The recent improvements and research on aviation have focused on the subject of aircraft safe flight even in the severe weather conditions. As one type of such weather conditions, aircraft icing considerably has negative effects on the aircraft flight performance. The risks of the iced aerodynamic surfaces of the flying aircraft have been known since the beginning of the first flights. Until recent years, as a solution for this event, the icing conditions ahead flight route are estimated from radars or other environmental sensors, hence flight paths are changed, or, if it exists, anti-icing/de-icing systems are used. This work aims at the detection and identification of airframe icing based on statistical properties of aircraft dynamics and reconfigurable control protecting aircraft from hazardous icing conditions. In this paper, aircraft icing identification based on neural networks is investigated. Following icing identification, reconfigurable control is applied for protecting the aircraft from hazardous icing conditions.