| US20140148980A1 | 2014-05-29 | |||
| US20090048724A1 | 2009-02-19 | |||
| US20080300741A1 | 2008-12-04 | |||
| EP1257987B1 | 2007-12-26 | |||
| US20100070115A1 | 2010-03-18 |
| CLAIMS 1. A system for warning a flight crew of potential runway overruns during landing, comprising: 1. a computer processor implementing at least one neural network comprising plural inputs and at least one output, wherein the inputs comprise at least one or more of the following: pressure altitude, air temperature, wind speed, weight, calibrated airspeed, autobrake setting, runway slope, and runway condition; and using the neural network to continually calculate at least the remaining distance on the runway and to generate an alert when the remaining distance is insufficient. 2. The system of claim 1 wherein the neural network calculates the air and ground landing distances, during the approach. 3. The system of claim 1 wherein the air landing distance during landing phase is obtained by 4. The system of claim 1 wherein the ground landing distance during air landing phase is obtained by is calculated by the neural network at SH position. 5. The system of claim 1 wherein the remaining ground distance during ground landing phase is obtained by where ground speed at the touchdown moment, GS ground speed in the moment of calculation, and curvature coefficient. 6. The system of claim 1 further including a further neural network for calculating remaining ground distance during ground landing phase. 7. The system of claim 1 wherein the processor calculates the position where the aircraft will touch the runway · TD, as 8. The system of claim 1 wherein the processor calculates the position where the aircraft will stop - SP, as defined by: 9. The system of claim 1 further including a control that allows the flight crew to shorten LDA length for operator to increase safety margin or to account for threshold displacement and LDA reduction. |
CROSS-REFERENCE TO RELATED APPLICATIONS
[0001] None.
STATEMENT REGARDING FEDERALLY SPONSORED RESEARCH OR DEVELOPMENT [0002] None.
FIELD
[0003]The present application relates to a method and a device to improving flight crew situational awareness during an aircraft final approach and landing, through landing distance dynamic calculation, and more particularly to the use of neural networks in such contexts.
BACKGROUND
[0004] Statistically, in the decade between 2004 and 2013, the third most common fatal aircraft accident type was due to runway overrun during landing (Statistical Summary of Commercial Jet Airplane Accidents - Worldwide Operations 1959 - 2013, Aviation Safety - Boeing Commercial Airplanes, Seattle - Washington - US 2014). Minimizing or avoiding such accidents is an important goal.
BRIEF DESCRIPTION OF THE DRAWINGS
[0005]The following detailed description of exemplary non-limiting illustrative embodiments is to be read in conjunction with the drawings of which:
[0006] Figure 1 shows a basic non- limiting system architecture, which represents the diagram of a system flowing the information up to flight crew and recorders;
[0007] Figure 2 shows an example final approach and landing schematic view, containing variables names and definitions;
[0008] Figure 3 shows an example non-limiting neural network schematic architecture; and [0009] Figure 4 shows an exampte non-limiting Landing Distance Monitor calculation procedure.
DETAILED DESCRIPTION OF EXAMPLE NON-LIMITING EMBODIMENTS
[0010]The reasons for the occurrence of runway overruns during landing can be better understood by making a closer analysis about landing maneuvers. This flight phase requires a high level of pilot attention, as many factors can contribute to an overrun event. As part of the performed tasks one can enumerate monitoring the glide slope, aircraft speed, aircraft configuration, atmosphere and runway conditions, and touch down point, among others. Additionally, some landings are made based on information obtained in a calculation which the inputs are the best information known at the moment when the aircraft took off.
[0011]A known procedure to reduce the possibility of the occurrence of overruns is to continuously recalculate the landing distance before landing the aircraft, in order to account for operational deviations at required landing distance. Some deviations are common contributing factors in landing overrun statistics, e.g., wind speed increase, aircraft overspeed, long flare, and delay on brakes application, among others. However, manual continuous calculation is generally not practicable, and the goal of increasing situational awareness had not been achievable.
[0012] A solution is to automate the calculation and advise pilots when required landing distance is greater than the landing distance available. By doing this, the flight crew is alerted only when deviations can result in an undesirable condition. Some systems exist to perform this function but they generally have either high acquisition costs or low precision due to the calculation method applied
[0013JThe non-limiting method and device presented herein minimizes the risks of a landing overrun and increases flight crew situational awareness. To accomplish these goals, the system actuates in the final approach and landing phases returning automatic alerts to the flight crew when necessary.
[0014] Figure 1 shows an example non-limiting embodiment of a system 10 providing a basic system architecture that flows information from left to right up to a flight crew and recorders. In this example system 10, input parameters 12 and a database 14 are provided to a parameters check 16. If the parameters check 16 indicates that the parameters are okay, the system is activated (block 18) and the system begins monitoring landing distance (block 20). The output of the landing distance monitor 20 is processed by block 22 and is provided to a flight data recorder 24, a display 26 and an aural indicator 28. Meanwhile, if the parameters check 16 indicates that the parameters are not okay (other exit to decision block 16), a signal is provided to alerts processor 22 to provide a visual and aural warning to the flight crew via the display 26 and aural indicator 28. In one example embodiment, the blocks 16, 18, 20, 22 are performed by one or more computer processors executing software instructions stored in non-transitory memory.
[0015]The final approach starts when aircraft is in air and configured to land, and ends where landing begins. In the example non-limiting embodiment, the beginning of the landing maneuver is defined by a reference height, called H REF . See Fig. 2. Horizontal distances SACFT* SRWY and AS are obtained through GPS coordinates distance calculations, using the equation below.
[0016]where
[0017] For a given aircraft position P2, knowing a previous position P1, the system calculates the horizontal distance up to SH - SAPP, as below.
[0018] where
[0019] in order to calculate distances SAIR and S G ND during final approach (from higher heights - for example, 1000ft - up to SH height), one example non-limiting embodiment of system 10 uses a neural network with a said number of inputs and a said number of outputs, based on the performance data for a given airplane (see Figure 3). The number of neurons can vary depending on which airplane the neural network is prepared for. A non-limiting example of inputs and outputs is demonstrated in the table below, as the neural network example in Figure 3:
[0020] During the landing phase (between SH and SP), the remaining air distance is calculated using the equation below.
[0021 ] where calculated by the neural network at SH position.
[0022] During the air landing phase (between SH and TD), the remaining ground distance is calculated as below:
[0023]where calculated by neural network at SH position.
[0024] During the ground landing phase (between TD and SP), two methods can be used to calculate the remaining ground distance. The first one is given by the equation below.
[0025] where
■ ground speed at the touchdown moment, GS = ground speed in the moment of calculation, and
a = curvature coefficient.
[0026]The second method is to obtain a neural network exclusively for remaining ground distance. A non-limiting example of inputs and outputs is demonstrated in the table below.
[0027]The position where landing initiates - SH, is calculated by
[0028]The position where the aircraft will touch the runway - TD, is defined by:
[0029]The position where the aircraft will stop - SP, is defined by:
[0030] The flight crew is alerted when SP trending indicates the stop position will pass LDA (Landing Distance Available) end in t REACT seconds. The parameter t REACT is a manufacturer choice and may be any number, as zero or one, for example.
[0031] Furthermore, the example non-limiting system allows the flight crew to shorten the LDA length for the operator to increase the safety margin (as its option) or accounts for threshold displacement and LDA reduction.
[0032]Moreover, the non-limiting system is prepared to alert the pilot when air landing distance is greater than expected. Timing from the moment in which aircraft fly over SH position, if spent time is higher than
t LF seconds (7s, for example) and aircraft didn't touch the runway yet, the device provides a "long flare" alert.
[0033] An example non-limiting procedure for Landing Distance Monitor 20 calculations can be seen on Figure 4. in this non-limiting embodiment, the processor gets input parameters (102) and calculates the current LDA (104). The processor calculates SP position (106) and its derivative (108). The processor then detects whether SP will exceed LDA in t react seconds. If SP will exceed LDA within the time period, the processor will generate an alert trigger (112). Otherwise, the processor tests whether the air time at landing will be greater than tu (114). If the air time at landing will exceed this value, the processor generates an alert trigger (116). Otherwise, no alerts are generated (118).
[0034] While the invention has been described in connection with what is presently considered to be the most practical and preferred embodiments, it is to be understood that the invention is not to be limited to the disclosed embodiments, but on the contrary, is intended to cover various modifications and equivalent arrangements included within the spirit and scope of the appended claims.