In recent years, both globally and in Poland, there has been an increased interest in air transport. This interest is visible not only among passengers but primarily among the constantly growing number of aviation personnel. For example, at the end of 2006, the number of active licenses was 3717 for airline, commercial, and private pilots, both for airplanes and helicopters. Over the next 4 years, the number of personnel increased by 376 pilots, while at the end of 2011, there were already 4717 valid licenses. As can be observed, this number doubled in the last year compared to the previous 4 years [1].
Along with the constantly growing number of pilots, air traffic intensity has increased, as has the number of newly registered aircraft types. Furthermore, with the developing air transport technology, requirements regarding safety and transport economics are increasing and must be met.
In summary, it becomes essential to map (with a high degree of accuracy) actual flights using a mathematical model. One of the methods allowing for such high accuracy is the use of artificial neural networks. In the conducted research, in order to obtain very good accuracy in mapping the actual flight, results of analyses conducted on real data, sourced from an onboard flight recorder, were used.
2. DATA FROM ONBOARD FLIGHT RECORDERS
The data used for modeling came from the “black box” records of an Embraer 170 aircraft. This is a passenger jet aircraft produced by the Brazilian company Empresa Brasileira de Aeronautica S.A. since February 19, 2002. In Poland, it has been in use since March 2004, mainly by PLL LOT, which operates ten EMB 170 units. Four of them are in the 170ST version, and the remaining six are in the 170LR version. (…)
The article contains 12007 characters.
Source: Logistyka Magazine 4/2012
Artificial neural networks applied to modeling the aircraft climb phase – PDF file for download







