Unmanned aerial vehicle attitude control abnormity monitoring method and system based on data driving

Abstract

The invention discloses an unmanned aerial vehicle attitude control abnormity monitoring method and system based on data driving, and relates to the technical field of unmanned aerial vehicle attitude control. The method comprises the steps of establishing an unmanned aerial vehicle data control center, and planning a flight path of an unmanned aerial vehicle; an attitude data acquisition chip is arranged in the unmanned aerial vehicle, and millisecond data acquisition is performed by using the attitude data acquisition chip; and the control server carries out fitting comparison on the actual flight attitude route of each time node and the path planning route of the corresponding time node, carries out comparative analysis on yaw data and a driving database, and finds a deviation correction route instruction for correcting the route. The system comprises a data control center, an attitude data acquisition chip in the unmanned aerial vehicle, a gyroscope, a magnetometer, an accelerometer, a GPS navigator and a radar which are arranged in the unmanned aerial vehicle, and a signal base station established on the ground; and a driving database is established in the control server. According to the method, the problems that the generalization ability of an existing model is reduced, the prediction result is unstable or the sensitivity to abnormal data is increased are solved.

Description

Technical Field

[0001] This invention belongs to the field of unmanned aerial vehicle (UAV) attitude control technology, and in particular relates to a data-driven method and system for monitoring UAV attitude control anomalies. Background Technology

[0002] In recent years, the rapid development of unmanned aerial vehicle (UAV) technology has led to its widespread application in both military and civilian fields, playing an increasingly important strategic role. However, with the increasing complexity of UAV system structures and the rising difficulty of mission execution, UAV malfunctions have become frequent, leading to serious safety accidents and causing huge economic losses. For UAVs, flight attitude control is a core technology to ensure stable and precise control. Flight data, as an indirect reflection of the UAV's operational status, is an important basis for evaluating flight performance and monitoring system health. Therefore, developing data-driven anomaly detection for UAV attitude control, providing accurate attitude data in real time, and enabling quadcopter UAVs to maintain stable flight attitude to avoid unnecessary casualties and losses, is of great practical significance for improving the safety and reliability of UAV flight. Currently, common anomaly detection methods include those based on prior knowledge, models, and data-driven methods.

[0003] While deep learning models possess powerful feature extraction and pattern recognition capabilities when processing noisy real-world data, they may exhibit insufficient model matching, especially when faced with a large amount of unpredictable noise or outliers. This insufficient matching typically manifests as decreased generalization ability, unstable prediction results, or increased sensitivity to outliers.

[0004] However, existing drone flight methods lack the ability to fit and compare actual flight paths with planned paths to identify route deviations. Furthermore, there's no precedent for designing extensive corrective flight path commands to adjust flight paths in real time, enabling timely adjustments to drone attitude control anomalies and subsequent operational control. Moreover, it fails to allow a monitoring center to simultaneously control multiple drones independently, achieving the effect of independent flight attitude adjustment for multiple drones. Summary of the Invention

[0005] The purpose of this invention is to provide a data-driven method and system for monitoring anomalies in UAV attitude control. By establishing a path planning route and configuring a driving database, the actual flight attitude route is fitted using real-time flight data of the UAV and compared with the path planning route to generate yaw data. The corresponding correction route command is then found in the driving database based on the yaw data to achieve attitude anomaly adjustment. This invention solves the problems of decreased generalization ability, unstable prediction results, or increased sensitivity to abnormal data in existing models.

[0006] To solve the above-mentioned technical problems, the present invention is achieved through the following technical solution:

[0007] This invention relates to a data-driven method for monitoring anomalies in UAV attitude control, the specific method of which is as follows:

[0008] A drone data control center is established, which includes an operating console, a control server, and a display device. A driver database is established in the control server, and the flight path of each drone is planned to form a path planning route in the control server.

[0009] A gyroscope is installed in the middle of each of the four sides of the bottom of the drone. A magnetometer is installed near the front and rear of the drone. An accelerometer is installed in the center of the drone. The drone is equipped with GPS navigation and an attitude data acquisition chip. The attitude data acquisition chip collects data in milliseconds and processes the collected data. The attitude data acquisition chip is equipped with a wireless module, which uploads the actual flight attitude and route data to the control server in real time in seconds. After the wireless module uploads the data to the control server, the control server processes the data in a timely manner and constructs the actual flight attitude and route.

[0010] The control server fits and compares the actual flight attitude route at each time point with the path planning route at the corresponding time point to find the yaw data of the actual flight attitude route. It compares and analyzes the yaw data with the drive database to find the correction route command for correcting the route of the UAV. The control server sends the correction route command to the UAV in a timely manner, and the UAV performs correction flight in a timely manner. After the correction flight is completed, the UAV continues to fly according to the path planning route.

[0011] The control server performs a fitting comparison of the actual flight attitude and route uploaded by the UAV each time, and issues a route correction command to achieve second-level monitoring of abnormal UAV attitude.

[0012] The present invention is further configured such that variance analysis is performed on the data measured by the four gyroscopes, and the data of the gyroscope with a large deviation is removed, and the data of the other three gyroscopes are taken as the effective data of the gyroscopes.

[0013] The attitude data acquisition chip performs millisecond data acquisition at intervals of 100-200ms, achieving 5-10 acquisitions per second. The time interval between each data upload for constructing the actual flight attitude route is 3-5s. The average roll axis, pitch axis, and yaw axis deflection angles of the three effective gyroscopes at time i are Ri, Pi, and Yi, respectively. The northward direction Ni of the UAV at time i is acquired through the magnetometer, and the velocity Vi of the UAV at time i is acquired through the accelerometer. The coordinates of the actual flight attitude route are then established.

[0014] Each time, the coordinates of the endpoint of the actual flight attitude route coordinate line are compared with the coordinates marked on the path planning route at that time point to determine the spatial coordinates (X1, Y1, Z1) of the UAV deviating from the actual position at that time point. Then, after receiving the correction route instruction, the flight coordinates according to the path planning route in the next time period are (X2, Y2, Z2), and the actual flight route is (X1, Y1, Z1) → (X2, Y2, Z2).

[0015] The three-axis deflection angles at the final moment of the actual flight attitude path are: (Ri+Ri-1+···+R1, Pi+Pi-1+···+P1, Yi+Yi-1+···+Y1). The three-axis deflection angles of the path planning route at this moment are: (Rz, Pz, Yz). The three-axis deflection angles that need to be corrected for the UAV attitude are: (Rz, Pz, Yz) - (Ri+Ri-1···+R1, Pi+Pi-1+···+P1, Yi+Yi-1+···+Y1).

[0016] The invention is further configured such that the control server will search the drive database for the optimal route instruction for the flight route (X1, Y1, Z1) → (X2, Y2, Z2), allocate the optimal route instruction and send it to the UAV, and the UAV will fly according to the new instruction when it receives the optimal route instruction.

[0017] The invention is further configured such that when the attitude data acquisition chip is used to acquire millisecond data, the average deflection angles Ri, Pi and Yi of the three effective gyroscopes at time i are compared with the actual deflection angle of the path planning route at that time. When the values ​​of Ri, Pi and Yi at a certain time deviate from the set value by more than 5%, the deflection angle should be corrected in time.

[0018] The invention is further configured to establish three signal base stations that are not in a straight line on the ground in the flight area of ​​the UAV, and to install a radar to detect the flight altitude in each UAV. The path planning route includes a distance route based on the changes in the positions of the three signal base stations and a route based on the changes in altitude during flight.

[0019] The position data fitted by connecting the actual flight attitude and route coordinates is compared with the position data measured by the UAV based on three signal base stations and radar. If the data deviation between the two exceeds the set error value, the position coordinates fitted by connecting the actual flight attitude and route coordinates will be revised and replaced with the position coordinate data measured by the UAV based on three signal base stations and radar.

[0020] The present invention is further configured to compare the yaw data obtained by fitting and comparing the actual flight attitude route established based on the gyroscope, magnetometer and accelerometer with the path planning route with the yaw data of the distance route based on the position changes of the three signal base stations.

[0021] If the yaw data deviation between the two is less than or equal to the deviation setting value, the yaw correction command will be executed first.

[0022] If the yaw data deviation between the two is greater than the deviation setting value, then the yaw data based on the distance route changes of the three signal base stations will be activated. The three signal base stations will be used to correct the position of the UAV, so that the UAV can be corrected to the planned route in time.

[0023] A system for implementing a data-driven method for monitoring anomalies in UAV attitude control, the system comprising:

[0024] The data control center is equipped with an operating console, a control server, and a display device. The operating console is used to input the path planning route, and the display device is used to display the flight path of the UAV in real time.

[0025] An attitude data acquisition chip is installed in each drone. The attitude data acquisition chip is equipped with a wireless module. A gyroscope is installed in the middle of the four sides of the bottom of the drone. A magnetometer is installed near the front and rear of the drone. An accelerometer is installed in the center of the drone. The drone is equipped with GPS navigation. Three signal base stations are established on the ground in the drone's flight area that are not in a straight line. A radar for detecting flight altitude is installed in each drone.

[0026] The control server has a driver database and a data processing system installed. The data control center has a wireless transceiver module, which is connected to the control server via a data cable.

[0027] The invention is further configured such that the data acquired by the attitude data acquisition chip is used to form a route map of the actual flight attitude path through a data fitting program and compared with the path planning route; the data processing system is equipped with a search module; the search module searches the drive database through yaw data to form a yaw correction route command; the yaw correction route command is sent to the UAV through a wireless transceiver module.

[0028] The invention is further configured such that the attitude data acquisition chip is equipped with a timing acquisition unit. The timing acquisition unit will issue a data acquisition command every 100-200ms, and simultaneously issue data acquisition commands to the gyroscope, magnetometer, accelerometer and GPS navigation to acquire a set of UAV attitude data. The attitude data acquisition chip processes each set of UAV attitude data, and the series of UAV attitude data formed after continuous acquisition for 3-5 seconds is a UAV attitude route unit group.

[0029] The invention is further configured such that the attitude data acquisition chip is also equipped with a gyroscope data comparison module. The gyroscope data comparison module compares the gyroscope data acquired after each data acquisition command is issued with the deflection angle set value of the UAV at the same moment on the path planning route. If the deviation of the comparison result obtained by the gyroscope data comparison module is greater than the deflection angle error set value, the UAV is immediately started to adjust the deflection angle.

[0030] The present invention has the following beneficial effects:

[0031] 1. This invention plans a path planning route within the control server and transmits the path planning route to the UAV's storage module. The UAV flies according to the path planning route. Due to limitations in sensor performance, algorithm limitations, environmental interference, and system coupling effects, the UAV's attitude adjustment is affected by factors such as sensor performance, algorithm limitations, environmental interference, and system coupling effects, resulting in the accumulation of errors during flight. Consequently, the degree of yaw during flight increases, requiring timely correction. This application designs a standard path planning route and configures a driving database. By fitting the actual flight attitude route with the path planning route using real-time flight data of the UAV, yaw data is generated. The corresponding correction route command is found in the driving database based on the yaw data, enabling rapid adjustment of attitude anomalies. This eliminates the need for complex weighted fusion technology and high-precision remote control, resulting in low control and monitoring costs and high efficiency.

[0032] 2. This invention only requires the UAV to collect its own flight attitude data and send it to the control server for processing. By fitting and comparing the data, the yaw data is determined, and the corrective flight path command matching the yaw data is searched in the drive database. There is no need for complex weighted fusion technology and high-precision monitoring. In this way, multiple UAVs can be remotely controlled and fly independently to adjust their attitude, so as to realize the independent transport flight monitoring attitude of multiple UAVs.

[0033] Of course, any product implementing this invention does not necessarily need to achieve all of the advantages described above at the same time. Attached Figure Description

[0034] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the 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.

[0035] Figure 1 This is a schematic diagram of a data-driven method for monitoring anomalies in UAV attitude control.

[0036] Figure 2 System module diagram for implementing a data-driven method for monitoring anomalies in UAV attitude control.

[0037] Figure 3 A schematic diagram of the route for executing the correction instruction (B1→C0).

[0038] Figure 4 This is a diagram illustrating the process of first returning to the normal position and then continuing flight along the planned route.

[0039] Figure 5 This is a schematic diagram of a route for timely deflection correction when a deflection angle error exceeds 5%. Detailed Implementation

[0040] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0041] Please see Figure 1-5 This invention relates to a data-driven method for monitoring anomalies in UAV attitude control, the specific method of which is as follows:

[0042] A drone data control center is established, which includes an operating console, a control server, and a display device. A driver database is established in the control server, and the flight path of each drone is planned to form a path planning route in the control server.

[0043] During drone flight, following a pre-defined path (including the drone's flight drive mode), the drone flies along the planned route. However, accumulated yaw can occur, necessitating timely adjustments and corrections. The control server issues commands and analyzes the drone's attitude monitoring data. The display device shows the drone's flight path.

[0044] A gyroscope is installed in the middle of each of the four sides of the bottom of the drone. A magnetometer is installed near the front and rear of the drone. An accelerometer is installed in the center of the drone. The drone is equipped with GPS navigation and an attitude data acquisition chip. The attitude data acquisition chip collects data in milliseconds and processes the collected data. The attitude data acquisition chip is equipped with a wireless module, which uploads the actual flight attitude and route data to the control server in real time in seconds. After the wireless module uploads the data to the control server, the control server processes the data in a timely manner and constructs the actual flight attitude and route.

[0045] The gyroscope is a MEMS gyroscope, the magnetometer is a MEMS magnetometer, and the accelerometer is a MEMS accelerometer.

[0046] A MEMS gyroscope is installed on each of the four sides of the bottom of the drone, which can achieve four-sided rotation (the four gyroscopes actually rotate the same, but each gyroscope has a measurement deviation, and the four can be averaged to improve accuracy). The accelerometer can measure the drone's acceleration and speed (indirectly), and then combine it with GPS navigation and positioning to accurately calculate the speed.

[0047] After simple processing, the data collected by the attitude data acquisition chip is sent to the control server every 3-5 seconds to fit the route and form an actual flight attitude route. The route is then compared with the path planning route in the same time period to form a yaw route.

[0048] The control server fits and compares the actual flight attitude route at each time point with the path planning route at the corresponding time point to find the yaw data of the actual flight attitude route. It compares and analyzes the yaw data with the drive database to find the correction route command for correcting the route of the UAV. The control server sends the correction route command to the UAV in a timely manner, and the UAV performs correction flight in a timely manner. After the correction flight is completed, the UAV continues to fly according to the path planning route.

[0049] The driver database contains a large number of yaw correction commands, and the search for these commands is based on yaw data, primarily position coordinate offset data, to bring the system back onto the planned route.

[0050] The control server performs a fitting comparison on each data uploaded by the UAV to fit the actual flight attitude and route, and searches for correction route instructions to achieve second-level monitoring of abnormal UAV attitude.

[0051] Each time the drone uploads data, it is equivalent to monitoring and identifying abnormal attitudes of the drone, enabling an opportunity to identify and correct abnormal attitudes every 3-5 seconds, thus reducing the problem of drone error accumulation.

[0052] A variance analysis was performed on the data measured by the four gyroscopes. The data from the gyroscope with the largest data deviation was removed, and the data from the other three gyroscopes were taken as the valid data from the gyroscopes.

[0053] The attitude data acquisition chip performs millisecond data acquisition at intervals of 100-200ms, achieving 5-10 acquisitions per second. The time interval between each data upload for constructing the actual flight attitude route is 3-5s. The average roll axis, pitch axis, and yaw axis deflection angles of the three effective gyroscopes at time i are Ri, Pi, and Yi, respectively. The northward direction Ni of the UAV at time i is acquired through the magnetometer, and the velocity Vi of the UAV at time i is acquired through the accelerometer. The coordinates of the actual flight attitude route are then established.

[0054] The coordinates of the endpoint of the actual flight attitude path coordinate line are compared with the coordinates marked on the path planning route at that time point to determine the spatial coordinates (X1, Y1, Z1) of the UAV's deviation from its actual position at that time point. Figure 3 If the coordinates are (B1), then after receiving the flight path correction instruction, the flight coordinates according to the planned route in the next time period will be (X2, Y2, Z2). Figure 3 (C0 coordinates), at this point the actual flight path is (X1, Y1, Z1) → (X2, Y2, Z2); continuing to fly according to the B0 → C0 route instruction will only lead to greater errors. At this point, executing the correction route instruction (B1 → C0) can not only achieve heading flight, but also return to the planned route, thereby eliminating the yaw correction caused by the previous flight.

[0055] The three-axis deflection angles at the final moment of the actual flight attitude path are: (Ri+Ri-1+···+R1, Pi+Pi-1+···+P1, Yi+Yi-1+···+Y1). The three-axis deflection angles of the path planning route at this moment are: (Rz, Pz, Yz). The three-axis deflection angles that need to be corrected for the UAV attitude are: (Rz, Pz, Yz) - (Ri+Ri-1···+R1, Pi+Pi-1+···+P1, Yi+Yi-1+···+Y1).

[0056] Adjusting the three-axis yaw angle is for UAV attitude adjustment. After adjusting its attitude, executing the (X1, Y1, Z1) → (X2, Y2, Z2) route will be more conducive to flight. Because the attitude change during flight from A0 to B0 to C0 is minimal, if attitude adjustment is not performed, it will be more troublesome to adjust the attitude during flight from B1 to C0. Therefore, attitude adjustment is performed first.

[0057] The control server will search the drive database for the optimal route instruction for the flight path (X1, Y1, Z1) → (X2, Y2, Z2), allocate the optimal route instruction and send it to the UAV. When the UAV receives the optimal route instruction, it will fly according to the new instruction.

[0058] like Figure 3 The flight path from B1 to C0 is close to a straight line, which is the optimal flight path. Therefore, the flight path correction command (B1 to C0) should be executed.

[0059] When using the attitude data acquisition chip to acquire millisecond data, the average deflection angles Ri, Pi, and Yi of the three effective gyroscopes at time i are compared with the actual deflection angle of the path planning route at that time. When the values ​​of Ri, Pi, and Yi at a certain time deviate from the set value by more than 5%, the deflection angle should be corrected in time.

[0060] like Figure 5 During the A0→B1 route, millisecond-level data collection will be performed. When at position A1, if the deflection angle (any one of the deflection angles Ri, Pi, and Yi deviates significantly from the deflection angle set in the path planning) deviates by more than 5%, immediate deflection angle correction will be initiated. After correction, the route will be A1→B1, and the route yaw will be significantly reduced.

[0061] Three signal base stations not on the same straight line are established on the ground in the UAV flight area. A radar for detecting flight altitude is installed in each UAV. The path planning route includes a distance route based on the changes in the positions of the three signal base stations and a route based on changes in altitude during flight.

[0062] Three signal base stations not aligned with a straight line can determine the drone's position on its flight path, but altitude determination is not precise enough and requires radar detection of ground altitude. Signals emitted by the drone in all directions are triggered by the signal base stations, and the drone promptly transmits a return signal. The time difference between these signals represents the wave propagation distance, thus determining spatial distance coordinates (X and Y axes). Radar determines the Z-axis coordinate (altitude). The signal base stations and radar are used to correct yaws. Correction commands include not only spatial coordinate relocation but also drone attitude adjustments (during flight) to achieve optimal attitude and reach a specific point within the planned route.

[0063] The position data fitted by connecting the actual flight attitude and route coordinates is compared with the position data measured by the UAV based on three signal base stations and radar. If the data deviation between the two exceeds the set error value, the position coordinates fitted by connecting the actual flight attitude and route coordinates will be revised and replaced with the position coordinate data measured by the UAV based on three signal base stations and radar.

[0064] The yaw data obtained by fitting the actual flight attitude route established based on gyroscopes, magnetometers, and accelerometers with the planned route is compared with the yaw data of the distance route based on the position changes of three signal base stations.

[0065] Because when the values ​​of Ri, Pi, and Yi deviate from the set value by more than 5% at a certain moment, yaw angle correction must be performed in a timely manner. If yaw angle correction is not implemented or not executed (but the actual system records the yaw operation, then the fitted actual flight attitude trajectory will be inaccurate), the following will occur: Figure 4 The phenomenon is that point B1 deviates significantly from position B0 (the deflection angle was not corrected in time). In this case, the aircraft will fly directly to that point in time (B0) without issuing a course correction command. Figure 3 (B1→C0), then continue flying according to the planned route (B0→C0) to quickly return to the planned route.

[0066] If the yaw data deviation between the two is less than or equal to the deviation setting value, the yaw correction command will be executed first. Figure 3 The indicated correction route command is B1→C0.

[0067] If the yaw data deviation between the two is greater than the deviation setting value, then the yaw data based on the distance route changes of the three signal base stations will be activated. The three signal base stations will be used to correct the position of the UAV, so that the UAV can be corrected to the planned route in time. Figure 4 The route shown is A0→B1→B0→C0. If the route A0→B1→C0 is adopted, since the distance between B1 and B0 is relatively far (i.e., it deviates significantly from the flight path), it is necessary to first directly correct the course to the planned route in order to quickly and efficiently achieve the regression to the planned route.

[0068] A system for implementing a data-driven method for monitoring anomalies in UAV attitude control, the system comprising:

[0069] The data control center is equipped with an operating console, a control server, and a display device. The operating console is used to input the path planning route, and the display device is used to display the flight path of the UAV in real time.

[0070] An attitude data acquisition chip is installed in each drone. The attitude data acquisition chip is equipped with a wireless module. A gyroscope is installed in the middle of the four sides of the bottom of the drone. A magnetometer is installed near the front and rear of the drone. An accelerometer is installed in the center of the drone. The drone is equipped with GPS navigation. Three signal base stations are established on the ground in the drone's flight area that are not in a straight line. A radar for detecting flight altitude is installed in each drone.

[0071] The control server has a driver database and a data processing system installed. The data control center has a wireless transceiver module, which is connected to the control server via a data cable.

[0072] The attitude data acquisition chip is a control chip (commonly available on the market) that effectively controls the drone's attitude flight (but cannot intelligently eliminate errors introduced by sensors). It can also perform basic data processing (yaw angle correction and data organization before sending to the wireless transceiver module). The wireless transceiver module acts as the antenna for the monitoring center (functioning as a base station), enabling multiple drones to connect to its signal. This allows for the simultaneous control of multiple drones flying independently.

[0073] The driving database consists of a yaw database and a yaw correction command database. The yaw database and the yaw correction command database are corresponding; each yaw data corresponds to a yaw correction command. When the yaw data obtained from actual measurement is compared with the yaw data in the driving database, if the similarity is above 90%, it can be used as the required yaw data, and the corresponding yaw correction command is the yaw correction command that needs to be issued and transmitted.

[0074] The attitude data acquisition chip collects data and uses a data fitting program to form a route map of the actual flight attitude path, which is then compared with the planned route. The data processing system is equipped with a search module, which searches the drive database using yaw data to form a yaw correction route command. The yaw correction route command is then sent to the UAV via a wireless transceiver module.

[0075] The yaw data searched by the search module mainly includes the yaw position coordinates (B1), the actual position coordinates to be flown to (B0), and the next point coordinates (C0). The yaw correction command is to execute the route from B1 to C0, which requires combining the positional relationship between the coordinates of point B1 and point C0 for navigation.

[0076] The attitude data acquisition chip is equipped with a timing acquisition unit. The timing acquisition unit will issue a data acquisition command every 100-200ms, and simultaneously issue data acquisition commands to the gyroscope, magnetometer, accelerometer and GPS navigation to acquire a set of UAV attitude data. The attitude data acquisition chip processes each set of UAV attitude data, and the series of UAV attitude data formed after continuous acquisition for 3-5 seconds is a UAV attitude route unit group.

[0077] Synchronization data is collected every 0.1-0.2 seconds, totaling 15-35 data points within 3-5 seconds. This allows for data fitting to form the actual flight attitude and path. A UAV attitude and path unit group uses a series of data points collected within 3-5 seconds as a single data unit to fit the actual flight attitude and path. If the data collection time is too long before uploading, the yaw will be significant, making it difficult to accurately correct the yaw using correction commands. Fine-tuning the heading minimizes the yaw, making corrections easier and ensuring the UAV stays within the planned path.

[0078] The timing acquisition unit is used to send acquisition signals at fixed points to receive data from the gyroscope, magnetometer, accelerometer, and GPS navigation. When not sending signals, these signals are not received, but they are still constantly uploading. When receiving signals are turned on simultaneously, data from the same moment can be received synchronously, enabling the simultaneous acquisition of data from the gyroscope, magnetometer, accelerometer, and GPS navigation, which can then be fitted to a series of data later.

[0079] The attitude data acquisition chip also includes a gyroscope data comparison module. The gyroscope data comparison module compares the gyroscope data acquired after each data acquisition command is issued with the yaw angle set value of the UAV at the same moment on the path planning route. If the deviation of the comparison result obtained by the gyroscope data comparison module is greater than the yaw angle error set value, the UAV is immediately started to adjust the yaw angle.

[0080] The gyroscope data comparison module compares the required flight angle of the UAV at the same point in time with the path planning route after each data acquisition. If a large deviation suddenly occurs (more than 5%), the attitude of the UAV is adjusted, but the flight path is not adjusted. Only after a series of data is uploaded to the control server for fitting, yaw data search and issuance of yaw correction command will yaw correction to the path planning route begin. Yaw angle correction is to prevent excessive yaw.

[0081] In the description of this specification, references to terms such as "an embodiment," "example," "specific example," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0082] The preferred embodiments of the present invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to the specific implementations described. Clearly, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention. The invention is limited only by the claims and their full scope and equivalents.

Claims

1. A data-driven method for monitoring anomalies in UAV attitude control, characterized in that: The specific method is as follows: A drone data control center is established, which includes an operating console, a control server, and a display device. A driver database is established in the control server, and the flight path of each drone is planned to form a path planning route in the control server. A gyroscope is installed in the middle of each of the four sides of the bottom of the drone. A magnetometer is installed near the front and rear of the drone. An accelerometer is installed in the center of the drone. The drone is equipped with GPS navigation and an attitude data acquisition chip. The attitude data acquisition chip collects data in milliseconds and processes the collected data. The attitude data acquisition chip is equipped with a wireless module, which uploads the actual flight attitude and route data to the control server in real time in seconds. After the wireless module uploads the data to the control server, the control server processes the data in a timely manner and constructs the actual flight attitude and route. The control server fits and compares the actual flight attitude route at each time point with the path planning route at the corresponding time point to find the yaw data of the actual flight attitude route. It compares and analyzes the yaw data with the drive database to find the correction route command for correcting the route of the UAV. The control server sends the correction route command to the UAV in a timely manner, and the UAV performs correction flight in a timely manner. After the correction flight is completed, the UAV continues to fly according to the path planning route. The control server performs a fitting comparison on each data uploaded by the UAV to fit the actual flight attitude and route, and searches for correction route instructions to achieve second-level monitoring of abnormal UAV attitude.

2. The data-driven UAV attitude control anomaly monitoring method according to claim 1, characterized in that, A variance analysis was performed on the data measured by the four gyroscopes. The data from the gyroscope with the largest data deviation was removed, and the data from the other three gyroscopes were taken as the valid data from the gyroscopes. The attitude data acquisition chip performs millisecond data acquisition at intervals of 100-200ms, achieving 5-10 acquisitions per second. The time interval between each data upload for constructing the actual flight attitude path is 3-5 seconds. The average yaw angles of the roll, pitch, and yaw axes of the three effective gyroscopes at time i are R... i P i and Y i The magnetometer is used to collect the northward direction N of the UAV at time i. i The accelerometer is used to collect the velocity V of the UAV at time i. i Establish the coordinates of the actual flight attitude and route; Each time, the coordinates of the endpoint of the actual flight attitude route coordinate line are compared with the coordinates marked on the path planning route at that time point to determine the spatial coordinates (X1, Y1, Z1) of the UAV deviating from the actual position at that time point. Then, after receiving the correction route instruction, the flight coordinates according to the path planning route in the next time period are (X2, Y2, Z2), and the actual flight route is (X1, Y1, Z1) → (X2, Y2, Z2). The three-axis deflection angles at the final moment of the actual flight attitude path are: (R i +R i-1 +···+R1,P i +P i-1 +···+P1, Y i +Y i-1 +···+Y1), the three-axis deflection angle of the path planning route at this moment is: (R) z P z Y z At this point, the three-axis deflection angles that need to be corrected for the drone's attitude are: (R) z P z Y z )-(R i +R i-1 ...+R1, P i +P i-1 +···+P1, Y i +Y i-1 +···+Y1).

3. The data-driven UAV attitude control anomaly monitoring method according to claim 2, characterized in that, The control server will search the drive database for the optimal route instruction for the flight path (X1, Y1, Z1) → (X2, Y2, Z2), allocate the optimal route instruction and send it to the UAV. When the UAV receives the optimal route instruction, it will fly according to the new instruction.

4. The data-driven UAV attitude control anomaly monitoring method according to claim 2, characterized in that, When using an attitude data acquisition chip to acquire millisecond data, the average deflection angle R of the three effective gyroscopes at time i is... i P i and Y i Compare the actual deflection angle of the planned route at that moment with the actual deflection angle at that moment. When R at a certain moment... i P i and Y i If the value deviates from the set value by more than 5%, the deflection angle should be corrected in time.

5. The data-driven UAV attitude control anomaly monitoring method according to claim 2, characterized in that, Three signal base stations not on the same straight line are established on the ground in the UAV flight area. A radar for detecting flight altitude is installed in each UAV. The path planning route includes a distance route based on the changes in the positions of the three signal base stations and a route based on changes in altitude during flight. The position data fitted by connecting the actual flight attitude and route coordinates is compared with the position data measured by the UAV based on three signal base stations and radar. If the data deviation between the two exceeds the set error value, the position coordinates fitted by connecting the actual flight attitude and route coordinates will be revised and replaced with the position coordinate data measured by the UAV based on three signal base stations and radar.

6. The data-driven UAV attitude control anomaly monitoring method according to claim 5, characterized in that, The yaw data obtained by fitting the actual flight attitude route established based on gyroscopes, magnetometers, and accelerometers with the planned route is compared with the yaw data of the distance route based on the position changes of three signal base stations. If the yaw data deviation between the two is less than or equal to the deviation setting value, the yaw correction command will be executed first. If the yaw data deviation between the two is greater than the deviation setting value, then the yaw data based on the distance and route changes of the three signal base stations will be activated. The three signal base stations will be used to correct the position of the UAV, so that the UAV can be corrected to the planned route in a timely manner.

7. A system for implementing the data-driven unmanned aerial vehicle attitude control anomaly monitoring method according to any one of claims 1-6, characterized in that, include: The data control center is equipped with an operating console, a control server, and a display device. The operating console is used to input the path planning route, and the display device is used to display the flight path of the UAV in real time. An attitude data acquisition chip is installed in each drone. The attitude data acquisition chip is equipped with a wireless module. A gyroscope is installed in the middle of the four sides of the bottom of the drone. A magnetometer is installed near the front and rear of the drone. An accelerometer is installed in the center of the drone. The drone is equipped with GPS navigation. Three signal base stations are established on the ground in the drone's flight area that are not in a straight line. A radar for detecting flight altitude is installed in each drone. The control server has a driver database and a data processing system installed. The data control center has a wireless transceiver module, which is connected to the control server via a data cable.

8. The system according to claim 7, characterized in that, The data processing system includes a data fitting program. The data collected by the attitude data acquisition chip is used by the data fitting program to form a route map of the actual flight attitude path and compare it with the planned route. The data processing system also includes a search module. The search module searches the drive database using yaw data to form a yaw correction route command. The yaw correction route command is sent to the UAV through a wireless transceiver module.

9. The system according to claim 7, characterized in that, The attitude data acquisition chip is equipped with a timing acquisition unit. The timing acquisition unit will issue a data acquisition command every 100-200ms, and simultaneously issue data acquisition commands to the gyroscope, magnetometer, accelerometer and GPS navigation to acquire a set of UAV attitude data. The attitude data acquisition chip processes each set of UAV attitude data, and the series of UAV attitude data formed after continuous acquisition for 3-5 seconds is a UAV attitude route unit group.

10. The system according to claim 9, characterized in that, The attitude data acquisition chip also includes a gyroscope data comparison module. The gyroscope data comparison module compares the gyroscope data acquired after each data acquisition command is issued with the yaw angle set value of the UAV at the same moment on the path planning route. If the deviation of the comparison result obtained by the gyroscope data comparison module is greater than the yaw angle error set value, the UAV is immediately started to adjust the yaw angle.

Images