Method and system for monitoring active safety of low-altitude economic airspace

Abstract

The invention discloses a low-altitude economic airspace active safety monitoring method and system, and relates to the related field of low-altitude safety technology.The method comprises the steps that before an aircraft takes off, a flight mission risk assessment mechanism is established, the risk level of a flight mission is determined through an entropy weight method and fuzzy comprehensive evaluation, and the flight mission of the aircraft is comprehensively arranged; performing flight control adaptation on the aircraft; a small target detection method based on a deep learning model is applied in the flight process of the aircraft to monitor and predict the obstacle distribution condition on the path, so that emergency obstacle avoidance of the aircraft is realized, and the flight path is dynamically optimized; in addition, a data management platform in the system carries out safety management on the whole life cycle of flight data, establishes an omnibearing authentication system, and combines quantum encryption, constructs a safety communication network and monitors hidden threats. According to the invention, a three-in-one active safety guarantee system of flight control safety, flight safety and flight data safety is established, and the intelligence and automation level of active safety of a low-altitude economic airspace is improved.

Description

Technical Field

[0001] This application relates to the field of low-altitude safety technology, and in particular to a monitoring method and system for active safety in low-altitude economic airspace. Background Technology

[0002] The low-altitude economy is a comprehensive new economic form that takes place in low-altitude airspace (usually below 1,000 meters, but can be extended to no more than 3,000 meters depending on actual needs). It is mainly based on civilian manned and unmanned aircraft, and is driven by low-altitude flight activities in various scenarios such as carrying people, cargo and other operations. It also radiates and drives the integrated development of commercial activities or public service sectors. The degree of openness and management efficiency of low-altitude airspace directly affects the development of the low-altitude economy.

[0003] The low-altitude economy involves various aircraft such as drones and helicopters. Without effective airspace safety management, collisions are highly likely, leading to serious consequences. Furthermore, low-altitude airspace is a crucial component of national airspace; its insecurity can allow illegal activities such as intrusion, smuggling, and human trafficking, posing a serious threat to national security and social stability. A safe low-altitude airspace environment is essential for the innovation and upgrading of the low-altitude economy, propelling the industry towards higher levels of development. Therefore, proactive safety technologies for the low-altitude economy are key to ensuring airspace safety. These technologies utilize a series of proactive, preventative, and dynamic measures and techniques to identify, assess, and control potential safety risks in advance, ensuring the safe, orderly, and efficient operation of low-altitude flight activities. However, existing proactive safety technologies have limited monitoring capabilities, insufficient positioning accuracy, and require further improvement in data processing capabilities. Summary of the Invention

[0004] To address the technical problems of the prior art, this application provides a monitoring method and system for active safety in low-altitude economic airspace, establishing a three-in-one active safety assurance system encompassing flight control safety, flight safety, and flight data security, thereby enhancing the intelligence and automation level of active safety in low-altitude economic airspace.

[0005] This application provides a method for monitoring active safety in low-altitude economic airspace, including: (1) Before the aircraft takes off, establish a flight mission risk assessment mechanism, conduct a comprehensive review of meteorological data, aircraft status and flight area safety, determine the risk level of the flight mission through entropy weight method and fuzzy comprehensive evaluation, comprehensively arrange the flight mission of the aircraft, and adapt the flight control of the aircraft. (2) During the flight mission, the aircraft acquires sensor data and low-altitude airspace environmental data in real time, and uses a small target detection method based on deep learning model to monitor and predict the distribution of obstacles on the path, so as to realize emergency obstacle avoidance and dynamically optimize the flight path. (3) Encryption algorithms are used to encrypt the acquired aircraft sensor data and low-altitude airspace environment data to achieve full life-cycle security management of flight data, establish a comprehensive authentication system, and combine quantum encryption to build a secure communication network to monitor hidden security threats and attack behaviors.

[0006] Furthermore, a safety assessment is conducted before the aircraft performs a flight mission to ensure that flight conditions meet safety requirements and to rationally allocate flight missions. This involves a comprehensive review of meteorological data, aircraft status, and safety indicators of the flight area to identify and quantify risk factors. The entropy weight method is used to determine the weight of each risk factor based on its entropy value: the smaller the entropy value, the greater the dispersion of the factor, the greater its contribution to the risk assessment, and the greater its weight. Fuzzy mathematics theory and methods are then used to perform fuzzy calculations on the quantified values ​​and weights of each risk factor to obtain the risk level of the flight mission. Based on the magnitude of the comprehensive risk value, the risk level of the flight mission is divided into low risk, medium risk, and high risk. For flight missions of different risk levels, appropriate aircraft are assigned to perform the missions, and corresponding risk control measures are taken to achieve unified flight control adaptation for the aircraft.

[0007] Furthermore, during the flight mission, the aircraft is monitored in real time. By applying a small target detection method based on a deep learning model and intelligently analyzing the aircraft's sensor data and environmental data, accurate obstacle location and risk prediction information can be provided for flight path planning and dynamic adjustment.

[0008] The low-altitude areas where aircraft perform flight missions are filled with various obstacles, including natural and man-made objects. These obstacles are relatively small in size within the aircraft's field of vision, and their flight speed and direction are constantly changing. By learning from a large amount of sample data and leveraging the powerful feature extraction capabilities of neural networks, the target detection model uses an attention mechanism to accurately capture the edge features of small targets, monitors dynamically changing obstacles in real time, and promptly captures their signals, providing a basis for emergency obstacle avoidance.

[0009] Furthermore, the data governance platform provides multiple measures to ensure flight data security: it adopts a combination of advanced encryption standards and symmetric encryption algorithms to ensure the high efficiency of data encryption; it conducts security management in all aspects of flight data collection, transmission, storage, processing, opening, and destruction; and it verifies the identity and permissions of users and devices through comprehensive authentication methods to ensure that only authorized users or devices can access specific resources or perform specific operations, thereby ensuring the safety and controllability of the aircraft and preventing illegal operations and data tampering.

[0010] This application also provides a monitoring system for active safety in low-altitude economic airspace, including: Flight control interface management platform: used to establish a flight mission risk assessment mechanism before the aircraft takes off, comprehensively review meteorological data, aircraft status and flight area safety, determine the risk level of the flight mission through entropy weight method and fuzzy comprehensive evaluation, comprehensively arrange the aircraft's flight mission, and perform flight control adaptation for the aircraft; Aircraft monitoring and operation platform: used to acquire real-time sensor data and low-altitude airspace environmental data of aircraft during flight missions, apply small target detection methods based on deep learning models, monitor and predict the distribution of obstacles on the path, realize emergency obstacle avoidance of aircraft, and dynamically optimize flight path; Data governance platform: Used to encrypt acquired aircraft sensor data and low-altitude airspace environmental data using encryption algorithms, realize full lifecycle security management of flight data, establish a comprehensive authentication system, and combine quantum encryption to build a secure communication network to monitor hidden security threats and attack behaviors.

[0011] The present invention discloses the following technical effects: This invention provides a monitoring method and system for proactive safety in low-altitude economic airspace. Through a series of proactive, preventative, and dynamic monitoring methods, it comprehensively safeguards low-altitude economic safety, ensures that low-altitude aircraft such as drones comply with safety regulations during flight, and avoids safety issues caused by flight control incompatibility or lack of authentication. This invention establishes a flight mission risk assessment mechanism to improve the reliability of flight mission allocation. Before flight, it comprehensively reviews meteorological data, aircraft status, and flight area safety. A fuzzy evaluation mathematical model is used to rationally classify risk levels, assigning suitable flight missions to the aircraft and reducing the possibility of dangerous accidents during mission execution. During mission execution, the invention monitors the aircraft's operational status and surrounding environment in real time, leveraging artificial intelligence's ability to efficiently and accurately detect emergencies and enable emergency obstacle avoidance, thus safeguarding flight safety. To reduce the possibility of privacy leaks in flight data, this invention uses a data governance platform to manage the entire lifecycle of flight data securely, employing comprehensive authentication methods to maximize data security. This invention establishes a three-in-one proactive safety system for low-altitude economic airspace, encompassing flight control safety, flight safety, and flight data security, enhancing the system's real-time flight monitoring capabilities, data processing and analysis capabilities, and intelligence level. Attached Figure Description

[0012] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings of the embodiments of the present invention will be briefly described below. Flowcharts are used in this application to illustrate the operations performed by the system according to the embodiments of the present application. It should be understood that the preceding or following operations are not necessarily performed precisely in sequence. Instead, various steps can be processed in reverse order or simultaneously as needed. Furthermore, other operations can be added to these processes, or one or more steps can be removed from these processes.

[0013] Figure 1 A schematic diagram of the monitoring method for active safety in low-altitude economic airspace provided in this application embodiment.

[0014] Figure 2 A schematic diagram of the monitoring system structure for active safety in low-altitude economic airspace provided in this application embodiment. Detailed Implementation

[0015] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, the following are specific embodiments of this application.

[0016] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description of this application will be provided in conjunction with the accompanying drawings. The described embodiments should not be considered as limitations on this application. All other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0017] In the following description, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those explicitly listed, but may include other steps or modules not explicitly listed or inherent to such processes, methods, products, or devices. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing embodiments of this application only.

[0018] Example 1: This application provides a method for monitoring active safety in low-altitude economic airspace, such as... Figure 1 As shown, the method includes: Step S10: Before the aircraft takes off, establish a flight mission risk assessment mechanism, comprehensively review meteorological data, aircraft status and flight area safety, determine the risk level of the flight mission through entropy weight method and fuzzy comprehensive evaluation, comprehensively arrange the aircraft's flight mission, and adapt the aircraft to flight control.

[0019] In this embodiment, a comprehensive review of meteorological data, aircraft status, and flight area safety is conducted 24 hours before flight, first identifying the risk factors for the aircraft to perform the flight mission: Meteorological factors include wind speed, wind direction, temperature, air pressure, precipitation probability, and visibility; aircraft condition factors include the aircraft's mechanical performance, electronic equipment functionality, flight control system reliability, and flight hours; and flight area safety factors include airspace congestion, airport facility conditions, and airspace restrictions.

[0020] Secondly, based on the identified flight mission risk factors, different types of risk factors are quantified: When quantifying meteorological factors, continuous variables such as wind speed and wind direction are classified into low wind speed, medium wind speed and high wind speed according to their impact on flight, and assigned corresponding quantitative values; for temperature and air pressure, they are quantified according to their deviation from the normal range, with a quantification range of [0,1]. The greater the deviation, the closer the quantification value is to 1; for precipitation probability and visibility, the original values ​​are directly used as quantification values. When quantifying the state factors of an aircraft, for mechanical performance and electronic equipment functionality, the state is categorized into three levels—good, average, and poor—by reviewing periodic inspection and maintenance records, and assigned corresponding quantitative values. For the reliability of the flight control system, quantification is based on the failure rate obtained from maintenance records; the lower the failure rate, the smaller the quantitative value. For the aircraft's flight hours, quantification is based on its ratio to the aircraft's design life: when the flight hours reach the design life... Quantization value ; When quantifying flight area safety factors, for airspace congestion, the airspace is divided into busy airspace, moderately busy airspace, and idle airspace based on the number of aircraft passing through the target mission airspace per unit time, and a corresponding quantitative value is assigned. For airport facility conditions, a comprehensive quantification is performed based on the accuracy of the airport's navigation equipment, runway conditions, and airspace clearance conditions: airports with high equipment accuracy, good runway conditions, and good airspace clearance conditions have lower quantification values, and vice versa. For airspace restrictions, quantification is performed based on whether the flight mission involves no-fly zones and restricted areas: the quantification value is 0.1 for no-fly zones and restricted areas at all, 0.5 for partially involved areas, and 0.9 for fully involved areas.

[0021] Then, the entropy weight method was used to determine the weight of each risk factor in the flight mission risk assessment model based on the entropy value of each risk factor. Before determining the weight, an initial aircraft was specified for each flight mission, and the quantitative value of each risk factor in meteorological data, aircraft status, and flight area safety was calculated. Let there be a total of m There are 1 flight mission, each mission has 1n For each risk factor, the quantified value of each risk factor is standardized to make it dimensionless:

[0022] in, It is the first The first flight mission The original quantitative values ​​of each risk factor, It is a standardized quantified value; and Let these represent the minimum and maximum value functions, respectively. A decision matrix is ​​constructed using the standardized quantized values. :

[0023] in, and Represent the total number of flight missions and risk factors, respectively; calculate the weight of each risk factor in each flight mission:

[0024] in, Indicates the first The first risk factor in the The proportion of each flight mission; Calculate the entropy value of each risk factor based on its proportion in the flight mission:

[0025] in, Indicates the first The entropy value of each risk factor. Here is the normalization constant; calculate the difference coefficient for each risk factor: , Indicates the first The difference coefficients of each risk factor are used to obtain the weight of each risk factor:

[0026] in, Indicates the first The weight of each risk factor in the risk assessment model.

[0027] Finally, a fuzzy comprehensive evaluation model was established to complete the flight mission risk assessment mechanism. Based on the standardized decision matrix Based on the risk factor weights, a fuzzy evaluation matrix is ​​constructed to classify the risk level of flight missions into three levels: low risk, medium risk, and high risk. Represented as:

[0028] in, Indicates the first The first flight mission Membership degree of each risk level, 2 and These represent low risk, medium risk, and high risk, respectively. Based on actual needs, determine the membership function of each risk factor to different risk levels. The membership functions for low-risk, medium-risk, and high-risk are defined as follows:

[0029]

[0030]

[0031] in, , and Indicates risk factors Membership degree to low-risk, medium-risk, and high-risk categories; , and The threshold is based on actual data. By calculating the statistical characteristics of each risk factor, the mean is used as the center point of medium risk, and the mean plus or minus the standard deviation is used as the upper and lower bounds of medium risk. Those below and above this range are respectively classified as low risk and high risk. Based on the membership function, calculate the fuzzy evaluation matrix for each flight mission. The first flight mission The membership degrees of each risk factor to low-risk, medium-risk, and high-risk are respectively... , and Then the fuzzy evaluation matrix The Line 1 The elements of the column are ; Risk factor weight vector With fuzzy evaluation matrix Perform fuzzy multiplication, where, They represent The weights of each risk factor are used to obtain a comprehensive evaluation result for each flight mission:

[0032] in, Indicates the first The overall classification of a flight mission as low-risk, medium-risk, and high-risk. This represents the fuzzy synthesis operation, using the maximum-minimum synthesis method, and the final result is:

[0033] Based on the comprehensive evaluation results, the risk level of each risk mission is determined, and the risk level corresponding to the maximum value of the comprehensive evaluation results is used as the level assessment result. Low-risk missions can be executed normally, medium-risk missions require further risk analysis and control, the aircraft to be assigned to the missions are re-assessed for risk assessment, and high-risk missions need to be postponed or canceled until the risk is reduced to an acceptable level.

[0034] Based on the risk assessment results of the flight mission, after assigning flight missions to each aircraft, a unified flight control system is adapted to achieve decoupling between flight control and payload, and between the management platform and communication module. Define a unified interface standard for payloads and communication modules to enable different types of payloads and communication modules to be compatible with flight control systems and management platforms. Develop dedicated drivers for each payload to enable communication and control between the payload and the flight control system and management platform. Flexibly configure payloads and communication modules according to flight mission requirements. Monitor the working status of payloads and communication modules in real time to ensure normal operation of the payload and stable and reliable communication links. Adapters are used to resolve interface differences, ensuring that all sensors of the aircraft are compatible with the flight control system interface. The frequency and pulse width of the PWM signal are adjusted to adapt to different types of actuators, ensuring that all actuators can receive and execute commands from the flight control system accurately. A load driver is used for load initialization and status monitoring. According to the requirements of the flight mission, update the firmware version of the flight control system, use the common communication protocol MAVLink for data transmission, adjust the PID parameters of the flight control system according to the characteristics of the flight mission, and regularly maintain the hardware and software systems of the aircraft to ensure their normal operation.

[0035] In step S20, during the flight mission performed by the aircraft according to the schedule, the aircraft's sensor data and low-altitude airspace environmental data are acquired in real time. A small target detection method based on a deep learning model is applied to monitor and predict the distribution of obstacles on the path, so as to realize emergency obstacle avoidance and dynamically optimize the flight path. In this embodiment, the aircraft is equipped with sensors including lidar and cameras to acquire environmental data around the aircraft, collect image, distance, speed and angle information; and acquire real-time weather data through meteorological services, acquire aircraft navigation data through the ADS-B system, and obtain the height and position of obstacles from the terrain database.

[0036] To address dynamic small-target obstacles encountered by aircraft during flight, including birds and small drones, a target detection model is established to automatically identify obstacle locations, predict obstacle trajectories, and enable emergency obstacle avoidance. The detailed steps for obtaining the pre-trained target detection model are as follows: Collect images and infrared imaging data of different types of small target obstacles in the flight airspace of the aircraft, including obstacles at different distances from the aircraft, at different speeds and with different shapes, label the location and type of the obstacles, and use them as the dataset for model training. The dataset is divided into training set, validation set and test set in a ratio of 8:1:1. Construct an object detection model, set the input and output dimensions of each convolutional and linear layer in the model, select the SGD optimizer, set the initial learning rate to 0.0001, and adjust the training period and batch size; During training, the learning rate is dynamically adjusted, cross-validation is used, and the change curve of the loss function is recorded in real time. After each training cycle, the error rate, recall rate, and F1 score are used to evaluate the model performance. When the loss function converges, the parameters of the best-performing model are saved as the pre-trained object detection model.

[0037] By integrating pre-trained target detection models, path planning algorithms, obstacle avoidance algorithms, and other software systems into the flight control system, and based on the obstacle positions output by the model, a target tracking algorithm is introduced to predict the obstacle's trajectory, assess the potential risks to the flight path, and combine current aircraft operating status data, meteorological data, and obstacle distribution information to make obstacle avoidance decisions and path planning in real time. The system calculates the distance between the aircraft and obstacles to determine if it has entered a safe distance range; assesses the relative speed of the obstacles to determine if they pose an immediate threat to the aircraft; based on the assessment results, it controls the aircraft to change its flight path, adjust its flight altitude, and adjust its speed; and it uses an A*-based path planning algorithm to optimize the flight path, dynamically updating it based on real-time data to ensure the real-time performance and safety of the path.

[0038] Step S30: Encryption algorithms are used to encrypt the acquired aircraft sensor data and low-altitude airspace environmental data to achieve full lifecycle security management of flight data, establish a comprehensive authentication system, and combine quantum encryption to build a secure communication network to monitor hidden security threats and attack behaviors.

[0039] In this embodiment, security management is implemented throughout the entire lifecycle of flight data acquisition, transmission, storage, processing, access, and destruction, including the following specific methods: During the data acquisition phase, symmetric encryption algorithms are used to encrypt the real-time acquired aircraft sensor data and low-altitude airspace environmental data, and key management services are applied to generate and manage encryption keys. During data transmission, the SSL / TLS protocol is used to encrypt the data, and quantum key distribution is used to generate unclonable keys to further improve the security of data transmission. During the data storage phase, encryption keys are distributed and stored in different storage media or devices to increase the difficulty of key theft. Data is backed up regularly and the backup data is stored in a secure location. During the data processing phase, a secure Trusted Execution Environment (TEE) is used to implement fine-grained access control policies, restricting the scope of data access based on user roles and permissions. During the data sharing phase, sensitive data is anonymized, data sharing agreements are established, and the purpose, scope, and responsibilities for data use are clearly defined to ensure the legal use of data. During the data destruction phase, secure data erasure tools are used to physically destroy the physical media storing sensitive data, ensuring data security.

[0040] A comprehensive authentication system includes three aspects: identity authentication, access control, and network authentication. A multi-factor authentication mechanism is adopted, combining username and password, digital certificate and biometric identification, and introducing dynamic passwords to improve identity authentication security; a role-based permission management approach is applied to assign different permissions according to the user's role, and access control lists are used to restrict the access permissions of users or devices to specific resources; the network access authorization technology 802.1X is used to verify the network access permissions of devices, and the network side is upgraded to support UAS network functions to complete the identification, mapping and authorization of the aircraft's trusted identifier.

[0041] Example 2: The small target detection method based on a deep learning model provided in this embodiment of the invention establishes a target detection model containing a multi-layer neural network, processes and analyzes the data, and automatically outputs prediction results.

[0042] In this implementation, the target detection model consists of a backbone network, a neck network, and a head network. To reduce network complexity and facilitate embedding into the flight control system, the model adopts a lightweight structural design. A parameter-free attention mechanism is applied within the convolutional units of the backbone and neck networks to filter small and dark features from the input infrared image, enhancing the relevance of local contextual information. This attention mechanism simulates the human brain's attention mechanism to generate three-dimensional weights, assigning a unique weight to each neuron. First, define an energy function for each neuron:

[0043] in, Represents the target neuron Energy value, and These represent the mean and standard deviation of the input feature within a single channel, respectively. The first term is a constant; secondly, the reciprocal of the energy value is taken to obtain the importance of each neuron. Then, a scaling operator is used to refine the input features, grouping all energy values ​​by their reciprocals in the channel and spatial dimensions to obtain the weight matrix. Furthermore, by limiting the range of values ​​in the matrix using the sigmoid function, the weight matrix is ​​multiplied point by point with the input features. Without increasing the network structure parameters, the attention weights of the feature map are obtained, which improves the detection accuracy and robustness of small targets.

[0044] To further reduce model parameters while ensuring high detection accuracy, a lightweight three-scale neck network is employed. A single aggregation module is used, combining standard convolution, depthwise separable convolution, and channel shuffling operations to design a grouped hybrid convolution, reducing inference time and improving inference speed. The grouped hybrid convolution first expands the channel dimension of the input features through convolutional layers and extracts feature information using lightweight depthwise separable convolution. Then, the output features of the depthwise separable convolution layer are concatenated with the original input features, and a channel shuffling operation is performed to mix the feature information from different channels, enhancing information exchange between channels. Finally, the single aggregation module, based on a residual structure, concatenates the feature maps of different scales output by the three-layer grouped hybrid convolution according to the channel dimension, continuously aggregating the original feature information, reducing the probability of overfitting, and preventing the loss of important information.

[0045] Example 3: The monitoring system for active safety in low-altitude economic airspace provided in this embodiment of the invention can execute the monitoring method for active safety in low-altitude economic airspace provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects of the method, such as... Figure 2 As shown, it includes the following modules: Flight control interface management platform: used to establish a flight mission risk assessment mechanism before the aircraft takes off, comprehensively review meteorological data, aircraft status and flight area safety, determine the risk level of the flight mission through entropy weight method and fuzzy comprehensive evaluation, comprehensively arrange the aircraft's flight mission, and perform flight control adaptation for the aircraft; Aircraft monitoring and operation platform: used to acquire real-time sensor data and low-altitude airspace environmental data of aircraft during flight missions, apply small target detection methods based on deep learning models, monitor and predict the distribution of obstacles on the path, realize emergency obstacle avoidance of aircraft, and dynamically optimize flight path; Data governance platform: Used to encrypt acquired aircraft sensor data and low-altitude airspace environmental data using encryption algorithms, realize full lifecycle security management of flight data, establish a comprehensive authentication system, and combine quantum encryption to build a secure communication network to monitor hidden security threats and attack behaviors.

[0046] Although this application makes various references to certain modules in the system according to the embodiments of this application, any number of different modules can be used and run on user terminals and / or servers. The various units and modules included are only divided according to functional logic, but are not limited to the above division, as long as the corresponding functions can be achieved; in addition, the specific names of each functional unit are only for easy distinction between each other and are not used to limit the scope of protection of this invention.

[0047] The specific embodiments described above do not constitute a limitation on the scope of protection of this application. Those skilled in the art should understand that various modifications, combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this application should be included within the scope of protection of this application. In some cases, the actions or steps described in this application can be performed in a different order than that shown in the embodiments and still achieve the desired results. Furthermore, the processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

Claims

1. A method for monitoring the active safety of low-altitude economic airspace, characterized in that, The method comprises: (1) Before the aircraft takes off, a flight task risk assessment mechanism is established, meteorological data, aircraft state and flight area safety are comprehensively audited, the risk level of the flight task is determined through the entropy weight method and fuzzy comprehensive evaluation, the flight task of the aircraft is comprehensively arranged, and the flight control of the aircraft is adapted; (2) In the process of the aircraft executing the flight task according to the arrangement, sensor data and low-altitude airspace environment data of the aircraft are acquired in real time, a small target detection method based on a deep learning model is applied, the distribution of obstacles on the path is monitored and predicted, the aircraft emergency obstacle avoidance is realized, and the flight path is dynamically optimized; (3) The acquired aircraft sensor data and low-altitude airspace environment data are encrypted by using an encryption algorithm, the whole life cycle safety management of flight data is realized, a comprehensive authentication system is established, a safe communication network is constructed by combining quantum encryption, and hidden security threats and attack behaviors are monitored.

2. The method of claim 1, wherein the low-altitude economic airspace active safety monitoring method is characterized by, In step (1), risk factors of the aircraft executing the flight task are identified from the aspects of meteorological data, aircraft state and flight area safety: The meteorological factors include wind speed, wind direction, air temperature, air pressure, precipitation probability and visibility; the aircraft state factors include mechanical performance, electronic equipment function, flight control system reliability and flight hours of the aircraft; the flight area safety factors include airspace busy degree, airport facility condition and airspace restriction; After identification, the risk factors are quantitatively processed according to the type, actual distribution and task demand of different types of flight tasks.

3. The method of claim 2, wherein the low-altitude economic airspace active safety monitoring method is characterized by, The weight of the risk factor in the flight task risk assessment model is determined by using the entropy value method, including the following steps: The quantitative value of each risk factor is normalized to obtain , which represents the standard quantitative value of the risk factor of the flight mission, and a decision matrix is constructed using the standard quantitative values ; The proportion of each risk factor in each flight task is calculated: wherein, represents the proportion of the risk factor in the flight mission, represents the total number of flight missions; The entropy value of each risk factor is calculated according to the proportion of the risk factor in the flight task: wherein, represents the entropy value of the risk factor, is a normalization constant; the difference coefficient of each risk factor is calculated: , represents the difference coefficient of the risk factor, and the weight of each risk factor is obtained: wherein, represents the weight of the th risk factor in the risk assessment model, represents the total number of risk factors.

4. The method of claim 3, wherein the low-altitude economic airspace active safety monitoring method is characterized by, The weight of the risk factor in the flight task risk assessment model is used to construct a fuzzy evaluation comprehensive model: According to the standardized decision matrix and risk factor weights, a fuzzy evaluation matrix is constructed to divide the risk level of flight tasks into three levels: low risk, medium risk, and high risk. Based on actual needs, determine the membership function of each risk factor to different risk levels. The membership functions for low-risk, medium-risk, and high-risk are defined as follows: wherein, , and represent risk factors membership to low, medium and high risk; , and are threshold values divided according to statistical characteristics of actual data; The fuzzy evaluation matrix of each flight task is calculated according to the low-risk, medium-risk and high-risk membership functions, and the weight vector is multiplied with the fuzzy evaluation matrix , wherein, , and the weight of each risk factor is represented by , to obtain the comprehensive evaluation result of each flight task. wherein, denotes the overall membership degree of the i-th flight mission to low, medium and high risk, denotes the overall membership degree of the i-th flight mission to low, medium and high risk, denotes the fuzzy composition operation, using the max-min composition method, and finally gives: in, and These represent the functions for finding the maximum and minimum values, respectively. Indicates the first The first risk factor in the The degree of membership in each flight mission is determined based on the comprehensive evaluation results, and the risk level of each risk mission is determined. The risk level corresponding to the maximum value of the comprehensive evaluation results is used as the evaluation result of the flight mission.

5. The method of claim 1, wherein the low-altitude economic airspace active safety monitoring method is characterized by, In step (2), for the dynamic small target obstacles encountered by the aircraft during flight, a target detection model is established to automatically identify the position of the obstacles and predict the action trajectory of the obstacles, realize emergency obstacle avoidance, and the steps of obtaining the pre-trained target detection model are as follows: Collect images and infrared imaging data of different types of small target obstacles in the flight area of the aircraft, including obstacles at different distances, different moving speeds and different shapes from the aircraft, label the obstacle position and type as the data set for model training, and divide them into training set, validation set and test set according to the ratio of 8:1:1; The target detection model is constructed, the input and output dimensions of each convolution and linear layer in the model are set, the SGD optimizer is selected, the initial learning rate is set to 0.0001, and the training period and batch size are adjusted; During the training process, the learning rate is dynamically adjusted, the cross-validation method is used, the change curve of the loss function is recorded in real time, the error rate, recall rate and F1 score are used to evaluate the model performance after each training period, and when the loss function converges, the model parameters with the best performance are saved as the pre-trained target detection model.

6. The method of claim 5, wherein the low-altitude economic airspace active safety monitoring method is characterized by, The target detection model is composed of a backbone network, a neck network and a head network; for embedding the flight control system, the model adopts a lightweight structure design, and a parameter-free attention mechanism is applied in the convolution unit of the backbone network and the neck network, small and dark feature information is screened from the input image and infrared imaging data, and the relevance of local context information is enhanced, and the attention mechanism simulates the human brain attention mechanism to generate a three-dimensional weight, and each neuron is assigned a unique weight.

7. The method of claim 6, wherein the low-altitude economic airspace active safety monitoring method is characterized by, The parameter-free attention mechanism defines an energy function for each neuron: wherein, represents the energy value of the target neuron , and respectively represent the mean and standard deviation of the input features in a single channel, is a constant term; taking the reciprocal of the energy value obtains the importance of each neuron, the input features are refined using the scaling operator, and all energy value reciprocals are grouped in the channel and spatial dimensions to obtain the weight matrix , and the value range of the numbers in the matrix is limited by the sigmoid function. The weight matrix is multiplied point by point with the input features to obtain the attention weight of the feature map without increasing the network structure parameters.

8. The method of claim 1, wherein the low-altitude economic airspace active safety monitoring method is characterized by, In step (3), the full life cycle of flight data collection, transmission, storage, processing, opening and destruction is managed, including the following methods: In the data collection stage, the real-time collected aircraft sensor data and low-altitude airspace environment data are encrypted using a symmetric encryption algorithm, and a key management service is applied to generate and manage encryption keys; In the data transmission process, the data is encrypted and transmitted using the SSL / TLS protocol, and a non-cloning key is generated using quantum key distribution; In the data storage stage, the encryption keys are stored in different storage media or devices to increase the difficulty of key theft, and the data is backed up regularly and stored in a secure location; In the data processing stage, a secure trusted execution environment is used to implement fine-grained access control policies, and access to data is limited according to the user's role and authority; In the data opening stage, sensitive data is desensitized, a data sharing agreement is formulated, the purpose, scope and responsibility of data use are clearly defined, and legal use of data is realized; In the data destruction stage, a secure data erasure tool is used to physically destroy the physical medium storing sensitive data.

9. The method of claim 1, wherein the low-altitude economic airspace active safety monitoring method is characterized by, In step (3), the comprehensive authentication system includes identity authentication, permission management and network authentication: A multi-factor authentication mechanism is adopted, combining username and password, digital certificate and biometric identification, and introducing dynamic password; a role-based permission management method is applied, different permissions are assigned according to the user's role, and the user or device's access to specific resources is limited through an access control list; the network access permission of the device is verified through network access authentication technology, and the network side is upgraded to support UAS network function, completing the identification, mapping and authentication authorization of the aircraft trusted identity.

10. A monitoring system for active safety of low-altitude economic airspace, characterized in that, The system is used to implement the low-altitude economic airspace active safety monitoring method of any one of claims 1-9, and the system comprises: A flight control interface management platform is used to establish a flight task risk assessment mechanism before the aircraft takes off, comprehensively review meteorological data, aircraft status and flight area safety, determine the risk level of the flight task through entropy weight method and fuzzy comprehensive evaluation, and comprehensively arrange the flight task of the aircraft, and perform flight control adaptation on the aircraft; An aircraft monitoring and operation platform is used to obtain real-time sensor data and low-altitude airspace environment data of the aircraft during the execution of the flight task arranged by the aircraft, apply a small target detection method based on a deep learning model, monitor and predict the obstacle distribution on the path, realize emergency obstacle avoidance of the aircraft, and dynamically optimize the flight path. Data governance platform: used to encrypt the obtained aircraft sensor data and low-altitude airspace environment data with encryption algorithms, realize the whole life cycle security management of flight data, establish a comprehensive authentication system, and combine quantum encryption to build a secure communication network to monitor hidden security threats and attack behavior.

Images