Unmanned aerial vehicle operation state fault early warning method and system based on data analysis

Abstract

The invention relates to the technical field of unmanned aerial vehicle operation state analysis, and discloses an unmanned aerial vehicle operation state fault early warning method and system based on data analysis, and the method comprises the steps: collecting dynamic parameter data, path control data, an actual track and real-time energy consumption when an unmanned aerial vehicle moves; according to the dynamic parameter data, carrying out direct evaluation on an operation state fault of the unmanned aerial vehicle to obtain a first evaluation result; according to the path control data, the actual trajectory and the real-time energy consumption, performing confrontation evaluation on the operation state fault of the unmanned aerial vehicle to obtain a second evaluation result; and summarizing and analyzing the first evaluation result and the second evaluation result, and generating a fault early warning strategy of the operation state of the unmanned aerial vehicle. The probability of misjudgment and missed judgment is remarkably reduced, so that the unmanned aerial vehicle has higher safety and reliability in complex tasks, and the operation efficiency and the task completion quality are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of unmanned aerial vehicle operation status analysis, and in particular to a method and system for early warning of unmanned aerial vehicle operation status failure based on data analysis. Background Art

[0002] With the rapid development of drone technology, drones have been widely used in logistics, inspection, agricultural monitoring, environmental protection and other fields. However, drones often face multiple challenges in the complex external environment (such as wind, humidity, temperature, etc.) and internal systems (such as batteries, power, flight control systems, etc.) during operation. The uncertainty of their operating status and potential failure risks have become important issues restricting their widespread application. Once the operating status of a drone is abnormal, it may lead to the failure of the flight mission or even cause a safety accident. Therefore, how to accurately monitor the operating status of drones and promptly detect and warn of potential failures has become a core research direction in the field of drone technology.

[0003] In the prior art, most fault assessment methods for the operation status of drones use single data source analysis, such as anomaly detection based on attitude data or energy consumption data. However, these methods are unable to fully integrate multidimensional data and ignore the interaction of multiple factors under complex operating conditions, resulting in low fault recognition rate and accuracy. In addition, traditional assessment methods are not able to distinguish between external environmental interference (such as strong winds) and drone system failures, which can easily lead to misjudgments or missed judgments. Therefore, there is an urgent need for a drone fault warning method based on multidimensional data fusion analysis, which combines direct evaluation and adversarial evaluation to achieve accurate fault identification and rapid warning. Summary of the invention

[0004] In view of the above-mentioned problems, the present invention is proposed.

[0005] Therefore, the technical problem solved by the present invention is that the existing UAV operation status fault warning method has the problem of being unable to clearly distinguish whether the fault is caused by the UAV itself or by wind force; and is unable to analyze a variety of control modes separately.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions: a method for early warning of UAV operation status failure based on data analysis, comprising: Collect dynamic parameter data, path control data, actual trajectory and real-time energy consumption of drones during movement; Directly evaluating the operating state failure of the UAV according to the dynamic parameter data to obtain a first evaluation result; According to the path control data, the actual trajectory, and the real-time energy consumption, a countermeasure evaluation of the operating state failure of the UAV is performed to obtain a second evaluation result; The first evaluation result and the second evaluation result are summarized and analyzed to generate a fault warning strategy for the operating status of the UAV.

[0007] As a preferred solution of the method for early warning of UAV operation status failure based on data analysis described in the present invention, wherein: the dynamic parameter data includes attitude data and power system data; The posture data includes posture angle data, angular velocity data, acceleration data, and vibration frequency; The power system data includes the speed of each motor, the temperature of each motor, and the implemented power of each motor; The path control data includes control data of the running trajectory of the UAV. If the UAV flies along a predetermined route, the path control data indicates the control of the predetermined route; if the UAV is under real-time control, the path control data indicates that the UAV receives a control instruction for real-time control; The actual trajectory includes real-time updating of the trajectory of the drone through real-time positioning.

[0008] As a preferred solution of the method for early warning of the operating state of a UAV based on data analysis described in the present invention, the direct evaluation includes comparing the real-time dynamic parameter data with the normal fluctuation range corresponding to the respective data types for the fluctuation range preset for each dynamic parameter data. If the real-time dynamic parameter data is not within the range, the operating state of the UAV is judged to be abnormal; otherwise, the operating state of the UAV is not judged. The judgment that the operating status of the drone is abnormal is taken as the first evaluation result.

[0009] As a preferred solution of the method for early warning of the operating state fault of the UAV based on data analysis described in the present invention, the confrontation evaluation includes, according to the path control data, respectively confronting the actual trajectory and the real-time energy consumption to realize the verification of the operating state fault of the UAV; and use the verification result as the second evaluation result; Among them, the confrontation process is divided into: trajectory confrontation analysis and energy consumption confrontation analysis; The goal of the trajectory confrontation analysis is to verify whether the flight trajectory of the drone meets the established target or real-time control instructions; The goal of the energy consumption confrontation analysis is to verify whether the current energy consumption of the drone matches its flight mission and environmental conditions.

[0010] As a preferred solution of the method for early warning of the operating status of a UAV based on data analysis described in the present invention, the trajectory confrontation analysis includes: Case 1: flying along a predetermined route, analyzing the real-time shortest distance between the UAV and the predetermined route through real-time positioning. ; When the drone deviates from the predetermined route, Record and Arrange in time series, when the drone returns to the predetermined path and stably moves the preset distance S, the recording ends; and outputs the records from the beginning of deviation from the predetermined path to the return to the predetermined path and the stable movement of the preset distance S. sequence ; The next time the drone deviates from the predetermined route, the sequence is recorded and generated again; in, Indicates the sequence of recording the shortest distance when the nth deviation from the predetermined path occurs; Use the trained generator 1 and discriminator 1 in the adversarial network for analysis and input: and The corresponding wind parameters in the time series ; Output: The probability of a drone failure causing a shift in the corresponding time series; in, express The maximum wind speed in the corresponding time series, express The minimum wind speed in the corresponding time series, express The average wind speed in the corresponding time series; Case 2: Use real-time control to fly, and calculate the degree of control behavior by analyzing each control signal; specifically: use the trained generator 2 and discriminator 2 in the adversarial network for analysis, input: and The wind force at the time of occurrence; Output: Predict the displacement vector of the drone under control ; exist In the sequence, randomly select the displacement vectors corresponding to L continuous control signals, add the L displacement vectors, and obtain L continuous control signals to predict the achievable displacement control results. At the same time, obtain L continuous control signals, the position information at the starting and ending points, and construct the actual displacement based on the position information ; The degree of realization of the control behavior is expressed as: In the second case, the maximum value of all sampling results is output , the average value of all sampling results ; in, represents the mth control signal; L represents the value of the fluctuation, which is any integer between the preset maximum value and the preset minimum value; Indicates the degree of realization of the control behavior corresponding to L continuous control signals; The energy consumption adversarial analysis includes using the trained generator 3 and discriminator 3 in the adversarial network to perform analysis, and inputting: the actual trajectory obtained by the drone in real time, the wind parameters from the start of movement to the current moment ; Output: Probability of excessive energy consumption anomaly; in, Indicates the maximum wind force from the start of movement to the current moment, Indicates the minimum value of the wind force from the beginning of the movement to the current moment, Indicates the average wind speed from the start of the movement to the current moment.

[0011] As a preferred solution of the method for early warning of the operating state fault of a UAV based on data analysis of the present invention, the summary analysis includes obtaining the first evaluation result and the second evaluation result respectively under the two types of flight states of the predetermined route and the real-time control; Comparing the obtained evaluation results with the corresponding preset thresholds respectively to obtain a determination result of each output element in the first evaluation result and the second evaluation result; Match the fault type according to the output judgment result; Get the fault label for evaluating the operating status of the drone.

[0012] As a preferred solution of the UAV operation status fault warning method based on data analysis described in the present invention, the fault warning strategy includes setting a corresponding warning strategy for each fault label, sending the control instructions in the warning strategy as the first priority to the UAV, and executing them.

[0013] A UAV operation status fault warning system based on data analysis using any method described in the present invention, characterized in that: The acquisition unit collects the dynamic parameter data, path control data, actual trajectory and real-time energy consumption of the UAV during movement; An evaluation unit, which directly evaluates the operating state failure of the UAV according to the dynamic parameter data to obtain a first evaluation result; The analysis unit performs a countermeasure evaluation of the operating state failure of the UAV according to the path control data, the actual trajectory, and the real-time energy consumption to obtain a second evaluation result; The early warning unit summarizes and analyzes the first evaluation result and the second evaluation result to generate a fault early warning strategy for the operation status of the UAV.

[0014] A computer device comprises: a memory and a processor; the memory stores a computer program, wherein: the processor implements the steps of any one of the methods of the present invention when executing the computer program.

[0015] A computer-readable storage medium stores a computer program, wherein: when the computer program is executed by a processor, the steps of any one of the methods of the present invention are implemented.

[0016] Beneficial effects of the present invention: The UAV operating status fault warning method based on data analysis provided by the present invention realizes accurate fault detection and real-time warning of the UAV operating status by combining direct evaluation of dynamic parameter data with adversarial evaluation of path and energy consumption. By using multi-source data fusion and adversarial network technology, the ability to distinguish between external interference and internal faults in complex environments is improved, and the probability of misjudgment and missed judgment is significantly reduced. At the same time, personalized warning strategies are generated in combination with fault labels, so that UAVs have higher safety and reliability in complex tasks, and improve operating efficiency and task completion quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use 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 ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0018] Figure 1 An overall flow chart of a method for early warning of a UAV operating status fault based on data analysis provided in the first embodiment of the present invention. DETAILED DESCRIPTION

[0019] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in the art without creative work should fall within the scope of protection of the present invention.

[0020] Example 1, reference Figure 1 , as an embodiment of the present invention, provides a method for early warning of a UAV operation status failure based on data analysis, comprising: S1: Collect dynamic parameter data, path control data, actual trajectory and real-time energy consumption of the UAV during movement.

[0021] Furthermore, the dynamic parameter data includes attitude data and power system data; the attitude data includes attitude angle data, angular velocity data, acceleration data, and vibration frequency; the power system data includes the rotation speed of each motor, the temperature of each motor, and the implemented power of each motor. The path control data includes control data for the trajectory of the UAV. If the UAV flies along a predetermined route, the path control data indicates control of the predetermined route; if the UAV is under real-time control, the path control data indicates that the UAV receives control instructions for real-time control. The actual trajectory includes real-time updating of the trajectory of the UAV through real-time positioning.

[0022] By collecting attitude data and power system data, the UAV's operating attitude (such as flight angle, rotation speed) and power performance (such as motor speed, temperature, power) can be monitored in real time, providing a core basis for judging the stability and health status of the UAV. The path control data combined with the actual trajectory of real-time positioning can fully understand whether the UAV is performing the flight mission as planned and identify possible trajectory deviation or path loss of control problems.

[0023] Combined with the established route or real-time control instructions, it can distinguish between trajectory deviations caused by external environmental interference (such as wind, obstacles) and problems caused by internal system failures (such as motor failure, control lag). Through the correlation analysis of real-time energy consumption data and trajectory deviation, it can be judged whether high energy consumption is caused by external factors (such as strong wind) or system abnormalities (such as battery or motor failure). By collecting and updating dynamic parameters and trajectory data in real time, it provides the system with real-time monitoring capabilities, which can quickly perceive abnormal changes during flight. Providing real-time input data enables subsequent warning algorithms to respond quickly under dynamic conditions to ensure the flight safety of drones. Various types of data collected in real time can not only be used for fault warning, but also provide a basis for optimizing the flight control algorithm of drones. For example, by analyzing the changes in attitude data and power system data, flight control parameters can be improved to improve flight performance and stability.

[0024] S2: Based on the dynamic parameter data, directly evaluate the operating status failure of the UAV to obtain a first evaluation result.

[0025] For the preset fluctuation range of each dynamic parameter data, the real-time dynamic parameter data is compared with the normal fluctuation range corresponding to each data type. If it is not within the range, it is judged that the operation state of the drone is abnormal; otherwise, the operation state of the drone is not judged. The judgment of the abnormal operation state of the drone is output as the first evaluation result.

[0026] It is worth mentioning that by comparing the real-time monitoring of the UAV's dynamic parameter data with the preset normal fluctuation range, it is possible to quickly identify and judge whether the UAV's operating status is abnormal, thereby achieving a preliminary assessment of potential faults. By real-time acquisition of dynamic parameters (such as attitude angle, angular velocity, motor speed, etc.) and comparing them with the normal range, it is possible to accurately capture fluctuations that occur during operation that exceed the normal range, thereby effectively detecting whether the UAV has abnormal attitude, power system abnormalities, or other abnormal operating states. This step uses static thresholds or dynamically adjusted normal fluctuation ranges as a benchmark to ensure the scientificity and operability of the evaluation criteria, while avoiding misjudgments due to slight interference or occasional noise in the external environment, and improving the accuracy and stability of the evaluation results. By directly evaluating the first evaluation result generated, it is possible to quickly screen out obvious abnormal operating states, providing input basis for subsequent more complex adversarial evaluations, which not only shortens the response time of fault identification, but also ensures the comprehensiveness and efficiency of system warnings.

[0027] S3: Based on the path control data, the actual trajectory, and the real-time energy consumption, a countermeasure evaluation of the operating status failure of the UAV is performed to obtain a second evaluation result.

[0028] It should be noted that the confrontation evaluation includes, according to the path control data, respectively confronting the actual trajectory and real-time energy consumption to verify the operation status failure of the UAV. And the verification result is used as the second evaluation result. Among them, the confrontation process is divided into: trajectory confrontation analysis and energy consumption confrontation analysis; Furthermore, the goal of the trajectory confrontation analysis is to verify whether the flight trajectory of the drone meets the established target or real-time control instructions. The goal of the energy consumption confrontation analysis is to verify whether the current energy consumption of the drone matches its flight mission and environmental conditions.

[0029] Specifically, the trajectory confrontation analysis includes: Case 1: Flying along a predetermined route, analyzing the real-time shortest distance between the drone and the predetermined route through real-time positioning ; When the UAV deviates from the predetermined route ( If it is greater than the preset value, it is considered as deviation. Record and Arrange in time series, when the drone returns to the predetermined path and stably moves the preset distance S, the recording ends; and outputs the records from the beginning of deviation from the predetermined path to the return to the predetermined path and the stable movement of the preset distance S. sequence ; The next time the drone deviates from the predetermined route, the sequence is recorded and generated again.

[0030] in, It indicates that when the vehicle deviates from the predetermined path for the nth time, the sequence of the shortest distance is recorded.

[0031] Use the trained generator 1 and discriminator 1 in the adversarial network for analysis and input: and The corresponding wind parameters in the time series ; Output: The probability of a drone failure causing an offset in the corresponding time series.

[0032] in, express The maximum wind speed in the corresponding time series, express The minimum wind speed in the corresponding time series, express The corresponding wind speed average in the time series.

[0033] What needs to be said is that by analyzing the trajectory deviation of the drone during flight on a given route, and combining the influence of external environmental factors such as wind force, the adversarial network model is used to accurately evaluate the probability of the drone deviating from the given path due to its own fault, thereby distinguishing the deviation caused by external interference from that caused by internal faults. Using wind parameters (such as maximum, minimum, and average wind force) as input features, combined with the trajectory deviation sequence, and training the adversarial network model can effectively model the degree of influence of the external environment (wind force) on trajectory deviation. On this basis, through the adversarial learning of the generator and the discriminator, the possible failure risk of the drone system itself during the deviation process is evaluated, and the probability of deviation caused by the fault is output. This design avoids the simple judgment of the traditional method that simply relies on the deviation trajectory, and can accurately identify whether it is a normal deviation caused by the external environment (such as strong wind) or an abnormal deviation caused by the internal system (such as motor failure or abnormal control response).

[0034] Case 2: Use real-time control to fly, and calculate the degree of control behavior by analyzing each control signal; specifically: use the trained generator 2 and discriminator 2 in the adversarial network for analysis, input: and The wind force at the time of occurrence; Output: Predict the displacement vector of the drone under control .

[0035] exist In the sequence, randomly select the displacement vectors corresponding to L continuous control signals, add the L displacement vectors, and obtain L continuous control signals to predict the achievable displacement control results. At the same time, obtain L continuous control signals, the position information at the starting and ending points, and construct the actual displacement based on the position information .

[0036] The degree of realization of the control behavior is expressed as: In the second case, the maximum value of all sampling results is output , the average value of all sampling results .

[0037] in, represents the mth control signal; L represents the value of the fluctuation, which is any integer between the preset maximum value and the preset minimum value; Indicates the degree of realization of the control behavior corresponding to L continuous control signals.

[0038] By inputting each control signal and its corresponding wind force, the adversarial network is used to generate the theoretical motion displacement vector of the drone under the control signal, and compare it with the actual motion displacement. This process effectively distinguishes the difference between the theoretical effect and the actual execution effect of the control signal, and helps to identify the control behavior deviation caused by internal faults (such as flight control system delay or motor failure). Take any L displacement vectors corresponding to the continuous control signals, and obtain the predicted results of the control signal within a period of time by cumulative calculation, and compare them with the actual displacement, so as to evaluate the degree of realization of the control behavior within a specific time period as a whole. This design overcomes the limitation that a single signal may be subject to accidental interference, and more comprehensively reflects the overall performance of the control system. By combining the input of real-time wind parameters (such as wind force), the control behavior realization deviation can be correlated with the influence of external environmental factors, and the deviation caused by external interference such as strong wind and the deviation caused by internal control system failure can be distinguished. This modeling method reduces misjudgment and missed judgment, and improves the accuracy of the evaluation of control behavior in complex flight environments.

[0039] By outputting the maximum and average values ​​of all sampling results, the fluctuation range and stability of the degree of control behavior realization are quantified. This provides data support for analyzing the overall performance of the control system (such as response delay, lack of precision), and provides optimization direction for subsequent improvements in flight control algorithms or hardware configurations.

[0040] The energy consumption adversarial analysis includes using the trained generator 3 and discriminator 3 in the adversarial network to perform analysis, and inputting: the actual trajectory obtained by the drone in real time, the wind parameters from the start of movement to the current moment ; Output: probability of excessive energy consumption anomaly.

[0041] in, Indicates the maximum wind force from the start of movement to the current moment, Indicates the minimum value of the wind force from the beginning of the movement to the current moment, Indicates the average wind speed from the start of the movement to the current moment.

[0042] S4: Summarize and analyze the first evaluation result and the second evaluation result to generate a fault warning strategy for the operating status of the UAV.

[0043] The first evaluation result and the second evaluation result are obtained respectively in the two types of flight states, namely, the predetermined route and the real-time control. The obtained evaluation results are compared with the corresponding preset thresholds respectively to obtain the determination results of each output element in the first evaluation result and the second evaluation result. The fault type is matched according to the output determination results. The fault label for evaluating the operation status of the UAV is obtained.

[0044] Specifically, in this embodiment, under the given route (under this condition, it is not necessary to output the probability of deviation caused by excessive wind): when the first evaluation result does not show abnormality, if the probability of deviation caused by drone failure is less than the corresponding threshold, and the probability of excessive energy consumption abnormality is less than the corresponding threshold, it is determined that there is no abnormality. When the first evaluation result does not show abnormality, if the probability of deviation caused by drone failure is greater than the corresponding threshold, and the probability of excessive energy consumption abnormality is less than the corresponding threshold, it is determined that the system is abnormal. When the first evaluation result does not show abnormality, if the probability of deviation caused by drone failure is less than the corresponding threshold, and the probability of excessive energy consumption abnormality is greater than the corresponding threshold, it is determined that the power system is abnormal. When the first evaluation result does not show abnormality, if the probability of deviation caused by drone failure is greater than the corresponding threshold, and the probability of excessive energy consumption abnormality is greater than the corresponding threshold, it is determined that there is a mixed abnormality. When the first evaluation result shows abnormality, if the probability of excessive energy consumption abnormality is less than the corresponding threshold, it is determined that the power system is abnormal; if the probability of excessive energy consumption abnormality is greater than the corresponding threshold, it is determined that there is a mixed abnormality.

[0045] During flight through real-time control: when the first evaluation result does not show abnormality, if the output of the second situation is less than the corresponding threshold value, and the probability of excessive energy consumption abnormality is less than the corresponding threshold value, it is determined that there is no abnormality. When the first evaluation result does not show abnormality, if any value of the output of the second situation is greater than the corresponding threshold value, and the probability of excessive energy consumption abnormality is less than the corresponding threshold value, it is determined that the system is abnormal. When the first evaluation result does not show abnormality, if any value of the output of the second situation is greater than the corresponding threshold value, and the probability of excessive energy consumption abnormality is greater than the corresponding threshold value, it is determined that there is a mixed abnormality. When the first evaluation result does not show abnormality, if the output of the second situation is less than the corresponding threshold value, and the probability of excessive energy consumption abnormality is greater than the corresponding threshold value, the power system is abnormal. When the first evaluation result shows abnormality, if the probability of excessive energy consumption abnormality is greater than the corresponding threshold value, it is determined that there is a mixed abnormality; if the probability of excessive energy consumption abnormality is less than the corresponding threshold value, it is determined that there is a power system abnormality.

[0046] When only the maximum value of all sampling results is output is greater than the corresponding threshold, and the average value of all sampling results When it is less than the corresponding threshold, it is judged as a sudden abnormality.

[0047] The fault warning strategy includes setting a corresponding warning strategy for each fault tag, sending the control instructions in the warning strategy as the first priority to the drone, and executing them.

[0048] In an optional embodiment, when a mixed abnormality occurs, the aircraft is grounded immediately; when a sudden abnormality occurs, it is temporarily ignored and recorded; when a system abnormality or a power system abnormality occurs, the aircraft is grounded after completing the current mission; all plans can be changed according to the preset.

[0049] It should be said that by combining the results of direct evaluation and confrontation evaluation, the abnormal operation status of the drone is divided into different types (no abnormality, system abnormality, power system abnormality, mixed abnormality, and sudden abnormality), and the judgment logic and conditions are set according to the flight mode. This refined classification design avoids the misjudgment or missed judgment that may be caused by the traditional single evaluation method, and ensures that the source of the problem (such as wind interference, power system failure or control system abnormality) can be accurately identified in a complex operating environment. According to the type of abnormality, multi-level response strategies are provided (such as immediate grounding, grounding after the mission is completed, and continued recording and observation, etc.) to achieve hierarchical processing of abnormalities. In particular, the high-priority response strategy (immediate grounding) for mixed abnormalities (coexistence of internal system and power system problems) and the low-priority strategy (temporary ignoring and recording) for sudden abnormalities not only ensure flight safety, but also avoid excessive intervention in flight missions due to non-critical problems.

[0050] Through the linkage design of fault labels and early warning strategies, the system can quickly generate and execute corresponding control instructions after fault determination, ensuring the operational stability of the drone in complex tasks. Under the premise of not affecting the completion of the task, the system abnormalities or power system abnormalities can be reasonably handled (such as stopping the flight after completing the task) to improve the efficiency of task execution.

[0051] On the other hand, this embodiment also provides a UAV operation status fault warning system based on data analysis, which includes: The acquisition unit collects the dynamic parameter data, path control data, actual trajectory and real-time energy consumption of the UAV during movement.

[0052] The evaluation unit directly evaluates the operating status failure of the UAV according to the dynamic parameter data to obtain a first evaluation result.

[0053] The analysis unit performs a countermeasure evaluation of the operating state failure of the UAV according to the path control data, the actual trajectory and the real-time energy consumption to obtain a second evaluation result.

[0054] The early warning unit summarizes and analyzes the first evaluation result and the second evaluation result to generate a fault early warning strategy for the operation status of the UAV.

[0055] If the above functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc., which can store program code.

[0056] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in conjunction with such instruction execution systems, devices or apparatuses. For the purposes of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in conjunction with such instruction execution systems, devices or apparatuses.

[0057] More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk case (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be a paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering or, if necessary, processing in another suitable manner, and then stored in a computer memory.

[0058] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0059] Example 2 is an embodiment of the present invention, which provides a UAV operating status fault warning method based on data analysis. In order to verify the beneficial effects of the present invention, scientific demonstration is carried out through economic benefit calculation and simulation experiments.

[0060] The test subjects are two drones of the same model: Type A drone adopts the fault warning method of the present invention, and Type B drone adopts the traditional single dynamic parameter analysis method as a comparison. The test environment is set in a complex flight mission scenario, including two modes: set route flight and real-time control. The specific conditions of the flight mission include a flight altitude of 50 meters, a wind speed range of 3-10 m / s, a total flight distance of 5 kilometers, and a load of 1 kg. During the experiment, the following data are collected in real time: attitude data (pitch angle, roll angle), power system data (motor speed, power consumption), path deviation data (shortest deviation distance, deviation time) and real-time energy consumption.

[0061] Before the experiment, the flight control parameters of each drone were set to the factory default values. For type A drone, the fault warning method of the invention content was loaded, including dynamic parameter direct evaluation, trajectory confrontation analysis, and energy consumption confrontation analysis modules. Type B drone only uses the traditional method to perform fault judgment by monitoring the fixed threshold of dynamic parameters.

[0062] The experiment is divided into the following steps: Predetermined route flight mission: The drone flies along the preset route and collects dynamic parameter data and trajectory deviation data in real time. When the deviation distance exceeds 1 meter or the energy consumption exceeds 20% of the expected value, the deviation time and the probability of abnormal energy consumption are recorded.

[0063] Real-time control task: The UAV completes the navigation flight of random target points through ground control commands, and monitors the execution of control signals and the degree of realization of corresponding displacement in real time. When the degree of realization of control signals is less than 80%, the frequency and scope of influence are recorded.

[0064] Data recording and analysis: After the experiment, the operating data of the two drones were collected, and their fault warning success rate, misjudgment rate, and impact on task completion efficiency were counted. The record table is shown in Table 1.

[0065] Table 1 Data record table It can be seen from the test data that the fault warning method of the present invention has obvious advantages in the accuracy of fault identification, flight stability and task completion efficiency. Through analysis and comparison, the following conclusions can be drawn: The failure warning success rate of the A-type UAV reached 98%, which is much higher than the 82% of the B-type UAV. This result shows that the present invention can effectively capture potential failures by combining direct evaluation of dynamic parameters with adversarial evaluation of paths and energy consumption. In addition, the misjudgment rate of the A-type UAV is only 2%, which is much lower than the 10% of the B-type UAV, showing that the present invention has a stronger ability to distinguish external interference and internal failures in complex environments.

[0066] The average deviation distance of the A-type UAV is 0.6 meters, which is significantly lower than the 1.5 meters of the B-type UAV; the deviation time is also reduced from 28 seconds of the traditional method to 12 seconds. This shows that the trajectory confrontation analysis module of the present invention can timely identify and correct trajectory deviation, ensuring the accuracy of flight.

[0067] In terms of energy consumption, the average power consumption of the A-type drone is 280 W, while that of the B-type drone is 310 W. The energy consumption countermeasure analysis module of the present invention reduces unnecessary power loss by adjusting the power system output in real time, while effectively avoiding the abnormal situation of excessive energy consumption.

[0068] In the real-time control mode, the control signal realization degree of the A-type UAV is 94%, which is higher than the 85% of the B-type UAV. This shows that the method of the present invention can accurately predict the displacement response through dynamic analysis of the control signal and optimize the execution effect of the control system.

[0069] The A-type drone completed all the test tasks, while the B-type drone had a task completion rate of only 85% due to multiple failures and deviations. This shows that the present invention can not only improve operational safety, but also significantly enhance the reliability of task execution.

[0070] The test results clearly show that the fault warning method of the present invention overcomes the limitations of the traditional method of single data source analysis through multi-source data fusion analysis and intelligent evaluation, and has significant innovation and advantages. Especially in complex flight missions, the present invention can more effectively identify and warn potential faults, ensure the safety of drones and the completion rate of missions, and provide important support for the practical application of drone technology.

[0071] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A method for early warning of UAV operating status failure based on data analysis, characterized in that: include: Collect dynamic parameter data, path control data, actual trajectory and real-time energy consumption of drones during movement; Directly evaluating the operating state failure of the UAV according to the dynamic parameter data to obtain a first evaluation result; According to the path control data, the actual trajectory, and the real-time energy consumption, a countermeasure evaluation of the operating state failure of the UAV is performed to obtain a second evaluation result; The first evaluation result and the second evaluation result are summarized and analyzed to generate a fault warning strategy for the operating status of the UAV.

2. The method for early warning of a UAV operating state failure based on data analysis according to claim 1, characterized in that: The dynamic parameter data includes posture data and power system data; The posture data includes posture angle data, angular velocity data, acceleration data, and vibration frequency; The power system data includes the speed of each motor, the temperature of each motor, and the implemented power of each motor; The path control data includes control data of the running trajectory of the UAV. If the UAV flies along a predetermined route, the path control data indicates the control of the predetermined route; if the UAV is under real-time control, the path control data indicates that the UAV receives a control instruction for real-time control; The actual trajectory includes real-time updating of the trajectory of the drone through real-time positioning.

3. The method for early warning of a UAV operating state failure based on data analysis as claimed in claim 2, characterized in that: The direct evaluation includes comparing the real-time dynamic parameter data with the normal fluctuation range corresponding to the respective data types for the fluctuation range preset for each dynamic parameter data, and if it is not within the range, determining that the operation state of the drone is abnormal; Otherwise, the UAV operation status will not be judged; The judgment that the operating status of the drone is abnormal is taken as the first evaluation result.

4. The method for early warning of a UAV operating state failure based on data analysis as claimed in claim 3, characterized in that: The confrontation evaluation includes, according to the path control data, respectively confronting the actual trajectory and the real-time energy consumption to verify the operation status failure of the UAV; and use the verification result as the second evaluation result; Among them, the confrontation process is divided into: trajectory confrontation analysis and energy consumption confrontation analysis; The goal of the trajectory confrontation analysis is to verify whether the flight trajectory of the drone meets the established target or real-time control instructions; The goal of the energy consumption confrontation analysis is to verify whether the current energy consumption of the drone matches its flight mission and environmental conditions.

5. The method for early warning of a UAV operating state failure based on data analysis according to claim 4, characterized in that: The trajectory confrontation analysis includes: Case 1: flying along a predetermined route, analyzing the real-time shortest distance between the drone and the predetermined route through real-time positioning ; When the drone deviates from the predetermined route, Record and Arrange in time series, when the drone returns to the predetermined path and stably moves the preset distance S, the recording ends; and outputs the records from the beginning of deviation from the predetermined path to the return to the predetermined path and the stable movement of the preset distance S. sequence ; The next time the drone deviates from the predetermined route, the sequence is recorded and generated again; in, Indicates the sequence of recording the shortest distance when the nth deviation from the predetermined path occurs; Use the trained generator 1 and discriminator 1 in the adversarial network for analysis and input: and The corresponding wind parameters in the time series ; Output: The probability of a drone failure causing a shift in the corresponding time series; in, express The maximum wind speed in the corresponding time series, express The minimum wind speed in the corresponding time series, express The average wind speed in the corresponding time series; Case 2: Use real-time control to fly, and calculate the degree of control behavior by analyzing each control signal; specifically: use the trained generator 2 and discriminator 2 in the adversarial network for analysis, input: and The wind force at the time of occurrence; Output: Predict the displacement vector of the drone under control ; exist In the sequence, randomly select the displacement vectors corresponding to L continuous control signals, add the L displacement vectors, and obtain L continuous control signals to predict the achievable displacement control results. At the same time, obtain L continuous control signals, the position information at the starting and ending points, and construct the actual displacement based on the position information ; The degree of realization of the control behavior is expressed as: In the second case, the maximum value of all sampling results is output , the average value of all sampling results ; in, represents the mth control signal; L represents the value of the fluctuation, which is any integer between the preset maximum value and the preset minimum value; Indicates the degree of realization of the control behavior corresponding to L continuous control signals; The energy consumption adversarial analysis includes using the trained generator 3 and discriminator 3 in the adversarial network to perform analysis, and inputs: the actual trajectory obtained by the drone in real time, the wind parameters from the start of movement to the current moment ; Output: Probability of excessive energy consumption anomaly; in, Indicates the maximum wind force from the start of movement to the current moment, Indicates the minimum value of the wind force from the beginning of the movement to the current moment, Indicates the average wind speed from the start of the movement to the current moment.

6. The method for early warning of UAV operation status failure based on data analysis according to claim 5, characterized in that: The summary analysis includes obtaining the first evaluation result and the second evaluation result under two types of flight states, the predetermined route and the real-time control, respectively; Comparing the obtained evaluation results with the corresponding preset thresholds respectively to obtain a determination result of each output element in the first evaluation result and the second evaluation result; Match the fault type according to the output judgment result; Get the fault label for evaluating the operating status of the drone.

7. The method for early warning of a UAV operating state failure based on data analysis according to claim 6, characterized in that: The fault warning strategy includes setting a corresponding warning strategy for each fault tag, sending the control instructions in the warning strategy as the first priority to the drone, and executing them.

8. A UAV operation status fault warning system based on data analysis using the method according to any one of claims 1 to 7, characterized in that: The acquisition unit collects the dynamic parameter data, path control data, actual trajectory and real-time energy consumption of the UAV during movement; An evaluation unit, which directly evaluates the operating state failure of the UAV according to the dynamic parameter data to obtain a first evaluation result; The analysis unit performs a countermeasure evaluation of the operating state failure of the UAV according to the path control data, the actual trajectory, and the real-time energy consumption to obtain a second evaluation result; The early warning unit summarizes and analyzes the first evaluation result and the second evaluation result to generate a fault early warning strategy for the operation status of the UAV.

9. A computer device comprising: A memory and a processor; the memory stores a computer program, wherein the processor implements the steps of any method as claimed in claim 1 when executing the computer program.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images