Abnormal flight overrun event identification method and system

Abstract

The invention provides a method and a system for identifying an abnormal flight overrun event, which are applied to the technical field of data processing. The method comprises the following steps: processing flight path data and flight attitude data based on historical abnormal flight overrun event data and geographical environment information and flight rule information of different airspaces to generate flight state characteristic parameter information; processing the aircraft performance parameter data to generate aircraft performance element associated information; processing the flight state characteristic parameter information and the airplane performance element association information to generate abnormal flight event preliminary identification information; processing the flight trajectory data based on the abnormal flight event preliminary identification information to generate abnormal flight overrun judgment information; and processing the abnormal flight event preliminary identification information and the abnormal flight overrun judgment information based on the target abnormal flight overrun event early warning model, and generating abnormal flight overrun event early warning result information.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and particularly to a method and system for identifying abnormal flight overlimit events. Background Art

[0002] In the aviation field, ensuring flight safety has always been of utmost importance. With the continuous improvement of aircraft performance and the increasing complexity of flight missions, accurately identifying abnormal flight overlimit events has become even more critical. Currently, the processing of flight data in the industry mostly focuses on the analysis of single dimensions or a few parameters, making it difficult to comprehensively capture abnormal conditions during flight. For example, traditional methods may only focus on obvious deviations in flight trajectories, but ignore the combined effects of flight attitudes, aircraft performance factors, and external environmental factors. When analyzing flight trajectories, the geographical environment information and flight rule information of different airspaces are not fully combined, resulting in limitations in the judgment of abnormal situations. At the same time, the utilization of historical abnormal flight overlimit event data is not sufficient, and the laws therein cannot be effectively mined and applied to the assessment of the current flight state. In addition, existing abnormal identification technologies also have deficiencies in data fusion and algorithm application, unable to accurately generate abnormal flight risk assessment values, making it difficult to provide reliable early warnings for flight safety in advance, posing potential threats to flight safety.

[0003] It should be noted that the information disclosed in the above background art section is only used to enhance the understanding of the background of the present disclosure, and thus includes information that does not constitute the prior art known to those of ordinary skill in the art. Summary of the Invention

[0004] The purpose of this application is to provide a method and system for identifying abnormal flight overlimit events, which can at least overcome the problems existing in the prior art to a certain extent. By focusing on the accurate identification of abnormal flight overlimit events, first collect information from multiple aspects such as flight trajectories, attitudes, performance parameters, historical events, geographical environment, and flight rules. Then, based on historical, geographical, and rule data, process flight trajectory, attitude, and performance parameter data to obtain the associated information between flight state characteristic parameters and aircraft performance factors. Next, use the random forest algorithm to process this information to obtain preliminary identification information of abnormal flight events, and combine flight trajectory data to generate abnormal flight overlimit judgment information. Finally, analyze relevant information through the target warning model, generate a risk assessment value, adjust the warning parameters, and convert to obtain a warning result, helping relevant personnel ensure flight safety and reduce risks.

[0005] Other features and advantages of this application will become apparent through the following detailed description, or be learned in part through the practice of the present invention.

[0006] According to one aspect of the present application, a method for identifying abnormal flight overlimit events is provided, including: obtaining flight trajectory data, flight attitude data, aircraft performance parameter data, historical abnormal flight overlimit event data, geographical environment information and flight rule information of different airspaces; processing the flight trajectory data and flight attitude data based on the historical abnormal flight overlimit event data, geographical environment information and flight rule information of different airspaces to generate flight state characteristic parameter information; processing the aircraft performance parameter data based on the historical abnormal flight overlimit event data, geographical environment information and flight rule information of different airspaces to generate aircraft performance element association information; processing the flight state characteristic parameter information and aircraft performance element association information to generate preliminary identification information of abnormal flight events; processing the flight trajectory data based on the preliminary identification information of abnormal flight events to generate abnormal flight overlimit judgment information; processing the preliminary identification information of abnormal flight events and abnormal flight overlimit judgment information based on the target abnormal flight overlimit event warning model to generate warning result information of abnormal flight overlimit events.

[0007] Another aspect of the present application provides an apparatus for identifying abnormal flight overlimit events, characterized by including: an acquisition module for obtaining flight trajectory data, flight attitude data, aircraft performance parameter data, historical abnormal flight overlimit event data, geographical environment information and flight rule information of different airspaces; a processing module for processing the flight trajectory data and flight attitude data based on the historical abnormal flight overlimit event data, geographical environment information and flight rule information of different airspaces to generate flight state characteristic parameter information; processing the aircraft performance parameter data based on the historical abnormal flight overlimit event data, geographical environment information and flight rule information of different airspaces to generate aircraft performance element association information; processing the flight state characteristic parameter information and aircraft performance element association information to generate preliminary identification information of abnormal flight events; processing the flight trajectory data based on the preliminary identification information of abnormal flight events to generate abnormal flight overlimit judgment information; processing the preliminary identification information of abnormal flight events and abnormal flight overlimit judgment information based on the target abnormal flight overlimit event warning model to generate warning result information of abnormal flight overlimit events.

[0008] According to yet another aspect of the present application, a computer-readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a second processor, the above-mentioned method for identifying abnormal flight overlimit events is implemented.

[0009] A method and system for identifying abnormal flight over-limit events provided by this application, where the server focuses on the accurate identification of abnormal flight over-limit events. First, collect information from multiple aspects such as flight trajectories, attitudes, performance parameters, historical events, geographical environments, and flight rules. Then, based on historical, geographical, and rule data, process the flight trajectory, attitude, and performance parameter data to obtain the correlation information between flight state characteristic parameters and aircraft performance elements. Next, use the random forest algorithm to process this information to obtain preliminary identification information of abnormal flight events, and generate abnormal flight over-limit judgment information in combination with flight trajectory data. Finally, analyze relevant information through the target warning model to generate a risk assessment value, adjust the warning parameters, and convert to obtain a warning result to assist relevant personnel in ensuring flight safety and reducing risks.

[0010] It should be understood that the above general description and the following detailed description are only exemplary and explanatory, and do not limit the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] Figure 1 The flowchart shows a method for identifying an abnormal flight over-limit event provided by an embodiment of this application; Figure 2 The structural schematic diagram shows a device for identifying an abnormal flight over-limit event provided by an embodiment of this application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0012] The following describes the preferred embodiments of the present invention with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are only for the purpose of illustrating and explaining the present invention, and are not used to limit the present invention.

[0013] Next, in combination with Figure 1 to describe the method for identifying an abnormal flight over-limit event according to an exemplary embodiment of this application. In one embodiment, this application also proposes a method and system for identifying an abnormal flight over-limit event. As Figure 1 shown, this method is applied to a server and includes: S101, obtain flight trajectory data, flight attitude data, aircraft performance parameter data, historical abnormal flight over-limit event data, geographical environment information of different airspaces, and flight rule information.

[0014] In one implementation, the actual path information of the aircraft during flight includes the changes in longitude and latitude, altitude over time, etc. These data directly reflect the flight route of the aircraft. By analyzing the trajectory data, it can be determined whether the aircraft deviates from the scheduled route and whether the flight path meets the airspace planning and navigation requirements, providing a basic basis for subsequent judgments on whether the flight is abnormal. The flight attitude data includes information such as the aircraft's pitch angle, roll angle, and yaw angle, which reflects the aircraft's attitude changes in the air. Whether the flight attitude is stable is crucial to flight safety. Abnormal attitude changes indicate that the aircraft's control system has failed or encountered special circumstances, and are an important indicator for evaluating the flight status.

[0015] Aircraft performance parameter data covers engine thrust, fuel consumption, avionics system operating parameters, fuselage structure stress conditions and other data. Engine thrust and fuel consumption reflect the performance of the power system, avionics system parameters are related to the normal operation of communication, navigation, monitoring and other functions, and fuselage structure stress data are related to the structural safety of the aircraft. These parameters can help understand the working status of each system of the aircraft and determine whether there are performance abnormalities. Historical abnormal flight limit event data records detailed information on abnormal flight limit events that occurred in the past, such as event type (altitude limit, speed limit, etc.), airspace where it occurred, flight phase, impact and severity. By analyzing historical data, the laws and patterns of abnormal events can be summarized, which can provide a reference for the analysis of the current flight situation and assist in determining whether there are potential risks in the current flight status.

[0016] The geographical environment information of different airspaces includes the topography (mountains, plains, etc.) and meteorological conditions (temperature, air pressure, wind speed and direction, etc.) of the airspace. Complex terrain affects the flight trajectory and performance of the aircraft, and severe weather conditions are an important threat to flight safety. Understanding geographical environment information helps to analyze its impact on the current flight and more accurately evaluate the flight status in combination with other data. Flight rules information involves flight restrictions, altitude regulations, speed limits, communication requirements and other rules in different airspaces. Flight rules are important guidelines to ensure flight safety and order, and aircraft must strictly abide by them during flight. Obtaining flight rules information can be used to compare the actual flight data of the aircraft, determine whether the aircraft is operating in violation of regulations, and provide a basis for identifying abnormal flights.

[0017] S102, based on the historical abnormal flight limit exceeding event data and the geographical environment information and flight rule information of different airspaces, the flight trajectory data and the flight attitude data are processed to generate the flight status characteristic parameter information.

[0018] In one implementation, feature extraction is performed on historical abnormal flight limit exceeding event data to generate event type features, occurrence airspace features, flight phase features, and event severity features. If there is a historical abnormal flight event, the event type is "altitude exceeding limit", and this type feature can clarify the nature of the abnormality; the occurrence airspace is "mountainous airspace near a certain airport", which determines the location of the event, and the complex terrain of the mountainous airspace may be related to the abnormality; the flight phase is in the "take-off and climb phase", and different flight phases have different requirements for aircraft performance and operation. There are specific reasons for the abnormality in this phase; the severity of the event is assessed as "serious", which affects the safe flight of the aircraft. These features provide key information for analyzing similar abnormal events. The geographical environment information and flight rules information of different airspaces are quantitatively analyzed and processed to generate quantitative factors for airspace terrain complexity, quantitative factors for airspace meteorological impact, and quantitative factors for strictness of rules. In a certain airspace, through the analysis of terrain data, the quantification factor of the airspace terrain complexity is obtained to be 0.8 (the value range is 0-1, and the larger the value, the more complex the terrain), which means that the terrain of the airspace is complex and may affect the flight trajectory and performance of the aircraft; based on meteorological data, the quantification factor of the airspace meteorological impact is calculated to be 0.6, indicating that meteorological conditions have a certain impact on flight, such as strong winds and low visibility; the flight rules in this airspace are strict, and after quantitative analysis, the quantification factor of the strictness of the rules is obtained to be 0.7, which means that aircraft flying in this airspace need to strictly abide by many rules, and any illegal operations may cause abnormalities.

[0019] The flight trajectory data is processed by feature extraction to generate trajectory deviation features, trajectory change frequency features, and trajectory speed change features. In a certain flight, the actual flight trajectory of the aircraft is compared with the scheduled route. The trajectory deviation feature shows that the average deviation distance is 5 kilometers, indicating that the aircraft has deviated from the normal route; over a period of time, the trajectory change frequency feature shows that there is an obvious trajectory change every 10 minutes, and the change is relatively frequent, which may be abnormal; the trajectory speed change feature shows that the speed of the aircraft fluctuates greatly in a certain period of time, and the maximum speed change reaches 20 knots, which may affect flight safety and stability. The flight attitude data is processed by feature extraction to generate attitude angle deviation features and attitude change rate features. During the flight, the pitch angle of the aircraft is compared with the standard attitude. The attitude angle deviation feature shows that the maximum deviation reaches 5 degrees, which exceeds the normal range and may affect the lift and flight direction of the aircraft; the attitude change rate feature shows that at a certain moment, the change rate of the aircraft's roll angle is 3 degrees per second, which changes too fast, which poses a challenge to the aircraft's controllability and stability, suggesting that there are problems with the aircraft's control system or encounters special circumstances.

[0020] Based on the airspace terrain complexity quantification factor, the airspace meteorological impact quantification factor, and the rule strictness quantification factor, the trajectory deviation characteristics, the trajectory change frequency characteristics, the trajectory speed change characteristics, as well as the attitude angle deviation characteristics and the attitude change rate characteristics are respectively fused to generate the trajectory comprehensive quantification characteristics and the attitude comprehensive quantification characteristics. Combining the quantification factors of a certain airspace with the flight trajectory and attitude characteristics, assuming that the airspace terrain complexity quantification factor is 0.8 and the average deviation distance in the trajectory deviation characteristics is 5 kilometers, after fusion processing through the weighted average algorithm, the trajectory comprehensive quantification characteristic value is 0.7 (the value range is 0 - 1, and the larger the value, the higher the probability of abnormality), indicating that there is a certain abnormal risk in the current trajectory deviation in this complex terrain airspace; similarly, the airspace meteorological impact quantification factor is 0.6, and the roll angle change rate in the attitude change rate characteristics is 3 degrees per second. After fusion processing, the attitude comprehensive quantification characteristic value is 0.65, indicating that the meteorological conditions and the current attitude change rate together show that there is a certain abnormal tendency in the aircraft attitude.

[0021] Based on the event type characteristics, the occurrence airspace characteristics, the flight phase characteristics, and the event severity characteristics, the trajectory comprehensive quantification characteristics and the attitude comprehensive quantification characteristics are analyzed and processed to generate the flight state characteristic parameter information. Combining the historical event characteristics and the fused quantification characteristics mentioned above, if the altitude overrun event in the historical events occurs during the takeoff and climb phase in a mountainous airspace and the severity is "severe", the current trajectory comprehensive quantification characteristic value is 0.7, and the attitude comprehensive quantification characteristic value is 0.65. Through analysis and processing, the generated flight state characteristic parameter information shows that the current flight state has a high risk, which is similar to the characteristics of the historical altitude overrun event. It is necessary to closely monitor the flight trajectory and attitude changes of the aircraft and take timely measures to avoid the occurrence of abnormal flight overrun events.

[0022] S103, based on the historical abnormal flight overrun event data and the geographical environment information and flight rule information of different airspaces, the aircraft performance parameter data is processed to generate the aircraft performance element association information.

[0023] In one implementation, the geographical environment information and flight rule information of different airspaces are quantitatively analyzed and processed to generate the airspace altitude quantification factor, the airspace meteorological complexity quantification factor, and the flight rule restriction intensity quantification factor. The altitude of the airspace has various impacts on aircraft flight. Assuming that the average altitude of this specific airspace is 3000 meters, in order to measure its impact on flight, it needs to be compared with the standard altitude. The standard altitude is the altitude reference value used for unified reference in the aviation field. Different flight missions and aircraft types have relatively stable performance at the standard altitude.

[0024] Standard altitude ; The average altitude H of this specific airspace is 3000 meters. The formula for the relationship between air density and altitude: , where is the air density at standard altitude (assumed to be known as ), is the air density at altitude H. The coefficient relating engine thrust to air intake volume (obtained through engine performance tests, =0.8, indicating that when the air intake volume decreases by 1%, the thrust decreases by 0.8%). The coefficient relating lift to air density and speed (obtained through research on the aircraft lift system, =0.6, indicating the degree of influence of air density change on lift). The coefficient relating the impact of speed adjustment on maintaining flight altitude (obtained through flight mechanics research, =0.5, indicating the importance of speed adjustment for maintaining altitude after lift reduction). The weight of engine thrust influence = 0.4, the weight of lift influence = 0.4, and the weight of speed adjustment influence = 0.2.

[0025] Calculate the air density ; The proportion of air density change ; Calculate the degree of influence on engine thrust: the proportion of engine thrust reduction ; Calculate the degree of influence on lift: the proportion of lift reduction ; Calculate the proportion of speed adjustment required to maintain flight altitude: the proportion of speed increase required ; The comprehensive influence value . The maximum value of the comprehensive influence value (determined through calculations and experience at a large number of different altitudes, assumed ), and the quantization factor of the airspace altitude.

[0026] After calculation, the quantization factor of the altitude in the airspace is 0.6 (value range 0 - 1). This means that the impact of the altitude in this airspace on flight is at a relatively high level. For example, a higher altitude will reduce the air density, resulting in a decrease in the air intake of the aircraft engine, thereby affecting the thrust output of the engine; at the same time, the decrease in air density will also reduce the lift generated by the aircraft wing, and the aircraft needs a higher speed to maintain the flight altitude, which poses higher requirements for the power system and flight operation of the aircraft.

[0027] Meteorological conditions are one of the key factors affecting flight safety. Severe convective weather often occurs in this specific airspace, making the meteorological situation complex and changeable. In order to quantify this complexity, it is necessary to analyze meteorological data using a specific quantization model. The diversity of meteorological data is reflected in many aspects, including the changes in elements such as temperature, humidity, air pressure, wind speed, wind direction, and precipitation. Severe convective weather will cause these meteorological elements to change violently in a short period of time. For example, in the severe convective area, the wind speed shear in the vertical direction may be very large, and the aircraft will be subjected to strong turbulence when crossing, seriously affecting the flight stability; the sudden change of wind direction in the horizontal direction will also bring difficulties to the navigation and control of the aircraft. By using statistical analysis methods, analyze the change frequency, amplitude of each element in the historical meteorological data, and the correlation between different elements, so as to obtain a quantization value that can reflect the meteorological complexity of this airspace. In this example, the quantization factor of the meteorological complexity in the airspace is 0.8, indicating that the meteorological conditions in this airspace have a greater impact on flight, and the aircraft faces higher meteorological risks when flying in this airspace. Pilots need to pay special attention to meteorological changes and adjust flight strategies in a timely manner.

[0028] Flight rules are important guidelines to ensure flight safety and order, and the degree of restriction of flight rules varies in different airspaces. This specific airspace has strict restrictions on the flight speed, altitude change, etc. of aircraft. Evaluating the intensity of flight rule restrictions requires comprehensive consideration of multiple factors, such as the traffic flow in this airspace, the layout of surrounding airports, and the impact of military activities on flight. If the traffic flow in this airspace is large, in order to avoid conflicts between aircraft, strict control will be imposed on the flight speed and altitude change; if there is a military activity area nearby, in order to ensure safety, many restrictions will also be imposed on the flight of civilian aircraft. Through professional evaluation methods, analyze and quantify these restrictive factors, and obtain a quantization factor of the flight rule restriction intensity of 0.7. This means that when flying in this airspace, the aircraft must strictly abide by these rules, and any violation may lead to a safety accident. For example, when flying in this airspace, the speed adjustment of the aircraft must be carried out at the specified rate, and the altitude change also needs to be within the specified range, which poses high requirements for the operation accuracy and rule-abiding awareness of pilots.

[0029] Quantification factors for airspace altitude, quantification factors for the complexity of airspace meteorology, and quantification factors for the intensity of flight rule restrictions. These quantification factors play a crucial role in subsequent analyses of the relationship between aircraft performance and flight environment. Airlines can formulate more reasonable flight plans for different flights based on these quantification factors. For flights bound for this airspace, in aircraft selection, models more suitable for high altitudes, complex meteorological conditions, and strict rule restrictions can be chosen; during the pre-flight preparation stage, flight crew can understand the difficulties that may be encountered during the flight based on the quantification factors and take preventive measures in advance, such as formulating emergency plans for special meteorological conditions and familiarizing themselves with the special flight rules of this airspace. During the flight, the real-time monitoring system can combine these quantification factors with the actual performance data of the aircraft to more accurately evaluate the operating state of the aircraft, timely detect potential safety hazards, and provide strong support for ensuring flight safety.

[0030] Feature extraction processing is performed on the aircraft performance parameter data to generate thrust features, fuel consumption features, avionics system performance features, and airframe structural stress features. During the flight of a certain flight, data is collected through sensors. The average thrust of the engine at a specific stage is 100 kN, which is the specific manifestation of the thrust feature; the fuel consumption of the aircraft per 100 km of flight at this stage is 500 liters, which is the fuel consumption feature. In terms of the avionics system, the communication signal strength is stable above 90%, and the navigation and positioning error is within the allowable range, which constitute the avionics system performance features. The stress on the key parts of the airframe monitored by the stress sensor is 50 MPa, which belongs to the airframe structural stress feature. These features reflect the real-time working state of each system of the aircraft during the flight.

[0031] Based on the airspace altitude quantization factor, the airspace meteorological complexity quantization factor, and the flight rule restriction intensity quantization factor, the thrust characteristics and fuel consumption characteristics are fused to generate the airspace and power correlation quantization characteristics. The airspace altitude quantization factor is 0.6, indicating that the impact of the airspace altitude on flight is above the medium level. Higher altitudes will reduce the air density, resulting in a decrease in the intake air volume of the aircraft engine, affecting the thrust output. At the same time, the lift of the aircraft will also decrease, and more power is required to maintain the flight altitude. The meteorological complexity quantization factor is 0.8, meaning that the meteorological conditions in this airspace are complex. Severe weather such as strong convective weather will cause the aircraft to encounter turbulence, wind shear, etc., which puts higher requirements on the stability and response speed of the aircraft power system. The flight rule restriction intensity quantization factor is 0.7, indicating that the airspace has strict restrictions on the speed and altitude changes of the aircraft. Under such rules, the aircraft frequently adjusts its speed and altitude, which will increase the workload of the power system. The aircraft thrust characteristic is 100 kN, which reflects the output ability of the engine; the fuel consumption characteristic is 500 liters per 100 kilometers, which reflects the fuel consumption of the aircraft at this stage. The principle and application of the fusion algorithm are to comprehensively consider these factors through a pre-set fusion algorithm (such as weighted average, etc.). The weighted average algorithm assigns different weights according to the importance of each factor's impact on the aircraft power system. Suppose that after a large number of experiments and data analyses, it is determined that the weight of the airspace altitude quantization factor is 0.2, the weight of the meteorological complexity quantization factor is 0.4, the weight of the rule restriction intensity quantization factor is 0.3, the weight of the thrust characteristic is 0.05, and the weight of the fuel consumption characteristic is 0.05. Then the calculation of the airspace and power correlation quantization characteristic value is as follows: This fusion algorithm integrates different factors, converts a variety of seemingly independent data into a quantization characteristic value that comprehensively reflects the airspace and power correlation, and can more intuitively evaluate the state of the aircraft power system in a specific environment.

[0032] The obtained airspace and power correlation quantization eigenvalue is 0.75, indicating that under the airspace environment and rule restrictions, there is a certain abnormal tendency in the comprehensive performance of the aircraft power system. From the data perspective, 0.75 is at a relatively high level, meaning that the combined impact of the airspace environment and rules on the power system is relatively large. In this airspace, the change in air density caused by high altitude, the additional power requirements under complex meteorological conditions, and the frequent power adjustments brought about by strict flight rules together exert great pressure on the aircraft power system. For example, due to the relatively high quantization factor of meteorological complexity, when the aircraft encounters severe convective weather, the engine needs to frequently adjust the thrust to maintain flight stability and safety, which may lead to increased fuel consumption. Currently, there is a certain deviation between the thrust and fuel consumption performance and the ideal state under the airspace environment and rule requirements. Therefore, it is necessary to pay attention to the rationality of thrust and fuel consumption. Airlines and flight crews can take corresponding measures in advance based on this quantization eigenvalue, such as checking the engine performance and optimizing the flight plan to reduce unnecessary power consumption, ensuring flight safety and cost control.

[0033] Based on the airspace altitude quantization factor, airspace meteorological complexity quantization factor, and flight rule restriction intensity quantization factor, the performance characteristics of the avionics system and the force characteristics of the airframe structure are fused to generate the airspace and system correlation quantization characteristics. The linear weighting method is used to fuse various indicators. Let the airspace altitude quantization factor be A, the airspace meteorological complexity quantization factor be B, the flight rule restriction intensity quantization factor be C, the communication signal strength of the avionics system be D (value range 0 - 1), the navigation positioning error situation be E (the error degree is normalized to (0 - 1, the smaller the value, the smaller the error)), and the airframe structure force be F (also normalized to 0 - 1). Assume the weights of each item are , , , , , . Then the calculation formula for the airspace and system correlation quantization eigenvalue S is: .

[0034] During the flight of the aircraft, accurately evaluating the adaptability of the avionics system and the airframe structure to the flight environment is crucial for ensuring flight safety. Taking the flight of a certain flight in a specific airspace as an example, the calculation process and significance of its airspace and system correlation quantization eigenvalue are analyzed in detail. The airspace altitude quantization factor A = 0.6 in this airspace, indicating that the altitude has a certain impact on flight; the airspace meteorological complexity quantization factor B = 0.8, indicating that the meteorological conditions are complex and there may be severe weather such as severe convection; the flight rule restriction intensity quantization factor C = 0.7, meaning that this airspace has relatively strict restrictions on the speed, altitude change, etc. of the aircraft flight.

[0035] The avionics system communication signal strength is stable at more than 90%, that is, D=0.9, which indicates that the communication link is relatively reliable and can ensure normal information exchange between the aircraft and the outside world. The navigation positioning error is within the allowable range. Assuming that it is normalized to E=0.9 (the smaller the error, the closer the value is to 1), it means that the aircraft can accurately determine its own position and fly according to the scheduled route. The fuselage structure is subjected to a force of 50 MPa. After normalization, F=0.8 (assuming that it is normalized according to the structural design and safety standards of the aircraft). This value is within the safe range and reflects the pressure condition of the fuselage in the current flight state.

[0036] Substitute the above indicators into the linear weighted fusion algorithm formula: . The calculated quantitative characteristic value of the airspace-system association is 0.6. The calculated quantitative characteristic value of the airspace-system association is 0.6, which indicates that under the current airspace conditions and rules, the adaptability of the avionics system and the airframe structure to the environment is at a medium level. Although the avionics system currently has good communication, accurate navigation and positioning, normal force on the airframe structure, and stable overall operation, there are still potential risks.

[0037] Due to the high quantitative factor of meteorological complexity in this airspace, severe weather conditions may interfere with avionics system communications. For example, lightning activity in severe convective weather may cause strong electromagnetic interference, resulting in interruption or distortion of communication signals, affecting communication between pilots and the ground and remote control of the aircraft. At the same time, strong winds and airflow changes under complex meteorological conditions may cause sudden changes in the force of the fuselage structure, exceeding the safety range and threatening the safety of the aircraft structure. Based on the medium adaptation level reflected by the quantitative eigenvalues, airlines and flight crews need to continue to pay attention to changes in environmental factors and formulate response measures in advance. For example, adjust the flight altitude or route before the weather conditions deteriorate, and strengthen the monitoring frequency of the avionics system and fuselage structure to ensure flight safety.

[0038] Based on the event type characteristics, occurrence airspace characteristics, flight phase characteristics, and event severity characteristics in the historical abnormal flight overlimit event data, analyze and process the quantitative characteristics of the association between airspace and power and the quantitative characteristics of the association between airspace and system to generate aircraft performance element association information. There was a serious abnormal event in the historical data. The event type was a power system failure, which occurred in an airspace similar to the current one, and the flight phase was the cruise phase. When the current flight is in this airspace, the quantitative characteristic value of the association between airspace and power is 0.75, and the quantitative characteristic value of the association between airspace and system is 0.6. Through analysis, it is found that the performance of the power system of the current flight in this airspace has certain similarities with the historical abnormal event. The generated aircraft performance element association information prompts that in the current flight phase, it is necessary to focus on the operating status of the power system. Although the avionics system and the airframe structure are currently within the normal range, continuous monitoring should also be carried out to prevent abnormal situations from occurring to ensure flight safety.

[0039] S104, process the flight state characteristic parameter information and the aircraft performance element association information to generate preliminary identification information of abnormal flight events.

[0040] In one implementation, extract and classify the flight state characteristic parameter information to generate abnormal flight trajectory change information, abnormal flight attitude fluctuation information, and abnormal flight state stability information. Suppose a certain flight shows that during the flight, the average deviation distance of its flight trajectory data from the predetermined route suddenly increases from the normal 0.5 km to 2 km, and the deviation distance fluctuates greatly for a period of time. This belongs to the abnormal flight trajectory change information. In terms of flight attitude, the pitch angle of the aircraft should originally be maintained within a relatively stable range, such as within plus or minus 3 degrees, but during a certain period, the pitch angle fluctuates greatly between plus and minus 5 degrees frequently. This is the abnormal flight attitude fluctuation information. The abnormal flight state stability information can be judged by comprehensively analyzing the flight trajectory and attitude data. For example, during the cruise phase of the aircraft, normally the speed and attitude should be relatively stable, but at this time, the aircraft speed fluctuates greatly, and at the same time, the attitude angle also changes frequently, indicating that the flight state stability of the aircraft is abnormal.

[0041] Extract and classify the associated information of aircraft performance elements to generate abnormal information on aircraft power performance, avionics system performance, and abnormal stress on the airframe structure. In terms of aircraft power performance, if the thrust of the engine drops from the rated 120 kN to 80 kN within a short period of time and cannot be restored to the normal level, this constitutes abnormal information on aircraft power performance. For the avionics system, if the communication signal strength drops suddenly from a stable 95% to 70%, and signal interruptions occur frequently, or the navigation positioning error exceeds the allowable range, such as the original positioning error being within 50 meters and now reaching 200 meters, these all belong to the abnormal information on avionics system performance. In terms of the stress on the airframe structure, when the stress on the key part of the aircraft wing monitored by sensors reaches 80 MPa, while the upper limit of the stress during normal operation of this part is 60 MPa, this indicates that the stress on the airframe structure is abnormal and belongs to the abnormal information on the stress of the airframe structure.

[0042] Based on the random forest algorithm, process the abnormal change information of the flight trajectory, abnormal fluctuation information of the flight attitude, abnormal information on flight state stability, abnormal information on aircraft power performance, abnormal information on avionics system performance, and abnormal information on the stress of the airframe structure to generate probability information on abnormal flight events and severity assessment information on abnormal flight events. The random forest algorithm will comprehensively consider the above various abnormal information. After being trained with a large amount of historical data, the random forest algorithm model performs calculations after receiving this abnormal information of the flight. If the probability information on abnormal flight events output by the model shows 0.8 (the value range is 0 - 1, and the larger the value, the higher the probability of the occurrence of abnormal events), this means that there is a high probability of an abnormal flight event for this flight. At the same time, the model assesses the severity of this abnormal flight event as "severe" because abnormalities have occurred in multiple aspects such as the flight trajectory, attitude, power, avionics, and airframe structure, which pose a greater threat to flight safety overall. This indicates that the current flight state of this flight is extremely unstable and immediate measures need to be taken, such as adjusting the flight strategy, checking the aircraft system, etc., to avoid possible serious flight accidents.

[0043] In another implementation, target data in the flight state characteristic parameter information and the aircraft performance element association information is marked and filtered based on the abnormal flight event probability information and the abnormal flight event severity assessment information to generate a potential abnormal flight event data screening result. In the aviation safety monitoring system, the accurate identification of abnormal flight events is a key link in ensuring flight safety. When the probability information of an abnormal flight event for a certain flight is 0.8 (a relatively high probability) and the severity assessment is "severe" after being processed by the random forest algorithm, the subsequent in-depth analysis and processing of relevant data become particularly important. As a powerful machine learning tool, the random forest algorithm integrates the prediction results of multiple decision trees. By learning from a large amount of historical flight data and the corresponding flight states and abnormal event situations, it can relatively accurately evaluate the probability and severity of an abnormal flight occurring in the current flight. In this example, an abnormal flight event probability of 0.8 means that there is an 80% chance of an abnormal flight event occurring under the current flight conditions, and the "severe" severity assessment indicates that once an abnormality occurs, it will pose a major threat to flight safety.

[0044] In terms of the flight state characteristic parameter information, the flight trajectory is an important basis for judging whether the flight is normal. Under normal circumstances, the average deviation distance of the aircraft from the flight path should be within 0.5 kilometers, which is the standard range to ensure flight safety and order. However, when it is monitored that the average deviation distance of the aircraft from the flight path reaches 2 kilometers, this significant deviation, combined with the previously obtained high abnormal probability and severity assessment, makes it a key potential abnormal indicator. A large trajectory deviation will cause the aircraft to enter a dangerous area, increasing the risk of collision with other aircraft. At the same time, it also means that there is a malfunction in the aircraft's navigation system or external interference. Therefore, based on the set abnormal judgment rules, this trajectory deviation data will be marked as potential abnormal data.

[0045] Among the aircraft performance element association information, the engine thrust is one of the core indicators for measuring the health status of the aircraft power system. For a normally operating engine, its thrust should remain stable under rated operating conditions. If the engine thrust suddenly drops from the rated 120 kN to 80 kN, this is an obvious performance abnormality. Considering the previously evaluated abnormal flight event severity of "severe", this thrust data is also marked as potential abnormal data. The sudden drop in engine thrust may cause the aircraft to be unable to maintain normal flight speed and altitude, affecting flight stability and safety, and may even lead to more serious flight accidents.

[0046] To comprehensively and accurately capture potential abnormal data, a series of reasonable thresholds and rules need to be set. These thresholds and rules are formulated based on a large amount of historical data, industry standards, and the experience of flight safety experts. For example, for flight trajectory deviation, an abnormal threshold is set where the average deviation distance exceeds 0.5 kilometers; for engine thrust, a thrust decrease exceeding a certain proportion (such as 20%) is regarded as abnormal. In the associated information of aircraft performance elements, for the avionics system communication signal, if the signal strength is lower than 90% and the duration exceeds a certain period (such as 1 minute), it is determined that the communication signal is unstable and is also included in the category of potential abnormal data. Through these set thresholds and rules, a comprehensive scan is conducted on the flight state characteristic parameter information and the associated information of aircraft performance elements. During the scan process, the system automatically compares the real-time data with the preset thresholds and rules, and when abnormal-condition data is found, it will be marked. After the marking is completed, all the marked data is screened out, and these data together constitute the screening result of potential abnormal flight event data. The screened potential abnormal data covers multiple aspects. In addition to the trajectory deviation data and engine thrust abnormal data within a specific time period mentioned above, it also includes avionics system communication signal instability data, etc. The instability of the avionics system communication signal may lead to communication interruption or errors between the aircraft and the ground control center as well as other aircraft, affecting flight command and coordination, and also posing a serious threat to flight safety. These potential abnormal data reflect the current abnormal conditions of the aircraft from different perspectives. They are interrelated and jointly point to the possibility of an abnormal flight event, providing an important basis for subsequent further analysis and decision-making.

[0047] Integrate and quantify the screening results of potential abnormal flight event data to generate preliminary identification information of abnormal flight events. The preliminary identification information of abnormal flight events is used to characterize the possibility of abnormal flight events, the flight systems involved, and the potential risk level. After obtaining the screening results of potential abnormal flight event data, integrate them. Classify different types of potential abnormal data according to flight systems. For example, classify trajectory deviation data into the flight trajectory system, thrust anomaly data into the power system, and unstable communication signal data into the avionics system, etc. Then perform quantification processing. Calculate a comprehensive anomaly index based on factors such as the number of abnormal data and the severity of the anomaly. The anomaly index of this flight is 0.7 (the value range is 0 - 1, and the larger the value, the more serious the abnormal situation). Based on this, the generated preliminary identification information of abnormal flight events shows that the possibility of this flight experiencing an abnormal flight event is relatively high (echoing the previous abnormal probability information), involving multiple flight systems such as flight trajectory, power, and avionics, and the potential risk level is relatively high. The specific manifestations are problems such as flight trajectory deviation, insufficient thrust in the power system, and unstable communication in the avionics system, which may pose a serious threat to flight safety, indicating that relevant personnel need to take timely measures for handling, such as further inspecting the aircraft system and adjusting the flight plan, etc.

[0048] S105, process the flight trajectory data based on the preliminary identification information of abnormal flight events to generate abnormal flight overrun judgment information.

[0049] In one implementation, extract and classify the preliminary identification information of abnormal flight events to generate information on the change in the possibility of abnormal events, the degree of abnormality of the flight systems involved, and the trend of potential risks. Assume that the preliminary identification information of abnormal flight events for a certain flight shows that the probability of an abnormal flight event is 0.8 (a relatively high probability) and involves multiple systems such as flight trajectory, power, and avionics. Extract the information on the change in the possibility of abnormal events from these information. For example, it is found that the probability of an abnormal flight event for this flight has rapidly increased from 0.3 to 0.8 over a certain period of time, indicating a relatively large change in the abnormal possibility. In terms of the information on the degree of abnormality of the flight systems involved, if the average deviation of the flight trajectory reaches 2 kilometers (normally within 0.5 kilometers), the engine thrust has decreased by 30% (far exceeding the normal fluctuation range), and the communication signal strength of the avionics system has dropped suddenly from a stable 95% to 70%, evaluate that the degree of abnormality of the flight trajectory system is relatively high, the degree of abnormality of the power system is relatively high, and the degree of abnormality of the avionics system is medium. For the information on the trend of potential risks, combined with the current state of the aircraft and environmental factors, if this flight is in an airspace with complex meteorological conditions and multiple of its own systems are abnormal, predict that the potential risk will show an upward trend in the next period of time.

[0050] Extract and classify the flight trajectory data to generate information on the change in trajectory deviation amplitude, abnormal change in trajectory speed, and sudden change in trajectory direction. During the flight of a certain flight, continuously monitor the flight trajectory data. The information on the change in trajectory deviation amplitude shows that the aircraft was originally deviating from the flight path by a small distance, but suddenly increased from a deviation of 0.5 km to 2 km at a certain moment, with a significant change in the deviation amplitude. In terms of the information on abnormal change in trajectory speed, the normal speed of the aircraft during the cruise phase should be stable around a certain value. However, the actual speed suddenly dropped from the standard 800 km / h to 700 km / h within a short period of time and then quickly rose to 850 km / h. The speed fluctuation exceeded the normal range, which belongs to the abnormal change in trajectory speed. The information on sudden change in trajectory direction is manifested as follows: when the aircraft is flying normally, it changes direction regularly according to the predetermined flight path, but at this time, there suddenly appears a situation where the turning angle far exceeds the normal turning range. For example, the normal turning angle is within 5 degrees, while this time the turning reaches 15 degrees. This is the information on sudden change in trajectory direction.

[0051] Based on the random forest algorithm, process the information on the change in the probability of abnormal events, the degree of abnormality of the flight system involved, the trend of potential risks, the change in trajectory deviation amplitude, the abnormal change in trajectory speed, and the sudden change in trajectory direction to generate information on the probability of abnormal flight overrun and information on the severity assessment of abnormal flight overrun. The random forest algorithm takes the above various information as input for analysis. After being trained with a large amount of historical data, the random forest algorithm model performs calculations after receiving this information of the flight. The model comprehensively considers various factors, such as a large change in the probability of abnormal events, a relatively high degree of abnormality in multiple flight systems, an upward trend in potential risks, and various abnormal change situations of the trajectory. It outputs the information on the probability of abnormal flight overrun as 0.9 (indicating a 90% probability of an abnormal flight overrun event) and assesses the severity of the abnormal flight overrun as "extremely severe". This is because multiple key indicators all show that the current state of the aircraft is extremely unstable, and once an abnormality occurs, it will have a devastating impact on flight safety.

[0052] Based on the abnormal flight overrun probability information and the evaluation information of the severity of abnormal flight overrun, mark and screen the target data in the preliminary identification information of abnormal flight events and the flight trajectory data to generate the screening results of potential abnormal flight overrun data. According to the previously obtained evaluation results of an abnormal flight overrun probability of 0.9 and a severity of "extremely severe", set the marking and screening rules. In the preliminary identification information of abnormal flight events, mark the relevant data involving a high degree of abnormality in the flight system, such as abnormal engine thrust data, unstable communication signal data of the avionics system, etc.; in the flight trajectory data, also mark the data with large changes in trajectory deviation amplitude, abnormal speed changes, and sudden direction changes. Then, screen out these marked data to form the screening results of potential abnormal flight overrun data. Screen out the data of continuous decline in engine thrust within a specific time period, the data of the signal strength of the avionics system below a specific threshold, and the data of the flight trajectory deviation amplitude exceeding the warning value, etc. These data are all closely related to abnormal flight overrun events.

[0053] Integrate and quantitatively process the screening results of potential abnormal flight overrun data to generate abnormal flight overrun judgment information. After obtaining the screening results of potential abnormal flight overrun data, the first thing to do is to systematically integrate these data. This means classifying and summarizing the data according to the flight system and the type of abnormality. The flight system mainly includes the flight trajectory system, the power system, the avionics system, the airframe structure system, etc., and the types of abnormality cover various situations such as deviation, abnormal change, and fault. For example, in the potential abnormal flight overrun data of a certain flight, regarding the flight trajectory system, classify all the data related to changes in trajectory deviation amplitude, abnormal trajectory speed changes, and sudden trajectory direction changes into one category; for the power system, integrate the data of abnormal engine thrust, such as the data record of the engine thrust suddenly dropping from the rated 120 kN to 80 kN, separately; in terms of the avionics system, classify the relevant data such as the sudden drop in communication signal strength and the navigation positioning error exceeding the allowable range into the category of avionics system abnormal data. Through this classification and summarization method, the abnormal situations existing in each flight system can be clearly sorted out, facilitating subsequent targeted analysis and processing.

[0054] After integrating the data, quantification processing is required to more precisely evaluate the severity of flight anomalies. The quantification process takes into account multiple key factors, among which the quantity of anomaly data and the severity of the anomalies are the core elements. Different anomaly data have different degrees of impact on flight safety, so corresponding weights need to be set for each type of anomaly. For severe deviations in the flight trajectory, since it is directly related to whether the aircraft will enter a dangerous area and collide with other aircraft, a higher weight is assigned; for some relatively minor anomalies, such as minor malfunctions of certain auxiliary equipment, the weight is set lower. The quantity of anomaly data, their respective severities, and the corresponding weights are comprehensively calculated. When calculating the anomaly index of a certain flight, through the weighted summation algorithm, the final anomaly index of this flight is 0.85 (the value range is 0 - 1, and the larger the value, the more severe the anomaly situation). This index intuitively reflects that the current anomaly situation of this flight is at a relatively high level, indicating a relatively high safety risk.

[0055] Based on the calculated comprehensive anomaly index of 0.85, the generated abnormal flight overlimit judgment information will provide a key basis for flight safety decisions. This information shows that the likelihood of this flight experiencing an abnormal flight overlimit event is extremely high. From the perspective of the flight systems involved, it covers multiple important systems such as flight trajectory, power, and avionics. These systems play an indispensable role in the normal operation and flight safety of the aircraft. The severity of the anomaly situation indicates that the current state of the aircraft has seriously deviated from the normal operation range. Specifically, the flight trajectory is severely deviated, and the aircraft has deviated far from the planned route. This not only increases the risk of conflict with other aircraft but also makes the aircraft enter dangerous areas such as no-fly zones or areas with complex terrain; the speed fluctuates abnormally, affecting the stability of the aircraft, increasing the difficulty of flight operations, and at the same time imposing additional stress on the aircraft structure, threatening the safety of the airframe; the engine thrust is insufficient, making the aircraft unable to maintain normal flight altitude and speed, and even causing the aircraft to stall, leading to serious accidents; the communication of the avionics system is unstable, interfering with the communication between the pilot and the ground control center as well as other aircraft, affecting flight command and coordination, and preventing the pilot from obtaining accurate flight information in a timely manner, further exacerbating the flight risk. In view of this, it poses a great threat to flight safety, and relevant personnel must immediately take emergency measures. For example, the air traffic control department should guide the aircraft to land as soon as possible, select a suitable alternate airport to ensure that the aircraft can land in a safe environment and avoid more serious consequences caused by continuing to fly in an abnormal state; or adjust the flight plan according to the actual situation, avoid dangerous areas, and re-plan the route to reduce flight risks and ensure the safety of passengers, crew, and the safe operation of the aircraft.

[0056] S106. Process the preliminary identification information of the abnormal flight event and the abnormal flight over-limit judgment information based on the target abnormal flight over-limit event warning model to generate the warning result information of the abnormal flight over-limit event.

[0057] In one implementation, analyze and process the preliminary identification information of the abnormal flight event and the abnormal flight over-limit judgment information based on the target abnormal flight over-limit event warning model to generate an abnormal flight risk assessment value. Suppose the preliminary identification information of an abnormal flight event of a certain flight shows that the abnormal change information of its flight trajectory indicates that the average deviation distance of the aircraft from the route reaches 2 kilometers (normally within 0.5 kilometers), and the abnormal fluctuation information of the flight attitude shows that the pitch angle frequently fluctuates greatly between plus and minus 5 degrees (normally within plus and minus 3 degrees); the abnormal flight over-limit judgment information shows that the probability information of the abnormal flight over-limit is 0.9 (indicating a 90% possibility of an abnormal flight over-limit event), and the severity assessment is "extremely severe". After receiving this information, the target abnormal flight over-limit event warning model analyzes factors such as the severity, occurrence probability of these abnormal situations, and the possible impact on flight safety, and calculates that the abnormal flight risk assessment value of this flight is 0.8 (the value range is 0 - 1, and the larger the value, the higher the risk). This means that the flight is currently in a high-risk state, and the possibility of an abnormal flight over-limit event is extremely high, and it is necessary to pay close attention and take corresponding measures.

[0058] Process the warning decision parameter vector inside the target abnormal flight over-limit event warning model based on the abnormal flight risk assessment value to generate a warning deviation correction vector. Some parameters originally set in the warning decision parameter vector inside the target abnormal flight over-limit event warning model, such as the sensitivity to the deviation of the flight trajectory and the assessment weight of the severity of the abnormal event, will be adjusted by the model after receiving the abnormal flight risk assessment value of 0.8. Suppose the original weight set for the deviation distance of the flight trajectory is 0.3, considering that the flight trajectory deviation of this flight is serious and the risk assessment value is high, its weight is adjusted to 0.4; the weight of a sub-parameter for the assessment of the severity of the abnormal event was originally 0.2 and is now adjusted to 0.3. Through these parameter adjustments, a new vector, that is, the warning deviation correction vector, is generated. This vector records the correction direction and amplitude of the original decision parameters by the model for the current flight situation, making the subsequent warning judgment of the model more in line with the actual flight situation.

[0059] Analyze and transform the warning deviation correction vector to generate the warning result information for abnormal flight overlimit events. After analyzing and transforming the warning deviation correction vector, the generated warning result information for abnormal flight overlimit events is shown as: "This flight is currently in a highly dangerous state and is extremely likely to have an abnormal flight overlimit event. The main abnormal situations include a serious deviation of the flight trajectory and extremely unstable flight attitude. It is expected that within the next 5 - 10 minutes, if no effective measures are taken, the aircraft may enter a no - fly zone and the risk of collision with other aircraft is extremely high. It is recommended to immediately guide the aircraft to make an emergency landing, select the nearest alternate airport, and closely monitor various parameters of the aircraft." Such warning result information clearly points out the problems, potential risks, and coping suggestions during the flight, providing a clear decision - making basis for relevant parties such as airlines and air traffic control departments, and helping to take timely measures to ensure flight safety.

[0060] In another implementation, the target abnormal flight overlimit event warning model receives the preliminary identification information of abnormal flight events and the overlimit judgment information of abnormal flights. The preliminary identification information includes various aspects such as abnormal changes in flight trajectory, abnormal fluctuations in flight attitude, and abnormal aircraft power performance; the overlimit judgment information covers key contents such as the probability of abnormal flight overlimit and severity assessment. The model comprehensively analyzes this information, considers factors such as the severity, occurrence probability of abnormal situations, and the possible impact on flight safety, and calculates the abnormal flight risk assessment value. For example, the preliminary identification information of an abnormal flight event of a certain flight shows a serious deviation of the flight trajectory and frequent large - amplitude fluctuations in flight attitude; the overlimit judgment information of abnormal flight overlimit indicates that the probability of abnormal flight overlimit is as high as 0.9 and the severity assessment is "extremely serious". Based on this, the model calculates that the abnormal flight risk assessment value of this flight is 0.8 (the value range is 0 - 1, and the larger the value, the higher the risk), meaning that this flight is in a high - risk state.

[0061] The overlimit events cover both real - occurring abnormal flight events and false - trigger events caused by problems such as data quality. The records of these events contain rich flight status information, which is crucial for understanding the nature of abnormal situations. When an overlimit event is triggered, the system automatically saves the relevant parameters at the triggering moment. Taking the touchdown moment as an example, in addition to speed and altitude, it may also include the touchdown angle, pitch angle, roll angle, engine thrust, etc. These parameters describe the flight state of the aircraft at a specific moment from different dimensions, providing comprehensive data support for subsequent analysis. For example, in a touchdown overlimit event, it is recorded that the speed at the touchdown moment is 250 knots, the altitude is 10 meters, the touchdown angle is 3 degrees, the pitch angle is 5 degrees, the roll angle is 2 degrees, and the engine thrust is 80%. These parameters form the feature vector of this event, reflecting the specific flight situation of the aircraft at that time.

[0062] The relevant parameters at the event trigger moment extracted are used as the input of the random forest algorithm. In the training phase, the random forest model is trained using known real and false over-limit event data. The model will learn the association patterns between different parameter combinations and the truth or falsehood of events. For example, when the touchdown speed is too high and the altitude is abnormal, the event is more likely to be a real abnormal event; while when there are obvious data fluctuations in some parameters or they do not conform to normal flight logic, it may be a false trigger event. For a new over-limit event, the random forest algorithm will judge the truth or falsehood of the event through the voting mechanism of multiple decision trees based on the input parameters. For example, for a new touchdown over-limit event, after inputting the parameters, through the voting of 100 decision trees, 80 decision trees consider the event to be real and 20 consider it to be false, then the model will determine that the event is a real abnormal event.

[0063] After identifying true and false events, the model will adjust the internal early warning decision parameter vector in combination with the abnormal flight risk assessment value. The abnormal flight risk assessment value is calculated by comprehensively considering factors such as the severity, occurrence probability of abnormal flight events, and the possible impact on flight safety. The early warning decision parameter vector contains multiple parameter weights related to abnormal flight judgment, and these weights determine the importance of each parameter in the early warning judgment. For example, parameters such as the deviation distance of the flight trajectory, abnormal flight speed, and change in flight attitude all have corresponding weights. When the abnormal flight risk assessment value of a certain flight is high, it indicates that the abnormal situation faced by the flight is relatively serious, and the early warning decision parameter vector needs to be adjusted to improve the accuracy of the early warning. For example, the weight of the deviation distance of the flight trajectory in the early warning judgment was originally set at 0.3. Considering that the flight trajectory deviation of this flight is serious and the risk assessment value is high, its weight is adjusted to 0.4, which means that in subsequent early warning judgments, the parameter of the deviation distance of the flight trajectory will play a more important role. Similarly, for the sub-parameters of the abnormal event severity assessment, such as the severity of the change in flight attitude, the original weight was 0.2, and now it is adjusted to 0.3 according to the risk assessment situation, so that the model pays more attention to the change in flight attitude when judging abnormal events.

[0064] By adjusting the early warning decision parameter vector, an early warning deviation correction vector is generated, which reflects the adjustment amplitude of the early warning judgment standard by the model according to the current flight situation. The role of the early warning deviation correction vector is to make the subsequent early warning judgment of the model more in line with the actual flight conditions. For example, in previous early warning judgments, due to the unreasonable setting of the weights of some parameters, the early warning of some abnormal situations was not timely or accurate enough. After generating the early warning deviation correction vector by adjusting the parameter weights, the model can more sensitively capture real abnormal situations, reduce false alarms, and improve the early warning ability for real abnormal events. The early warning deviation correction vector will be continuously updated according to new over-limit events and abnormal flight risk assessment values to ensure that the model always maintains an accurate judgment of the flight safety situation. The process of optimizing decision parameters by combining over-limit events and the random forest algorithm, through data collection, algorithm analysis, and parameter adjustment, enables the abnormal flight event early warning model to more intelligently and accurately warn of the flight safety situation, providing strong support for ensuring aviation safety.

[0065] Analyze and transform the early warning deviation correction vector to generate detailed early warning result information for abnormal flight over-limit events. This information not only clearly points out problems in flight, such as a serious deviation of the flight trajectory and extremely unstable flight attitude, but also explains potential risks. For example, it is predicted that within the next 5 - 10 minutes, if no effective measures are taken, the aircraft may enter a no-fly zone and the risk of collision with other aircraft is extremely high. At the same time, specific countermeasures are given, such as suggesting immediately guiding the aircraft to make an emergency landing, choosing the nearest alternate airport, and closely monitoring various parameters of the aircraft, providing clear decision-making basis for relevant parties such as airlines and air traffic control departments, and helping to ensure flight safety in a timely manner.

[0066] This application aims to achieve the precise identification of abnormal flight over-limit events. First, multi-source information such as flight trajectory, attitude, performance parameters, historical events, geographical environment, and flight rules is obtained. Then, based on historical data, geographical environment, and flight rules, the flight trajectory and attitude data, as well as the aircraft performance parameter data, are processed respectively to generate flight state characteristic parameter information and aircraft performance element association information. Next, the above two types of information are processed, and preliminary identification information for abnormal flight events is generated by means of the random forest algorithm, and then combined with flight trajectory data to obtain abnormal flight over-limit judgment information. Finally, the target abnormal flight over-limit event early warning model analyzes the preliminary identification and over-limit judgment information to generate an abnormal flight risk assessment value, adjusts the early warning decision parameter vector accordingly, and obtains the early warning result information through parsing and conversion, providing strong guarantee for flight safety, helping relevant personnel to timely grasp flight abnormal situations and take measures to reduce flight risks.

[0067] In one implementation, such as Figure 2As shown in the figure, the present application also provides an identification device for abnormal flight overlimit events, including: An acquisition module 201, configured to acquire flight trajectory data, flight attitude data, aircraft performance parameter data, historical abnormal flight overlimit event data, geographical environment information of different airspaces, and flight rule information; A processing module 202, configured to process the flight trajectory data and flight attitude data based on the historical abnormal flight overlimit event data, geographical environment information of different airspaces, and flight rule information to generate flight state characteristic parameter information; process the aircraft performance parameter data based on the historical abnormal flight overlimit event data, geographical environment information of different airspaces, and flight rule information to generate aircraft performance element association information; process the flight state characteristic parameter information and aircraft performance element association information to generate preliminary identification information of abnormal flight events; process the flight trajectory data based on the preliminary identification information of abnormal flight events to generate abnormal flight overlimit judgment information; process the preliminary identification information of abnormal flight events and the abnormal flight overlimit judgment information based on the target abnormal flight overlimit event warning model to generate warning result information of abnormal flight overlimit events.

[0068] Each embodiment in the present application is described in a related manner. For the same or similar parts among the embodiments, reference can be made to each other. Each embodiment focuses on the differences from other embodiments. In particular, for the embodiments of the identification method, electronic device, electronic equipment, and readable storage medium for evaluating abnormal flight overlimit events, since they are basically similar to the embodiment of the identification method for abnormal flight overlimit events described above, the description is relatively simple. For the relevant parts, reference can be made to the partial description of the embodiment of the identification method for abnormal flight overlimit events described above.

Claims

1. A method for identifying abnormal flight limit exceeding events, characterized in that: include: Obtain flight trajectory data, flight attitude data, aircraft performance parameter data, historical abnormal flight limit event data, geographical environment information of different airspaces and flight rules information; Based on historical abnormal flight limit exceeding event data and geographical environment information and flight rule information of different airspaces, the flight trajectory data and flight attitude data are processed to generate flight status characteristic parameter information; Based on historical abnormal flight limit exceeding event data and geographical environment information and flight rule information of different airspaces, the aircraft performance parameter data is processed to generate aircraft performance factor correlation information; Process the flight status characteristic parameter information and aircraft performance factor correlation information to generate preliminary identification information of abnormal flight events; Process the flight trajectory data based on the preliminary identification information of abnormal flight events to generate abnormal flight limit judgment information; Based on the target abnormal flight over-limit event warning model, the preliminary identification information of abnormal flight events and the abnormal flight over-limit judgment information are processed to generate abnormal flight over-limit event warning result information.

2. The method according to claim 1, characterized in that Based on historical abnormal flight limit exceeding event data and geographical environment information and flight rule information of different airspaces, the flight trajectory data and flight attitude data are processed to generate flight status characteristic parameter information, including: Perform feature extraction and processing on historical abnormal flight limit exceeding event data to generate event type features, occurrence airspace features, flight phase features, and event severity features; Quantitatively analyze and process the geographical environment information and flight rules information of different airspaces to generate quantitative factors of airspace terrain complexity, airspace meteorological impact, and rule strictness; Perform feature extraction on the flight trajectory data to generate trajectory deviation features, trajectory change frequency features, and trajectory speed change features; Perform feature extraction and processing on the flight attitude data to generate attitude angle deviation features and attitude change rate features; Based on the quantitative factors of airspace terrain complexity, airspace meteorological impact, and rule strictness, the trajectory deviation characteristics, trajectory change frequency characteristics, trajectory speed change characteristics, attitude angle deviation characteristics, and attitude change rate characteristics are fused to generate trajectory comprehensive quantitative characteristics and attitude comprehensive quantitative characteristics. Based on the event type characteristics, occurrence airspace characteristics, flight phase characteristics, and event severity characteristics, the trajectory comprehensive quantitative characteristics and attitude comprehensive quantitative characteristics are analyzed and processed to generate flight status characteristic parameter information.

3. The method according to claim 1, characterized in that Based on the historical abnormal flight limit exceeding event data and the geographical environment information and flight rules information of different airspaces, the aircraft performance parameter data is processed to generate the aircraft performance factor correlation information, including: Quantitatively analyze and process the geographical environment information and flight rules information of different airspaces to generate quantitative factors of airspace altitude, airspace meteorological complexity, and flight rules restriction intensity; Perform feature extraction and processing on aircraft performance parameter data to generate thrust characteristics, fuel consumption characteristics, avionics system performance characteristics, and airframe structure stress characteristics; Based on the quantitative factors of airspace altitude, airspace meteorological complexity, and flight rules restriction intensity, the thrust characteristics and fuel consumption characteristics are fused to generate quantitative characteristics of airspace and power correlation. Based on the quantitative factors of airspace altitude, airspace meteorological complexity, and flight rules restriction intensity, the avionics system performance characteristics and the aircraft structure stress characteristics are integrated to generate quantitative characteristics associated with airspace and system. Based on the event type characteristics, occurrence airspace characteristics, flight phase characteristics, and event severity characteristics in the historical abnormal flight limit event data, the quantitative characteristics of the association between airspace and power and the quantitative characteristics of the association between airspace and system are analyzed and processed to generate aircraft performance element association information.

4. The method according to claim 1, characterized in that Process the flight status characteristic parameter information and aircraft performance factor correlation information to generate preliminary identification information of abnormal flight events, including: Extract and classify the characteristic parameter information of the flight status to generate information on abnormal changes in the flight trajectory, abnormal fluctuations in the flight attitude, and abnormal stability of the flight status; Extract and classify the related information of aircraft performance elements to generate abnormal information of aircraft power performance, abnormal information of avionics system performance, and abnormal information of fuselage structure stress; Based on the random forest algorithm, the information on abnormal changes in flight trajectory, abnormal fluctuations in flight attitude, abnormal flight state stability, abnormal aircraft power performance, abnormal avionics system performance, and abnormal airframe structure stress are processed to generate abnormal flight event probability information and abnormal flight event severity assessment information.

5. The method according to claim 4, characterized in that Process the flight status characteristic parameter information and aircraft performance factor correlation information to generate preliminary identification information of abnormal flight events, including: Based on the abnormal flight event probability information and the abnormal flight event severity assessment information, target data in the flight status characteristic parameter information and the aircraft performance element association information are marked and screened to generate potential abnormal flight event data screening results; The screening results of potential abnormal flight event data are integrated and quantified to generate preliminary identification information of abnormal flight events. The preliminary identification information of abnormal flight events is used to characterize the possibility of abnormal flight events, the flight systems involved and the potential risk level.

6. The method according to claim 1, characterized in that Based on the preliminary identification information of abnormal flight events, the flight trajectory data is processed to generate abnormal flight limit judgment information, including: Extract and classify the preliminary identification information of abnormal flight events to generate information on the possibility of abnormal events, the degree of abnormality of the flight system involved, and potential risk trend information; Extract and classify flight trajectory data to generate information on trajectory deviation amplitude changes, trajectory speed abnormal changes, and trajectory direction mutations; Based on the random forest algorithm, the information on the possibility change of abnormal events, the degree of abnormality of the flight system involved, the potential risk trend, the trajectory deviation amplitude change, the trajectory speed abnormal change, and the trajectory direction mutation are processed to generate abnormal flight over-limit probability information and abnormal flight over-limit severity assessment information; Based on the abnormal flight over-limit probability information and the abnormal flight over-limit severity assessment information, the target data in the preliminary identification information of abnormal flight events and the flight trajectory data are marked and screened to generate a potential abnormal flight over-limit data screening result; The screening results of potential abnormal flight limit-exceeding data are integrated and quantified to generate abnormal flight limit-exceeding judgment information.

7. The method according to claim 6, characterized in that Based on the target abnormal flight over-limit event warning model, the preliminary identification information of abnormal flight events and the abnormal flight over-limit judgment information are processed to generate abnormal flight over-limit event warning result information, including: Based on the target abnormal flight over-limit event warning model, the preliminary identification information of abnormal flight events and the abnormal flight over-limit judgment information are analyzed and processed to generate an abnormal flight risk assessment value; Based on the abnormal flight risk assessment value, the warning decision parameter vector inside the target abnormal flight limit exceeding event warning model is processed to generate a warning deviation correction vector; The warning deviation correction vector is analyzed and converted to generate abnormal flight over-limit event warning result information.

8. A device for identifying abnormal flight limit exceeding events, characterized in that: The device comprises: An acquisition module is used to acquire flight trajectory data, flight attitude data, aircraft performance parameter data, historical abnormal flight limit exceeding event data, geographical environment information of different airspaces and flight rules information; The processing module is used to process the flight trajectory data and the flight attitude data based on the historical abnormal flight over-limit event data and the geographical environment information and flight rule information of different airspaces to generate the flight status characteristic parameter information; based on the historical abnormal flight over-limit event data and the geographical environment information and flight rule information of different airspaces, process the aircraft performance parameter data to generate the aircraft performance element correlation information; process the flight status characteristic parameter information and the aircraft performance element correlation information to generate the abnormal flight event preliminary identification information; based on the abnormal flight event preliminary identification information, process the flight trajectory data to generate the abnormal flight over-limit judgment information; based on the target abnormal flight over-limit event warning model, process the abnormal flight event preliminary identification information and the abnormal flight over-limit judgment information to generate the abnormal flight over-limit event warning result information.

9. An electronic device, characterized in that: include: a first processor; and a memory for storing executable instructions of the first processor; The first processor is configured to execute the method for identifying abnormal flight limit exceeding events as described in any one of claims 1 to 7 by executing the executable instructions.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the second processor, the method for identifying abnormal flight limit exceeding events described in any one of claims 1 to 7 is implemented.

Images