Optimal attitude and flight vector recovery for large transport aircraft

Engelbrecht, J. J. K. (2017-12)

Thesis (MScEng)--Stellenbosch University, 2017.

Thesis

ENGLISH ABSTRACT: Loss of control (LOC) is the largest contributor to commercial jet aircraft fatal accidents worldwide. Aircraft upset conditions are a primary cause leading to LOC situations. Despite flight envelope protection systems, a need exists for an automatic system to assist the pilot in recovering from a flight envelope upset condition. This thesis presents the design and implementation of an attitude and flight vector recovery system for large transport aircraft. The upset recovery system consists of two major components, namely an optimal trajectory planning component and a practical trajectory execution component. For the optimal trajectory planning, the upset recovery problem is formulated as an optimal control problem and is solved using two different optimal control algorithms, namely dynamic programming (DP) and sequential quadratic programming (SQP). For the trajectory execution, four different control schemes are investigated that use a conventional fly-by-wire flight control system in different configurations to control the aircraft to practically execute the planned optimal trajectory. The attitude and flight vector recovery system was verified in simulation on the NASA Generic Transport Model (GTM), a wide-envelope aircraft model that is able to model the flight mechanics of large transport aircraft in out-of-envelope conditions. The simulation results show that the trajectory planning component generates kinematically feasible optimal upset recovery trajectories, and that the trajectory execution component successfully controls the aircraft to follow the planned trajectories using a representative flight control system. The SQP trajectory optimisation algorithm proposed in this thesis also improves on the dynamic programming algorithm used in previous research, because it is able to use a more representative model of the aircraft dynamics that includes the inner-loop controller dynamics and the engine lag dynamics.

AFRIKAANSE OPSOMMING: Verlies van beheer (“Loss of control” of LOC) is wêreldwyd die grootste bydraende faktor tot noodlottige ongelukke van kommersiële vliegtuie. Ongewone vlugtoestande is ’n primêre oorsaak van verlies van beheer (LOC) situasies. Ten spyte van vlug-omhullende beveiligingstelsels, bestaan daar steeds ’n behoefte vir ’n outomatiese stelsel om die vlieënier te help om die vlug-omhullende te herstel. Hierdie tesis beskryf die ontwerp en implementering van ’n oriëntasie en vlugvektor herstel stelsel vir groot kommersiële passassiersvliegtuie aan. Die vlugherstel stelsel bestaan uit twee hoofkomponente, naamlik ’n optimale trajekbeplanning komponent en ’n praktiese trajekuitvoering komponent. Vir die optimale trajekbeplanning word die probleem as ’n optimale beheer probleem geformuleer, en word deur twee verskillende optimale beheer algoritmes opgelos, naamlik dinamiese programmering (“dynamic programming (DP)”) en sekwensiële kwadratiese programmering (“sequential quadratic programming (SQP)”). Vir die trajekuitvoering, word vier verskillende beheerskemas ondersoek. Hierdie beheerskemas maak gebruik van ’n konvensionele “fly-by-wire” vlugbeheerstelsel in verskillende konfigurasies om die vliegtuig te beheer om prakties die beplande optimale trajek uit te voer. Die outomatiese oriëntasie en vlugvektor herstel stelsel is in simulasie op die NASA “Generic Transport Model (GTM)” geverifeer. Die GTM ’n wye-omhullende vliegtuigmodel wat in staat is om die vlugmeganika van groot transport vliegtuie in toestande buite die normale vlug-omhullende te modelleer. Die simulasie resultate wys dat die trajekbeplanning komponent realistiese optimale hersteltrajekte genereer, en dat die trajekuitvoering komponent die vliegtuig suksesvol beheer om die beplande trajekte uit te voer deur gebruik te maak van ’n verteenwoordigende vlugbeheerstelsel. Die SQP trajekoptimisering algoritme wat in hierdie tesis voorgestel word, verbeter ook op die dinamiese programmering algoritme wat in vorige navorsing gebruik is, omdat dit ’n meer verteenwoordigende model van die vliegtuig dinamika kan gebruik wat die binnelus beheerder dinamika en die enjin naloop dinamika insluit.

Please refer to this item in SUNScholar by using the following persistent URL: http://hdl.handle.net/10019.1/102934
This item appears in the following collections: