TY - GEN
T1 - A computational model for prediction of pilot manual control behavior in air transport concepts of operation
AU - Dickson, David C.
AU - Pritchett, Amy
AU - Bozan, Anil E.
PY - 2016/1/1
Y1 - 2016/1/1
N2 - This paper describes a computational model of pilot manual control of a transport aircraft in both problematic simple maneuvers and realistic, complex concepts of operation. In this model, the pilot is represented by the combination of an adaptive controller, and a state estimator informed by pilot expectation defined by a previously-developed model-based observer (MBO) that simulates vestibular and visual feedback. As a demonstration, this model will be used to examine both problematic simple maneuvers and NextGen operations that apply new methods of air traffic management during optimized profile descent (OPD). The NextGen operations will specifically examine aspects of air traffic control clearances and flight conditions that might be precursors to flight upset. For purposes of comparison, this work will contrast a representation of the pilot as a manual controller against a simulated autoflight system. Both controllers will apply a model reference adaptive controller (MRAC), with the difference between the two comprising the reference model within the MRAC. Specifically, the manual controller will use the classical crossover model as its reference model on all control loops, whereas the autoflight system will use first-order and second-order reference models. Additionally, this work will employ these models in the cases of pilot feedback being represented as either a.) “perfect” and full-state, and b.) imperfect, with pilot distractions being simulated via the MBO. These models will then be applied in a.) simple, potentially spatial disorientation (SD)-inducing maneuvers, and b.) simulations of OPD with the following succeeding levels of complexity: i.) Nominal descent, ii.) Descent with “problematic” air traffic controller commands, and iii.) Extreme variants of descent with “problematic” air traffic controller commands, where the total energy state of the aircraft varies significantly over the course of the simulation run.
AB - This paper describes a computational model of pilot manual control of a transport aircraft in both problematic simple maneuvers and realistic, complex concepts of operation. In this model, the pilot is represented by the combination of an adaptive controller, and a state estimator informed by pilot expectation defined by a previously-developed model-based observer (MBO) that simulates vestibular and visual feedback. As a demonstration, this model will be used to examine both problematic simple maneuvers and NextGen operations that apply new methods of air traffic management during optimized profile descent (OPD). The NextGen operations will specifically examine aspects of air traffic control clearances and flight conditions that might be precursors to flight upset. For purposes of comparison, this work will contrast a representation of the pilot as a manual controller against a simulated autoflight system. Both controllers will apply a model reference adaptive controller (MRAC), with the difference between the two comprising the reference model within the MRAC. Specifically, the manual controller will use the classical crossover model as its reference model on all control loops, whereas the autoflight system will use first-order and second-order reference models. Additionally, this work will employ these models in the cases of pilot feedback being represented as either a.) “perfect” and full-state, and b.) imperfect, with pilot distractions being simulated via the MBO. These models will then be applied in a.) simple, potentially spatial disorientation (SD)-inducing maneuvers, and b.) simulations of OPD with the following succeeding levels of complexity: i.) Nominal descent, ii.) Descent with “problematic” air traffic controller commands, and iii.) Extreme variants of descent with “problematic” air traffic controller commands, where the total energy state of the aircraft varies significantly over the course of the simulation run.
UR - http://www.scopus.com/inward/record.url?scp=84957991237&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84957991237&partnerID=8YFLogxK
M3 - Conference contribution
SN - 9781624103872
T3 - AIAA Modeling and Simulation Technologies Conference
BT - AIAA Modeling and Simulation Technologies Conference
PB - American Institute of Aeronautics and Astronautics Inc, AIAA
T2 - AIAA Modeling and Simulation Technologies Conference, 2016
Y2 - 4 January 2016 through 8 January 2016
ER -