet Airliner and Fighter Aircraft Classification
Pp. 133-171 (39)
Bento S. de Mattos, Jose A. T. G. Fregnani and Paulo Eduardo C. S. Magalhaes
The present work describes the entropy statistics theory and its application
to analyze the commercial aviation market and fighter aircraft. Airliners are designed to
comply with certification requirements and to provide operational profitability to
airlines. They need to be more economical, lighter, faster and better than its
predecessors. Understanding the technological evolution of aviation is extremely useful
when one is designing new aircraft. Saviotti (1984) and later, Frenken (1997) proposed
a method of analysis of aircraft technological evolution. This method, based on
information theory developed by Shannon (1948), specially the concept of entropy
statistics, has proved to be very effective for product classification. Thus, a
methodology based on entropy statistics was employed to analyze the jet airliners since
the 1950s and fighter aircraft since World War I. The results show the impact of
notable events such as the 1978 Airline Deregulation Act on the airplane designs and
naturally, the world wars. The airplanes can be classified into four groups: niche, fuzzy,
scaled trajectories and breakthrough designs. Some airplanes were chosen for a deeper
analysis and the entropy statistics methodology could correct label some airplanes like
the ERJ 145 family, the Boeing 777-300, and Airbus 340-500. Fighter airplanes like
the Vought Cutlass, Westland Wyvern, and IA-54 Pucará also could be adequately
categorized and rated.
Aircraft design, Airliner, Airplane evolution, Entropy statistics,
Fighter aircraft, Information theory.
Aircraft Design Department Instituto Tecnologico de Aeronautica (ITA) Sao Jose dos Campos, Sao Paulo Brazil.