CN118736904A - A civil aircraft stability assessment method based on real-time flight parameters - Google Patents

Abstract

The present invention relates to the field of flight safety technology, specifically a civil aircraft stability assessment method based on real-time flight parameters, comprising the following steps: real-time acquisition of aircraft parameter data; filtering, denoising, correction, calibration, cleaning and data correction of the parameter data collected in real time to ensure the accuracy and reliability of the parameter data collected in real time. The present invention uses real-time flight parameter data to establish a stability assessment model, learns the stability characteristics of the aircraft through a machine learning algorithm, and can more accurately assess the stability of the aircraft. Combined with flight dynamics theory and advanced algorithms, the stability of the aircraft can be quantitatively assessed to provide an objective basis for judgment. At the same time, the real-time flight data of the aircraft can be collected and processed in real time, and an alarm can be issued according to the evaluation results, which enables the pilot to accurately and quickly understand the stability of the aircraft and take corresponding countermeasures to reduce flight risks.

Description

A civil aircraft stability assessment method based on real-time flight parameters

Technical Field

The invention relates to the technical field of flight safety, and in particular to a civil aircraft stability assessment method based on real-time flight parameters.

Background Art

In the field of civil aviation, the stability of an aircraft is a key factor in flight safety. Currently, civil aircraft are facing great challenges in flying under extreme weather conditions, which often have an adverse effect on the stability of the aircraft, which may lead to flight accidents. Therefore, evaluating the stability of an aircraft in extreme weather conditions is of great significance to flight safety.

However, in actual use, the existing technology mainly uses the stability assessment method based on the pilot's experience and feeling, or uses an offline analysis method. This method has problems of subjectivity and delay, cannot meet the needs of real-time assessment, and cannot accurately reflect the stability status of the aircraft in extreme weather in real time.

Summary of the invention

The purpose of the present invention is to provide a civil aircraft stability assessment method based on real-time flight parameters to solve the problem that the stability state of the aircraft under extreme weather conditions cannot be accurately reflected in real time.

To achieve the above object, the present invention provides the following technical solution: a method for evaluating the stability of a civil aircraft based on real-time flight parameters, comprising the following steps:

S1. Obtaining real-time flight parameter data during flight: collecting aircraft parameter data in real time;

S2, parameter data preprocessing: filtering, denoising, correcting, calibrating, cleaning and data correction of the parameter data collected in real time in step S1 to ensure the accuracy and reliability of the parameter data collected in real time in step S1;

S3. Obtaining meteorological data: Obtaining meteorological data of the current location and flight altitude through meteorological radar, meteorological satellite, and meteorological station;

S4, establishing a flight parameter model: Based on the flight parameter data preprocessed in step S2, a flight parameter model is established to describe the stability of the aircraft under different flight conditions, and the real-time flight parameters are correlated with the meteorological data obtained in step S3. According to the flight parameters and meteorological data, the stability index of the aircraft is calculated, and at the same time, the flight environment of the aircraft is evaluated in combination with the meteorological data;

S5. Evaluate the stability of the aircraft: using the flight parameter model established in step S4, evaluate the stability of the aircraft under extreme weather conditions in real time;

S6. Flight safety warning: determine the state of the aircraft according to the stability index evaluated in step S5, and set the corresponding alarm threshold.

Preferably, the parameter data in step S1 includes real-time flight parameter data of flight attitude, acceleration, aerodynamic parameters, wind speed and air pressure.

Preferably, the preprocessing of the parameter data in step S2 is implemented using a BF neural network.

Preferably, the meteorological data in step S3 includes but is not limited to wind speed, wind direction, precipitation and temperature.

Preferably, the stability indicators in step S4 include but are not limited to: sideslip angle, angle of attack, overload factor and stall boundary.

Preferably, in step S5, the stability state of the aircraft under extreme weather conditions is evaluated, and the evaluation criteria include but are not limited to: control capability evaluation, flight performance evaluation and flight limit evaluation.

Compared with the prior art, the present invention has the following beneficial effects:

1. The present invention uses real-time flight parameter data to establish a stability assessment model. By learning the stability characteristics of the aircraft through a machine learning algorithm, the stability of the aircraft can be assessed more accurately. Combining flight dynamics theory and advanced algorithms, the stability of the aircraft can be quantitatively assessed, providing an objective basis for judgment. At the same time, the real-time flight data of the aircraft can be collected and processed in real time, and an alarm can be issued according to the assessment results, which enables the pilot to accurately and quickly understand the stability of the aircraft and take corresponding countermeasures to reduce flight risks.

2. The stability assessment model of the present invention can be trained and learned according to different flight conditions, and is therefore suitable for flights under various extreme weather conditions, including severe weather, airflow disturbances, and the like.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG1 is an overall flow chart of a method for evaluating the stability of a civil aircraft based on real-time flight parameters according to the present invention;

FIG2 is a flow chart of a flight stability assessment system of a civil aircraft stability assessment method based on real-time flight parameters according to the present invention;

FIG3 is a diagram showing a meteorological model construction under extreme weather conditions for a method for evaluating the stability of a civil aircraft based on real-time flight parameters according to the present invention;

FIG. 4 is a complete stable flight index system of a civil aircraft stability evaluation method based on real-time flight parameters according to the present invention.

DETAILED DESCRIPTION

The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

Embodiment 1

Referring to FIGS. 1-4 , the present invention provides a technical solution: a method for evaluating the stability of a civil aircraft based on real-time flight parameters, comprising the following steps:

S1. Acquiring real-time flight parameter data during flight: collecting aircraft parameter data in real time, wherein the aircraft parameter data includes data of real-time flight parameters such as flight attitude, acceleration, angular velocity, aerodynamic parameters, wind speed and air pressure collected during flight through the flight data recorder or sensor equipment on the aircraft.

S2. Parameter data preprocessing: preprocessing the real-time flight parameter data collected in step S1, including filtering, denoising, correcting, calibrating, cleaning, data correction and correlation analysis of the collected real-time flight parameters through machine learning to ensure the accuracy and reliability of the data;

Machine learning includes but is not limited to RBF neural networks. First, the QAR data is cleaned, including processing missing values, outliers, and duplicate values, and useful features are extracted from the original QAR data, such as extracting keywords from text and extracting features from images. Then the extracted features are standardized so that the distribution of the data meets the input requirements of the RBF neural network. The data is then encoded, such as converting text data into a bag of words model and converting image data into a feature vector. The data is then divided into a training set and a test set for training and evaluating the performance of the RBF neural network model. The data is then balanced so that the number of positive and negative samples is similar to avoid deviations during model training. Finally, the data is subjected to dimensionality reduction processing to reduce the number of features and improve the training efficiency and generalization ability of the model. Through the above preprocessing steps, the performance and accuracy of the RBF neural network on QAR data can be effectively improved.

S3. Obtain meteorological data: Obtain meteorological data of the current location and flight altitude through meteorological radar, meteorological satellite, meteorological station and other equipment. The meteorological data includes but is not limited to wind speed, wind direction, precipitation, temperature, etc.

S4. Establishing a flight parameter model: Based on the pre-processed flight parameter data, a flight parameter model is established. The establishment of this model is a mathematical model based on physical principles or a mathematical model based on statistical methods. This model is a mathematical model that describes the movement and response of the aircraft. It takes into account the dynamic characteristics of the aircraft under different flight conditions. In the present invention, the flight dynamics model is used to evaluate the stability of the aircraft, and the real-time flight parameters are correlated with the meteorological data. That is, the known physical characteristics of the aircraft and the dynamic model under extreme weather conditions are combined, and the stability index of the aircraft, such as the sideslip angle, angle of attack, overload coefficient, roll angle, pitch angle, and yaw angle deviation value, etc., are calculated using a pre-set algorithm and model. This model is based on historical data and aircraft dynamics theory, and can accurately predict the stability of the aircraft. At the same time, combined with meteorological data, the flight environment of the aircraft is evaluated, such as whether the wind speed exceeds the aircraft's ultimate bearing capacity, etc.;

The flight dynamics model of an aircraft may need to consider the following situations when affected by extreme weather conditions:

Wind field effects: The strength and direction of wind may change dramatically, which can have a significant impact on the stability of the aircraft. The flight dynamics model needs to consider the effects of the wind field, including crosswinds, tailwinds, and headwinds.

Atmospheric density changes: Atmospheric density may change, such as atmospheric turbulence, temperature changes, etc. These changes will affect the lift and drag of the aircraft, and thus the stability of the aircraft. The flight dynamics model needs to consider the changes in atmospheric density and incorporate them into the stability assessment.

Turbulence effects: Turbulence can be more intense and unstable, affecting the stability of the aircraft. Turbulence can cause changes in the aircraft's attitude and turbulence, and the flight dynamics model needs to take turbulence effects into account and simulate its impact on the aircraft.

Thunderstorms and precipitation: Thunderstorms and precipitation are common manifestations of extreme weather, which have a direct impact on the aerodynamic characteristics and aerodynamic performance of the aircraft. The flight dynamics model needs to consider the impact of thunderstorms and precipitation on the aircraft, including the accumulation of ice and snow on the aircraft surface, the impact of rainfall on the wings and fuselage, etc.

The maneuverability and stability of an aircraft are important considerations in the flight dynamics model of an aircraft, especially in extreme weather conditions. Here is how the flight dynamics model considers the stability of an aircraft in extreme weather conditions:

Maneuverability modeling: Flight dynamics models usually include the modeling of the aircraft's control surfaces and control systems. For extreme weather conditions, the model needs to consider factors such as the response speed, torque, and control efficiency of the aircraft's control surfaces. These parameters are affected by factors such as wind speed, wind direction, and changes in atmospheric density. By introducing these influencing factors into the model, the aircraft's maneuverability under extreme weather conditions can be evaluated.

Stability Modeling: Stability is the ability of an aircraft to maintain balance and stability during flight. Flight dynamics models typically include modeling of both static and dynamic stability of an aircraft. In extreme weather conditions, factors such as wind fields and turbulence may have a significant impact on the stability of an aircraft. Flight dynamics models need to account for these influencing factors and evaluate the stability characteristics of an aircraft in extreme weather conditions.

Flight control system modeling: The flight dynamics model also considers the aircraft's flight control system, including the autopilot and flight control system. In extreme weather conditions, the flight control system needs to be able to accurately control the aircraft's attitude and trajectory to maintain the aircraft's stability and maneuverability. The flight dynamics model considers the characteristics and responses of the flight control system and evaluates its control capabilities in extreme weather conditions;

The above calculation of the stability index of the aircraft refers to the calculation by analyzing the real-time flight parameters and meteorological data of the aircraft. The specific calculation methods of these stability indexes vary depending on the type of aircraft, aircraft sensors and data processing algorithms. Usually, the sensor data of the aircraft and the relevant physical models are used to obtain the values of the stability index through mathematical calculation and analysis, including but not limited to: sideslip angle, angle of attack, load factor, stall margin, etc.

Sideslip angle: refers to the angle between the flight direction of the aircraft and the orientation of the aircraft body. It indicates whether the aircraft is in a horizontal flight state and whether the aircraft is affected by lateral forces. The calculation of the sideslip angle can be calculated using the aircraft's attitude sensor data or through the aircraft's airspeed meter and compass data.

Angle of attack: refers to the angle between the airflow on the aircraft wing and the chord line of the wing. It indicates the relative attitude of the aircraft wing and the airflow, and affects the lift and drag of the aircraft. The calculation of the angle of attack can be inferred from the aircraft's attitude sensor and airspeed meter data.

Load factor: refers to the ratio of the load on the aircraft to the acceleration due to gravity. It indicates the magnitude of the acceleration on the aircraft and has a significant impact on the aircraft's structure and handling characteristics. The load factor can be calculated by comparing the aircraft's accelerometer with the acceleration due to gravity.

Stall margin: refers to the margin of the aircraft from stalling under given flight conditions. Stalling means that the aircraft wing loses lift due to excessive angle of attack, causing the aircraft to lose control. The calculation of the stall margin can be analyzed through the aircraft's angle of attack sensor and lift coefficient measurement data.

The stability index of an aircraft is the preset stability threshold, which can be dynamically adjusted according to the characteristics of different aircraft models. Differentiated thresholds also need to be set for different flight phases, such as takeoff, cruising, and landing. You can refer to the relevant flight safety standards formulated by the civil aviation authorities, such as the regulations of ICAO and FAA. For example, for large passenger aircraft and small general aviation aircraft, their stability requirements and thresholds are usually different. In addition to the aircraft model, you can also consider the current flight status, such as cruising, takeoff, landing and other different stages, and set appropriate thresholds respectively. The requirements for stability at different stages will also vary. You can refer to industry standards, safety assessment indicators and historical data to determine a reasonable threshold range. At the same time, in practical applications, you can also dynamically optimize the threshold to improve detection accuracy through methods such as machine learning.

The proposed civil aircraft stability assessment method based on real-time assessment parameters can be further improved by selecting an anomaly detection algorithm and dynamically setting thresholds in combination with factors such as aircraft model and flight status. The algorithm includes but is not limited to anomaly detection algorithms.

The anomaly detection algorithm uses an unsupervised anomaly detection algorithm, Isolation Forest, to learn the distribution pattern of normal flight data. According to the detection probability or distance of the anomaly point, the threshold range of each parameter index is dynamically adjusted to improve the accuracy of anomaly identification. Combined with the correlation between parameters, anomaly detection based on the covariance matrix is used to capture multi-dimensional joint anomalies.

S5. Evaluate the stability status of the aircraft: Use the constructed flight parameter model to evaluate the stability of the aircraft under extreme weather conditions in real time. The method of real-time evaluation of the stability of the aircraft under extreme weather conditions is implemented using fuzzy comprehensive evaluation. By comparing and analyzing the real-time flight parameter data with the model, it can be identified whether the aircraft has abnormal stability behavior, such as attitude changes, acceleration fluctuations, etc. The use of fuzzy comprehensive evaluation methods can well solve the fuzzy and difficult to quantify problem of flight quality evaluation. By establishing a fuzzy comprehensive evaluation set, selecting appropriate membership functions and algorithms, and using fuzzy comprehensive evaluation to calculate the results of the flight quality evaluation, and further analyzing and processing the evaluation results, a more scientific and effective flight stability evaluation result can be obtained;

To judge the stability of an aircraft, we need to analyze flight parameters and compare and analyze the real-time collected flight parameters with the preset safety range. According to different stability indicators, we use corresponding algorithms and models to calculate and evaluate the stability of the aircraft. We also use meteorological data to evaluate the flight environment of the aircraft, such as whether the wind speed exceeds the aircraft's limit tolerance.

The aircraft stability assessment method combined with the mathematical calculation and simulation technology of the flight dynamics model can provide a quantitative assessment of the aircraft's control ability and flight performance under extreme weather conditions. These assessment results are of great significance to the pilot's decision-making and flight operations, as well as the improvement of aircraft design and flight control systems. The assessment criteria include but are not limited to: control ability assessment, flight performance assessment, and flight limit assessment.

Control capability assessment: Simulate the aircraft's control response under extreme weather conditions. By applying different input signals to the flight dynamics model, such as rudder, elevator, and aileron, the aircraft's control response characteristics can be evaluated. These characteristics include control surface efficiency, response speed, stability, and control torque. By analyzing the aircraft's control response under different extreme weather conditions, the adaptability and reliability of its control capability can be evaluated.

Flight performance evaluation: Evaluate the flight performance of an aircraft under extreme weather conditions. This includes parameters such as the aircraft's lift, drag, thrust, and speed. The model can take into account the effects of extreme weather conditions on these parameters, such as increased drag from turbulence, changes in atmospheric density, etc. By analyzing the aircraft's flight performance under extreme weather conditions, its acceleration, climb, glide, and landing performance can be evaluated.

Flight limit assessment: Evaluate the flight limits of an aircraft under extreme weather conditions. The flight dynamics model can simulate the dynamic characteristics of an aircraft under various flight conditions, such as stall, dangerous pitch angle, dangerous roll angle, etc. By analyzing the extreme conditions that an aircraft may encounter under extreme weather conditions, its safety margins and flight limits can be assessed, thereby helping to formulate appropriate operating procedures and flight restrictions.

S6. Flight safety warning: Determine the status of the aircraft based on the evaluation results and set the corresponding alarm threshold. The preset threshold can be determined by factors such as the aircraft model and flight status. If the stability index of the aircraft exceeds the preset safety range, the system will trigger a warning or alarm mechanism. The level and method of the alarm can be determined according to the severity and urgency of the stability index. Based on the results of the real-time evaluation and the alarm information, the pilot can take appropriate measures to adjust the aircraft's attitude, speed or heading, or automatically adjust the aircraft's control parameters to maintain the stability and safety of the aircraft. The warning mechanism includes but is not limited to: audio warnings, visual warnings, vibration warnings, alarm messages;

Audio warnings are when the system issues an audible alert through the pilot's headphones or speakers inside the aircraft to draw the pilot's attention. Different stability issues can be distinguished using different sounds or tones to help pilots identify the type and severity of the problem.

Visual warnings refer to the system that can display warning information on the aircraft's instrument panel or display screen to attract the pilot's visual attention. These warning messages are usually presented in eye-catching colors, flashing fonts or graphics to increase attention and recognition speed.

Vibration warning means that the system can generate vibrations on the pilot's seat or control stick by controlling the vibration device in the aircraft cockpit to remind the pilot of stability problems. This physical perception warning method can play an effective role when the pilot is not paying attention or the flight environment is noisy.

Alarm information means that the system can send alarm information to the ground dispatch center through the aircraft's communication equipment or data link so that ground operators can understand the stability of the aircraft and take corresponding measures. This method can provide additional support and decision-making basis to ensure the safety of the aircraft.

Embodiment 2

Taking the wind field effect as an example, a technical method is provided, including the following steps:

Machine learning-radial basis function: RBF neural network is used to pre-process QAR data, identify and exclude abnormal and invalid data, interpolate missing data, reduce system errors, and reconstruct the actual curves of key parameters in the flight phase so that the parameters meet the accuracy requirements.

RBF neural network is a three-layer network structure. The input layer nodes pass the input information to the hidden layer, and they are connected through radial basis functions. The basis functions in the hidden layer nodes will respond locally to the input signal. The output nodes are connected to the hidden layer through weights.

Where x is the n-dimensional input vector. _σi is the variable of the i-th perception, which determines the width of the center point of this basis function. m is the number of perception units. _ci is the center of the i-th basis function, ^a vector with the same dimension as x. _‖xci‖2 represents the distance between x and _ci , and h _i (x) has a unique maximum at _ci .

The input layer is a nonlinear mapping from x to _hi (x), and the output layer is a linear mapping from _hi (x) to _yk , that is:

Where r is the number of output nodes and _wik is the weight.

Under wind conditions, there are many meteorological environment-related QAR parameters that affect flight quality factors. The following will combine the real QAR data of the B737-800 model provided by Shandong Airlines to briefly analyze the impact of wind field changes on pilots' approach and landing phases of flight.

A change in wind direction will cause the approach path to deviate, and the pilot needs to make heading adjustments based on the actual wind direction to maintain the correct path and glide path. An increase in wind speed will increase the aircraft ground speed and shorten the approach segment. Conversely, a decrease in wind speed will reduce the aircraft ground speed, and measures need to be taken to maintain the correct speed and energy management.

Crosswinds can create lateral forces on the aircraft, causing it to skid on the approach path. The pilot needs to counteract the effects of the crosswind by using the skid rudder and roll control to keep the aircraft on the correct level. Skidding can cause the aircraft to deviate from the intended glide path, and the pilot needs to readjust the aircraft by using the skid and roll control to get it back on the correct glide path.

In actual flight, the aircraft will also be affected by gusts and crosswinds. Gusts are instantaneous winds with sudden changes in wind speed and direction. When the aircraft encounters strong gusts during the approach phase, the pilot needs to react quickly and maintain the stability and control of the aircraft by adjusting thrust, attitude and speed. Crosswind refers to wind that is perpendicular to the heading. Crosswinds will generate lateral forces, causing the aircraft to skid on the approach path. The pilot needs to control the sideslip and roll according to the size and direction of the crosswind to keep the aircraft on the predetermined course. During the approach phase, you must remain vigilant and adjust the flight controls in time according to changes in wind speed and direction to keep the aircraft aligned with the runway and land safely.

Therefore, considering the combined effects of multiple factors under wind field conditions, it is necessary to model the aircraft's flight maneuverability, stability, and flight control system. When modeling, in the selection of flight stability evaluation indicators, combined with the aforementioned situations that may be encountered under wind field conditions, the sideslip angle, angle of attack, overload factor, stall boundary, etc. are selected as the first-level evaluation indicators.

In order to obtain relatively reasonable secondary indicators for flight stability evaluation, the QAR parameter set affecting the flight quality of the B737-800 model during the approach phase was selected based on the QAR data recorded by the B737-800 series aircraft provided by Shandong Airlines, and with reference to the Boeing aircraft performance manual and flight quality monitoring standards. If all of them are used to establish a model to study the flight quality during the approach phase, the calculation speed will be slow, and the interaction and information overlap between the parameters will also affect the evaluation results.

Therefore, before establishing the evaluation index system, it is necessary to simplify the QAR parameters and extract the QAR key parameter set that affects the flight quality in the approach phase, retaining only the QAR parameters that have a great impact on the flight quality and ignoring the QAR parameters with a small impact.

The present invention adopts an unsupervised anomaly detection algorithm - isolation forest in the classification of QAR parameters: a decision tree-based strategy is used to find isolated points, and the data is divided into multiple non-overlapping intervals by setting a hyperplane, and whether it is abnormal is determined according to the data division result. Then, t isolated trees are formed, and a feature point is randomly selected at the node to divide the data into two intervals, and then random feature points are continuously selected to divide each interval, and the cycle continues until the data is indivisible or reaches a limited height. Since the interval division is random, it is necessary to use an integrated method to obtain the convergence value, that is, repeatedly constructing new trees, and finally integrating the results of all trees.

After generating t trees, the training is finished. At this time, it is necessary to summarize the impact of each tree on the sample. For a test sample x, let it traverse each tree and calculate the height of x in each tree to get the average height of x in the tree. If x contains multiple training data in a node, the height of x needs to be corrected. The specific formula is:

Where p(x, y) is the anomaly score, x is the test sample, y is the number of samples, and d(y) is the average depth of the samples. is Euler's constant.

After obtaining the anomaly score of each test sample, a threshold can be set as the boundary for detecting abnormal samples. In this paper, the value range of p(x, y) is [0, 1]. The closer p(x, y) is to 1, the more abnormal the sample is. The threshold is 0.5.

According to the FOQA monitoring manual and the indicator screening of random forest, 12 secondary indicators were selected as the benchmark, including pressure altitude, lateral acceleration, indicated airspeed, wind speed, wind direction, pressure altitude, vertical acceleration, rudder, roll angle, elevator, dangerous pitch angle and total atmosphere. When the spatial dimension d=12 after dimensionality reduction, the reconstruction threshold t>80%, so the first 12 principal components were selected for dimensionality reduction. After dimensionality reduction, the feature matrix V was calculated as the sample data set of the low-altitude approach prediction model.

To prevent overfitting, the feature matrix V is randomly sampled into 12 sets of data sets for subsequent k-fold cross validation, that is, the data set is randomly divided into 12 parts, one part is taken as the test sample X-test each time, and the remaining subsets are used as training samples X-train, and this is repeated 12 times until all subsets are taken as test samples. In order to consider the impact of the wind field, X-train with a radio altitude of 800 feet, a headwind speed greater than 10m/s, and a crosswind speed greater than 8m/s is selected for analysis, and X-test does not need to be screened according to wind conditions.

The key parameters of the random forest algorithm are the number of trees and the number of samples when training the trees. Usually, when the number of trees is 120 and the number of samples is 234, the prediction results of the algorithm reach the optimal value, and no complicated parameter adjustment process is required.

The relevant flight parameters of the approach and landing phases in the QAR original data are selected to obtain the data feature matrix, and then the training set and test set are randomly sampled from them. The test set data are labeled with reference to the flight quality monitoring standards, and finally the screening and deletion of the second-level indicators are completed.

Finally, the overall construction of the aircraft stability evaluation method is completed. The aircraft stability evaluation method combines the mathematical calculation and simulation technology of the flight dynamics model to provide a quantitative evaluation of the aircraft's control ability and flight performance under extreme weather conditions.

The evaluation criteria include: control capability assessment, flight performance assessment, and flight limit assessment.

An unstable approach can cause the aircraft to shake slightly or violently when landing, and even affect the pilot's driving posture. According to the relevant monitoring items in flight quality monitoring, QAR data is used to determine whether there is an unstable approach during the aircraft landing process. According to the three evaluation indicators, the segment data of the low-altitude approach phase of 902 flights on a certain day in March 2021 were screened to obtain a sample of unstable events, and then a stability assessment score was performed.

According to the established flight quality evaluation index system, the fuzzy comprehensive evaluation factor set U can be obtained, U = {U ₁ ,U ₂ ,…U _i }, (i = 1,2,…,4), among which U ₁ ,U ₂ ,U ₃ ,U ₄ are used as the middle layer (first-level rating index), and the corresponding bottom layer (second-level rating index) is U _i = {U _i1 ,U _i2 ,…U _in }, (n = 1,2,3,…,s), and must satisfy the following conditions:

The present invention adopts the hierarchical analysis method to determine the weight of each evaluation index, and obtains the judgment matrix by adopting the form of a questionnaire.

When determining the relative importance levels of the factors for pairwise comparison, the 10 returned questionnaires were used to determine the major comparative trends using the majority voting method based on the relative importance expert scoring table of the hierarchical analysis method. Then, after statistical processing, the importance levels of each evaluation index factor were determined to form a judgment matrix.

Next, we use MATLAB software to find the eigenvector and maximum eigenvalue of the matrix and perform a consistency test.

In order to find out the influence of all evaluation indicators in the flight quality evaluation index system on the flight quality of the approach phase of the target layer, the overall ranking of the system should be carried out on the basis of the ranking of each level. The ranking weight of each factor on the flight quality of the approach phase is obtained through the total ranking of the levels.

Finally, a fuzzy comprehensive evaluation set is established. The fuzzy comprehensive evaluation set of flight quality can be represented by V = {V ₁ , V ₂ , ... V _n }. According to the commonly used method for flight quality evaluation, n = 5 is taken, that is, the evaluation results are divided into 5 levels. Excellent (V ₁ ), good (V ₂ ), medium (V ₃ ), poor (V ₄ ), and very poor (V ₅ ) are used to form the evaluation set V = {V ₁ , V ₂ , V ₃ , V ₄ , V ₅ }.

Flight quality monitoring sets monitoring items and standards. Each monitoring item has a preset threshold. Scoring is based on the degree of deviation from the threshold and expert experience. Five flight safety experts scored each evaluation indicator for the flight stability during the approach phase of the B737-800 flight.

In the feedback phase after the scoring, when a flight parameter does not meet the established standard or seriously exceeds the limit, the warning device will be triggered. The system will sound an alarm through the pilot's headset to remind the pilot of unreasonable operation; the aircraft's dashboard will flash red, and the pilot's seat will vibrate to remind the pilot of stability problems; finally, the system will send an alarm message to the ground dispatch center through the aircraft's data link. After the ground operator understands the stability of the aircraft, he will issue corrective instructions to the pilot.

It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims (6)

1. A method for evaluating the stability of a civil aircraft based on real-time flight parameters, characterized in that it comprises the following steps: S1. Obtaining real-time flight parameter data during flight: collecting aircraft parameter data in real time; S2, parameter data preprocessing: filtering, denoising, correcting, calibrating, cleaning and data correction of the parameter data collected in real time in step S1 to ensure the accuracy and reliability of the parameter data collected in real time in step S1; S3. Obtaining meteorological data: Obtaining meteorological data of the current location and flight altitude through meteorological radar, meteorological satellite, and meteorological station; S4, establishing a flight parameter model: Based on the flight parameter data preprocessed in step S2, a flight parameter model is established to describe the stability of the aircraft under different flight conditions, and the real-time flight parameters are correlated with the meteorological data obtained in step S3. According to the flight parameters and meteorological data, the stability index of the aircraft is calculated, and at the same time, the flight environment of the aircraft is evaluated in combination with the meteorological data; S5. Evaluating the stability of the aircraft: using the flight parameter model established in step S4, evaluating the stability of the aircraft under extreme weather conditions in real time; S6. Flight safety warning: determine the state of the aircraft according to the stability index evaluated in step S5, and set the corresponding alarm threshold. 2. A civil aircraft stability assessment method based on real-time flight parameters according to claim 1, characterized in that the parameter data in step S1 includes real-time flight parameter data of flight attitude, acceleration, aerodynamic parameters, wind speed and air pressure. 3. A civil aircraft stability assessment method based on real-time flight parameters according to claim 1, characterized in that the preprocessing of parameter data in step S2 is implemented using a BF neural network. 4. A civil aircraft stability assessment method based on real-time flight parameters according to claim 1, characterized in that: the meteorological data in step S3 includes but is not limited to wind speed, wind direction, precipitation and temperature. 5. A civil aircraft stability assessment method based on real-time flight parameters according to claim 1, characterized in that the stability indicators in step S4 include but are not limited to: sideslip angle, angle of attack, overload factor and stall boundary. 6. A civil aircraft stability assessment method based on real-time flight parameters according to claim 1, characterized in that: the stability state of the aircraft under extreme weather conditions is assessed in step S5, and its assessment criteria include but are not limited to: control capability assessment, flight performance assessment and flight limit assessment.