Netjasov, Fedja T. - Prvi autor
Vujosevic, Mirko - Drugi autori
BABIC, Obrad J. - Drugi autori
Janic, Milan M. - Drugi autori
Blom, Henk - Drugi autori
Tosic, Vojin S. - Mentor

656.7

Monografska publ./Monography

English/engleski

[F. Netjasov]

2010

Beograd

XII, 179 listova : graf. prikazi, tabele ; 30 cm

Umnozeno za odbranu. - Univerzitet u Beogradu, Saobracajni fakultet, odbranjeno 20.07.2010. - Bibliografija: listovi 59-64. - Summary.
- - Risk analysis and safety assessment of air traffic control system [Elektronski izvor]. - 1 elektronski opticki disk (CD-ROM) ; 12 cm

Existing international regulations and policies in the field of air transport do not consider the effect of an airborne safety net in the analysis of safety risks. This widely accepted practice tends to create significant tension between the realization of the ambitious safety improvement targets of SESAR and NextGEN, and the current regulations. In order to close this gap there is need for the systematic development of risk and safety assessment of airborne safety nets within the specific Air Traffic Control/Management (ATC/ATM) context, which may range from current practices to advanced ATC/ATM concepts. The aim of the research described in this dissertation is to make a contribution through thesystematic development of an unambiguous model of ACAS operations, i.e. TCAS II version 7,together with its interactions with pilots and air traffic controllers. The model is intended forrisk/safety assessment of ACAS operations, which would allow for the assessment of the possiblebenefit of ACAS in risk reduction in current and advanced ATC/ATM. The specific modelling formalism used for this purpose is Stochastically and Dynamically Coloured Petri Nets (SDCPN). It was shown that the SDCPN representation is very powerful and allows the modeller to represent all elements of such a complex system (technical elements, pilots, air traffic controllers, procedures in force), dynamical and stochastic system, as well as interactions between them, in a flexible and modular way. The model is built in a way to be easily added to certain previously developed SDCPN model of existing or future operational concept, enabling for the assessment of the possible benefit of the safety net in it. The model also could be used as standalone. The fact that the model is stochastic and dynamic enables its use in Monte Carlo simulation of rare events such as aircraft collisions, repeating the experiments many times (order of magnitude 1010) and tracing back in order to determine the chain of events leading to collision. However, in this research this kind of application, i.e. risk assessment itself was not performed. Apart from model development, attention was given to model validation in order to build a confidence in the model and to use it as a credible tool for safety assessment purposes of current and future ATM operational concepts. A validation process is proposed in this research and is applied to historical data, in which a set of different real life encounters were compared to simulation outputs of the developed SDCPN-based model. Although the number of real life encounters was not great, this kind of validation was very valuable due to the fact that it is rarely made for such a kind of models. It was concluded that the developed model could be used for the intended purpose, i.e. as part of the models for the risk/safety assessment of operational concepts, although only the first two levels of validation (of proposed four) were evaluated as satisfactory. The model application is illustrated on a real life accident. Namely, a collision between Inex Adria DC9 and British Airways Trident 3, which occurred on September 10, 1976, over VOR Zagreb (former Yugoslavia) at FL330. ACAS was not in use at the time of collision. The SDCPN model is demonstrated to work well - if TCAS II had existed at the time of the accidents, it could have prevented a collision. Further research steps are proposed. The developed SDCPN-based model should be applied as standalone in certain characteristic cases (encounters) for risk/safety assessment with the aim of estimating possible benefit of risk reduction in airborne safety nets. Additionally, further improvement of TCAS SDCPN-based model and Conceptual model of TCAS Logic should be considered. Finally, the developed model should be integrated with broader models for risk and safety assessment of existing or future operational concepts, which would show the full benefit of airborne safety nets.