Low-altitude aircraft safety assessment method and system based on multi-source data fusion

Abstract

The invention discloses a multi-source data fusion-based low-altitude aircraft safety assessment method and system, and the method comprises the steps: analyzing the real-time flight data of a low-altitude aircraft, determining key flight parameters, determining the data characteristics of the key flight parameters, and assessing the safety state of the low-altitude aircraft based on the real-time flight data and the data characteristics, thereby obtaining a flight safety state assessment value; constructing a prediction model for prediction based on the historical flight environment data and a preset model to obtain flight environment prediction data; analyzing the flight environment prediction data, determining an environment state coefficient, determining an influence coefficient, and determining an environment influence coefficient based on the two coefficients; and correcting the flight safety state evaluation value based on the environmental influence coefficient to obtain a comprehensive safety evaluation value, and determining the safety level of the low-altitude aircraft based on the comprehensive safety evaluation value. According to the method, the capability of comprehensively monitoring and evaluating the safety state of the low-altitude aircraft is improved, the safety state of the aircraft can be accurately determined, the flight safety level of the aircraft is improved, and the flight risk is reduced.

Description

Technical Field

[0001] This invention relates to the field of flight safety technology, and in particular to a method and system for safety assessment of low-altitude aircraft based on multi-source data fusion. Background Technology

[0002] With the rapid rise of the low-altitude economy, applications such as drone logistics, urban air mobility (UAM), emergency rescue, and agricultural plant protection are rapidly moving from concept to reality. The low-altitude airspace, once a tranquil area, is evolving into a highly dynamic, dense, and complex emerging digital domain. However, the most critical bottleneck for the large-scale commercial application of low-altitude aircraft, especially unmanned aerial vehicles, is safety.

[0003] However, traditional safety assessment methods are proving inadequate because they often rely on a single or limited number of data sources (such as GPS location) and pre-set static rules (such as geofencing). They are essentially passive responses and alarms, lacking the ability to predict potential risks, which cannot meet the stringent safety requirements of low-altitude aircraft for real-time performance, autonomy, and high reliability. Summary of the Invention

[0004] To address the aforementioned technical problems, this invention provides a method and system for safety assessment of low-altitude aircraft based on multi-source data fusion, comprising: Acquire real-time flight data of low-altitude aircraft and analyze the real-time flight data to determine key flight parameters related to aircraft safety; The data characteristics of each key flight parameter are determined, and the safety status of the low-altitude aircraft is assessed based on the key flight parameters and data characteristics to obtain the flight safety status assessment value. Historical flight environment data is acquired, and a flight environment prediction model is constructed based on the historical flight environment data and a preset neural network model to make predictions and obtain flight environment prediction data. Analyze flight environment prediction data, determine the environmental state coefficient, determine the environmental impact coefficient on flight safety, and determine the environmental impact coefficient based on the environmental state coefficient and the impact coefficient; The flight safety status assessment value is corrected based on the environmental impact coefficient to obtain a comprehensive safety assessment value, and the safety level of the low-altitude aircraft is determined based on the comprehensive safety assessment value.

[0005] Furthermore, the acquisition of real-time flight data of the low-altitude aircraft and the analysis of this real-time flight data to determine key flight parameters related to aircraft safety include: Acquire real-time flight data of low-altitude aircraft and divide the real-time flight data into multiple flight parameter data groups according to parameter type; The correlation between each flight parameter data set and aircraft safety was determined, and the parameters corresponding to the flight parameter data sets with a correlation greater than a preset correlation threshold were identified as key flight parameters related to aircraft safety.

[0006] Furthermore, the process of determining the data characteristics of each key flight parameter, and assessing the safety status of the low-altitude aircraft based on these key flight parameters and data characteristics, yields a flight safety status assessment value, including: Identify the flight parameter data sets corresponding to each key flight parameter, and construct time progress data change curves based on each flight parameter data set; Determine the values ​​of preset data statistical indicators in each data change curve, and select the preset data statistical indicators that exceed the corresponding preset thresholds to identify them as data characteristics of each key flight parameter. Calculate the difference between the value of each data feature and the corresponding preset threshold, and evaluate the difference to obtain the deviation evaluation value of each data feature; The deviation evaluation values ​​of all data features corresponding to each key flight parameter are standardized, and the standardized deviation evaluation values ​​are added together to obtain the comprehensive deviation evaluation value of each key flight parameter. The preset weights of each key flight parameter are determined, and the flight safety status assessment value of the low-altitude aircraft is obtained by calculating based on the preset weights of each key flight parameter and the comprehensive deviation assessment value.

[0007] Furthermore, the formula for calculating the flight safety status assessment value of the low-altitude aircraft is as follows: , Where P is the flight safety status assessment value of the low-altitude aircraft, ai is the preset weight of the i-th key flight parameter, Gi is the comprehensive deviation assessment value of the i-th key flight parameter, and n is the number of key flight parameters.

[0008] Furthermore, the process of acquiring historical flight environment data and constructing a flight environment prediction model based on the historical flight environment data and a preset neural network model to obtain flight environment prediction data includes: Historical flight environment data is acquired, including historical meteorological data sets, historical airspace traffic data sets, and historical geographic data sets, and the data characteristics of each data set are extracted. Data sets are constructed based on each data group and its corresponding key data features, and the data sets are input into the corresponding preset neural network models to construct the corresponding initial prediction models. Each dataset is divided into a training set and a test set according to a preset ratio, and the training set and test set are input into the corresponding prediction initial model; Each initial prediction model is trained and tested until it meets the preset convergence condition, thus obtaining each prediction model. The various prediction models are fused to obtain a flight environment prediction model. Real-time flight environment data is then acquired and input into the flight environment prediction model for prediction, resulting in flight environment prediction data for a future period of time.

[0009] Furthermore, the analysis of flight environment prediction data to determine environmental state coefficients includes: Determine the meteorological forecast data group, airspace traffic forecast data group, and geographical forecast data group from the flight environment forecast data; Identify the meteorological parameter data sets within the meteorological forecast data set, and calculate the average value of each meteorological parameter data set; Determine the baseline average value for each meteorological parameter data set, and calculate the difference between the average value of each meteorological parameter data set and the corresponding baseline average value to obtain the data deviation of each meteorological parameter data set; Determine the preset coefficients for each meteorological parameter data set, and multiply the data deviations of each meteorological parameter data set by the preset coefficients to obtain the first state coefficients of the meteorological forecast data set. Based on the airspace traffic prediction data set, the predicted intrusion targets are identified, and the distance and time of arrival at the closest point of the predicted trajectory between each predicted intrusion target and the low-altitude aircraft are determined. The sub-safety factor between each predicted intrusion target and the low-altitude aircraft is calculated based on the closest point distance and closest point time. The formula for calculating the sub-safety factor is as follows: R = α*(1 / A) + β*(1 / B), Where R is the sub-safety coefficient, α is the first preset weight, A is the distance to the closest point, β is the second preset weight, and B is the time to the closest point; The sub-safety coefficients between each predicted intrusion target and the low-altitude aircraft are added together to obtain the second state coefficient of the airspace traffic prediction data set. Obstacles were identified based on a set of geographic prediction data, and the distances between the low-altitude aircraft and each obstacle were calculated. The maximum distance between the low-altitude aircraft and each obstacle is determined, and the maximum distance is multiplied by a preset safety conversion coefficient to obtain the third state coefficient of the geographic prediction data set. The first state coefficient of the meteorological forecast data group, the second state coefficient of the airspace traffic forecast data group, and the third state coefficient of the geographical forecast data group were normalized respectively, and the results of the normalization were added together to obtain the environmental state coefficient.

[0010] Furthermore, determining the influence coefficient of the environment on flight safety includes: Identify historical flight environment data and historical flight safety data, and divide the historical flight environment data into multiple historical flight environment parameter data groups according to parameter type; For the i-th historical flight environment parameter data set, determine the historical flight safety data when other flight environment parameters are the same, and determine the first change amount of the i-th historical flight environment parameter data set and the second change amount of the corresponding historical flight safety data. Calculate the ratio of the second change to the first change to obtain the sub-influence coefficient of each flight environment parameter on flight safety, and add the sub-influence coefficients of each flight environment parameter on flight safety to obtain the influence coefficient of the environment on flight safety.

[0011] Furthermore, the determination of the environmental impact coefficient based on the environmental state coefficient and the impact coefficient includes: The environmental impact coefficient is obtained by multiplying the environmental state coefficient and the impact coefficient.

[0012] Furthermore, the process of correcting the flight safety status assessment value based on the environmental impact coefficient to obtain a comprehensive safety assessment value, and determining the safety level of the low-altitude aircraft based on the comprehensive safety assessment value, includes: The comprehensive safety assessment value of the low-altitude aircraft is obtained by multiplying the environmental impact coefficient by the flight safety status assessment value. A preset security level-comprehensive security assessment value range correspondence is set in advance. For each comprehensive security assessment value range, a corresponding preset security level is associated with it. Obtain the comprehensive safety assessment value of the low-altitude aircraft, and based on the mapping relationship between the comprehensive safety assessment value interval to which the comprehensive safety assessment value belongs and the preset safety level-comprehensive safety assessment value interval correspondence, select the preset safety level corresponding to the comprehensive safety assessment value interval as the safety level of the low-altitude aircraft.

[0013] This invention also provides a low-altitude aircraft safety assessment system based on multi-source data fusion, comprising: The acquisition module is used to acquire real-time flight data of low-altitude aircraft and analyze the real-time flight data to determine key flight parameters related to aircraft safety. The evaluation module is used to determine the data characteristics of each key flight parameter, and to evaluate the safety status of the low-altitude aircraft based on each key flight parameter and data characteristics, thereby obtaining the flight safety status evaluation value. The prediction module is used to acquire historical flight environment data and construct a flight environment prediction model based on the historical flight environment data and a preset neural network model to make predictions and obtain flight environment prediction data. The calculation module is used to analyze flight environment prediction data, determine the environmental state coefficient, determine the environmental impact coefficient on flight safety, and determine the environmental impact coefficient based on the environmental state coefficient and the impact coefficient. The determination module is used to correct the flight safety status assessment value based on the environmental impact coefficient to obtain a comprehensive safety assessment value, and to determine the safety level of the low-altitude aircraft based on the comprehensive safety assessment value.

[0014] Compared with existing technologies, the safety assessment method and system for low-altitude aircraft based on multi-source data fusion, as described in this invention, have the following advantages: This invention acquires and analyzes real-time flight data of low-altitude aircraft, determines the data characteristics of key flight parameters, establishes the correlation and regularity between these parameters, and can more accurately assess the status of the aircraft, providing a basis for subsequent safety assessments. This invention acquires and analyzes real-time flight data of low-altitude aircraft to determine the data characteristics of each key flight parameter. Based on the key flight parameters and their data characteristics, it assesses the safety status of the aircraft and obtains a comprehensive flight safety status assessment value. This allows for a more accurate assessment of the aircraft's status and timely understanding of its safety condition. This invention utilizes historical flight environment data and a pre-set neural network model to construct a flight environment prediction model, which can predict future flight environment conditions, determine the environmental state coefficient and the environmental impact coefficient on flight safety, and quantify the degree of influence of environmental factors on flight safety. This invention corrects the flight safety status assessment value based on the environmental impact coefficient to obtain a comprehensive safety assessment value. This can more comprehensively consider the impact of environmental factors on flight safety and more accurately assess the overall safety of the aircraft. In summary, this invention integrates real-time flight data analysis and environmental prediction data, and uses multiple technical means such as flight safety status assessment, determination of environmental impact coefficients, and safety level evaluation to assess the safety status of low-altitude aircraft. This comprehensively improves the ability to monitor and assess the safety status of low-altitude aircraft, helps to accurately determine the safety status of aircraft, thereby improving the flight safety level and reducing flight risks. Attached Figure Description

[0015] Figure 1 This is a schematic diagram of the process structure of the low-altitude aircraft safety assessment method based on multi-source data fusion in an embodiment of the present invention; Figure 2This is a schematic diagram of the composition of a low-altitude aircraft safety assessment system based on multi-source data fusion in an embodiment of the present invention. Detailed Implementation

[0016] The specific embodiments of this application will be described in further detail below with reference to the accompanying drawings and examples. The following examples are for illustrative purposes only and are not intended to limit the scope of the invention.

[0017] In the description of this application, it should be understood that the terms "center", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this application and simplifying the description, and do not indicate or imply that the platform or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this application.

[0018] The terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this application, unless otherwise stated, "a plurality of" means two or more.

[0019] like Figure 1 As shown in the embodiments of this application, a method for safety assessment of low-altitude aircraft based on multi-source data fusion is provided, including: S100: acquiring real-time flight data of the low-altitude aircraft, analyzing the real-time flight data, and determining key flight parameters related to aircraft safety; S200: determining the data characteristics of each key flight parameter, assessing the safety status of the low-altitude aircraft based on each key flight parameter and data characteristics, and obtaining a flight safety status assessment value; S300: acquiring historical flight environment data, constructing a flight environment prediction model based on the historical flight environment data and a preset neural network model, and obtaining flight environment prediction data; S400: analyzing the flight environment prediction data, determining the environmental state coefficient, and determining the environmental impact coefficient on flight safety, and determining the environmental impact coefficient based on the environmental state coefficient and the impact coefficient; S500: correcting the flight safety status assessment value based on the environmental impact coefficient, obtaining a comprehensive safety assessment value, and determining the safety level of the low-altitude aircraft based on the comprehensive safety assessment value.

[0020] Furthermore, this invention acquires and analyzes real-time flight data of low-altitude aircraft to determine and analyze the data characteristics of key flight parameters, establishing the correlation and regularity between these parameters. This allows for a more accurate assessment of the aircraft's status and provides a foundation for subsequent safety assessments. This invention also acquires and analyzes real-time flight data of low-altitude aircraft to determine the data characteristics of key flight parameters, and assesses the aircraft's safety status based on these key parameters and their data characteristics, obtaining a comprehensive flight safety status assessment value. This allows for a more accurate assessment of the aircraft's status and timely understanding of its safety condition. Finally, this invention utilizes historical flight environment data and a pre-set neural network model to construct a flight environment prediction model, which can predict future flight environments. This invention quantifies the impact of environmental factors on flight safety by determining environmental state coefficients and environmental impact coefficients. Based on the environmental impact coefficients, the invention modifies the flight safety state assessment value to obtain a comprehensive safety assessment value. This allows for a more comprehensive consideration of the impact of environmental factors on flight safety and a more accurate assessment of the overall safety of the aircraft. In summary, this invention integrates real-time flight data analysis and environmental prediction data, and uses multiple technical means such as flight safety state assessment, determination of environmental impact coefficients, and safety level evaluation to assess the safety status of low-altitude aircraft. This comprehensively improves the ability to monitor and assess the safety status of low-altitude aircraft, helps to accurately determine the safety status of aircraft, thereby improving the flight safety level and reducing flight risks.

[0021] In the embodiments of this application, a method for safety assessment of low-altitude aircraft based on multi-source data fusion is provided. The method involves acquiring real-time flight data of the low-altitude aircraft and analyzing the real-time flight data to determine key flight parameters related to aircraft safety. This includes: acquiring real-time flight data of the low-altitude aircraft and dividing the real-time flight data into multiple flight parameter data groups according to parameter type; determining the correlation between each flight parameter data group and aircraft safety; and determining the parameters corresponding to the flight parameter data groups with a correlation greater than a preset correlation threshold as key flight parameters related to aircraft safety.

[0022] Specifically, real-time flight data is acquired by continuously collecting massive amounts of raw data streams generated during aircraft operation through airborne sensors, flight control systems, and data links. This data includes: navigation data (GPS / GNSS position, IMU inertial measurement unit data, airspeed, altitude), attitude data (roll angle, pitch angle, yaw angle), power system data (battery voltage, current, remaining charge, motor speed, ESC temperature), control data (control surface deflection, control commands), and mission payload status. Based on the physical properties and functional scope of the parameters, data groups are categorized by parameter type: Navigation status group: latitude, longitude, altitude, speed, heading; Flight attitude group: roll angle, pitch angle, yaw angle and their angular rates; Power health group: battery voltage, current, charge, main motor temperature, vibration frequency; Control response group: deviation between control commands and actual attitude, servo response delay; Communication link group: signal strength, data packet loss rate, link delay. A "safety relevance" is calculated for each parameter data group. The quantitative indicator, this correlation, is calculated by analyzing historical fault data to determine the statistical correlation between abnormal parameters and the eventual accident. A preset correlation threshold is set, and the calculated correlation is compared with this threshold to filter out core parameters that have a significant impact on flight safety. For example, the correlation between battery voltage and motor temperature may be as high as 0.9 (out of 1.0), far exceeding the threshold, and is therefore identified as a key parameter. On the other hand, the correlation of certain payload equipment status parameters may only be 0.2, and are therefore filtered out by the system and not used as core safety monitoring indicators. This step transforms massive, heterogeneous raw data streams into clearly categorized and meaningful data sets, focusing on a few key parameters. This significantly reduces the data processing load of subsequent computing modules, meeting the stringent real-time requirements of low-altitude aircraft. It can complete a safety assessment within milliseconds, reserving valuable time for timely decision-making and intervention. Through scientific correlation analysis, the system can eliminate interference from irrelevant or weakly correlated data, avoiding the problem of "garbage in, garbage out," ensuring that the safety assessment model is always based on the most sensitive and effective indicators, thereby significantly improving the signal-to-noise ratio and accuracy of risk assessment. This step is not just a simple data classification, but a feature selection strategy based on prior knowledge and real-time analysis, laying a solid data foundation for subsequent efficient, accurate, and real-time safety status assessments. It is an indispensable "intelligent front end" in the entire safety chain.

[0023] In the embodiments of this application, a method for safety assessment of low-altitude aircraft based on multi-source data fusion is provided. The method involves determining the data characteristics of each key flight parameter, assessing the safety status of the low-altitude aircraft based on the key flight parameters and data characteristics, and obtaining a flight safety status assessment value. This includes: determining the flight parameter data group corresponding to each key flight parameter, and constructing a time-progress data change curve based on each flight parameter data group; determining the values ​​of preset data statistical indicators in each data change curve, and selecting preset data statistical indicators that exceed corresponding preset thresholds as data characteristics of each key flight parameter; calculating the difference between the value of each data characteristic and the corresponding preset threshold, and evaluating the difference to obtain a deviation assessment value for each data characteristic; standardizing the deviation assessment values ​​of all data characteristics corresponding to each key flight parameter, and summing the standardized deviation assessment values ​​to obtain a comprehensive deviation assessment value for each key flight parameter; determining the preset weight of each key flight parameter, and calculating the flight safety status assessment value of the low-altitude aircraft based on the preset weight of each key flight parameter and the comprehensive deviation assessment value.

[0024] Specifically, for each identified key flight parameter, a data curve is constructed based on its time-series data to reflect its changes over time. Pre-defined statistical indicators are then calculated on these curves, such as: instantaneous value (current reading); average value (representing the overall level); standard deviation (representing the degree of fluctuation); and rate of change (representing the speed of deterioration or improvement). By comparing each indicator with a pre-defined safety threshold, abnormal features are precisely captured, such as "voltage value below the threshold" or "vibration standard deviation far exceeding historical normal levels." These captured anomalies are the data features characterizing safety. For each exceeding data feature, the system does not simply perform a "yes / no" judgment but calculates its value against the threshold. The specific difference is used to quantify how much it exceeds the standard. This difference is mapped to a unified scoring scale for evaluation (e.g., a small difference is assigned a value of 0.1, and a large difference is assigned a value of 0.9), resulting in a deviation assessment value. Deviations of different physical units (such as volts, degrees per second) and different orders of magnitude are uniformly converted into comparable, unitless risk scores. The deviation assessment values ​​of all data features under a key parameter (after standardization) are added together to obtain the comprehensive deviation assessment value of that parameter. Preset weights are introduced, as different key parameters have different importance to overall safety. Through a weighted summation model, the comprehensive deviation assessment values ​​of all key parameters are merged to finally obtain a unique flight safety status assessment value. This step represents a leap from "judgment" to "assessment." Traditional methods only trigger binary alarms (0 / 1) when parameters exceed limits, while this method can quantify the severity of risks. It not only knows that "there is a problem with the battery," but also accurately assesses that "the current battery problem causes a risk level of 75 points (out of 100)," thus providing richer gradient information for decision-making. Multi-feature fusion ensures a three-dimensional evaluation of a parameter's state, while weighted fusion ensures that the final assessment result conforms to the objective importance of various factors in flight safety, avoiding "false alarms" and "missed alarms."

[0025] In an embodiment of this application, a safety assessment method for low-altitude aircraft based on multi-source data fusion is provided, wherein the calculation formula for the flight safety status assessment value of the low-altitude aircraft is as follows: , Where P is the flight safety status assessment value of the low-altitude aircraft, ai is the preset weight of the i-th key flight parameter, Gi is the comprehensive deviation assessment value of the i-th key flight parameter, and n is the number of key flight parameters.

[0026] In the embodiments of this application, a method for safety assessment of low-altitude aircraft based on multi-source data fusion is provided. The method involves acquiring historical flight environment data and constructing a flight environment prediction model based on the historical flight environment data and a preset neural network model to obtain flight environment prediction data. The method includes: acquiring historical flight environment data, which includes historical meteorological data, historical airspace traffic data, and historical geographical data, and extracting data features from each data group; constructing datasets based on each data group and corresponding key data features, and inputting the datasets into corresponding preset neural network models to construct corresponding initial prediction models; dividing each dataset into training and testing sets according to a preset ratio, and inputting the training and testing sets into the corresponding initial prediction models; training and testing each initial prediction model until it meets a preset convergence condition to obtain each prediction model; fusing the prediction models to obtain a flight environment prediction model, and acquiring real-time flight environment data, inputting the real-time flight environment data into the flight environment prediction model for prediction to obtain flight environment prediction data for a future period of time.

[0027] Specifically, a large amount of historical flight environment data is collected and divided into three major data groups according to their source: historical meteorological data (such as wind speed, precipitation, and visibility), historical airspace traffic data (such as the trajectories and density of surrounding aircraft), and historical geographical data (such as the distribution of static obstacles and terrain undulations). Key data features are extracted from these raw data (for example, instead of directly using all wind speed data, features such as maximum wind speed, wind speed variation trends, and turbulence intensity are extracted). A dedicated initial prediction model is constructed for each data group and its features. These pre-set neural network models (such as using CNN to process geographic raster data, and LSTM or Transformer to process meteorological and traffic data) are used. The time-series data is structurally best suited to its corresponding data type. Each dataset is divided into a training set and a test set, used to teach the model to learn patterns and evaluate its performance, respectively. Through iterative training and testing, each sub-model reaches the preset convergence condition in its respective domain (such as achieving sufficient prediction accuracy and the loss function no longer decreasing), resulting in three high-performance expert models: a meteorological prediction model, a traffic prediction model, and a geographical environment prediction model. The three trained expert models are then fused into a powerful flight environment prediction model. By inputting real-time flight environment data into this fused model, a comprehensive prediction result for the environment over a future period can be output, i.e., flight environment prediction data. This step fully leverages the advantages of various models in processing specific data by tailoring dedicated neural networks for different data types, thereby achieving the highest individual prediction accuracy in each of the meteorological, traffic, and geographical sub-categories. The prediction results provided by the fused model are no longer vague trends, but precise and quantitative snapshots of the future environment. Unlike traditional prediction methods based on physical formulas or simple statistics, neural networks can automatically learn extremely complex nonlinear relationships and hidden patterns in the environment from massive historical data.

[0028] In the embodiments of this application, a method for safety assessment of low-altitude aircraft based on multi-source data fusion is provided. The analysis of flight environment prediction data and determination of environmental state coefficients includes: determining a meteorological prediction data group, an airspace traffic prediction data group, and a geographical prediction data group from the flight environment prediction data; determining each meteorological parameter data group in the meteorological prediction data group and calculating the average value of each meteorological parameter data group; determining the baseline average value of each meteorological parameter data group and calculating the difference between the average value of each meteorological parameter data group and the corresponding baseline average value to obtain the data deviation of each meteorological parameter data group; determining the preset coefficient of each meteorological parameter data group, multiplying the data deviation of each meteorological parameter data group by the preset coefficient to obtain the first state coefficient of the meteorological prediction data group; determining the predicted intrusion target based on the airspace traffic prediction data group, and determining the closest point distance and the time to reach the closest point between each predicted intrusion target and the low-altitude aircraft; calculating the sub-safety coefficient between each predicted intrusion target and the low-altitude aircraft based on the closest point distance and the closest point time, the calculation formula of the sub-safety coefficient is as follows: R = α*(1 / A) + β*(1 / B), Where R is the sub-safety factor, α is the first preset weight, A is the closest point distance, β is the second preset weight, and B is the closest point time; the sub-safety factors between each predicted intrusion target and the low-altitude aircraft are added together to obtain the second state coefficient of the airspace traffic prediction data set; obstacles are identified based on the geographic prediction data set, and the distance between the low-altitude aircraft and each obstacle is calculated; the maximum distance between the low-altitude aircraft and each obstacle is determined, and the maximum distance is multiplied by the preset safety conversion factor to obtain the third state coefficient of the geographic prediction data set; the first state coefficient of the meteorological prediction data set, the second state coefficient of the airspace traffic prediction data set, and the third state coefficient of the geographic prediction data set are normalized respectively, and the results of the normalization are added together to obtain the environmental state coefficient.

[0029] Specifically, meteorological forecast data sets (such as wind speed, precipitation, and visibility) are extracted from the forecast data, and the average value of each parameter over a future period is calculated. This forecast average value is compared with the historical baseline average value for the same period to obtain the data deviation, which reflects the degree of abnormality in meteorological conditions. The deviation of each meteorological parameter is multiplied by its unique preset coefficient (i.e., weight), which represents the degree of influence of the parameter on flight safety (e.g., wind speed has a greater weight than temperature). The results of all parameters are combined by weighting or other methods to obtain the first state coefficient. The larger the coefficient value, the more severe the meteorological environment. All potential predicted intrusion targets (such as other drones and aircraft) are identified from the airspace traffic forecast data set. For each target, the nearest point distance (CPA) between it and the predicted trajectory of the pilot is calculated. And the Time of Closest Arrival (TCPA); based on CPA and TCPA (usually inversely proportional to risk), a sub-safety factor is calculated to quantify the threat level of a single target. Then, the sub-safety factors of all targets are added together to obtain the second state factor. Threats in the airspace are cumulative; the more targets, the closer they are, and the shorter the encounter time, the higher the risk represented by this factor. Static obstacles (such as tall buildings and cables) are identified from the geographic prediction data set, and the distance between the machine and each obstacle is calculated. Among all distances, the largest distance (i.e., the distance between the machine and the nearest obstacle) is selected. In the most dangerous scenario, this minimum distance is multiplied by a preset safety conversion coefficient to obtain the third state coefficient. This conversion coefficient transforms the physical distance (meters) into a risk value; the smaller the distance, the higher the risk value. The state coefficients (first, second, and third) obtained from the above three dimensions are normalized to unify them into the same numerical range (such as 0-1) to eliminate the influence of dimensions and orders of magnitude. The three normalized values ​​are added together to obtain the final environmental state coefficient. This coefficient is a comprehensive indicator; the higher its value, the higher the overall risk of the predicted flight environment and the worse the safety. This step transforms three completely different categories of environmental threats (meteorological, dynamic traffic, and static geography) into comparable numerical values ​​using mathematical methods, and ultimately merges them into a "total score." This makes the complex flight environment safety situation clear, measurable, and comparable. In the assessment of each dimension, the method focuses on the most critical risk indicators: meteorology uses "deviation" (abnormality equals risk), traffic uses "CPA / TCPA" (proximity and speed equal risk), and geography uses "minimum distance" (proximity equals risk). This ensures that the assessment results directly address the key issues and effectively reflect the main risks. Since all calculations are based on predictive data, this method can calculate the trend of the environmental state in advance before the aircraft actually encounters a risk.

[0030] In the embodiments of this application, a method for safety assessment of low-altitude aircraft based on multi-source data fusion is provided. The method for determining the influence coefficient of the environment on flight safety includes: determining historical flight environment data and historical flight safety data, and dividing the historical flight environment data into multiple historical flight environment parameter data groups according to parameter type; for the i-th historical flight environment parameter data group, determining the historical flight safety data when other flight environment parameters are the same, and determining the first change amount of the i-th historical flight environment parameter data group and the second change amount of the corresponding historical flight safety data; calculating the ratio of the second change amount to the first change amount to obtain the sub-influence coefficient of each flight environment parameter on flight safety, and adding the sub-influence coefficients of each flight environment parameter on flight safety to obtain the influence coefficient of the environment on flight safety.

[0031] Specifically, historical flight environment data and corresponding historical flight safety data are collected. The environmental data is divided into multiple historical flight environment parameter data groups according to their physical and functional attributes. For the i-th environmental parameter data group (e.g., "wind speed"), flight data records in the vast historical database that are identical or extremely similar to all other environmental parameters except for that parameter are searched. After finding the data sample set of the aforementioned "control variables", the first change is calculated, which is the change of the environmental parameter (wind speed) itself, and the second change is calculated, which is the change of the corresponding historical flight safety data. The ratio of the two is calculated: sub-influence coefficient = second change / first change. When other environmental conditions remain unchanged, how much the flight safety value changes when the environmental parameter increases by one unit. The sub-influence coefficients calculated for each environmental parameter are added together to obtain the total influence coefficient of the environment on flight safety. This total coefficient represents the maximum potential impact on flight safety when all environmental parameters change together. This step removes the coupling and confusion between environmental factors, enabling precise measurement of the independent and purely marginal impact of each environmental parameter (such as wind speed, visibility, and traffic density) on flight safety.

[0032] In an embodiment of this application, a safety assessment method for low-altitude aircraft based on multi-source data fusion is provided. The method for determining the environmental impact coefficient based on the environmental state coefficient and the impact coefficient includes: multiplying the environmental state coefficient and the impact coefficient to obtain the environmental impact coefficient.

[0033] Specifically, the environmental impact coefficient = environmental state coefficient × impact coefficient. If a certain environment is very harsh, and my aircraft happens to be highly sensitive to this environment, then the actual risk it brings will be extremely high. If the environment is equally harsh, but my aircraft is specially designed for this type of environment and has strong anti-interference capabilities, then its actual risk will be significantly reduced. Conversely, in a very good environment, regardless of the impact coefficient, the final actual risk will be very low. This step enables personalized and precise safety assessment, breaking the traditional "one-size-fits-all" safety assessment model. The actual risks posed by the same thunderstorm weather to a large, heavy logistics drone and a small, lightweight surveying drone are completely different. The multiplicative model accurately captures and quantifies this difference, making the assessment results no longer a general environmental warning, but strongly correlated with the current operating entity.

[0034] In the embodiments of this application, a method for safety assessment of low-altitude aircraft based on multi-source data fusion is provided. The method involves correcting the flight safety status assessment value based on the environmental impact coefficient to obtain a comprehensive safety assessment value, and determining the safety level of the low-altitude aircraft based on the comprehensive safety assessment value. The method includes: multiplying the environmental impact coefficient by the flight safety status assessment value to obtain the comprehensive safety assessment value of the low-altitude aircraft; pre-setting a preset safety level-comprehensive safety assessment value interval correspondence, wherein each comprehensive safety assessment value interval is associated with a corresponding preset safety level; obtaining the comprehensive safety assessment value of the low-altitude aircraft, and selecting the preset safety level corresponding to the comprehensive safety assessment value interval as the safety level of the low-altitude aircraft based on the mapping relationship between the comprehensive safety assessment value interval to which the comprehensive safety assessment value belongs and the preset safety level-comprehensive safety assessment value interval correspondence.

[0035] Specifically, the flight safety status assessment value, representing the internal health status of the aircraft, is multiplied by the environmental impact coefficient, representing the actual threat from the external environment. A preset lookup table is used to map the corresponding safety level and comprehensive safety assessment value ranges. This table discretizes continuous comprehensive safety assessment values ​​into several distinct safety levels. For example: [0.9, 1.0] -> safe (green), [0.7, 0.9] -> caution (yellow), [0.3, 0.7] -> warning (orange), [0.0, 0.3] -> danger (red). After calculating the comprehensive safety assessment value, it is compared with this lookup table to determine its range, and finally, the preset safety level associated with that range is output. This step achieves deep integration and precise quantification of internal and external risks, avoiding the risk masking problem that may result from simple weighted averaging. It ensures that any high risk from any party will be significantly reflected in the final result, making the assessment results extremely sensitive and reliable, and able to truly reflect "the overall risk faced by an aircraft in a specific health condition under a specific environment." The final output is not a complex number that requires further interpretation, but a clear and internationally recognized safety level label, reducing the cognitive load of human-machine interaction. Whether it is the pilot, the ground station operator, or the automated system, they can understand the current safety situation instantly and trigger the preset standardized response process, saving valuable time in dealing with emergencies.

[0036] like Figure 2 As shown in the embodiments of this application, a low-altitude aircraft safety assessment system based on multi-source data fusion is provided, comprising: an acquisition module for acquiring real-time flight data of the low-altitude aircraft and analyzing the real-time flight data to determine key flight parameters related to aircraft safety; an assessment module for determining the data characteristics of each key flight parameter, assessing the safety status of the low-altitude aircraft based on each key flight parameter and data characteristics, and obtaining a flight safety status assessment value; a prediction module for acquiring historical flight environment data, constructing a flight environment prediction model based on the historical flight environment data and a preset neural network model, and obtaining flight environment prediction data; a calculation module for analyzing the flight environment prediction data, determining the environmental state coefficient, determining the environmental impact coefficient on flight safety, and determining the environmental impact coefficient based on the environmental state coefficient and the impact coefficient; and a determination module for correcting the flight safety status assessment value based on the environmental impact coefficient to obtain a comprehensive safety assessment value, and determining the safety level of the low-altitude aircraft based on the comprehensive safety assessment value.

[0037] In summary, this invention provides a method and system for safety assessment of low-altitude aircraft based on multi-source data fusion. The method includes: analyzing real-time flight data of the low-altitude aircraft, determining key flight parameters and their data characteristics; assessing the safety status of the low-altitude aircraft based on these parameters and data characteristics to obtain a flight safety status assessment value; constructing a prediction model based on historical flight environment data and a preset model to obtain flight environment prediction data; analyzing the flight environment prediction data to determine environmental state coefficients and influence coefficients; determining an environmental influence coefficient based on both; correcting the flight safety status assessment value based on the environmental influence coefficient to obtain a comprehensive safety assessment value; and determining the safety level of the low-altitude aircraft based on this comprehensive safety assessment value. This invention improves the ability to comprehensively monitor and assess the safety status of low-altitude aircraft, helps to accurately determine the safety status of aircraft, thereby improving the flight safety level and reducing flight risks.

[0038] Finally, it should be noted that those skilled in the art can obviously make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims and their equivalents, this invention also intends to include these modifications and variations.

[0039] The above description is merely one embodiment of the present invention, and should not be construed as limiting the scope of the invention. Any structural changes made based on the present invention, as long as they do not depart from the essence of the invention, should be considered as falling within the protection scope of the present invention and subject to its restrictions. Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process and related descriptions of the platform described above can be referred to the corresponding processes in the foregoing platform embodiments, and will not be repeated here.

[0040] The term "comprising" or any other similar term is intended to cover non-exclusive inclusion, such that a process, platform, article, or device / platform that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to those processes, platforms, articles, or devices / platforms.

[0041] The technical solutions of the present invention have been described in conjunction with the accompanying drawings and further embodiments. However, it will be readily understood by those skilled in the art that the scope of protection of the present invention is obviously not limited to these specific embodiments. Without departing from the principles of the present invention, those skilled in the art can make equivalent changes or substitutions to closely related technical features, and the technical solutions resulting from such changes or substitutions will all fall within the scope of protection of the present invention.

[0042] The above description is merely a preferred embodiment of the present invention and is not intended to limit the scope of protection of the present invention.

Claims

1. A safety assessment method for low-altitude aircraft based on multi-source data fusion, characterized in that, include: Acquire real-time flight data of low-altitude aircraft and analyze the real-time flight data to determine key flight parameters related to aircraft safety; The data characteristics of each key flight parameter are determined, and the safety status of the low-altitude aircraft is assessed based on the key flight parameters and data characteristics to obtain the flight safety status assessment value. Historical flight environment data is acquired, and a flight environment prediction model is constructed based on the historical flight environment data and a preset neural network model to make predictions and obtain flight environment prediction data. Analyze flight environment prediction data, determine the environmental state coefficient, determine the environmental impact coefficient on flight safety, and determine the environmental impact coefficient based on the environmental state coefficient and the impact coefficient; The flight safety status assessment value is corrected based on the environmental impact coefficient to obtain a comprehensive safety assessment value, and the safety level of the low-altitude aircraft is determined based on the comprehensive safety assessment value.

2. A method for safety assessment of low-altitude aircraft based on multi-source data fusion as described in claim 1, characterized in that, The process of acquiring real-time flight data of low-altitude aircraft and analyzing this data to determine key flight parameters related to aircraft safety includes: Acquire real-time flight data of low-altitude aircraft and divide the real-time flight data into multiple flight parameter data groups according to parameter type; The correlation between each flight parameter data set and aircraft safety was determined, and the parameters corresponding to the flight parameter data sets with a correlation greater than a preset correlation threshold were identified as key flight parameters related to aircraft safety.

3. A method for safety assessment of low-altitude aircraft based on multi-source data fusion as described in claim 2, characterized in that, The process involves determining the data characteristics of each key flight parameter, assessing the safety status of the low-altitude aircraft based on these key flight parameters and data characteristics, and obtaining a flight safety status assessment value, including: Identify the flight parameter data sets corresponding to each key flight parameter, and construct time progress data change curves based on each flight parameter data set; Determine the values ​​of preset data statistical indicators in each data change curve, and select the preset data statistical indicators that exceed the corresponding preset thresholds to identify them as data characteristics of each key flight parameter. Calculate the difference between the value of each data feature and the corresponding preset threshold, and evaluate the difference to obtain the deviation evaluation value of each data feature; The deviation evaluation values ​​of all data features corresponding to each key flight parameter are standardized, and the standardized deviation evaluation values ​​are added together to obtain the comprehensive deviation evaluation value of each key flight parameter. The preset weights of each key flight parameter are determined, and the flight safety status assessment value of the low-altitude aircraft is obtained by calculating based on the preset weights of each key flight parameter and the comprehensive deviation assessment value.

4. A method for safety assessment of low-altitude aircraft based on multi-source data fusion as described in claim 3, characterized in that, The formula for calculating the flight safety status assessment value of the low-altitude aircraft is as follows: ， Where P is the flight safety status assessment value of the low-altitude aircraft, ai is the preset weight of the i-th key flight parameter, Gi is the comprehensive deviation assessment value of the i-th key flight parameter, and n is the number of key flight parameters.

5. A method for safety assessment of low-altitude aircraft based on multi-source data fusion as described in claim 3, characterized in that, The process of acquiring historical flight environment data and constructing a flight environment prediction model based on the historical flight environment data and a preset neural network model to make predictions, thereby obtaining flight environment prediction data, includes: Historical flight environment data is acquired, including historical meteorological data sets, historical airspace traffic data sets, and historical geographic data sets, and the data characteristics of each data set are extracted. Data sets are constructed based on each data group and its corresponding key data features, and the data sets are input into the corresponding preset neural network models to construct the corresponding initial prediction models. Each dataset is divided into a training set and a test set according to a preset ratio, and the training set and test set are input into the corresponding prediction initial model; Each initial prediction model is trained and tested until it meets the preset convergence condition, thus obtaining each prediction model. The various prediction models are fused to obtain a flight environment prediction model. Real-time flight environment data is then acquired and input into the flight environment prediction model for prediction, resulting in flight environment prediction data for a future period of time.

6. A method for safety assessment of low-altitude aircraft based on multi-source data fusion as described in claim 5, characterized in that, The analysis of flight environment prediction data, and the determination of environmental state coefficients, includes: Determine the meteorological forecast data group, airspace traffic forecast data group, and geographical forecast data group from the flight environment forecast data; Identify the meteorological parameter data sets within the meteorological forecast data set, and calculate the average value of each meteorological parameter data set; Determine the baseline average value for each meteorological parameter data set, and calculate the difference between the average value of each meteorological parameter data set and the corresponding baseline average value to obtain the data deviation of each meteorological parameter data set; Determine the preset coefficients for each meteorological parameter data set, and multiply the data deviations of each meteorological parameter data set by the preset coefficients to obtain the first state coefficients of the meteorological forecast data set. Based on the airspace traffic prediction data set, the predicted intrusion targets are identified, and the distance and time of arrival at the closest point of the predicted trajectory between each predicted intrusion target and the low-altitude aircraft are determined. The sub-safety factor between each predicted intrusion target and the low-altitude aircraft is calculated based on the closest point distance and closest point time. The formula for calculating the sub-safety factor is as follows: R = α*(1 / A) + β*(1 / B), Where R is the sub-safety coefficient, α is the first preset weight, A is the distance to the closest point, β is the second preset weight, and B is the time to the closest point; The sub-safety coefficients between each predicted intrusion target and the low-altitude aircraft are added together to obtain the second state coefficient of the airspace traffic prediction data set. Obstacles were identified based on a set of geographic prediction data, and the distances between the low-altitude aircraft and each obstacle were calculated. The maximum distance between the low-altitude aircraft and each obstacle is determined, and the maximum distance is multiplied by a preset safety conversion coefficient to obtain the third state coefficient of the geographic prediction data set. The first state coefficient of the meteorological forecast data group, the second state coefficient of the airspace traffic forecast data group, and the third state coefficient of the geographical forecast data group were normalized respectively, and the results of the normalization were added together to obtain the environmental state coefficient.

7. A method for safety assessment of low-altitude aircraft based on multi-source data fusion as described in claim 6, characterized in that, The determination of the environmental impact coefficient on flight safety includes: Identify historical flight environment data and historical flight safety data, and divide the historical flight environment data into multiple historical flight environment parameter data groups according to parameter type; For the i-th historical flight environment parameter data set, determine the historical flight safety data when other flight environment parameters are the same, and determine the first change amount of the i-th historical flight environment parameter data set and the second change amount of the corresponding historical flight safety data. Calculate the ratio of the second change to the first change to obtain the sub-influence coefficient of each flight environment parameter on flight safety, and add the sub-influence coefficients of each flight environment parameter on flight safety to obtain the influence coefficient of the environment on flight safety.

8. A method for safety assessment of low-altitude aircraft based on multi-source data fusion as described in claim 7, characterized in that, The determination of the environmental impact coefficient based on the environmental state coefficient and the impact coefficient includes: The environmental impact coefficient is obtained by multiplying the environmental state coefficient and the impact coefficient.

9. A method for safety assessment of low-altitude aircraft based on multi-source data fusion as described in claim 8, characterized in that, The process of correcting the flight safety status assessment value based on the environmental impact coefficient to obtain a comprehensive safety assessment value, and determining the safety level of the low-altitude aircraft based on the comprehensive safety assessment value, includes: The comprehensive safety assessment value of the low-altitude aircraft is obtained by multiplying the environmental impact coefficient by the flight safety status assessment value. A preset security level-comprehensive security assessment value range correspondence is set in advance. For each comprehensive security assessment value range, a corresponding preset security level is associated with it. Obtain the comprehensive safety assessment value of the low-altitude aircraft, and based on the mapping relationship between the comprehensive safety assessment value interval to which the comprehensive safety assessment value belongs and the preset safety level-comprehensive safety assessment value interval correspondence, select the preset safety level corresponding to the comprehensive safety assessment value interval as the safety level of the low-altitude aircraft.

10. A safety assessment system for low-altitude aircraft based on multi-source data fusion, characterized in that, include: The acquisition module is used to acquire real-time flight data of low-altitude aircraft and analyze the real-time flight data to determine key flight parameters related to aircraft safety. The evaluation module is used to determine the data characteristics of each key flight parameter, and to evaluate the safety status of the low-altitude aircraft based on each key flight parameter and data characteristics, thereby obtaining the flight safety status evaluation value. The prediction module is used to acquire historical flight environment data and construct a flight environment prediction model based on the historical flight environment data and a preset neural network model to make predictions and obtain flight environment prediction data. The calculation module is used to analyze flight environment prediction data, determine the environmental state coefficient, determine the environmental impact coefficient on flight safety, and determine the environmental impact coefficient based on the environmental state coefficient and the impact coefficient. The determination module is used to correct the flight safety status assessment value based on the environmental impact coefficient to obtain a comprehensive safety assessment value, and to determine the safety level of the low-altitude aircraft based on the comprehensive safety assessment value.

Images