Behavior recognition method and device, equipment, storage medium and program product

Abstract

The invention provides a behavior recognition method and device, equipment, a storage medium and a program product, relates to the field of aviation safety, and is used for improving the recognition accuracy of aircraft flight events. The method comprises the following steps: acquiring a navigation data segment of a target aircraft in a target flight stage; based on the navigation data fragment, obtaining a navigation feature vector of the target aircraft; wherein the navigation feature vector is used for reflecting the flight state of the target aircraft; determining an abnormal flight event of the target aircraft in the target flight stage based on the navigation feature vector, a rule base and a navigation template; wherein the rule base is used for recording judgment rules of various abnormal flight events of the aircraft; the navigation template is used for indicating the navigation data of the aircraft when the abnormal flight event occurs.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of aviation safety, and particularly relates to a behavior recognition method and device, equipment, a storage medium and a program product. BACKGROUND

[0002] Flight Data Monitoring (FDM) and Quick Access Recorder (QAR) data analysis system is an important guarantee for modern aviation safety. In the system, flight event recognition is the core link of extracting safety information from parameter data and realizing early warning of risks, and the precision and effectiveness of the technology are the premise of implementing fine aviation safety management.

[0003] At present, the flight event recognition technology is generally based on a "static threshold determination" scheme, which presets fixed thresholds for specific flight parameters (such as airspeed, altitude, slope, etc.) of a target aircraft (such as a civil aviation aircraft), and scans the QAR data second by second. When the instantaneous value of the parameter exceeds the threshold, it is considered that a corresponding flight event (such as attitude instability, large take-off slope, etc.) has occurred.

[0004] However, the static threshold determination method is unstable in the critical state, and the recognition result is prone to fluctuation, which restricts the further improvement of the event recognition accuracy. SUMMARY

[0005] The purpose of the present application is to provide a behavior recognition method, device, equipment, storage medium and program product, aiming at solving the problem of how to improve the accuracy of aircraft flight event recognition.

[0006] In a first aspect, the present application provides a behavior recognition method, comprising: obtaining a navigation data segment of a target aircraft in a target flight phase; obtaining a navigation feature vector of the target aircraft based on the navigation data segment; wherein the navigation feature vector is used to reflect the flight state of the target aircraft; determining an abnormal flight event existing in the target flight phase of the target aircraft based on the navigation feature vector, a rule library and a navigation template; wherein the rule library is used to record the judgment rules of various abnormal flight events of the aircraft; and the navigation template is used to indicate the navigation data of the aircraft when the abnormal flight event occurs.

[0007] The technical scheme provided by the application brings at least the following beneficial effects: by obtaining the navigation data segment of the target aircraft in the target flight phase, data information basis is provided for subsequent flight event identification; then, the navigation feature vector capable of comprehensively reflecting the flight state of the target aircraft is extracted from the navigation data segment, and the accurate flight state of the target aircraft is obtained; subsequently, based on the rule base recording the judgment rules of various abnormal flight events of the aircraft and the navigation template used for indicating the navigation data of the aircraft when the abnormal flight event occurs, the abnormal flight event existing in the target flight phase of the target aircraft is determined, so that the accuracy of flight event identification is improved.

[0008] Optionally, the rule base comprises a judgment rule of a target abnormal flight event, and the navigation template comprises a navigation template of the target abnormal flight event; the judgment rule of the target abnormal flight event comprises a threshold value of target navigation data; the navigation template of the target abnormal flight event is used for indicating the navigation data of the aircraft when the target abnormal flight event occurs; based on the navigation feature vector, the rule base and the navigation template, the abnormal flight event existing in the target flight phase of the target aircraft is determined, comprising: the judgment rule of the target abnormal flight event is used to judge the navigation feature vector; in the case that the navigation feature vector is greater than the threshold value of the target navigation data, the similarity between the navigation feature vector and the navigation template of the target abnormal flight event is calculated; in the case that the similarity is greater than a similarity threshold value, it is determined that the target abnormal flight event exists in the target flight phase of the target aircraft.

[0009] Optionally, the navigation data segment of the target aircraft in the target flight phase is obtained, comprising: obtaining a navigation data sequence of the target flight phase of the target aircraft; wherein the target flight phase comprises at least one of the following: a take-off phase, an approach phase and a landing phase; at least one navigation data segment is extracted from the navigation data sequence by using a sliding time window.

[0010] Optionally, the navigation data comprises at least one of the following: the pitch angle of the target aircraft, the slope of the target aircraft, the vertical speed of the target aircraft, the thrust of the target aircraft and the configuration parameter of the target aircraft.

[0011] Optionally, the navigation feature vector comprises at least one of the following: the change rate of the navigation data, the average value of the navigation data, the standard deviation of the navigation data, the peak-valley distribution of the navigation data, the slope of the navigation data and the oscillation feature of the navigation data.

[0012] Optionally, the method further comprises: in the case that it is determined that the target aircraft has abnormal behavior in the target flight phase, outputting an abnormal event identification result, and the abnormal event identification result comprises an event type, event start and end time and navigation data information.

[0013] Secondly, this application provides a behavior recognition device, comprising: an acquisition module and a processing module; the acquisition module is used to acquire flight data segments of a target aircraft during the target flight phase; the processing module is used to obtain a flight feature vector of the target aircraft based on the flight data segments; wherein the flight feature vector is used to reflect the flight state of the target aircraft; the processing module is used to determine abnormal flight events existing in the target aircraft during the target flight phase based on the flight feature vector, a rule base, and a flight template; wherein the rule base is used to record the judgment rules for various abnormal flight events of the aircraft; the flight template is used to indicate the flight data of the aircraft when an abnormal flight event occurs.

[0014] Optional, processing module. Specifically, it is used to judge the flight feature vector according to the judgment rules of the target abnormal flight event; if the flight feature vector is greater than the target flight data threshold, it calculates the similarity between the flight feature vector and the flight template of the target abnormal flight event; if the similarity is greater than the similarity threshold, it determines that the target aircraft has a target abnormal flight event in the target flight phase.

[0015] Optionally, the acquisition module is specifically used to acquire the flight data sequence of the target aircraft during the target flight phase; wherein the target flight phase includes at least one of the following: takeoff phase, approach phase, and landing phase; and at least one flight data segment is extracted from the flight data sequence using a sliding time window.

[0016] Optionally, the processing module is specifically used to output abnormal event identification results when it is determined that the target aircraft has abnormal behavior during the target flight phase. The abnormal event identification results include event type, event start and end time, and flight data information.

[0017] Thirdly, this application provides an electronic device comprising: a processor and a memory; the memory storing processor-executable instructions; when the processor is configured to execute the instructions, causing the electronic device to implement the method of the first aspect described above.

[0018] Fourthly, this application provides a computer-readable storage medium comprising: computer software instructions; which, when executed in an electronic device, cause the electronic device to implement the method described in the first aspect.

[0019] Fifthly, this application provides a computer program product comprising a computer program; when the computer program is run in an electronic device, the electronic device performs the method described in the first aspect.

[0020] The beneficial effects of the second to fifth aspects mentioned above are described in the corresponding description of the first aspect and will not be repeated here. Attached Figure Description

[0021] To more clearly illustrate the technical solutions of the embodiments of this application, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0022] Figure 1 This is a flowchart illustrating a current behavior recognition method; Figure 2 A schematic diagram illustrating the application environment of a behavior recognition method provided in this application; Figure 3 A flowchart illustrating a behavior recognition method provided in this application; Figure 4 A flowchart illustrating another behavior recognition method provided in this application; Figure 5 A flowchart illustrating another behavior recognition method provided in this application; Figure 6 A schematic diagram of the composition of a behavior recognition device provided in this application; Figure 7 This is a schematic diagram of the structure of an electronic device provided in this application. Detailed Implementation

[0023] In the embodiments of this application, the terms "first," "second," "third," "fourth," "fifth," and "sixth" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined with "first," "second," "third," "fourth," "fifth," and "sixth" may explicitly or implicitly include one or more of that feature.

[0024] In embodiments of this application, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0025] "A and / or B" includes the following three combinations: A only, B only, and a combination of A and B.

[0026] In the embodiments of this application, "parallel," "perpendicular," and "equal" include the described situation and situations similar to the described situation, where the range of similarity is within an acceptable deviation range, which is determined by those skilled in the art taking into account the measurement under discussion and the error associated with the measurement of a particular quantity (i.e., the limitations of the measurement system). For example, "parallel" includes absolute parallelism and approximate parallelism, where the acceptable deviation range for approximate parallelism can be, for example, a deviation within 5°; "perpendicular" includes absolute perpendicularity and approximate perpendicularity, where the acceptable deviation range for approximate perpendicularity can also be, for example, a deviation within 5°. "Equal" includes absolute equality and approximate equality, where the acceptable deviation range for approximate equality can be, for example, a difference between the two equals being less than or equal to 5% of either one.

[0027] With the development of aviation safety management, flight event identification methods have become a key technical component of Flight Operations Quality Assurance (FOQA) systems. Analyzing the time-series flight data of the target aircraft recorded by QAR (Quality Assurance Recorder), the core objective is to automatically and accurately identify abnormal flight events deviating from normal flight conditions from massive amounts of data. Flight event identification methods can be used for post-flight safety audits and operational trend analysis of target aircraft to identify systemic risks and optimize flight training programs. They can also be embedded in airborne or ground-based FOQA systems to achieve real-time or near-real-time flight status monitoring and risk warnings for critical flight phases such as takeoff and approach.

[0028] Currently, flight event identification methods are mainly based on the core logic of static rules and threshold determination. For example... Figure 1 As shown, the process is as follows: First, flight data containing parameters such as speed, altitude, and attitude is extracted from the QAR device. The flight data undergoes preprocessing such as time alignment and unit conversion. Then, predefined static rules are configured for flight events. Each rule explicitly specifies parameters, thresholds, and time conditions (e.g., "airspeed greater than 250 knots within 30 seconds before landing" is defined as a high-speed landing event, and "vertical speed less than -600 ft / min and pitch angle greater than 10 degrees" is defined as a tailscratching risk). In the identification phase, the system scans the preprocessed data frame by frame, determining whether the data parameter values ​​at each time point meet the trigger conditions of the rules. If they do, the moment is marked as the corresponding event, and a result report containing the event category is output. Taking the "large takeoff bank angle event identification" event as an example, this method only needs to determine whether the bank angle parameter has an instantaneous value exceeding 10° within the period from takeoff to an altitude of 11 meters. If it exceeds 10 degrees, it is determined to be a large takeoff bank angle event.

[0029] However, with the increasing complexity of aircraft operations and the growing richness of QAR data dimensions, current event recognition methods based on static threshold judgments have shown significant limitations in the following aspects: First, their reliance on single-point instantaneous values ​​of parameters cannot effectively capture the temporal evolution of events (such as changing trends and fluctuation patterns), resulting in insufficient accuracy in recognizing critical states or complex abnormal behaviors (such as "pitch attitude instability"); second, the rule configuration method based on independent events often divides complex events composed of multiple behavioral stages (such as accelerating climb to steep slope, and then pitch fluctuations) into multiple independent event records, failing to reflect the continuity and evolution path between events; finally, the rule base is highly dependent on manual maintenance based on expert experience, and the system's adaptability and intelligence level urgently need to be improved when facing massive amounts of data and new aircraft models.

[0030] To address the aforementioned technical problems, this application provides a behavior recognition method. The method's approach involves: acquiring flight data segments of the target aircraft during the target flight phase to provide a data foundation for subsequent flight event recognition; then, extracting flight feature vectors from the flight data segments that comprehensively reflect the target aircraft's flight status to obtain the target aircraft's precise flight status; subsequently, based on a rule base that records judgment rules for various abnormal flight events of the aircraft and a flight template used to indicate the flight data of the aircraft when abnormal flight events occur, determining the abnormal flight events present in the target aircraft during the target flight phase, thereby improving the accuracy of flight event recognition.

[0031] The embodiments provided in this application will now be described in detail with reference to the accompanying drawings.

[0032] The behavior recognition method provided in this application can be applied to, for example... Figure 2 The application environment shown. For example... Figure 2 As shown, the application environment includes a data acquisition device 10 and a processing device 20. The data acquisition device 10 and the processing device 20 are interconnected.

[0033] In some embodiments, the acquisition device 10 is used to receive and acquire input data information. In this application, the acquisition device 10 is used to acquire in real time the flight data generated by the target aircraft (such as a civil airliner) during flight, including but not limited to key flight data parameters such as attitude, speed, thrust, and altitude, to provide the raw data basis for subsequent flight event identification and analysis.

[0034] In some embodiments, the data acquisition device 10 can be a device with wireless transceiver capabilities, such as an airborne data logger (e.g., QAR), wearable device, in-vehicle device, augmented reality (AR) / virtual reality (VR) device, laptop computer, ultra-mobile personal computer (UMPC), netbook, personal digital assistant (PDA), etc. This application does not limit the specific device form of the data acquisition device 10.

[0035] In some embodiments, the processing device 20 is used to receive data information collected by the acquisition device 10 and to analyze and process the data information. In this application, the processing device 20 is used to receive navigation data information collected by the acquisition device 10, convert the navigation data into feature vectors (such as rate of change, oscillation amplitude, etc.) representing the flight state of the target aircraft, and finally determine whether there are abnormal flight events that deviate from the normal pattern by comparing the feature vectors with the rule base and calculating the similarity between the feature vectors and the navigation template.

[0036] In some embodiments, the processing device 20 may be a server cluster consisting of multiple servers, a single server, a computer, or a processor or processing chip in a server or computer. This application does not limit the specific device form of the processing device 20.

[0037] It should be noted that the system architecture described in the embodiments of this application is for the purpose of more clearly illustrating the technical solutions of the embodiments of this application, and does not constitute a limitation on the technical solutions provided in the embodiments of this application. As those skilled in the art will know, with the evolution of system architecture, the technical solutions provided in the embodiments of this application are also applicable to similar technical problems.

[0038] See Figure 3 This is a flowchart of a behavior recognition method provided in an embodiment of this application. Figure 3 As shown, the behavior recognition method provided in this application specifically includes the following steps S301 to S303.

[0039] S301. Acquire flight data segments of the target aircraft during the target flight phase.

[0040] Among them, the target aircraft refers to the aircraft object to be identified (such as civil aircraft, general aviation aircraft, etc.).

[0041] In some embodiments, the target flight phase includes at least one of the following: takeoff phase, approach phase, and landing phase.

[0042] The takeoff phase, or transition from ground to air, involves dramatic changes in energy state, frequent pilot maneuvers, and extremely limited time and space for handling situations due to low altitude. The approach phase, the preparation and adjustment phase before landing, requires the aircraft to maintain precise track, airspeed, and descent rate, which is fundamental to ensuring a safe landing. The landing phase, from final approach to touchdown and the entire deceleration and taxiing process on the runway, involves precise leveling, touchdown, and energy management. These are the most critical, dangerous, and complex phases of flight. Successfully identifying abnormal behavior during these phases is the most direct and effective way to prevent accidents and improve aviation safety.

[0043] In some embodiments, the flight data includes at least one of the following: the target aircraft's pitch angle, the target aircraft's bank angle, the target aircraft's vertical speed, the target aircraft's thrust, and the target aircraft's configuration parameters.

[0044] In some embodiments, the pitch angle of a target aircraft refers to the angle between the longitudinal axis of the aircraft body and the horizontal plane, which reflects the aircraft's climb or descent attitude and is a key indicator for judging whether the operation is smooth and whether it may cause a stall or overload.

[0045] In some embodiments, the bank angle of a target aircraft refers to the angle formed by the wings on both sides of the aircraft and the horizontal plane when the aircraft rotates around its own longitudinal axis. It is used to describe the degree of tilt when the aircraft turns or maintains balance. Excessive bank angle may cause safety risks.

[0046] In some embodiments, the vertical speed of a target aircraft refers to the rate of ascent or descent of the aircraft in the vertical direction, typically expressed in feet per minute (ft / min), and is used to reflect the rate of change of the aircraft's altitude. An unstable vertical speed indicates a poor state of approach.

[0047] In some embodiments, the thrust of a target aircraft refers to the output power generated by the aircraft engine, which is usually characterized by parameters such as engine pressure ratio (EPR) and torque (N1) to reflect the pilot's energy management intentions. The mismatch between thrust and attitude (such as insufficient thrust at large pitch angles) is an important basis for judging whether the flight status is normal.

[0048] In some embodiments, the configuration parameters of a target aircraft refer to the external shape and device status of the aircraft that change to adapt to different flight phases, including the extension angle of flaps and slats and the landing gear (retracted / lowered) status. There are strict configuration requirements at different phases (such as approach and landing), and an incorrect configuration is a major safety hazard.

[0049] As one feasible approach, a flight data sequence of the aircraft's target flight phase is acquired; and at least one flight data segment is extracted from the flight data sequence using a sliding time window.

[0050] For example, when analyzing the landing attitude during the "approach phase", a time window of 30 seconds can be set, with a step size of 1 second, and the flight data sequence from 2 minutes before landing to the moment of touchdown can be slid over. Each time window will capture a continuous flight data segment that includes parameters such as pitch angle and bank angle.

[0051] S302. Based on the flight data fragments, obtain the flight feature vector of the target aircraft.

[0052] Among them, the navigation feature vector is used to reflect the flight status of the target aircraft.

[0053] In some embodiments, the flight feature vector refers to a set of statistical features calculated based on the time series of flight parameters within a sliding window, used to describe the changing trend and behavioral characteristics of the parameters within that time period.

[0054] In some embodiments, the navigation feature vector includes at least one of the following: the rate of change of navigation data, the mean value of navigation data, the standard deviation of navigation data, the peak-valley distribution of navigation data, the slope of navigation data, and the oscillation characteristics of navigation data. The navigation feature vector transforms the original time-series data segments into a set of quantitative features characterizing behavioral patterns, providing a basis for subsequent abnormal behavior identification.

[0055] In some embodiments, the rate of change of navigation data refers to the average change of a parameter per unit time, reflecting the drasticness or trend strength of the parameter change. For example, if the pitch angle increases by 2 degrees in 1 second, its rate of change is +2 degrees / second. A high rate of change usually indicates drastic or sudden maneuvers.

[0056] In some embodiments, the average value of flight data refers to the arithmetic mean of data points over a period of time, reflecting the overall level or central trend of the parameter during that period. For example, if the average slope is close to 0 degrees over a period of time, it indicates that the aircraft is in a horizontal attitude for most of the time.

[0057] In some embodiments, the standard deviation of flight data is used to measure the dispersion of each data point relative to its mean. A larger standard deviation indicates greater data fluctuation and a more unstable state. For example, a small standard deviation of pitch angle indicates a stable flight attitude, while a large standard deviation indicates frequent up-and-down swings in pitch angle.

[0058] In some embodiments, the peak-valley distribution of navigation data refers to the difference between the maximum value (peak) and the minimum value (valley) in the data (i.e., the range). Sometimes it also includes information on the frequency and location of extreme values, reflecting the range of parameter fluctuations during that period. A large peak-valley difference usually means drastic state changes.

[0059] In some embodiments, the slope of the flight data is obtained by linearly fitting the data points to a straight line, which describes the overall trend of the parameter over a period of time (such as continuous increase, decrease, or stability). For example, a positive slope for altitude indicates that the aircraft is continuously climbing.

[0060] In some embodiments, the oscillation characteristics of flight data are used to describe the periodic or regular fluctuations of parameters around their average values, including oscillation frequency (the number of complete fluctuations per unit time) and oscillation amplitude (the average intensity of each fluctuation). High-frequency or large-amplitude oscillations indicate that the target aircraft's flight status is unstable.

[0061] One feasible approach to converting navigation data segments into navigation feature vectors includes: representing time-series segments as behavioral feature vectors using characteristic equations.

[0062] For example, for a flight data segment captured by a sliding time window (such as a pitch angle sequence 30 seconds before landing), feature calculations are first performed using Python scientific computing libraries (such as NumPy and SciPy) to extract the mean and standard deviation of the segment to describe the basic distribution of pitch angles; the mean of the absolute values ​​of the first-order difference sequence is calculated as the rate of change feature; the oscillation frequency is obtained by finding extreme points and counting the number of values ​​per unit time; and the oscillation amplitude is calculated by combining the peak and trough values. Finally, these statistical features are combined into a multi-dimensional vector in a predetermined order, which constitutes the flight feature vector characterizing the dynamic characteristics of pitch attitude within that time period, providing standardized input for subsequent pattern matching.

[0063] As another possible approach, the flight data segments are converted into flight feature vectors, which involves calculating a series of statistics for each parameter of the flight data, such as maximum value (to detect whether the limit is exceeded), average value (overall trend), standard deviation (stability), slope (pitch rate), etc., and then concatenating these statistics into a flight feature vector.

[0064] For example, to identify "excessive takeoff pitch angle", extract a pitch angle data segment 10 seconds after takeoff. Calculate the maximum value, mean value, standard deviation, slope, etc. of this segment. Combine these data into a vector (e.g., [max=12.5, mean=8.1, std=2.3, slope=0.15]), which is a flight feature vector describing the pitch angle behavior of this takeoff.

[0065] S303. Based on navigation feature vectors, rule bases, and navigation templates, identify abnormal flight events that exist in the target aircraft during the target flight phase.

[0066] The rule base is used to record the judgment rules for various abnormal flight events of aircraft, and the navigation template is used to indicate the navigation data of aircraft when abnormal flight events occur.

[0067] In some embodiments, the rule base includes rules for determining abnormal flight events of a target, and the navigation template includes a navigation template for abnormal flight events of a target; the rules for determining abnormal flight events of a target include thresholds for target navigation data; the navigation template for abnormal flight events of a target is used to indicate the navigation data of an aircraft when an abnormal flight event of a target occurs.

[0068] For example, a rule base is a set of structured behavioral rule expressions derived from the experience of flight experts or data mining, used to determine whether a certain flight feature vector meets the conditions for the occurrence of a certain type of event.

[0069] For example, a flight template refers to a set of parameter behavior feature vector patterns extracted from typical events confirmed in a large amount of historical flight data. Each type of flight event can correspond to one or more standard behavior templates, which are used to describe the typical variation characteristics of the event at the parameter level. The flight template serves as a "reference benchmark" for system identification and judgment, and is used to match and compare with behavior fragments of real-time or new data.

[0070] In some embodiments, abnormal behavior can be attitude abnormality, such as excessive takeoff / landing bank angle (bank angle exceeding limits) or excessive pitch angle (excessive pitching or slamming attitude); abnormal behavior can also be energy abnormality, such as high-speed approach / landing (airspeed exceeding limits) or low-altitude speed (near stall risk); abnormal behavior can also be trajectory abnormality, such as excessive vertical speed (excessive rate of descent or climb) or hard landing (excessive vertical acceleration upon touchdown); abnormal behavior can also be configuration abnormality, such as landing gear not deployed or incorrect flap settings.

[0071] For example, in terms of attitude, a "high takeoff bank angle" event may be identified, characterized by a bank angle value that continuously exceeds the safety template range; in terms of system configuration, a "flap configuration error" may be found, that is, the actual configuration parameter characteristics do not match the standard procedure template.

[0072] In some embodiments, when it is determined that the target aircraft exhibits abnormal behavior during the target flight phase, an abnormal event identification result is output, which includes the event type, event start and end time, and flight data information.

[0073] For example, when an abnormal event of "large takeoff bank angle" is identified, an identification result will be generated, specifically including: the event type is "large takeoff bank angle," the event start and end time is "UTC time 2023-11-05 08:15:30 to 08:15:45," and key flight data information includes "maximum bank angle value 13.5 degrees, duration of bank angle exceeding 8 degrees 3.2 seconds, average airspeed during the trigger phase 145 knots, and associated pitch angle change range +5 to +9 degrees, etc." This identification result fully describes the abnormal behavior through quantitative indicators, providing data support for subsequent risk assessment and operational optimization.

[0074] Based on the above embodiments, by acquiring flight data segments of the target aircraft during the target flight phase, a data information foundation is provided for subsequent flight event identification. Then, flight feature vectors that can comprehensively reflect the flight status of the target aircraft are extracted from the flight data segments to obtain the accurate flight status of the target aircraft. Subsequently, based on the rule base that records the judgment rules of various abnormal flight events of the aircraft and the flight template used to indicate the flight data of the aircraft when abnormal flight events occur, the abnormal flight events existing in the target flight phase of the target aircraft are determined, thereby improving the accuracy of flight event identification.

[0075] In some embodiments, such as Figure 4 As shown, step S303 above, which determines the abnormal flight events of the target aircraft during the target flight phase based on the flight feature vector, rule base, and flight template, can be specifically implemented as the following steps S401~S403: S401. Use the judgment rules for abnormal flight events of the target to judge the flight feature vector.

[0076] In some embodiments, the navigation feature vector is compared with the judgment rules for abnormal flight events in the rule base for rapid event identification, which can serve as a preliminary screening mechanism or as a basis for generating training samples.

[0077] For example, for the "excessive takeoff bank" event, the pre-defined judgment rule in the rule base is: (pitch_angle_max>10) AND (pitch_angle_slope>2.5) AND (duration>=3), where pitch_angle_max represents the maximum pitch angle feature, pitch_angle_slope represents the pitch angle change slope feature, and duration represents the duration of the feature segment.

[0078] S402. When the navigation feature vector is greater than the target navigation data threshold, calculate the similarity between the navigation feature vector and the navigation template of the target abnormal flight event.

[0079] As an feasible approach, cosine similarity can be used to calculate the similarity between the flight feature vector and the flight template of the target's abnormal flight event. The directional consistency is evaluated by measuring the cosine value of the angle between the flight feature vector and the flight template of the target's abnormal flight event in space. It is not sensitive to the absolute value of the vector and is suitable for measuring the trend similarity of flight behavior patterns.

[0080] As another feasible approach, Euclidean distance can be used to calculate the similarity between the navigation feature vector and the navigation template of the target's abnormal flight event. The degree of closeness can be evaluated by calculating the straight-line distance between the navigation feature vector and the navigation template of the target's abnormal flight event in space. The smaller the distance value, the closer the overall numerical characteristics of the two vectors are.

[0081] As another feasible approach, a dynamic warping algorithm can be used to calculate the similarity between the navigation feature vector and the navigation template of the target's abnormal flight event. This algorithm can effectively handle the nonlinear deformation of the navigation feature vector and the navigation template of the target's abnormal flight event on the time axis. By finding the optimal alignment path, the morphological similarity can be evaluated. This algorithm is suitable for analyzing flight behaviors with similar time-series patterns but abnormal phases or velocities.

[0082] S403. If the similarity is greater than the similarity threshold, it is determined that there is an abnormal flight event of the target aircraft during the target flight phase.

[0083] The similarity threshold setting can be dynamically adjusted according to the security sensitivity of different event types, thereby achieving accurate identification.

[0084] In some embodiments, when the similarity between the navigation feature vector and the event template of the target abnormal flight event exceeds a preset similarity threshold, it indicates that the current flight behavior is highly similar to the target abnormal flight event, and the target aircraft may have a target abnormal flight event during the target flight phase.

[0085] In some embodiments, based on the recognition results generated from the rule base and flight templates, the system automatically constructs a high-quality labeled training sample set: behavioral segments that match the rule or template similarity calculation are marked as positive samples, and segments that do not match are marked as negative samples. Using these accurately labeled samples, supervised learning algorithms such as LightGBM and Random Forest are employed to train a machine learning model for flight event recognition, enabling the machine model to learn the complex mapping relationship from flight feature vectors to event types. After the machine model training is complete, it enters the deployment and application phase. The feature vectors generated after feature extraction from flight data can be directly input into the trained classification model for analysis. The machine model outputs the confidence or probability of various events and automatically determines abnormal flight events based on preset thresholds, outputting and recording the abnormal event recognition results, thus improving recognition efficiency.

[0086] The behavior recognition method of this application embodiment is described below with reference to a specific example. The specific implementation process of this method is as follows: Figure 5 As shown.

[0087] S501. Obtain the flight data sequence of the target aircraft during the target flight phase.

[0088] S502. Using a sliding time window, extract at least one flight data segment from the flight data sequence.

[0089] S503. Based on the flight data fragments, obtain the flight feature vector of the target aircraft.

[0090] S504. Determine whether the navigation feature vector is greater than the target navigation data threshold of the judgment rule for abnormal flight events.

[0091] For example, if the navigation feature vector is greater than the target navigation data threshold of the judgment rule for the abnormal flight event, the process jumps to step S505; if the navigation feature vector is less than the target navigation data threshold of the judgment rule for the abnormal flight event, the process ends.

[0092] S505. Calculate the similarity between the navigation feature vector and the navigation template of the target's abnormal flight event.

[0093] S506. Determine whether the similarity is greater than the similarity threshold.

[0094] For example, if the similarity is greater than the similarity threshold, proceed to step S507; if the similarity is less than the similarity threshold, end the process.

[0095] S507. Confirm that the flight characteristics are not within the range of the flight template, determine that the target aircraft is experiencing a flight anomaly, and record the anomaly event identification results.

[0096] As can be seen, the above mainly describes the solutions provided by the embodiments of this application from a methodological perspective. To achieve the above functions, the embodiments of this application provide corresponding hardware structures and / or software modules for executing each function. Those skilled in the art should readily recognize that, in conjunction with the modules and algorithm steps of the various examples described in the embodiments disclosed herein, the embodiments of this application can be implemented in hardware or a combination of hardware and computer software. Whether a function is executed by hardware or by computer software driving hardware depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this invention.

[0097] In some embodiments, this application also provides a behavior recognition device. The behavior recognition device may include one or more functional modules for implementing the behavior recognition method of the above method embodiments.

[0098] For example, Figure 6 This is a schematic diagram illustrating the composition of a behavior recognition device provided in an embodiment of this application. Figure 6 As shown, the behavior recognition device 600 includes an acquisition module 601 and a processing module 602.

[0099] The acquisition module 601 is used to acquire flight data segments of the target aircraft during the target flight phase; the processing module 602 is used to obtain the flight feature vector of the target aircraft based on the flight data segments; wherein, the flight feature vector is used to reflect the flight status of the target aircraft; the processing module 602 is used to determine the abnormal flight events existing in the target aircraft during the target flight phase based on the flight feature vector, the rule base, and the flight template; wherein, the rule base is used to record the judgment rules for various abnormal flight events of the aircraft; the flight template is used to indicate the flight data of the aircraft when an abnormal flight event occurs.

[0100] In some embodiments, the processing module 602 is specifically used to judge the navigation feature vector using the judgment rules for the abnormal flight event of the target; if the navigation feature vector is greater than the target navigation data threshold, calculate the similarity between the navigation feature vector and the navigation template of the abnormal flight event of the target; if the similarity is greater than the similarity threshold, determine that there is an abnormal flight event of the target aircraft in the target flight phase.

[0101] In some embodiments, the acquisition module 601 is specifically used to acquire a flight data sequence of the target flight phase of the target aircraft; wherein the target flight phase includes at least one of the following: takeoff phase, approach phase, and landing phase; and at least one flight data segment is extracted from the flight data sequence using a sliding time window.

[0102] In some other embodiments, the processing module 602 is specifically used to output an abnormal event identification result when it is determined that the target aircraft has abnormal behavior during the target flight phase. The abnormal event identification result includes the event type, the start and end time of the event, and flight data information.

[0103] In the case of implementing the functions of the integrated modules described above in hardware, this embodiment of the invention provides a possible structural schematic diagram of the electronic device involved in the above embodiments. For example... Figure 7 As shown, the electronic device 700 includes: a processor 702, a communication interface 703, and a bus 704. Optionally, the electronic device 700 may also include a memory 701.

[0104] Processor 702 may implement or execute various exemplary logic blocks, modules, and circuits described in conjunction with the disclosure of this application. Processor 702 may be a central processing unit, a general-purpose processor, a digital signal processor, an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. It may implement or execute various exemplary logic blocks, modules, and circuits described in conjunction with the disclosure of this application. Processor 702 may also be a combination that implements computing functions, such as including one or more microprocessor combinations, a combination of a DSP and a microprocessor, etc.

[0105] The communication interface 703 is used to connect to other devices via a communication network. This communication network can be Ethernet, wireless access network, wireless local area network (WLAN), etc.

[0106] The memory 701 may be a read-only memory (ROM) or other type of static storage device capable of storing static information and instructions, random access memory (RAM) or other type of dynamic storage device capable of storing information and instructions, or electrically erasable programmable read-only memory (EEPROM), disk storage medium or other magnetic storage device, or any other medium capable of carrying or storing desired program code in the form of instructions or data structures and accessible by a computer, but is not limited thereto.

[0107] In one possible implementation, the memory 701 can exist independently of the processor 702. The memory 701 can be connected to the processor 702 via a bus 704 and is used to store instructions or program code. When the processor 702 calls and executes the instructions or program code stored in the memory 701, it can implement the behavior recognition method provided in this embodiment of the invention.

[0108] In another possible implementation, the memory 701 can also be integrated with the processor 702.

[0109] The 704 bus can be an extended industry standard architecture (EISA) bus, etc. The 704 bus can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 7 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.

[0110] Through the above description of the implementation methods, those skilled in the art can clearly understand that, for the sake of convenience and brevity, only the division of the above functional modules is used as an example. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the service calling device can be divided into different functional modules to complete all or part of the functions described above.

[0111] This application also provides a computer-readable storage medium. All or part of the processes in the above method embodiments can be executed by computer instructions instructing related hardware. The program can be stored in the aforementioned computer-readable storage medium, and when executed, it can include the processes of the above method embodiments. The computer-readable storage medium can be any of the foregoing embodiments or memory. The aforementioned computer-readable storage medium can also be an external storage device of the aforementioned service invocation device, such as a plug-in hard drive, smart media card (SMC), secure digital (SD) card, flash card, etc., equipped on the aforementioned service invocation device. Further, the aforementioned computer-readable storage medium can include both internal storage units of the aforementioned service invocation device and external storage devices. The aforementioned computer-readable storage medium is used to store the aforementioned computer program and other programs and data required by the aforementioned service invocation device. The aforementioned computer-readable storage medium can also be used to temporarily store data that has been output or will be output.

[0112] This application also provides a computer program product, which includes a computer program that, when run on a computer, causes the computer to perform any of the behavior recognition methods provided in the above embodiments.

[0113] The above are merely specific embodiments of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A behavior recognition method, characterized by, The method comprises: obtaining a navigation data segment of a target aircraft in a target flight phase; based on the navigation data segment, obtaining a navigation feature vector of the target aircraft; wherein the navigation feature vector is used to reflect the flight state of the target aircraft; based on the navigation feature vector, a rule base and a navigation template, determining an abnormal flight event existing in the target flight phase of the target aircraft; wherein the rule base is used to record the judgment rules of various abnormal flight events of the aircraft; the navigation template is used to indicate the navigation data of the aircraft when the abnormal flight event occurs.

2. The method of claim 1, wherein, The rule base includes the judgment rule of the target abnormal flight event, and the navigation template includes the navigation template of the target abnormal flight event; the judgment rule of the target abnormal flight event includes the threshold value of the target navigation data; the navigation template of the target abnormal flight event is used to indicate the navigation data of the aircraft when the target abnormal flight event occurs; The determination of the abnormal flight event existing in the target flight phase of the target aircraft based on the navigation feature vector, the rule base and the navigation template comprises: using the judgment rule of the target abnormal flight event to judge the navigation feature vector; in the case that the navigation feature vector is greater than the target navigation data threshold, calculating the similarity between the navigation feature vector and the navigation template of the target abnormal flight event; in the case that the similarity is greater than the similarity threshold, determining that the target aircraft exists the target abnormal flight event in the target flight phase.

3. The method of claim 1, wherein, The method comprises: obtaining a navigation data sequence of the target aircraft in the target flight phase; wherein the target flight phase comprises at least one of the following: take-off phase, approach phase, landing phase; using a sliding time window to extract at least one navigation data segment from the navigation data sequence.

4. The method of claim 1, wherein, The navigation data comprises at least one of the following: the pitch angle of the target aircraft, the slope of the target aircraft, the vertical speed of the target aircraft, the thrust of the target aircraft, and the configuration parameters of the target aircraft.

5. The method of claim 1, wherein, The navigation feature vector comprises at least one of the following: the change rate of the navigation data, the average value of the navigation data, the standard deviation of the navigation data, the peak-valley distribution of the navigation data, the slope of the navigation data, and the oscillation characteristics of the navigation data.

6. The method of claim 1, wherein, The method further comprises: in the case that it is determined that the target aircraft exists abnormal behavior in the target flight phase, outputting an abnormal event identification result, the abnormal event identification result comprising an event type, an event start and end time and the navigation data information.

7. A behavior recognition apparatus characterized by comprising: The method comprises: an acquisition module and a processing module; The acquisition module is used to obtain a navigation data segment of a target aircraft in a target flight phase; The processing module is used to obtain a navigation feature vector of the target aircraft based on the navigation data segment; wherein the navigation feature vector is used to reflect the flight state of the target aircraft; The processing module is configured to determine an abnormal flight event of the target aircraft existing in the target flight phase based on the navigation feature vector, a rule base, and a navigation template; the rule base is configured to record judgment rules of various abnormal flight events of the aircraft; and the navigation template is configured to indicate navigation data of the aircraft when the abnormal flight event occurs.

8. An electronic device, comprising: The computer device comprises a processor and a memory, the processor is coupled to the memory; the memory is configured to store computer instructions, the computer instructions are loaded and executed by the processor to enable the computer device to implement the behavior identification method according to any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that, The computer readable storage medium comprises computer execution instructions, when the computer execution instructions run on the computer, the computer execution instructions enable the computer to execute the behavior identification method according to any one of claims 1 to 6.

10. A computer program product, characterised in that, The computer program product comprises a computer program, when the computer program runs on the electronic device, the computer program enables the electronic device to execute the behavior identification method according to any one of claims 1 to 6.

Images