Aircraft Damage Pattern Recognition Using Aerodynamic Coefficients and Fuzzy Logic

Uncertainties in aircraft structures modeling and measurements always lead to unreliable results of damage detection and estimation due to the lack of sufficient information of the reality. In this paper, a fuzzy logic system is designed and formulated to monitor and assess the structural damage of the aircraft in a way that even though the information is influenced by uncertainties, it can still find a balance between accuracy and practical use. The structural damage is modeled as different scales of tip loss of one of the three main aerodynamic surfaces: horizontal stabilizers, vertical tail and wings. For normal aircraft there are no onboard sensors that are directly related to the integrity of the structure, therefore in this paper the identified aerodynamic coefficients are adopted as damage indicators as well as the inputs to the fuzzy inference system (FIS). The implicit relationship between these indicators and structural damage has been investigated by a series of wind-tunnel and computational experiments. The performance of the proposed method is evaluated at different levels of noises, and it is shown that even the success rate of the damage pattern recognition decrease when higher noise level is applied, the fuzzy logic system can still give satisfactory results.