Terrain awareness and warning system excessive descent rate warning constant false alarm rate detection method

Abstract

The application provides a terrain awareness and warning system excessive descent rate warning constant false alarm rate detection method, comprising the following steps: step 1: calculating the no-alarm trajectory data of an airplane according to the flight parameters of the airplane; calculating the post-alarm trajectory data of the airplane; performing collision detection on the no-alarm flight trajectory and the post-alarm flight trajectory with a first-order Markov random terrain to determine whether a successful alarm or an unnecessary alarm is generated; step 2: using a Monte Carlo method to perform a large number of simulations, counting the successful alarm rate and the false alarm rate to generate an SOC curve, and selecting an optimal alarm threshold point; step 3: selecting a constant false alarm rate range according to the obtained optimal alarm threshold point; obtaining the relationship between the descent rate and the alarm height from the selected constant false alarm rate; and step 4: repeating step 3 to obtain an optimal constant false alarm rate. The application can effectively reduce the false alarm rate of the excessive descent rate mode alarm, increase the safe flight probability, and adapt to more flight conditions.

Description

Technical Field

[0001] This invention belongs to the field of terrain perception and alarm technology, and relates to a method for detecting the false alarm rate of an alarm with excessive drop rate in a terrain perception and alarm system. Background Technology

[0002] The Terrain Awareness and Warning System (TAWS) is an airborne avionics system developed to help pilots handle controlled flight collisions. TAWS' core functions include six warning modes and a forward-looking capability. These functions work together to ensure flight safety and help pilots take timely evasive action in various potentially hazardous situations.

[0003] In terrain awareness and warning systems, excessive descent rate alarms are also known as Mode 1 alarms. Mode 1 is a key configuration in aircraft warning systems, used to monitor the aircraft's descent rate based on radio altitude and trigger an alarm when it exceeds a predetermined threshold. To prevent aviation accidents, this system monitors radio altitude and descent rate, and if a warning is triggered, it emits audible and visual signals.

[0004] However, Mode 1 also suffers from false alarms, where the system incorrectly identifies a normal descent rate as an excessive one, triggering unnecessary alarms. These false alarms can be caused by various factors, such as sensor data fluctuations, environmental interference, or improper algorithm settings. Frequent false alarms can not only distract pilots but also trigger unnecessary emergency maneuvers, increasing the tension during flight. Therefore, optimizing and improving data processing capabilities to reduce false alarms is crucial for ensuring flight safety and enhancing pilot confidence. Summary of the Invention

[0005] Purpose of the Invention: The technical problem this invention aims to solve is to address the shortcomings of existing technologies by providing a constant false alarm rate (CFAR) detection method for terrain perception and warning systems with excessive descent rates. In the field of target detection using radar technology, the CFAR algorithm, as a target detection method, has evolved from traditional mean-based algorithms to modern neural networks. This invention focuses on using obstacle avoidance data from aircraft under different terrain and flight speed conditions to establish the basis of the warning model, making the warning threshold more adaptable to actual flight conditions. By adjusting the warning altitude, the probability of safe flight can be effectively increased while maintaining a constant CFAR rate.

[0006] The method of the present invention specifically includes the following steps:

[0007] Step 1: Calculate the flight trajectory data without alarm and the flight trajectory data after alarm. Perform collision detection between the flight trajectory without alarm and the flight trajectory after alarm and the first-order Markov random terrain to determine whether the alarm is successful or unnecessary. Unnecessary alarm is a false alarm.

[0008] Step 2: Use the Monte Carlo method to perform simulation, statistically analyze the successful alarm rate and false alarm rate to generate the System Operating Characteristic (SOC) curve, and select the optimal alarm threshold point;

[0009] Step 3: Based on the obtained optimal alarm threshold point, the false alarm rate P(UA) at the optimal alarm threshold point is processed in different ways to obtain the constant false alarm rate; from the obtained constant false alarm rate, the relationship between the decrease rate and the alarm height is obtained.

[0010] Step 4: Repeat step 3 to compare the effects of different treatments on the constant false alarm rate and obtain the optimal constant false alarm rate.

[0011] Furthermore, step 1 includes:

[0012] Step 1.1, Generate first-order Markov random terrain:

[0013] In a first-order Gaussian Markov terrain, the terrain height at any point depends only on the height of the previous location, and this dependency is described by a normal distribution, which produces a height with a mean of 0 and a variance of σ. 2 A one-dimensional random terrain model, with the current terrain height y n and the terrain height y at the next moment n+1 The following relationship must be satisfied:

[0014] y n+1 =e -β y n +ξ n

[0015] Where the intermediate parameter β = 1 / l0, and l0 is the correlation length; ξ n It satisfies the condition that the mean is 0 and the variance is σ. 2 (1-e -2β A normally distributed random variable is represented as:

[0016] ξ n ~N(0,σ 2 (1-e -2β ))

[0017] Where e represents the natural constant;

[0018] Changing σ and l0 can produce different terrains;

[0019] Step 1.2: Calculate the no-alarm trajectory data of the aircraft based on the aircraft's flight parameters;

[0020] Step 1.3: Calculate the aircraft trajectory data after the alarm by combining the pilot delay time and the rate of change of the flight path angle;

[0021] Step 1.4: Perform collision detection between the flight trajectory and the first-order Markov random terrain to determine whether the alarm was successfully triggered.

[0022] Furthermore, step 1.2 includes:

[0023] Assume the airplane is a point mass moving at a constant velocity, with an initial state of (v0, h0, l0) and a state at time t of (v...). t ,h t ,l t ), where v0, v t Let h0 and ht represent the airspeed of the aircraft at time 0 and time t, respectively; h0, ht t Let l0 and l represent the radio altitude of the aircraft at time 0 and time t, respectively; l0, l t Let v0 represent the horizontal distance the aircraft travels at time 0 and time t, respectively; since the aircraft is moving at a constant speed, then v0 = vt. t =v, where v is a constant velocity; θ0 is the aircraft's flight path angle;

[0024] After time t during normal flight, the aircraft's radio altitude is: h t =h0-v·sinθ0·t, the horizontal flight distance of the aircraft is: l t =l0+v·cosθ0·t.

[0025] Furthermore, step 1.3 includes:

[0026] Step 1.3.1, Generate the reaction delay phase trajectory:

[0027] The altitude h(t) of the aircraft at any time during the reaction delay phase is:

[0028] h(t) = h0 - v·sinθ0·t

[0029] Set the delay time t for the pilot to take pull-up action after the alarm is set. delay The calculated altitude h1 of the aircraft when the action is taken is:

[0030] h1=h0-v·sinθ0·t delay ;

[0031] Step 1.3.2, Generate the trajectory of the upward phase:

[0032] The aircraft's angle of attack α and the engine's angle of attack ω are relatively small. To simplify the model, we set both the aircraft's angle of attack α and the engine's angle of attack ω to 0. Simultaneously, during pull-up, we assume the aircraft's pull-up motion is stable and uniform, neglecting the bank angle. This means the aircraft will maintain a constant pitch velocity in the vertical direction, flying along an arc of radius R. The vertical track angle gradually changes from θ0 to θ1, while the rate of change of the track angle remains constant. Then the time t experienced in this stage pullup for:

[0033]

[0034] The altitude h(t) of the aircraft at any time t during the climb phase is:

[0035]

[0036] Where h1 represents the aircraft's altitude when the action is taken, d represents the integral sign, and τ represents the variable being integrated;

[0037] The aircraft's altitude h2 at the end of the climb phase is:

[0038]

[0039] Step 1.3.3, Generate the stable holding phase trajectory:

[0040] The aircraft climbs at a constant vertical track angle θ1. At any time t, the aircraft's altitude h(t) is:

[0041] h(t) = h2 + v·sinθ1·t.

[0042] Furthermore, step 1.4 includes:

[0043] During each flight simulation, if a normal, alarm-free trajectory collides with the terrain, the number of collisions for the alarm-free aircraft is incremented by 1; if the trajectory collides with the terrain at any time after an alarm, it is considered an alarm failure, and the number of collisions after an alarm is incremented by 1.

[0044] Furthermore, step 2 includes:

[0045] Repeat steps 1.1 to 1.4, using the Monte Carlo method to simulate near-ground flight events, and statistically analyze the successful alarm rate P(SA) and the false alarm rate P(UA):

[0046] P(SA) = 1 - X1 / X3,

[0047] P(UA) = 1 - X² / X³,

[0048] Where X1 represents the number of times the aircraft crashes to the ground without alarms, X2 represents the number of times the aircraft crashes to the ground after alarms, and X3 represents the total number of simulations.

[0049] Plot the SOC curve with P(UA) on the horizontal axis and P(SA) on the vertical axis. Use the value of P(SA)-P(UA) as the alarm benefit and the maximum value of the alarm benefit as the optimal alarm threshold point.

[0050] Furthermore, step 3 includes:

[0051] Using P(UA) obtained from the optimal alarm threshold point as the constant false alarm rate, repeat steps 1 to 2 to obtain the warning height under the constant false alarm rate; by changing the descent speed, the complete alarm envelope is obtained.

[0052] Furthermore, the present invention also provides an electronic device, including a processor and a memory, wherein the memory stores program code that, when executed by the processor, causes the processor to perform the steps of the method described.

[0053] Furthermore, the present invention also provides a storage medium storing a computer program or instructions that, when the computer program or instructions are run on a computer, execute the steps of the method described.

[0054] Compared with the prior art, the beneficial effects that the embodiments of this specification can achieve by using the above technical solutions include: by setting a constant false alarm rate, the present invention ensures that the false alarm rate is kept within a set threshold, thereby dynamically adjusting the warning height and improving the accuracy and reliability of the alarm. Attached Figure Description

[0055] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments, and the advantages of the present invention in the above and / or other aspects will become clearer.

[0056] Figure 1 This is a trajectory modeling diagram for an embodiment of the present invention.

[0057] Figure 2 This is a flowchart of an embodiment of the present invention.

[0058] Figure 3 This is a schematic diagram of the alarm envelope of Mode 1 finally generated in an embodiment of the present invention. Detailed Implementation

[0059] like Figure 1 and Figure 2 As shown, this embodiment of the invention provides a method for detecting the constant false alarm rate of an alarm with excessively high descent rate in a terrain perception and alarm system, comprising the following steps:

[0060] Step 1: Calculate the no-alarm trajectory data of the aircraft based on the aircraft's flight parameters; then, calculate the trajectory data after the alarm by combining the pilot delay time, the rate of change of the track angle, and other parameters; perform collision detection between the no-alarm flight trajectory and the flight trajectory after the alarm and the first-order Markov random terrain to determine whether the alarm was successful or unnecessary (false alarm).

[0061] Step 2: Use the Monte Carlo method to perform a large number of simulations, statistically analyze the successful alarm rate and false alarm rate to generate the SOC curve, and select the optimal alarm threshold point;

[0062] Step 3: Based on the obtained optimal alarm threshold point, select the range of constant false alarm rate; from the selected constant false alarm rate, obtain the relationship between the decrease rate and the alarm height;

[0063] Step 1 includes the following steps:

[0064] Step 1.1: Generate first-order Markov random terrain:

[0065] In a first-order Gaussian Markov terrain, the height of a point depends only on the height of its preceding point, and this dependency is described by a normal distribution, which produces a height with a mean of 0 and a variance of σ. 2 One-dimensional stochastic terrain model:

[0066] y n+1 =e -β y n +ξ n (1)

[0067] Where β = 1 / l0, and l0 is the correlation length; ξ n It satisfies the condition that the mean is 0 and the variance is σ. 2 (1-e -2β Normally distributed random variables:

[0068] ξ n ~N(0,σ 2 (1-e -2β ))(2)

[0069] Changing σ and l0 can produce different terrains;

[0070] Step 1.2: Calculate the no-alarm trajectory data of the aircraft based on the aircraft's flight parameters;

[0071] Step 1.3: Calculate the aircraft trajectory data after the alarm by combining parameters such as pilot delay time and track angle change rate;

[0072] Step 1.4: Perform collision detection between the flight trajectory and the first-order Markov random terrain to determine whether the alarm was successfully triggered;

[0073] Step 1.2 specifically includes:

[0074] Assume the airplane is a point mass moving at a constant velocity, with an initial state of (v0, h0, l0) and a state at time t of (v...). t ,h t ,l t ), where v0, v t Let h0 and ht represent the airspeed of the aircraft at time 0 and time t, respectively; h0, ht t Let l0 and l represent the radio altitude of the aircraft at time 0 and time t, respectively; l0, l t Let v0 represent the horizontal distance the aircraft travels at time 0 and time t, respectively; since the aircraft is moving at a constant speed, then v0 = vt. t =v, where v is a constant velocity; θ0 is the aircraft's flight path angle;

[0075] Then, after time t, the aircraft's radio altitude is:

[0076] h t =h0-v·sinθ0·t (3);

[0077] The horizontal flight distance of the aircraft is:

[0078] l t =l0+v·cosθ0·t (4)

[0079] Step 1.3 further includes:

[0080] Step 1.3.1: Generate the reaction delay phase trajectory

[0081] The altitude h(t) of the aircraft at any time during the reaction delay phase is:

[0082] h(t)=h0-v·sinθ0·t (5)

[0083] Set the delay time t for the pilot to take pull-up action after the alarm is set. delay The calculated altitude h1 of the aircraft when the action is taken is:

[0084] h1=h0-v·sinθ0·t delay (6)

[0085] Step 1.3.2: Generate the trajectory of the upward phase

[0086] The aircraft's angle of attack α and the engine's angle of attack ω are relatively small, so for the sake of model simplification, they are assumed to be 0. Simultaneously, during pull-up, it is assumed that the aircraft's pull-up motion is stable and uniform, and the bank angle is neglected. Thus, the aircraft will maintain a constant pitch velocity in the vertical direction, flying along an arc of radius R, with the vertical track angle gradually changing from θ0 to θ1, and the rate of change of the track angle remaining constant. Then the time t experienced in this stage pullup for:

[0087]

[0088] The altitude h(t) of the aircraft at any moment during the climb phase is:

[0089]

[0090] The aircraft's altitude h2 at the end of the climb phase is:

[0091]

[0092] Step 1.3.3: Generate the stable holding phase trajectory

[0093] The aircraft climbs at a constant vertical track angle θ1, and the altitude h(t) of the aircraft at any given time is:

[0094] h(t)=h2+v·sinθ1·t (10)

[0095] Step 1.4 further includes:

[0096] During each flight simulation, if a normal, alarm-free trajectory collides with the terrain, the number of times the alarm-free aircraft has collided with the ground is incremented by 1; if the trajectory collides with the terrain at any time after an alarm, it is considered an alarm failure, and the number of times the aircraft has collided with the ground after the alarm is incremented by 1.

[0097] The method for detecting excessive descent rate alarms and constant false alarms in a near-ground alarm system according to claim 2, characterized in that step 2 includes:

[0098] Repeat steps 1.1 to 1.4, and use the Monte Carlo method to simulate near-ground flight events, and calculate the successful alarm rate P(SA) and false alarm rate P(UA); where P(SA) = 1 - number of aircraft crashes without alarm / total number of simulations; P(UA) = 1 - number of aircraft crashes after alarm / total number of simulations; plot the SOC curve with P(UA) as the horizontal axis and P(SA) as the vertical axis, and use the value of P(SA) - P(UA) as the alarm benefit, and use its maximum value as the optimal alarm threshold point.

[0099] The method for detecting excessive descent rate alarms and constant false alarms in a near-ground alarm system according to claim 3, characterized in that step 3 includes:

[0100] By selecting an appropriate P(UA) as the constant false alarm rate through the optimal alarm threshold point, and repeating steps 1 to 2, the warning height under the optimal constant false alarm rate is obtained; by changing the descent speed, the complete alarm envelope can be obtained.

[0101] In this embodiment, under the conditions of terrain variance of 269 and airspeed range of 1000–5000 feet / minute, the constant false alarm rate alarm envelope and maximum gain alarm envelope obtained in steps 1–3 are as follows: Figure 3 As shown.

[0102] In this embodiment, under the conditions of maximum alarm benefit and constant false alarm rate, the overall P(SA) decreased from 80.43% to 73.54%, a decrease of 8.57%; the overall P(UA) decreased from 21.16% to 16%, a decrease of 24.4%, achieving a greater reduction in P(UA) with a smaller loss in P(SA).

[0103] Experimental results show that the method in this embodiment can obtain a reasonable and effective alarm envelope. The final generated alarm envelope effectively reduces the false alarm rate, allowing pilots to better handle alarm events with excessive descent rates.

[0104] This invention provides a method for detecting the constant false alarm rate of an alarm with excessive descent rate in a terrain perception and alarm system. Many methods and approaches exist for implementing this technical solution; the above description is merely a preferred embodiment of the invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of this invention, and these improvements and modifications should also be considered within the scope of protection of this invention. All components not explicitly stated in this embodiment can be implemented using existing technologies.

Claims

1. A method for detecting excessive false alarm rate in terrain perception and alarm systems, characterized in that: Includes the following steps: Step 1: Calculate the flight trajectory data without alarm and the flight trajectory data after alarm. Perform collision detection between the flight trajectory without alarm and the flight trajectory after alarm and the first-order Markov random terrain to determine whether the alarm is successful or unnecessary. Unnecessary alarm is a false alarm. Step 2: Use the Monte Carlo method to perform simulation, statistically analyze the successful alarm rate and false alarm rate to generate the system operating performance (SOC) curve, and select the optimal alarm threshold point; Step 3: Based on the obtained optimal alarm threshold point, the false alarm rate P(UA) at the optimal alarm threshold point is processed in different ways to obtain the constant false alarm rate; from the obtained constant false alarm rate, the relationship between the decrease rate and the alarm height is obtained. Step 4: Repeat step 3 and compare the effects of different treatments on the constant false alarm rate to obtain the optimal constant false alarm rate. Step 1 includes: Step 1.1, Generate first-order Markov random terrain: In a first-order Gaussian Markov terrain, the terrain height at any point depends only on the height of the previous location, and this dependency is described by a normal distribution, which produces a height with a mean of 0 and a variance of σ. 2 A one-dimensional random terrain model, with the current terrain height y n and the terrain height y at the next moment n+1 The following relationship must be satisfied: and n+1 =e -β and n +ξ n Where the intermediate parameter β = 1 / l0, and l0 is the correlation length; ξ n It satisfies the condition that the mean is 0 and the variance is σ. 2 (1-e -2β A normally distributed random variable is represented as: x n ~N(0,σ 2 (1-e -2β )) Where e represents the natural constant; Changing σ and l0 can produce different terrains; Step 1.2: Calculate the no-alarm trajectory data of the aircraft based on the aircraft's flight parameters; Step 1.3: Calculate the aircraft trajectory data after the alarm by combining the pilot delay time and the rate of change of the flight path angle; Step 1.4: Perform collision detection between the flight trajectory and the first-order Markov random terrain to determine whether the alarm was successfully triggered. Step 1.2 includes: Assume the airplane is a point mass moving at a constant velocity, with an initial state of (v0, h0, l0) and a state at time t of (v...). t ,h t ,l t ), where v0, v t Let h0 and ht represent the airspeed of the aircraft at time 0 and time t, respectively; h0, ht t Let l0 and l represent the radio altitude of the aircraft at time 0 and time t, respectively; l0, l t Let v0 represent the horizontal distance the aircraft travels at time 0 and time t, respectively; since the aircraft is moving at a constant speed, then v0 = vt. t =v, where v is a constant velocity; θ0 is the aircraft's flight path angle; After time t during normal flight, the aircraft's radio altitude is: h t =h0-v·sinθ0·t, the horizontal flight distance of the aircraft is: l t =l0+v·cosθ0·t; Step 1.3 includes: Step 1.3.1, Generate the reaction delay phase trajectory: The altitude h(t) of the aircraft at any time during the reaction delay phase is: h(t) = h0 - v·sinθ0·t Set the delay time t for the pilot to take pull-up action after the alarm is set. delay The calculated altitude h1 of the aircraft when the action is taken is: h1=h0-v·sinθ0·t delay ; Step 1.3.2, Generate the trajectory of the upward phase: Set the aircraft's angle of attack α and the engine's angle of attack ω to 0; simultaneously, during the pull-up, assume the pull-up action is stable and uniform, neglecting the bank angle. The aircraft will maintain a constant pitch speed in the vertical direction, flying along an arc of radius R, with the vertical track angle gradually changing from θ0 to θ1, and the rate of change of the track angle remaining constant. Then the time t pullup for: The altitude h(t) of the aircraft at any time t during the climb phase is: Where h1 represents the aircraft's altitude when the action is taken, d represents the integral sign, and τ represents the variable being integrated; The aircraft's altitude h2 at the end of the climb phase is: Step 1.3.3, Generate the stable holding phase trajectory: The aircraft climbs at a constant vertical track angle θ1. At any time t, the aircraft's altitude h(t) is: h(t) = h² + v·sinθ₁·t; Step 1.4 includes: During each flight simulation, if a normal, alarm-free trajectory collides with the terrain, the number of times the alarm-free aircraft has collided with the ground is incremented by 1; if the trajectory collides with the terrain at any time after an alarm is triggered, the alarm is considered to have failed, and the number of times the aircraft has collided with the ground after the alarm is triggered is incremented by 1. Step 2 includes: Repeat steps 1.1 to 1.4, using the Monte Carlo method to simulate near-ground flight events, and statistically analyze the successful alarm rate P(SA) and the false alarm rate P(UA): P(SA) = 1 - X1 / X3, P(UA) = 1 - X² / X³, Where X1 represents the number of times the aircraft crashes to the ground without alarms, X2 represents the number of times the aircraft crashes to the ground after alarms, and X3 represents the total number of simulations. Plot the SOC curve with P(UA) as the horizontal axis and P(SA) as the vertical axis. Use the value of P(SA)-P(UA) as the alarm benefit and the maximum value of the alarm benefit as the optimal alarm threshold point. Step 3 includes: Using P(UA) obtained from the optimal alarm threshold point as the constant false alarm rate, repeat steps 1 to 2 to obtain the warning height under the constant false alarm rate; by changing the descent speed, the complete alarm envelope is obtained.

2. An electronic device, characterized in that, It includes a processor and a memory, the memory storing program code that, when executed by the processor, causes the processor to perform the steps of the method as described in claim 1.

3. A storage medium, characterized in that, It stores a computer program or instructions that, when run on a computer, perform the steps of the method as described in claim 1.

Images