Data processing: vehicles – navigation – and relative location – Vehicle control – guidance – operation – or indication – Aeronautical vehicle
Reexamination Certificate
2003-05-06
2004-05-11
Jeanglaude, Gertrude A. (Department: 3661)
Data processing: vehicles, navigation, and relative location
Vehicle control, guidance, operation, or indication
Aeronautical vehicle
C701S001000, C701S013000, C244S017130, C244S075100
Reexamination Certificate
active
06735500
ABSTRACT:
FIELD OF THE INVENTION
This invention relates to tactile cueing flight control systems, and, more particularly, to flight control systems incorporating a plurality of neural networks to predict the most limiting flight envelope parameters and provide tactile cueing to a pilot.
BACKGROUND OF THE INVENTION
Aircraft warning systems generally require some form of state prediction. The states of an aircraft will vary greatly depending, however, upon the type and model of the aircraft. For example, the states of a rotorcraft may include torque parameters, thrust parameters, turbine temperatures, etc. Typically, the estimates of the states of an aircraft are derived from measurements provided by various sensors and knowledge of the dynamics of the aircraft.
One application of aircraft warning systems is to identify state limits such that the aircraft remains within a desired flight envelope. In this regard, tactile cueing systems have been used to avoid state limits in selected aircraft, but with limited applicability. Some cueing systems cue only with respect to avoidance of a single limiting factor or for only single maneuvers. While one state may be a limiting factor for many flight operations, other states may be more limiting in particular maneuvers. Single factor cueing systems, therefore, are not readily applicable to all aircraft maneuvers. Furthermore, some cueing systems depend only upon current aircraft states without respect to future estimated states. Without the lead time, provided by consideration of future estimated states these systems often fail to adequately ensure avoidance of operating limits.
Therefore, there is a need in the art to achieve sufficient accuracy in predicting aircraft states in dynamic maneuvers with enough lead time to provide the pilot with an effective tactile cueing system. In particular, a tactile cueing system should be robust enough to predict aircraft states and provide limit avoidance cues for maneuvers throughout different flight regimes.
SUMMARY OF THE INVENTION
A flight control system, a method, and a computer program product for determining tactile cueing of a flight control input apparatus are provided. Generally, pluralities of neural networks provide predicted and expected performance results of an aircraft in real time. These predictions and expected performance results are compared to the limits of the flight operating envelope for a particular aircraft. The most limiting of these parameters is identified and a corresponding position of the flight control input apparatus may be determined in order to provide a “stop” position to prevent the pilot from exceeding the flight envelope. In this regard, stop positions may be “softstops”, wherein further flight control movement is impeded, but not prevented, in order to provide the pilot with a cue that the flight envelope may be exceeded with further movement. Alternatively, a “stop” position may also comprise a hardstop, beyond which no further movement is permitted.
Neural networks are particularly advantageous to use when implementing a tactile cue to a flight control input apparatus because they can perform the sophisticated, nonlinear, multivariable, real-time calculations required to generate smooth and effective tactile cues within the processing constraints of current generation on-board flight control computers. It has been found that neural networks predict aircraft performance more accurately than alternative techniques, such as using simplified “classical” rotor performance equations, and are possible to implement in an on-board flight control computer. Neural networks are especially effective for envelope limit protection by tactile cueing because they do not exhibit the gross inaccuracies that are characteristic of classical equations during edge-of-the-envelope operation. In fact, neural network predicted state information supplemented by sensor measurements allows envelope limit encroachment to be predicted with adequate lead time to allow either the pilot or flight control system to act to prevent or mitigate the consequences of limit impingement.
To this end, a system, a method and a computer program product for determining tactile cueing of a flight control input apparatus are described herein. The system includes an interface for receiving observed parameters relating to the flight envelope of an aircraft. These observed parameters relate to an aircraft state and other states requiring limit avoidance. Upstream processing elements receive the observed states for filtering and provide the data to predictive neural networks. Several of the upstream processing elements convert dimensional aircraft state parameters into nondimensional aircraft state parameters. The predictive neural networks receive combinations of observed parameters and nondimensional rotor state parameters and predict flight envelope limiting parameters based thereupon. The neural networks are advantageous in this regard as they are generally capable of being trained to predict the limiting output based on a set of training data. A downstream processing element determines the most limiting flight envelope limiting parameter provided from the neural networks. In determining the most limiting parameter, a tactile cueing position is then provided for a flight control input apparatus, thereby providing limit avoidance.
REFERENCES:
patent: 5403155 (1995-04-01), Head et al.
patent: 5499785 (1996-03-01), Roberts et al.
Schaeffer et al. (US 2003/0094539 disclose a power lever tactile cueing system.*
Miller, Einthoven, Morse & Wood, “HACT Flight Control System (HFCS) Control Law Overview”, Presented at the American Helicopter Society 58thAnnual Forum, Montreal, Quebec, Canada, Jun. 11-13, 2002.
Nicholas, Clark, Ward, Jeram, & Prasad, “Development of Artificial Neural Networks for the HACT Flight Control System”, Presented at the American Helicopter Society 58thAnnual Forum, Montreal, Canada, Jun. 11-13, 2002.
Jeram, “Open Design For Helicopter Active Control Systems”, Paper written for the Georgia Institute of Technology, School of Aerospace Engineering, Atlanta, GA.
Horn, Calise, Prasad & O'Rourke, “Flight Envelope Cueing on a Tilt-Rotor Aircraft Using Neural Network Limit Prediction”, Presented at the American Helicopter Society 54thAnnual Forum, Washington, DC, May 20-22, 1998.
Einthoven, Miller, Nicholas & Margetich, “Tactile Cueing Experiments with a Three Axis Sidestick”, Presented at the American Helicopter Society 57thAnnual Forum, Washington, DC, May 9-11, 2001.
Calise, Kim, Kam and Azam, “Neural Networks for Feedback Linearization in Aircraft Control”,.
Ward, “Generalized Networks for Complex Function Modeling”, Presented at the 1994 IEEE Systems, Man, & Cybernetics (SMC-94) Conference, Oct. 2-5, 1994; San Antonio, Texas.
Miller David G.
Nicholas Jeffrey S.
Alston & Bird LLP
Jeanglaude Gertrude A.
The Boeing Company
LandOfFree
Method, system, and computer program product for tactile... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method, system, and computer program product for tactile..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method, system, and computer program product for tactile... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3206300