A low-altitude unmanned aerial vehicle route deviation monitoring method

Abstract

The application discloses a low-altitude unmanned aerial vehicle route deviation monitoring method and relates to the technical field of aircraft monitoring. The method comprises the following steps: defining a target area and an unmanned aerial vehicle; arranging a multi-source sensor on the unmanned aerial vehicle, acquiring multi-source positioning data, fusing the data according to the multi-source positioning data, and acquiring high-precision positioning data of the unmanned aerial vehicle; constructing a route deviation identification channel, collecting historical data of the target area and the unmanned aerial vehicle, and supervising and training the route deviation identification channel through the historical data; identifying deviation through the route deviation identification channel, acquiring a deviation identification result; and building a cloud platform, inputting the deviation identification result into the cloud platform, processing the deviation identification result, and displaying the deviation identification result. The application solves the technical problems that the prior art has deficiencies in real-time monitoring and accuracy and is difficult to meet the safety requirements in a complex flight environment, and achieves the technical effect of accurately monitoring and warning the route deviation of the unmanned aerial vehicle.

Description

Technical Field

[0001] This invention relates to the field of aircraft monitoring technology, specifically to a method for monitoring flight path deviation of low-altitude unmanned aerial vehicles. Background Technology

[0002] With the continuous advancement of aviation technology and the widespread application of low-altitude unmanned aerial vehicles (UAVs), they are playing an increasingly important role in many fields such as aerial photography, cargo transportation, and environmental monitoring. However, with the diversification of UAV flight missions and the increasing complexity of flight routes, the problem of flight path deviation is becoming increasingly prominent. Traditional flight path deviation monitoring methods often rely on pilot visual observation and simple positioning equipment, which are clearly insufficient in terms of real-time monitoring and accuracy. Summary of the Invention

[0003] This application provides a method for monitoring flight path deviation of low-altitude unmanned aerial vehicles, which addresses the technical problem that existing technologies are insufficient in terms of real-time monitoring and accuracy, making it difficult to meet safety requirements in complex flight environments.

[0004] In view of the above problems, this application provides a method for monitoring flight path deviation of low-altitude unmanned aerial vehicles.

[0005] A first aspect of this application provides a method for monitoring flight path deviation of a low-altitude unmanned aerial vehicle, the method comprising:

[0006] The process involves: defining the target area and the unmanned aerial vehicle (UAV), whereby the target area includes the flight environment and weather environment, and the UAV includes basic UAV information; deploying multi-source sensors on the UAV, acquiring multi-source positioning data based on these sensors, and fusing the multi-source positioning data to obtain high-precision positioning data for the UAV; constructing a flight path deviation identification channel, collecting historical data of the target area and the UAV, and conducting supervised training on the flight path deviation identification channel using the historical data; inputting the target area and the high-precision positioning data into the flight path deviation identification channel for deviation identification, and obtaining the deviation identification result; and building a cloud platform, which includes the target area, the UAV, the flight path deviation identification channel, and a control center, and inputting the deviation identification result into the cloud platform for processing and display.

[0007] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0008] This application clearly defines the target area and the unmanned aerial vehicle (UAV). The target area includes the flight environment and weather environment, and the UAV includes basic UAV information. Multi-source sensors are deployed on the UAV, and multi-source positioning data is acquired based on these sensors. Data fusion is performed on the multi-source positioning data to obtain high-precision positioning data for the UAV. A flight path deviation identification channel is constructed, and historical data of the target area and the UAV are collected. The flight path deviation identification channel is trained under supervised supervision using the historical data. The target area and the high-precision positioning data are input into the flight path deviation identification channel for deviation identification, and the deviation identification result is obtained. A cloud platform is built, which includes the target area, the UAV, the flight path deviation identification channel, and a control center. The deviation identification result is input into the cloud platform for processing and display. This invention addresses the shortcomings of existing technologies in real-time monitoring and accuracy, making it difficult to meet the safety requirements of complex flight environments. By integrating advanced positioning technology, sensor data acquisition, and real-time processing algorithms, it achieves the technical effect of accurate monitoring and early warning of UAV flight path deviations. Attached Figure Description

[0009] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0010] Figure 1 This is a schematic flowchart of a method for monitoring flight path deviation of a low-altitude unmanned aerial vehicle provided in an embodiment of this application. Detailed Implementation

[0011] This application provides a method for monitoring flight path deviation of low-altitude unmanned aerial vehicles (UAVs). It addresses the technical problem that existing technologies are insufficient in terms of real-time monitoring and accuracy, making it difficult to meet safety requirements in complex flight environments. By integrating advanced positioning technology, sensor data acquisition, and real-time processing algorithms, it achieves the technical effect of accurate monitoring and early warning of UAV flight path deviation.

[0012] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0013] It should be noted that 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 steps or units that are explicitly listed, but may include other steps or modules that are not explicitly listed or that are inherent to such process, method, product, or device. Example

[0014] like Figure 1 As shown, this application provides a method for monitoring flight path deviation of a low-altitude unmanned aerial vehicle, the method comprising:

[0015] Step S100: Define the target area and the unmanned aerial vehicle (UAV), wherein the target area includes the flight environment and weather environment, and the UAV includes basic UAV information;

[0016] In this embodiment, UAV aerial surveying technology is used to perform high-precision terrain mapping of the target area using high-definition cameras and sensors mounted on the UAV, identifying terrain features, buildings, utility poles, and other potential obstacles. Radar, ADS-B, and other technologies are employed to monitor other flight activities within the target area in real time, ensuring that the UAV avoids collisions with other aircraft during flight.

[0017] By connecting with meteorological stations, real-time key meteorological information such as wind speed, wind direction, temperature, and humidity of the target area can be obtained.

[0018] By using detailed aircraft technical specifications, we can obtain basic information about unmanned aerial vehicles (UAVs), such as model, size, weight, flight speed, flight altitude, and endurance.

[0019] Step S200: Deploy multi-source sensors on the unmanned aerial vehicle, acquire multi-source positioning data based on the multi-source sensors, and perform data fusion based on the multi-source positioning data to obtain high-precision positioning data of the unmanned aerial vehicle;

[0020] In this embodiment, various types of sensors are installed on the unmanned aerial vehicle (UAV). High-resolution images of the ground are acquired via optical cameras and high-definition cameras to aid visual positioning and environmental perception. An IMU provides angular velocity, acceleration, and attitude information for the UAV. A GPS receiver provides global positioning information. LiDAR is used to determine distance by measuring the reflection time of the laser beam, generating 3D point cloud data for environmental modeling and obstacle avoidance. Ultrasonic and infrared sensors are used to detect and avoid nearby obstacles.

[0021] These sensors continuously collect data during flight, such as GPS providing latitude and longitude, IMU providing flight attitude, and lidar scanning the surrounding environment. To improve positioning accuracy and robustness, data fusion technology is employed to process multi-source sensor data. Kalman filtering or extended Kalman filtering combines data from multiple sensors, and through prediction and update steps, the aircraft's state is optimally estimated. Since the sampling frequencies and data outputs of different sensors may have delays or asynchronies, time synchronization and spatial alignment processing are required. For potential sensor anomalies, such as GPS signal loss or lidar noise points, data cleaning and removal are performed.

[0022] After data fusion processing, high-precision positioning data of the unmanned aerial vehicle is obtained.

[0023] Step S300: Construct a flight path deviation identification channel, collect historical data of the target area and the unmanned aerial vehicle, and conduct supervised training of the flight path deviation identification channel using the historical data;

[0024] In this embodiment of the application, when constructing the flight path deviation identification channel, a flight path deviation identification model based on deep learning or other machine learning algorithms is adopted. These models are capable of learning and identifying flight path deviation patterns.

[0025] The database collects geographic information, meteorological data, obstacle locations, etc. of the target area, and also collects historical flight data of unmanned aerial vehicles, including flight trajectory, speed, altitude, attitude, etc.

[0026] The collected historical data was then cleaned, labeled, and formatted for training. The labeled historical data was used to train the flight path deviation detection model. The model parameters were continuously adjusted to ensure accurate identification of flight path deviations. A validation dataset was used to evaluate the model's performance, and necessary adjustments and optimizations were made based on the evaluation results.

[0027] After training is completed, a route deviation recognition channel is obtained.

[0028] Step S400: Input the target area and the high-precision positioning data into the flight path deviation identification channel to perform deviation identification and obtain the deviation identification result;

[0029] In this embodiment, detailed geographic information of the target area is integrated, including terrain, obstacles, airspace restrictions, etc., as well as real-time high-precision positioning data of the unmanned aerial vehicle. The integrated data is then formatted into a format acceptable to the flight path deviation identification channel through data cleaning, transformation, and standardization steps.

[0030] The integrated data is then input into the flight path deviation identification channel. The flight path deviation identification model trained in the channel is used to analyze the input target area information and high-precision positioning data to obtain the deviation identification results.

[0031] Step S500: Build a cloud platform, which includes the target area, the unmanned aerial vehicle, the flight path deviation identification channel and the control center. Input the deviation identification result into the cloud platform to process and display the deviation identification result.

[0032] In this embodiment, a reliable cloud service provider, such as Amazon AWS, Alibaba Cloud, or Huawei Cloud, is selected based on business needs and technical requirements. Appropriate server resources, including CPU, memory, storage, and network bandwidth, are configured on the cloud service to meet the demands of real-time data processing and storage.

[0033] Geographic information, meteorological data, and real-time flight data of unmanned aerial vehicles (UAVs) for the target area are connected to the cloud platform via API or SDK. The trained flight path deviation detection model is deployed to the cloud platform to ensure it can process the input data in real time. The cloud platform receives the deviation detection results output from the flight path deviation detection channel via the API interface.

[0034] The received results are further processed and analyzed, such as calculating the specific value of the deviation and judging the severity of the deviation.

[0035] Finally, an intuitive user interface was designed to display the flight deviation identification results and other relevant information, such as a graphical representation of the deviation, the specific numerical value of the deviation, and suggested corrective measures. Interactive functions were also provided, allowing users to view real-time flight data and receive deviation identification results.

[0036] Furthermore, step S200 in the method provided in the application embodiment further includes:

[0037] Multi-source sensors are deployed on the unmanned aerial vehicle, and the communication status of the multi-source sensors is adjusted to ensure smooth communication between the multi-source sensors and the control center.

[0038] The unmanned aerial vehicle is started and the multi-source sensors are activated to collect basic positioning data and obtain multi-source positioning data.

[0039] Data preprocessing is performed on the multi-source positioning data to obtain processed positioning data;

[0040] The accuracy of the multi-source sensors is determined, weight parameters are allocated based on the accuracy, and the processed positioning data is weighted and calculated to determine high-precision positioning data.

[0041] In this embodiment, various sensors, such as optical cameras, GPS receivers, and IMUs, are installed on the unmanned aerial vehicle (UAV) according to flight mission requirements to achieve multi-dimensional environmental monitoring and positioning. The communication protocol between the sensors and the control center is configured to ensure the stability and real-time performance of data transmission. The transmission frequency and bandwidth are adjusted to adapt to the characteristics of different sensor data, reducing transmission delays and the possibility of data loss.

[0042] Safely start the unmanned aerial vehicle according to the operating procedures, activate all multi-source sensors, and begin data acquisition. Each sensor collects data according to the set frequency and parameters, such as GPS location information and IMU attitude data. Timestamps and other methods are used to ensure that the data collected from different sensors are synchronized in time, thus obtaining multi-source positioning data.

[0043] The multi-source positioning data is then cleaned to remove noise and outliers, improving data quality. Filtering algorithms, such as Kalman filtering, are applied to further smooth the data and reduce errors. The positioning accuracy and stability of each sensor are evaluated by referring to the technical specifications.

[0044] Finally, based on the accuracy evaluation results of the sensors, different weights are assigned to each sensor. A weighted calculation method is then used to fuse the data from different sensors to generate high-precision positioning data.

[0045] Furthermore, step S300 in the method provided in the application embodiment further includes:

[0046] Collect historical data of the target area and the unmanned aerial vehicle. The historical data includes normal flight data and deviation flight data, wherein the deviation flight data includes environmental deviation and aircraft deviation.

[0047] A flight path deviation identification channel is constructed, which includes an environmental identification channel and an aircraft identification channel;

[0048] The target area is input into the environment recognition channel to determine whether the environment of the target area is suitable for the start of the unmanned aerial vehicle (UAV). If not, the UAV is directly determined to have deviated from its flight path. If so, the UAV is started, and the flight path deviation is determined based on the UAV.

[0049] In this embodiment of the application, historical data of the target area and the unmanned aerial vehicle are collected from a historical database. The collected data is divided into normal flight data and deviation flight data. The deviation data is further subdivided into deviations caused by environmental factors and deviations caused by problems of the aircraft itself.

[0050] A flight path deviation detection channel is constructed using machine learning algorithms and deep learning models, such as Convolutional Neural Networks (CNNs) and Long Short-Term Memory Networks (LSTMs). This channel comprises an environmental detection channel and an aircraft detection channel. The environmental detection channel uses weather prediction models, terrain analysis models, and obstacle detection algorithms to assess the safety of the flight environment. The aircraft detection channel is used to predict the performance and status of the aircraft.

[0051] Before flight, environmental data of the target area, such as weather forecasts and topographic elevation maps, are input into the environmental recognition channel for analysis. Streaming data processing technologies, such as Apache Flink and Storm, are used to process real-time sensor data to support rapid decision-making.

[0052] If the environmental recognition channel determines that the current environment is not suitable for flight, such as strong wind shear or thunderstorms, it will directly determine that the UAV has deviated from its flight path.

[0053] If the environment is suitable for flight, the unmanned aerial vehicle is activated, and its flight status is monitored in real time through the aircraft identification channel to determine whether it deviates from the flight path.

[0054] Furthermore, the method also includes constructing a route deviation identification channel:

[0055] A flight path deviation identification model is constructed, wherein the input data of the flight path deviation identification model is the target area and the high-precision positioning data, and the output data is the deviation identification result;

[0056] Acquire historical data of the target area and the unmanned aerial vehicle, as well as the corresponding flight path deviation results;

[0057] A sample set is constructed based on the target area, the historical data of the unmanned aerial vehicle, and the flight path deviation results;

[0058] The initial route deviation identification model is trained and validated using the sample set to obtain the route deviation identification model, which is embedded in the route deviation identification channel.

[0059] In this embodiment, the target area is first determined. The target area is a specific geographical region, such as a specific airspace or sea area. Historical data of the unmanned aerial vehicle (UAV) is collected, including its flight trajectory, timestamp, speed, and heading. Simultaneously, the flight path deviation results corresponding to this historical data are obtained.

[0060] The collected historical data and flight path deviation results are integrated to construct a sample set. Each sample includes information about the target area, real-time data from the UAV, and the corresponding flight path deviation label.

[0061] For complex flight deviation detection tasks, a convolutional neural network model is chosen as the initial flight deviation detection model. The initial flight deviation detection model is trained using a constructed sample set. The sample set is divided into a training set and a validation set using methods such as random partitioning or stratified sampling. The training set is used to train the model, and the validation set is used to evaluate the model's performance and prevent overfitting.

[0062] The model is trained using a training set, and its parameters are tuned using backpropagation and gradient descent optimizers. Techniques such as batch training and learning rate decay are employed to improve training efficiency and model performance. The model's performance is evaluated using a validation set, calculating metrics such as accuracy, recall, and F1 score. Based on the validation results, model parameters are adjusted or different model architectures are explored.

[0063] Finally, the trained flight deviation recognition model is integrated into the flight deviation recognition channel.

[0064] Furthermore, the method also includes:

[0065] Obtain a preset data partitioning ratio, and divide the sample set into a training set and a validation set according to the preset data partitioning ratio;

[0066] The initial flight path deviation identification model is trained under supervision using the training set. When the model output tends to converge, the output of the initial flight path deviation identification model is verified using the validation set.

[0067] Obtain a preset model validation accuracy index. When the output accuracy of the initial route deviation identification model meets the preset model validation accuracy index, the route deviation identification model is obtained.

[0068] In this embodiment, the preset data partitioning ratio is determined based on experience, the size of the dataset, and the complexity of the problem. This application uses a partitioning ratio of 70:30, employing random sampling techniques to divide the sample set into a training set and a validation set.

[0069] Similar to the process described above, supervised training of the initial model is performed using a training set. The model parameters are adjusted to minimize prediction error by inputting training samples and corresponding flight path deviation labels. Advanced techniques such as backpropagation and gradient descent optimizers are used to efficiently train the model, and metrics such as loss function and accuracy are monitored during the training process.

[0070] Once the model's performance on the training set has stabilized, it is validated using a validation set. The validation set samples are input into the model to obtain its predictions, which are then compared to the actual flight path deviation labels to determine the model's accuracy. The model's accuracy is then compared to a preset model validation accuracy metric to determine if the initial flight path deviation identification model's output accuracy meets the preset metric. This preset model validation accuracy metric is set by technical experts based on actual needs and industry standards.

[0071] If the model's accuracy meets or exceeds the preset target, the model is considered successfully trained, and a flight path deviation recognition model is obtained.

[0072] Step S500 in the method provided in the further application embodiments further includes:

[0073] Plan the overall architecture of the cloud platform, build the cloud platform, which includes the target area, the unmanned aerial vehicle, the flight path deviation identification channel, and the control center;

[0074] Develop a communication interface between the unmanned aerial vehicle and the cloud platform to ensure that the deviation identification results can be transmitted to the cloud platform in real time;

[0075] A feedback adjustment system is embedded in the control center to adjust the flight path of the unmanned aerial vehicle in real time and perform feedback adjustment.

[0076] In this embodiment, the overall architecture of the cloud platform is divided into a target area management module, an unmanned aerial vehicle management module, a flight path deviation identification channel, a control center, a data storage and analysis module, and a communication interface module.

[0077] The target area management module is responsible for storing and managing information about target areas, including geographical location, boundaries, and restrictions. It provides interfaces for other modules to query and update target area information.

[0078] The UAV management module tracks and manages the status, location, and flight plans of all UAVs. The flight path deviation detection channel integrates a flight path deviation detection model and receives high-precision positioning data from the UAVs. The control center receives the flight path deviation detection results and makes decisions based on preset rules and algorithms. The data storage and analysis module stores historical data from all UAVs, flight path deviation records, and control center command records. The communication interface module is responsible for data transmission and communication between the cloud platform and the UAVs.

[0079] When building a cloud platform, choose a suitable cloud service provider, such as Amazon Web Services, Google Cloud Platform, or Microsoft Azure. Set up the infrastructure, configuring virtual machines, databases, storage, and networks. Develop the planned modules into microservices or monolithic applications and deploy them to the cloud.

[0080] When developing the communication interface, define the communication protocol and determine the communication standards and data formats between the UAV and the cloud platform. Use MQTT, WebSocket, or other real-time communication protocols to achieve bidirectional data transmission. For security reasons, encrypt and sign the data to prevent data leakage and tampering. The aforementioned process ensures that deviation identification results can be transmitted to the cloud platform in real time.

[0081] Finally, the control center receives the output results from the flight path deviation identification channel in real time. Based on the deviation identification results and other relevant information, such as the UAV's status and environmental factors, a decision algorithm determines how to adjust the flight path. The decision results are converted into specific control commands and sent to the UAV via the communication interface to achieve feedback adjustment.

[0082] Furthermore, the method also includes:

[0083] High-risk areas are identified in the control center based on the target area;

[0084] For the high-risk areas, the cloud platform increases its sensitivity to flight path deviations and allocates corresponding computing power to process the deviation identification results.

[0085] In this embodiment, big data analytics is used to deeply mine historical flight route data and identify areas with frequent route deviations or accidents. By analyzing multi-source information such as flight logs and meteorological data, potential risk patterns and trends are discovered. Through the above steps, high-risk areas within the target region are determined.

[0086] In high-risk areas, the cloud platform increases its sensitivity to flight path deviations and lowers the deviation threshold. This means that a deviation alarm is triggered only when the drone's actual flight path deviates from the planned path by a smaller margin. For example, in general areas, an alarm is triggered only if the drone's flight path deviates by 50 meters from the planned path, while in high-risk areas, this threshold is lowered to 20 meters or less.

[0087] Finally, the cloud platform utilizes the dynamic resource allocation function of cloud computing to dynamically allocate corresponding computing power to process the deviation identification results based on the need for flight path deviation identification in high-risk areas.

[0088] Furthermore, the method also includes:

[0089] Select a map based on the target area and receive the location data of the unmanned aerial vehicle in real time, updating the map accordingly;

[0090] When the route deviation identification channel detects a deviation, it acquires the deviation identification result and uses visual elements to mark the deviation point on the map.

[0091] The actual flight path of the unmanned aerial vehicle is plotted and compared with the preset flight path, and deviations from the path are displayed.

[0092] In this embodiment, a map of appropriate scale and detail, such as satellite imagery, topographic maps, or other specialized maps, is selected based on the target area of ​​the flight mission. The drone's longitude, latitude, and altitude data are acquired in real time via its GPS or other positioning system. This data is transmitted to the ground control station in real time via a wireless communication link. The received drone position data is then overlaid onto the selected map in real time.

[0093] Before flight, a precise preset flight path is set for the drone using professional flight path planning software. When a deviation is detected by the deviation detection channel, a deviation detection result is generated, determining the precise location of the deviation point. On the map, using AR technology or dynamic layers, eye-catching visual elements such as flashing icons and warning colors are used to mark the deviation point in real time.

[0094] Finally, big data analytics is used to record and process the drone's flight data in real time, generate its actual flight path, and display deviations from the path.

[0095] In summary, the embodiments of this application have at least the following technical effects:

[0096] This application clearly defines the target area and the unmanned aerial vehicle (UAV). The target area includes the flight environment and weather environment, and the UAV includes basic UAV information. Multi-source sensors are deployed on the UAV, and multi-source positioning data is acquired based on these sensors. Data fusion is performed on the multi-source positioning data to obtain high-precision positioning data for the UAV. A flight path deviation identification channel is constructed, and historical data of the target area and the UAV are collected. The flight path deviation identification channel is trained under supervised supervision using the historical data. The target area and the high-precision positioning data are input into the flight path deviation identification channel for deviation identification, and the deviation identification result is obtained. A cloud platform is built, which includes the target area, the UAV, the flight path deviation identification channel, and a control center. The deviation identification result is input into the cloud platform for processing and display. This invention addresses the shortcomings of existing technologies in real-time monitoring and accuracy, making it difficult to meet the safety requirements of complex flight environments. By integrating advanced positioning technology, sensor data acquisition, and real-time processing algorithms, it achieves the technical effect of accurate monitoring and early warning of UAV flight path deviations.

[0097] It should be noted that the order of the embodiments described above is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, the above description focuses on specific embodiments of this specification. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some implementations, multitasking and parallel processing are possible or may be advantageous.

[0098] The above description is only a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

[0099] This specification and accompanying drawings are merely illustrative examples of this application and are intended to cover any and all modifications, variations, combinations, or equivalents within the scope of this application. Clearly, those skilled in the art can make various alterations and modifications to this application without departing from its scope. Therefore, if such modifications and modifications fall within the scope of this application and its equivalents, this application intends to include such modifications and modifications.

Claims

1. A method for monitoring flight path deviation of a low-altitude unmanned aerial vehicle, characterized in that, The method includes: The target area and the unmanned aerial vehicle (UAV) are clearly defined, wherein the target area includes the flight environment and weather environment, and the UAV includes basic information about the UAV; The unmanned aerial vehicle is equipped with multi-source sensors, multi-source positioning data is acquired based on the multi-source sensors, and data fusion is performed based on the multi-source positioning data to obtain high-precision positioning data of the unmanned aerial vehicle. A flight path deviation identification channel is constructed, and historical data of the target area and the unmanned aerial vehicle are collected. The flight path deviation identification channel is trained under supervision using the historical data. The method for constructing a flight path deviation identification channel further includes: A flight path deviation identification model is constructed, wherein the input data of the flight path deviation identification model is the target area and the high-precision positioning data, and the output data is the deviation identification result; Acquire historical data of the target area and the unmanned aerial vehicle, as well as the corresponding flight path deviation results; A sample set is constructed based on the target area, the historical data of the unmanned aerial vehicle, and the flight path deviation results; The initial flight deviation identification model is trained and validated using the sample set to obtain the flight deviation identification model, which is embedded in the flight deviation identification channel. The target area and the high-precision positioning data are input into the flight path deviation identification channel to perform deviation identification and obtain the deviation identification result. A cloud platform is established, which includes the target area, the unmanned aerial vehicle, the flight path deviation identification channel, and a control center. The deviation identification results are input into the cloud platform for processing and display.

2. The method as described in claim 1, characterized in that, The method further includes deploying multi-source sensors on the unmanned aerial vehicle (UAV), acquiring multi-source positioning data based on the multi-source sensors, and fusing the multi-source positioning data to obtain high-precision positioning data for the UAV. Multi-source sensors are deployed on the unmanned aerial vehicle, and the communication status of the multi-source sensors is adjusted to ensure smooth communication between the multi-source sensors and the control center. The unmanned aerial vehicle is started and the multi-source sensors are activated to collect basic positioning data and obtain multi-source positioning data. Data preprocessing is performed on the multi-source positioning data to obtain processed positioning data; The accuracy of the multi-source sensors is determined, weight parameters are allocated based on the accuracy, and the processed positioning data is weighted and calculated to determine high-precision positioning data.

3. The method as described in claim 1, characterized in that, Constructing a flight path deviation identification channel, collecting historical data of the target area and the unmanned aerial vehicle, and performing supervised training on the historical data for the flight path deviation identification channel, the method further includes: Collect historical data of the target area and the unmanned aerial vehicle. The historical data includes normal flight data and deviation flight data, wherein the deviation flight data includes environmental deviation and aircraft deviation. A flight path deviation identification channel is constructed, which includes an environmental identification channel and an aircraft identification channel; The target area is input into the environment recognition channel to determine whether the environment of the target area is suitable for the start of the unmanned aerial vehicle (UAV). If not, the UAV is directly determined to have deviated from its flight path. If so, the UAV is started, and the flight path deviation is determined based on the UAV.

4. The method as described in claim 3, characterized in that, The method further includes: Obtain a preset data partitioning ratio, and divide the sample set into a training set and a validation set according to the preset data partitioning ratio; The initial flight path deviation identification model is trained under supervision using the training set. When the model output tends to converge, the output of the initial flight path deviation identification model is verified using the validation set. Obtain a preset model validation accuracy index. When the output accuracy of the initial route deviation identification model meets the preset model validation accuracy index, the route deviation identification model is obtained.

5. The method as described in claim 1, characterized in that, A cloud platform is established, comprising the target area, the unmanned aerial vehicle, the flight path deviation identification channel, and a control center. The deviation identification results are input into the cloud platform for processing and display. The method further includes: Plan the overall architecture of the cloud platform, build the cloud platform, which includes the target area, the unmanned aerial vehicle, the flight path deviation identification channel, and the control center; Develop a communication interface between the unmanned aerial vehicle and the cloud platform to ensure that the deviation identification results can be transmitted to the cloud platform in real time; A feedback adjustment system is embedded in the control center to adjust the flight path of the unmanned aerial vehicle in real time and perform feedback adjustment.

6. The method as described in claim 5, characterized in that, The method further includes: High-risk areas are identified in the control center based on the target area; For the high-risk areas, the cloud platform increases its sensitivity to flight path deviations and allocates corresponding computing power to process the deviation identification results.

7. The method as described in claim 6, characterized in that, The method further includes: Select a map based on the target area and receive the location data of the unmanned aerial vehicle in real time, updating the map accordingly; When the route deviation identification channel detects a deviation, it acquires the deviation identification result and uses visual elements to mark the deviation point on the map. The actual flight path of the unmanned aerial vehicle is plotted and compared with the preset flight path, and deviations from the path are displayed.

Images