CN-119445906-B - Method for detecting alarm constant false alarm rate of overlarge decline rate of terrain sensing and alarm system

Abstract

The invention provides a method for detecting a constant false alarm rate of an oversized descent rate alarm of a terrain sensing and alarming system, which comprises the steps of 1, calculating alarm-free track data of an airplane according to flight parameters of the airplane, calculating the track data after the airplane alarms, carrying out collision detection on the alarm-free flight track and first-order Markov random terrains, judging whether a successful alarm or an unnecessary alarm is carried out, 2, carrying out a large number of simulations by using a Monte Carlo method, counting the successful alarm rate and the false alarm rate to generate an SOC curve, selecting an optimal alarm threshold point, 3, selecting a range of the constant false alarm rate according to the obtained optimal alarm threshold point, obtaining a relation between the descent rate and the alarm height according to the selected constant false alarm rate, and 4, repeating the step 3 to obtain the optimal constant false alarm rate. The invention 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.

Inventors

• YAN TIANKUN
• PAN WENPING
• DONG XINYI
• ZOU ZHIQIU
• WU WEIKANG
• CHEN GUO

Assignees

• 南京航空航天大学

Dates

Publication Date

20260508

Application Date

20241031

Claims (3)

1. 1. The method for detecting the constant false alarm rate of the overlarge descent rate alarm of the terrain sensing and alarm system is characterized by comprising the following steps of: Step 1, calculating alarm-free track data and alarm-post track data of an airplane, performing collision detection on the alarm-free flight track and the alarm-post flight track and first-order Markov random terrains, and judging whether the alarm is successful or unnecessary, wherein the unnecessary alarm is a false alarm; step 2, simulating by using a Monte Carlo method, counting the successful alarm rate and the false alarm rate to generate a system operation performance SOC curve, and selecting an optimal alarm threshold point; Step 3, according to the obtained optimal alarm threshold point, carrying out different treatments by the false alarm rate P (UA) of the optimal alarm threshold point to obtain a constant false alarm rate; Step 4, repeating the step 3, and comparing the effects of the constant false alarm rates under different treatments to obtain the optimal constant false alarm rate; The step 1 comprises the following steps: step 1.1, generating first-order Markov random topography: In first-order gaussian Markov terrain, the value of the terrain height at any point depends only on the height value of the previous position, and the dependency is described by normal distribution, and is used for generating a one-dimensional random terrain model with the average height value of 0 and the variance of sigma 2 , wherein the current terrain height y n and the terrain height y n+1 at the next moment meet the following relations: y n+1 ＝e -β y n +ξ n Where the intermediate parameter β=1/l 0 ,l 0 is the correlation length and ζ n is a random variable satisfying a normal distribution with a mean of 0 and a variance of σ 2 (1-e -2β ), expressed as: ξ n ～N(0,σ 2 (1-e -2β )) Wherein e represents a natural constant; different terrains can be obtained by changing sigma and l 0 ; Step 1.2, calculating alarm-free track data of the aircraft according to flight parameters of the aircraft; step 1.3, calculating track data after aircraft warning by combining pilot delay time and track angle change rate; Step 1.4, collision detection is carried out on the flight track and the first-order Markov random topography, and whether the warning is successful or not is judged; step 1.2 comprises: Setting the aircraft to make uniform motion, wherein the initial state is (v 0 ,h 0 ,l 0 ), and the state at the time t is (v t ,h t ,l t ), wherein v 0 ,v t respectively represents the airspeed of the aircraft at the time 0 and the airspeed at the time t, h 0 ,h t respectively represents the radio altitude of the aircraft at the time 0 and the radio altitude at the time t, l 0 ,l t respectively represents the flight distance of the aircraft in the horizontal direction at the time 0 and the flight distance in the horizontal direction at the time t, v 0 ＝v t =v is a speed constant value because the aircraft makes uniform motion, and θ 0 is the track angle of the aircraft; after the normal flight time t, the radio height of the aircraft is h t ＝h 0 -v·sinθ 0 & t, and the flight distance of the aircraft in the horizontal direction is l t ＝l 0 +v·cosθ 0 & t; Step 1.3 comprises: Step 1.3.1, generating a reaction delay stage track: The altitude h (t) of the aircraft at any moment in the reaction delay phase is: h(t)=h 0 -v·sinθ 0 ·t setting delay time of taking a pull-up measure by a pilot after alarming as t delay , and calculating the height h 1 of the aircraft when taking the measure as follows: h 1 ＝h 0 -v·sinθ 0 ·t delay ; Step 1.3.2, generating a track of a pull-up stage: Setting the attack angle alpha of the airplane and the attitude angle omega of the engine to be 0, setting the pulling action of the airplane to be stable and uniform during the pulling, and enabling the airplane to fly along an arc with a radius R along the vertical direction with a constant pitching speed regardless of the inclination angle, wherein the vertical track angle is gradually changed from theta 0 to theta 1 , and the track angle change rate is constant The time t pullup elapsed is: The height h (t) of the aircraft at any time t in the lifting stage is as follows: Wherein h 1 denotes the altitude of the aircraft when taking measures, d denotes the sign of the integral, τ denotes the variable to be integrated; The altitude h 2 of the aircraft at the end of the pull-up phase is: Step 1.3.3, generating a stable maintenance stage track: The aircraft climbs at a constant speed at a fixed vertical track angle theta 1 , and the height h (t) of the aircraft at any time t is as follows: h(t)=h 2 +v·sinθ 1 ·t; Step 1.4 comprises: In each flight simulation process, if the normal no-alarm track collides with the terrain, the number of times of collision of the no-alarm aircraft with the terrain is increased by 1, if the track collides with the terrain at any moment after the alarm, the warning is considered as failure, and the number of times of collision of the warning aircraft with the terrain is increased by 1; the step 2 comprises the following steps: repeating the steps 1.1-1.4, simulating the near-earth flight event by using a Monte Carlo method, and counting the successful alarm rate P (SA) and the false alarm rate P (UA): P(SA)=1-X 1 /X 3 , P(UA)=1-X 2 /X 3 , Wherein X 1 represents the number of times of airplane collision during no warning, X 2 represents the number of times of airplane collision after warning, and X 3 represents the total simulation number of times; Taking P (UA) as a horizontal axis, taking P (SA) as a vertical axis as an SOC curve, taking the values of P (SA) -P (UA) as alarm benefits, and taking the maximum value of the alarm benefits as an optimal alarm threshold point; The step 3 comprises the following steps: and (3) taking P (UA) obtained through the optimal alarm threshold point as a constant false alarm rate, repeating the steps 1-2 to obtain an early warning height under the constant false alarm rate, and obtaining a complete alarm envelope by changing the descending speed.
2. 2. An electronic device comprising 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 of claim 1.
3. 3. A storage medium storing a computer program or instructions which, when run on a computer, performs the steps of the method of claim 1.

Description

Method for detecting alarm constant false alarm rate of overlarge decline rate of terrain sensing and alarm system Technical Field The invention belongs to the technical field of terrain awareness and warning, and relates to a method for detecting a warning constant false alarm rate of an oversized descent rate of a terrain awareness and warning system. Background The terrain awareness and warning system (TERRAIN AWARENESS AND WARNING SYSTEM, TAWS) is an on-board avionics device developed to assist pilots in resolving controlled flight ground-impact incidents. The core functions of the TAWS comprise six alarm modes and a forward looking function, which together ensure the safety of flight and help pilots take timely danger avoidance measures under various potential dangerous situations. In the terrain awareness and warning system, the excessive drop rate warning becomes a mode 1 warning. Mode 1 is a critical configuration in an aircraft alert system for monitoring the descent rate of an aircraft based on radio altitude and triggering an alert when it exceeds a predetermined threshold. To avoid aeronautical accidents, the system monitors the radio altitude and the descent rate, and if a warning is triggered, sound and light signals are released. However, mode 1 also has a false alarm problem, meaning that the system erroneously recognizes a normal descent rate as an excessive descent rate, resulting in unnecessary alarm triggering. Such false alarms may be caused by a variety of factors, such as sensor data fluctuations, environmental disturbances, or algorithm misassembly. Frequent false alarms may not only cause distraction to the pilot, but may also cause unnecessary emergency operations, increasing the stressful atmosphere during flight. Therefore, optimizing and improving data processing capabilities to reduce the occurrence of false alarms becomes an important measure to ensure flight safety and improve pilot confidence. Disclosure of Invention The invention aims to solve the technical problem of providing a method for detecting the constant false alarm rate of the overlarge decline rate alarm of a terrain sensing and alarm system aiming at the defects of the prior art. In the field of target detection, a Constant false alarm rate (Constant FALSE ALARM RATE, CFAR) algorithm is used as a target detection means, and has undergone a development process from a traditional mean value class to a modern neural network. The invention focuses on the obstacle avoidance data of the airplane under different terrains and different flying speeds to establish the basis of the warning model, so that the warning threshold is more suitable for actual flying, and the safety flying probability can be effectively increased while the constant false warning rate is maintained by adjusting the warning height. The method specifically comprises the following steps: Step 1, calculating alarm-free track data and alarm-post track data of an airplane, performing collision detection on the alarm-free flight track and the alarm-post flight track and first-order Markov random terrains, and judging whether the alarm is successful or unnecessary, wherein the unnecessary alarm is a false alarm; step 2, simulating by using a Monte Carlo method, counting the successful alarm rate and the false alarm rate to generate a system operation performance (System Operating Characteristic, SOC) curve, and selecting an optimal alarm threshold point; Step 3, according to the obtained optimal alarm threshold point, carrying out different treatments by the false alarm rate P (UA) of the optimal alarm threshold point to obtain a constant false alarm rate; and 4, repeating the step 3, and comparing the effects of the constant false alarm rate under different treatments to obtain the optimal constant false alarm rate. Further, step 1 includes: step 1.1, generating first-order Markov random topography: In first-order gaussian Markov terrain, the value of the terrain height at any point depends only on the height value of the previous position, and the dependency is described by normal distribution, and is used for generating a one-dimensional random terrain model with the average height value of 0 and the variance of sigma 2, wherein the current terrain height y n and the terrain height y n+1 at the next moment meet the following relations: yn+1＝e-βyn+ξn Where the intermediate parameter β=1/l 0,l0 is the correlation length and ζ n is a random variable satisfying a normal distribution with a mean of 0 and a variance of σ 2(1-e-2β), expressed as: ξn～N(0,σ2(1-e-2β)) Wherein e represents a natural constant; different terrains can be obtained by changing sigma and l 0; Step 1.2, calculating alarm-free track data of the aircraft according to flight parameters of the aircraft; step 1.3, calculating track data after aircraft warning by combining pilot delay time and track angle change rate; And 1.4, performing collision detection on the flight track and the first-o