A system and a method for automatic air collision avoidance

IS 2 300 684 T3

DESCRIPTION

System and procedure for the automatic avoidance of aerial collisions.

Field of the invention and prior art

The present invention relates to a method for avoiding collisions between aircraft according to the preamble of claim 1.

The invention also relates to a system for automatically avoiding collisions between aircraft according to the preamble of claim 14.

The purpose of an On-Board Anti-Collision System (ACAS) is to prevent collisions in the air between airplanes, each of which incorporates the system. An avoidance is an automatic maneuver carried out to avoid a collision with another aircraft. Once the avoidance maneuver is activated, the maneuvers commanded by the pilot are ignored. Each aircraft in the system continuously computes an escape angle and a load factor to be used by the aircraft during an avoidance maneuver in the event of a looming collision. The escape angle is a relative roll angle. At the same time the plane computes separation trajectories in the air. The separation paths are computed by means of a response model of the aircraft. The separation flight path is a prediction of the space within which the airplane will be located with a certain probability if an avoidance maneuver occurs.

The separation flight path is an airborne path surrounded by a cone-shaped gap. The size of the cone-shaped space surrounding the trajectory depends on imponderables in the prediction of the separation trajectory. The imponderables in the prediction are due for example to the imprecision of the aircraft response model, to the imprecision of the timing of the avoidance activation due to the imprecision in the assumption of when the avoidance maneuver will start, and the maneuver of the last moment. The separation flight path is sent to the other aircraft. When the other aircraft receive a separation flight path, the path is recorded. Thus, the recorded separation paths are known to all neighboring aircraft that incorporate the system.

The aircraft continuously receives the separation paths from other aircraft. The system detects an approaching collision based on the separation flight path itself and the separation paths received from the other aircraft, and after detecting an approaching collision it activates the automatic avoidance maneuver. During the avoidance maneuver the aircraft is ordered to take the escape angle and the computed load factor at the same time as the last recorded separation flight path. A collision is detected when the system detects that its own recorded breakaway flight path traverses a recorded breakaway flight path of another aircraft. The avoidance maneuver must take place within a logged space that is known to the other aircraft. If no collision is detected, the system computes a new escape angle and load factor to be used during an avoidance maneuver, and a new separation flight path based on the separation paths received from the other aircraft. . The new separation flight path is sent to the other aircraft.

A problem in connection with automatic air collision avoidance systems is that in some situations the prediction uncertainties are large. To be sure that the avoidance occurs within the recorded gap of the separation flight path, the width of the recorded path is increased. When the width of the recorded path increases, the risk of disturbance increases. By the term disturbance is meant an event that produces an unintended or unforeseen response or that an activation system to automatically avoid collisions is deactivated and consequently the availability of the system is reduced.

One of the factors affecting the uncertainties is last-minute maneuvering of the aircraft during the time delay between the computation of a detach flight path and the receipt and recording of the detach flight path by the other aircraft. The time that elapses from the computation of the separation flight path until the other aircraft have received the path lasts approximately 3 seconds. This problem is partially addressed in automatic air collision avoidance systems, carrying out a prediction of the aircraft movements in the next 0.3 s, and considering this prediction when computing the separation flight path. This prediction assumes that the pilot's maneuvers during the next 0.3 s are the same as at the time of prediction, that is, that the position of the control lever is fixed. However, this is not always the case; instead, the pilot is free to carry out whatever maneuvers he wants until a collision is detected and the avoidance maneuver begins. In particular, there is a problem if the pilot performs the maneuvers that counteract the recorded separation maneuvers during said time delay. In an extreme circumstance it may no longer be possible for the aircraft to follow the recorded separation flight path, due to the pilot's maneuvers.

Document US-A-6133867 discloses a procedure and a system for avoiding collisions between airplanes, in which airplanes calculate and send their position data every few seconds. The procedure and the system recognize an instruction sent to the local aircraft to carry out some evasive maneuver dictated by an aircraft or

ES 2 300 684 T3 remote ground crew and respond with audible or visual or autopilot instructions to the cockpit.

Objects and summary of the invention

The object of the present invention is, therefore, to provide a solution to the problem described above, which reduces the risk that the aircraft will not be able to follow a recorded trajectory, when detecting an approaching collision and keeping the number low. of disturbances.

According to one aspect of the invention, this object is achieved by a method comprising the characterizing features of claim 1.

According to the invention, an approaching collision is detected based on the separation flight path itself, the separation trajectories received from the other aircraft, as well as the maneuvers of the aircraft itself commanded during a period of time between sending your own computed separation flight path to other aircraft and detecting the upcoming collision. The maneuvers dictated during the period of time between the sending of one's own computed separation flight path to the other aircraft and the detection of a collision that is looming, is hereinafter referred to as last-minute maneuvers. Thus, the point in time when the avoidance maneuver is activated is made dependent on the last moment maneuver of the aircraft itself. Preferably, the activation of the automatic collision avoidance system is advanced when the activation is close in time and the controlled maneuvers of the aircraft itself counteract the separation flight path itself. Thus, the avoidance maneuver is activated when the aircraft is maneuvered in such a way that the possibility that the aircraft can fly within a recorded avoidance path if a collision is detected is improved.

An advantage of this solution is that the uncertainties due to unpredictable movements of the aircraft can be reduced, and in this way the width of the recorded path is reduced, and thus the risk of disturbances is reduced.

According to an embodiment of the invention, the separation trajectories are computed based on the predicted maneuvers of the aircraft itself, and a looming collision is detected based on the deviations between the predicted maneuvers of the aircraft itself and the maneuvers commanded by the aircraft itself that counteracts its own separation flight path. Preferably, an oncoming collision is detected based on the deviations produced between the anticipated roll angles, and the commanded roll angles during said period of time. Last minute maneuvers can cause serious position errors compared to predicted positions. Roll guidelines in particular can cause serious position errors because the roll angle of avoidance is directly affected. The greatest influence on the possibility of following an avoidance path appears when roll guidelines are applied during last-minute maneuvers of the aircraft in a direction that counteracts the roll directive during the avoidance maneuver. This most certainly means shifting the flight path towards the flight path of others. Thus, it is advantageous to base collision detection on roll angle deviations.

According to another embodiment of the invention an approaching collision is detected based on the deviations between the predicted longitudinal tilt angles and the commanded tilt angles during said period of time. Although commanded roll angles have a greater influence on the ability to follow a recorded breakaway flight path, than commanded bank angles the influence of breakaway flight path on longitudinal bank angles should be ignored. This embodiment also improves the ability to fill in the recorded avoidance path.

The invention could be applied in manned aircraft as well as in unmanned aircraft. According to a further embodiment of the invention an approaching collision is detected based on the maneuvers commanded by a pilot, and the method comprises the reception of the movements of the control lever, and the detection of a collision that looms based on the joystick movements received during said time. The object of the invention is that if a collision is detected, there is the assurance that the aircraft itself can adjust to the separation flight path sent to the other aircraft. If the movements of the joystick during the period of time between sending its own separation flight path to the other aircraft and the detection of the impending collision impairs the ability to complete the recorded avoidance path, the maneuver avoidance is immediately activated. This embodiment is suitable for manned aircraft incorporating pilots who issue instructions by means of a joystick.

According to an embodiment of the invention the commanded maneuvers supporting the separation flight path are ignored. Thus, maneuvers commanded by the aircraft itself are not allowed to delay the activation of the collision avoidance system; they can only advance the activation of the system. A roll guideline in the same direction as the escape angle causes the relative roll angle to be used during the avoidance maneuver to be reduced. The load factor is then pre-commanded, which means that a lower load factor is needed to follow the avoidance path. In this way, the commanded maneuvers that support the separation flight path should not contribute to advancing the activation of the avoidance action.

IS 2 300 684 T3

According to an embodiment of the invention, said detection of a looming collision comprises computing an evasion condition based on the separation flight path itself, the separation trajectories received from the other aircraft and the maneuvers of the aircraft itself commanded during said period of time, and the activation of said automatic collision avoidance system, if the avoidance situation is met. The avoidance situation is computed as the sum of a first contribution term, computed based on the separation trajectories received and the separation flight path itself, and a second contribution term, computed based on the maneuvers of the aircraft itself. commanded during that period of time. Whether or not the evasion condition is met is determined depending on whether the evasion condition is below or above a threshold value. The value of the first contribution term depends on the risk of collision. The value of the second contribution term depends on whether or not the airplane has received instructions to carry out unpredictable, unfavorable maneuvers, which may lead to an avoidance maneuver not taking place within a registered space, that is, within of the recorded separation flight path. According to this embodiment, a term computed based on last-minute maneuvers of the aircraft itself is added to the evasion condition. Thus, the moment in which the evasion condition is fulfilled, that is, the moment in which the approaching collision is detected, is made dependent on the last-minute maneuvers of the aircraft itself.

According to an embodiment of the invention, the value of said second term, which is computed based on the last moment maneuvers of the aircraft itself, is made dependent on the risk of collision. This value of the second contribution term is reduced if there is no risk of collision. Whereby, the maximum contribution of this term is limited and the risk of disturbances is reduced.

According to an embodiment of the invention, the invention comprises the addition of roll instructions commanded during said period of time that counteracts the separation flight path itself, and the addition of all the longitudinal inclination instructions commanded during said period of time. time that counteracts the separation flight path itself, and the calculation based on said instructions of the second contribution term. This embodiment further improves the ability to fill in the recorded avoidance path.

According to another object of the invention, this object is achieved by means of a system comprising the characterizing features of claim 14.

According to a further aspect of the invention, the object is achieved by means of a computer program capable of being loaded directly into an internal memory of a computer or of a processor, which comprises portions of software code to carry out the steps of according to the invention, when said program is executed on a computer. The computer program is arranged either on a means of reading by computer or by means of a network, like for example Internet.

According to another object of the invention, the object is achieved by means of a computer reading means that has a program registered on it, when the program consists of getting a computer to carry out the steps of the procedure according to the invention, and said program is executed on the computer.

Brief description of the drawings

The invention will now be explained in greater detail by describing the different embodiments of the invention and with reference to the attached figures.

Fig. 1 shows the separation trajectories of two planes.

Fig. 2 shows a block diagram through a system for automatically avoiding collisions according to the invention.

Fig. 3 shows a flow chart through a method for automatically avoiding collisions according to an embodiment of the invention.

Detailed description of preferred embodiments of the invention

Figure 1 shows two aircraft 1 and 2 each incorporating an automatic collision avoidance system. Each aircraft computes a separation flight path 3, 4. The separation flight path is a path 5, 6 made in the air, which is surrounded by a cone-shaped space 7, 8. The width w of the space cone-shaped 7, 8 surrounding the trajectory 5,6 depends on the uncertainties in the computation of the separation flight path. The uncertainties, and consequently the width of the separation flight path, increases with time. The separation flight path consists of two parts. A first part of the separation flight path comprises a prediction of the aircraft movements based on the assumption that the pilot's maneuvers during the next 0.3 s are the same as at the time of the prediction, that is, that the position of the control lever is fixed. This prediction takes care that the elapsed time between when the break-away flight path is calculated and the other aircraft receives the break-away flight path. The second part of the separation flight path predicts the avoidance maneuver and is based on the computed escape angle.

IS 2 300 684 T3

When one aircraft has computed a separation flight path, information about the path is sent to the other aircraft. The other aircraft records the separation flight path space after receiving it. The trajectory 5, 6 can be described as a function of time, and the information about the separation flight path that is sent to the other aircraft comprises three points in the air, two velocity vectors in the function, and a value representing the uncertainties of computation. The other aircraft's control systems then model a spline function using the received information.

According to this embodiment, the detection of a looming collision means the detection that a computed and recorded separation maneuver cannot be carried out. An oncoming collision is detected when it is detected that the computed breakaway flight path itself collides with any breakaway flight path of the other aircraft. The distance between the separation flight paths is time dependent, and is mathematically described in the following equation:

SSD (t) = | H (t) - l (t) | - DSD (t) (1)

H (t) is the predicted avoidance path 3 of the airplane itself, and l (t) is the predicted separation flight path 4 of an intruding airplane. DSD is the desired spacing distance between the parting paths, and is the sum of the space recorded between the parting paths. The DSD is computed from the uncertainties and represents the size of the recorded space. Uncertainties include computational uncertainties, such as uncertainties in the aircraft response model.

A minimum safety separation distance, hereinafter referred to as MSSD, is defined as the minimum separation distance between the separation paths and appears at time tmm after activation of the avoidance maneuver.

A collision is detected depending on whether or not an evasion condition, designated as EC, is below or above a threshold value. The EC is computed using the MSSD and its rate of change according to the following equation:

EC = MSSD + d (MSSD) '^ -prediction + PIC) (2)

T_prediction is the period of time between two iterations in the computation loop.

A term called Pilot Input Contribution, designated below as PIC, which depends on the aircraft maneuvers is added to the EC computation in order to punish the pilot input that counteracts the avoidance maneuver. The EC comprises a first contribution term that includes the MSSD and its time derivative times T_prediction and a second contribution term computed based on the maneuvers of the aircraft itself commanded during a period of time between the dispatch between the separation flight path itself to the other aircraft and the detection of the collision that is coming. An oncoming collision is detected if the EC is below a threshold value, designated as TV.

It is important that the PIC is conservative, which means that the commanded maneuvers that support the separation flight path do not allow a smaller MSSD, that is, that the maneuvers that support the separation flight path are not allowed to influence. in the EC value and, consequently, influence the moment in which the avoidance maneuver is activated. Thus, one condition is that the ICP must be equal to or equal to zero:

ICP <0 (3)

In order to avoid reducing the effect of disturbances during the avoidance path the upper limit of the ICP should be less than zero. The lower limit of ICP is a function F₁ of t_min and possibly other variables as well. The function F₁ it is close to or equal to zero when there is no risk of collision and has a large negative value in situations of risk of collision. The idea of ​​Fi is to limit the maximum magnitude of the variable PI. If PI is defined to have a maximum value of zero, the PIC can be expressed as:

PIC = max [Fi (tmin), PI) (4)

The PI is the sum total of all pilot input during the delay of last-minute pilot maneuvers that impair the ability to complete the recorded separation flight path. Most of the PI depends on the pilot's roll instructions, and a smaller part depends on the pitch instructions. The PI can be divided into two parts, one that handles the unfavorable roll instructions, that is, the roll instructions that counteract the roll instruction of the avoidance maneuver, and one that handles all the unfavorable roll instructions. If the

ES 2 300 684 T3 roll instructions designated as PIR, as well as longitudinal tilt instructions, designated as PIP, the PI can be expressed as:

PI = PIR + PIP (5)

To fulfill condition 3 above, that is, the PICs must be below zero, both the PIRs and the PIPs must have a maximum value of zero. PIPs can be set to zero if it is considered that the pilot's pitch instructions can be ignored.

In rolling, the idea is to collect all the roll instructions in the wrong direction compared to the roll instruction used during computation of the avoidance path. PIRs can for example be expressed as:

PIR = Ki Σ min [(Pess - Parm). sign (EA), 0]

P_ESS = roll angles commanded by the pilot

P_ARM = roll angles used during computation of the separation flight path

EA = escape angle

Ki = a scale factor.

Here the difference between the current roll instruction and the roll instruction used during computation of the separation flight path is multiplied by the direction of the escape angle and is added over time. The escape angle is a relative roll angle. If the difference is positive, that is, the pilots' roll instruction is greater than the roll instruction used in the prediction, and if the roll instruction is in the direction of the escape angle, a zero is added. The sum is multiplied by a positive scale factor Ki in order to balance the number of PIRs in the PI.

In the longitudinal inclination the idea is to collect instructions of the pilot's longitudinal inclination greater than those used in the prediction, in occasions in which the escape angle is greater. PIPs can for example be expressed as:

PIP = K₂ · (1 - cos (EA)) · Σ max [| NZ_ess - cos θ | - | NZ_ARM - cosP |, 0]

NZ_ess = the angles of longitudinal inclination commanded

NZARM = load factor used during computation of the separation flight path

K2 = a scale factor.

Here both the longitudinal tilt instruction and the longitudinal tilt instruction used for the prediction are compared with the optimal load balance factor (1 x cosP). The magnitude of the comparison is compared, and the difference in magnitude of the pilot part and the predicted part is estimated. The maximum function is used to find only those occasions when the pilot part is greater than the predicted part. In other words, the second part of the expression punishes pilot pitch instructions that are greater in magnitude than those used during prediction compared to the optimal load roll factor.

The first part of the expression is to graduate the result of the second part. K₂ it is a positive scale factor. 1 - cos (EA) is close to zero when the escape angle is smaller and increases when the escape angles are increased.

K size₁, K₂ and the lower limit of the PI is a balance between the risk of disturbance and the position width w of the separation flight path during the avoidance maneuver. For a worst-case scenario, the time before early activation should probably be close to 0.1 seconds.

Figure 2 shows a system for automatic collision avoidance according to an embodiment of the invention. The system comprises a computing unit 10 adapted to compute an escape angle and a load factor to be used during an avoidance maneuver, and a separation flight path to be sent to the other aircraft. The escape angle, load factor, and parting paths are computed based on the aircraft's response model and the parting paths received from the other aircraft. The first 0.3 seconds of the separation flight path is a prediction based on an assumption that the aircraft

ES 2 300 684 T3 will continue to be maneuvered according to the last maneuver commanded. The second part of the avoidance path is computed based on the computed escape angle and load factor.

The system also comprises a transmitter 12 that sends the computed separation flight path to the other aircraft and a receiver 14 that receives the separation paths of the other aircraft. The system comprises a data storage system 16 adapted to store the computed escape angle, the load factor and the separation flight path. The system is preferably implemented in an aircraft's control system, and uses the aircraft's computer.

The system comprises collision management means 18 adapted to detect an approaching collision based on the separation flight path itself and the separation paths received from the other aircraft and the activation of a separation maneuver after detection. of an oncoming collision. A collision management means 18 is adapted to receive the commanded maneuvers of the aircraft itself and to detect an approaching collision based on the aircraft maneuvers commanded during a period of time between sending the own computed separation flight path. up to the other plane and detection of the coming collision.

Figure 3 shows a flow chart and a process and a computer program product in accordance with an embodiment of the present invention. It should be understood that each block of the flow chart can be implemented by the instructions of a computer program. The system receives the separation trajectories from the other aircraft, block 20. The separation trajectories are stored until the new separation trajectories are received. The newly received separation paths supersede previously received and stored separation flight paths. As shown in block 22 an escape angle, a load factor, and a separation flight path designated FAP are computed for the aircraft itself based on the separation paths received from the other aircraft. The computation comprises the predicted movements of the aircraft during the next 0.3 seconds. The computed escape angle, load factor, and separation flight path are stored. The computed separation flight path is sent to the other aircraft for its registration, block 23. From that moment, the movements of the joystick are registered, block 24, and based on this, an evasion condition is calculated, designated as EC, according to equation 2.

Before calculating the evasion condition, the PIR and PIP are calculated according to equations 6 and 7 based on the recorded movements of the pilot lever and the roll and longitudinal angles predicted in block 20. Then the PI is computed according to equation 5 and ICP is computed according to equation 4. The MSSD and its derivative are determined based on the separation flight path itself and the separation flight path itself received from the other aircraft. The EC is computed according to Equation 2, block 26. At block 28 it is determined whether or not there is an upcoming collision, based on the computed value of the EC. If the EC is below the threshold value TV an approaching collision is detected and the avoidance maneuver is activated, block 32; otherwise, no collision is detected and the procedure is repeated by calculating a new escape angle, a load factor and a new separation flight path, block 20. If new separation flight paths are received from the other aircraft then they are stored and the next computation of the separation path itself is based on those, block 29, 30. Next, the computation loop comprising blocks 22 to 28 repeats with a predetermined frequency.

The present invention is not limited to the disclosed embodiments but can be varied and modified within the scope of the claims that follow.