Unmanned aerial vehicle test method, device and equipment

Abstract

The invention provides an unmanned aerial vehicle testing method, device and equipment, belongs to the technical field of unmanned aerial vehicles, and solves the problems that a traditional unmanned aerial vehicle is low in testing efficiency, long in time consumption and high in testing cost, and the accuracy is difficult to guarantee. The method comprises the steps of generating a test instruction and sending the test instruction to the unmanned aerial vehicle; acquiring flight state data of the unmanned aerial vehicle; processing the flight state data to obtain performance data, the performance data including at least one of endurance time, a climbing rate, a maximum voyage, a maximum flight height and a maximum flight speed; and inputting the flight state data and the performance data into a flight state test model for processing to obtain test data. According to the scheme, the testing efficiency and accuracy of the unmanned aerial vehicle are improved, and the testing cost is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of unmanned aerial vehicles, and particularly to a method, device and equipment for testing unmanned aerial vehicles. Background Art

[0002] With the global low-altitude economy entering a stage of rapid development, the application boundaries of unmanned aerial vehicles (UAVs) are constantly expanding. In commercial scenarios, emerging models such as urban air mobility (UAM) and emergency disaster relief are gradually taking shape. Statistical data shows that the average annual growth rate of the number of globally registered commercial drones reaches 34%, but the accident rate has increased by 21% during the same period, exposing the technical limitations of the existing testing system.

[0003] The current mainstream UAV testing system still takes manual operation as the core, and its technical framework includes three levels:

[0004] (1) Physical environment testing: It relies on open-air test flight fields and wind tunnel laboratories to obtain flight data through repeated takeoffs and landings. A certain model of logistics UAV needs to complete more than 300 takeoff and landing tests, which takes more than 45 days and cannot reproduce extreme weather conditions.

[0005] (2) Function verification testing: A semi-automated test bench controlled by scripts is adopted, but manual intervention is still required in key links such as sensor calibration and control parameter debugging. Test data shows that the skill differences of operators will cause the fluctuation range of navigation accuracy test results to reach ±18%.

[0006] (3) Environmental adaptability testing: A simple simulation device is used to generate single-factor environments such as vibration and temperature change, lacking the ability to simulate multi-physical field coupling. A certain plateau scientific research UAV failed to detect the power attenuation under the composite condition of "low temperature + low air pressure", resulting in a mission failure rate as high as 37%.

[0007] It can be seen that the traditional testing method mainly relies on manual operation, and there are problems such as low testing efficiency, long duration, the test results being affected by human factors, and difficulty in ensuring accuracy. Therefore, there is an urgent need for an automated and high-precision method for testing unmanned aerial vehicles. Summary of the Invention

[0008] The present invention provides a method, device and equipment for testing unmanned aerial vehicles, which solves the problems that the traditional testing method for unmanned aerial vehicles mainly relies on manual operation, has low testing efficiency, long duration, high testing cost, the test results are affected by human factors, and it is difficult to ensure accuracy.

[0009] To solve the above technical problems, the technical solution of the present invention is as follows:

[0010] An embodiment of the present invention provides a method for testing an unmanned aerial vehicle, including:

[0011] Generate test instructions and send them to the unmanned aerial vehicle;

[0012] Obtain the flight status data of the unmanned aerial vehicle;

[0013] Process the flight status data to obtain performance data, where the performance data includes at least one of endurance time, climb rate, maximum range, maximum flight altitude, and maximum flight speed;

[0014] Input the flight status data and the performance data into a flight status test model for processing to obtain test data.

[0015] Optionally, the generating test instructions and sending them to the unmanned aerial vehicle includes:

[0016] Obtain waypoint coordinate data according to preset route data;

[0017] Obtain flight speed control data and flight attitude angle control data according to the waypoint coordinate data;

[0018] Generate test instructions according to the waypoint coordinate data, the flight speed control data, and the flight attitude angle control data and send them to the unmanned aerial vehicle.

[0019] Optionally, the obtaining the flight status data of the unmanned aerial vehicle includes:

[0020] Obtain at least one of the attitude data, altitude data, speed data, and battery status data of the unmanned aerial vehicle according to multiple sensors installed on the unmanned aerial vehicle;

[0021] Perform anomaly detection processing on at least one of the attitude data, altitude data, speed data, and battery status data to obtain multiple detection data;

[0022] Merge and process the multiple detection data according to a preset format to obtain the flight status data of the unmanned aerial vehicle.

[0023] Optionally, the processing the flight status data to obtain performance data includes:

[0024] Obtain the endurance time and the climb rate according to the battery status data and the speed data;

[0025] Obtain the maximum range according to the endurance time data and the speed data;

[0026] Obtain the maximum flight altitude according to the altitude data, the battery status data, and the speed data;

[0027] Obtain the maximum flight speed according to the battery status data and the attitude data.

[0028] Optionally, input the flight state data and the performance data into a flight state test model for processing to obtain test data, including:

[0029] Process the flight state data and the performance data to obtain an input vector;

[0030] Input the input vector into the flight state test model for processing to obtain test data; the flight state test model processes the input vector through a node function to obtain an intermediate result, and performs deviation correction on the intermediate result to obtain test data.

[0031] Optionally, the flight state test model is trained through the following process:

[0032] Obtain historical flight state data, historical performance data, and historical fault type data;

[0033] Perform induction processing on the historical flight state data, the historical performance data, and the historical fault type data to obtain sample data;

[0034] Perform random sampling processing on the sample data to obtain a training data set and a test sample data set;

[0035] Input the training data set into the flight state test model for processing to obtain a training result data set;

[0036] Compare the training result data set with the test sample data set to obtain a comparison result;

[0037] Adjust the node function parameters of the flight state test model according to the comparison result to obtain a trained flight state test model.

[0038] Optionally, the method further includes:

[0039] Output a test report for the unmanned aerial vehicle, where the test report for the unmanned aerial vehicle includes: the performance data of the unmanned aerial vehicle and the test data.

[0040] An embodiment of the present invention further provides a test device for an unmanned aerial vehicle, including:

[0041] A generation module, configured to generate a test instruction and send it to the unmanned aerial vehicle;

[0042] An acquisition module, configured to acquire the flight state data of the unmanned aerial vehicle;

[0043] A processing module, configured to process the flight state data to obtain performance data; input the flight state data and the performance data into a flight state test model for processing to obtain test data.

[0044] An embodiment of the present invention further provides a computing device, including: a processor and a memory storing a computer program. When the computer program is run by the processor, the above method is executed.

[0045] An embodiment of the present invention further provides a computer-readable storage medium storing instructions. When the instructions are run on a computer, the computer is caused to execute the above method.

[0046] The technical solution of the present invention at least includes the following effects:

[0047] The above solution of the present invention generates test instructions and sends them to the unmanned aerial vehicle; obtains the flight state data of the unmanned aerial vehicle; processes the flight state data to obtain performance data, where the performance data includes at least one of endurance time, climb rate, maximum range, maximum flight altitude, and maximum flight speed; inputs the flight state data and performance data into a flight state test model for processing to obtain test data. Compared with the traditional manual test method, this solution improves the test efficiency and coverage through digital processing and model-based evaluation, can dynamically adapt to the requirements of different test scenarios, improves the efficiency and accuracy of unmanned aerial vehicle testing, and reduces the test cost. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 is a flowchart of the unmanned aerial vehicle test method provided by an embodiment of the present invention;

[0049] Figure 2 is a schematic structural diagram of the flight state test model in an embodiment of the present invention;

[0050] Figure 3 is a structural diagram of the unmanned aerial vehicle test device provided by an embodiment of the present invention;

[0051] Figure 4 is a schematic structural diagram of the computing device provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0052] Hereinafter, the exemplary embodiments of the present invention will be described in more detail with reference to the drawings. Although the exemplary embodiments of the present invention are shown in the drawings, it should be understood that the present invention can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided so that the present invention can be more thoroughly understood and the scope of the present invention can be completely conveyed to those skilled in the art.

[0053] As Figure 1 shown, an embodiment of the present invention proposes an unmanned aerial vehicle test method, including:

[0054] Step 11, generating test instructions and sending them to the unmanned aerial vehicle;

[0055] Step 12, obtain the flight state data of the unmanned aerial vehicle;

[0056] Step 13, process the flight state data to obtain performance data, where the performance data includes at least one of endurance time, climb rate, maximum range, maximum flight altitude, and maximum flight speed;

[0057] Step 14, input the flight state data and the performance data into a flight state test model for processing to obtain test data.

[0058] In this embodiment, first, according to the design requirements and expected performance indicators of the unmanned aerial vehicle, determine the specific objectives of this test. For example, if testing the endurance of the unmanned aerial vehicle, the instructions may focus on making the vehicle fly for a long time at a specific speed and altitude; if testing the climb rate, the instructions will require the vehicle to perform rapid climbing operations under different initial conditions; based on the test objectives, design detailed test instructions. The instruction content usually includes the type of flight mission (such as cruising, climbing, landing, etc.), flight parameters (such as speed, altitude, heading, etc.), mission execution time, etc. For example, the instructions may stipulate that the vehicle cruises straight at a speed of 50 m / s at an altitude of 1000 m for 30 min, while collecting flight state data every 5 s; encode and encapsulate the designed instructions according to the communication protocol that the unmanned aerial vehicle can recognize to ensure that the instructions can be transmitted to the vehicle accurately; establish a reliable communication link with the unmanned aerial vehicle through a wireless communication device; before sending the instructions, it is necessary to test the communication link to ensure that the signal strength, stability, and data transmission rate meet the requirements; send the encapsulated test instructions to the unmanned aerial vehicle through the communication link; during the transmission process, adopt technical means such as checksum and retransmission mechanism to ensure the integrity and accuracy of the instructions.

[0059] The unmanned aerial vehicle is equipped with a variety of sensors for real-time collection of flight state data. Common sensors include inertial measurement units (IMUs, used to measure acceleration and angular velocity), barometers (used to measure altitude), pitot tubes (used to measure airspeed), gyroscopes (used to measure attitude angles), etc. According to the test requirements and the dynamic characteristics of the aircraft, an appropriate data collection frequency is set. For flight state parameters that change rapidly, such as acceleration and angular velocity, a relatively high collection frequency (such as 100 Hz or higher) may be required. For some parameters that change relatively slowly, such as altitude and airspeed, the collection frequency can be appropriately reduced (such as 10 Hz). The flight state data collected by the sensors is transmitted to the ground test equipment through the data link on the aircraft. The data link can use the same communication method as the command transmission or a dedicated data transmission channel. After receiving the flight state data, the ground test equipment performs preliminary processing and storage on it. The processing content includes data verification, format conversion, etc., to ensure the accuracy and availability of the data. The storage method can use databases, file systems, etc., for subsequent data processing and analysis.

[0060] Since the flight state data collected by the sensors may contain noise and interference, it needs to be filtered. Common filtering methods include low-pass filtering, high-pass filtering, Kalman filtering, etc., to remove noise and improve the data quality. The data collected by the sensors is calibrated to eliminate sensor errors. The calibration method can use laboratory calibration, on-site calibration, etc., and an appropriate calibration method is selected according to the characteristics of the sensor and the use environment. After the flight state data undergoes the above processing, it can be used to calculate performance data, where the performance data includes at least one of endurance, climb rate, maximum range, maximum flight altitude, and maximum flight speed. For example, the climb rate is calculated based on the change data of the aircraft's altitude over time. The endurance of the aircraft is calculated by recording the takeoff and landing times of the aircraft.

[0061] The test model is trained using historical flight data and corresponding performance data; during the training process, the parameters of the model need to be adjusted to enable the model to accurately predict the performance of the aircraft; the training data should cover various flight conditions and aircraft states as much as possible to improve the generalization ability of the model; after the model training is completed, the flight state data and performance data are integrated to form a data format suitable for model input; for example, the data from different sensors are arranged in a time series, and the performance data is added as a label to the data; to improve the training effect and prediction accuracy of the model, the input data is normalized and mapped to a specific interval; the integrated and normalized data is input into the trained flight state test model, and the model makes predictions based on the input data and outputs the test data; the test data includes performance evaluation indicators of the aircraft (such as flight quality scores, reliability evaluations, etc.) and fault warning information, etc.; after obtaining the performance data and test data, they can also be summarized and sorted into a measurement report and output through a printer or a display to timely reflect the current test results of the unmanned aircraft.

[0062] This solution improves the test efficiency and coverage through digital processing and model-based evaluation, can dynamically adapt to the requirements of different test scenarios, improves the efficiency and accuracy of unmanned aircraft testing, and reduces the test cost.

[0063] In an optional embodiment proposed by the present invention, step 11 may include:

[0064] Step 111, obtaining waypoint coordinate data according to the preset route data;

[0065] Step 112, obtaining flight speed control data and flight attitude angle control data according to the waypoint coordinate data;

[0066] Step 113, generating a test instruction according to the waypoint coordinate data, the flight speed control data and the flight attitude angle control data and sending it to the unmanned aircraft.

[0067] In this embodiment, the preset route data refers to the route data designed in advance before the test of the unmanned aerial vehicle, including the starting point, ending point and key waypoints in the middle of the flight mission. These data are usually represented in the form of coordinates such as longitude and latitude, altitude, etc.; according to the preset route data, through map services or flight planning software, the accurate coordinate data of each waypoint can be extracted; according to the distance between waypoints, the time requirements of the flight mission and the performance parameters of the unmanned aerial vehicle, the flight speed between each waypoint is calculated, including parameters such as average speed, maximum speed and acceleration; the flight attitude angles include pitch angle, yaw angle and roll angle, which determine the flight direction and attitude of the unmanned aerial vehicle; according to the relative position relationship between waypoints and the flight speed, the required change in flight attitude angle between each waypoint can be calculated.

[0068] Specifically, according to the geographic coordinate transformation, the longitude and latitude coordinates are converted into plane coordinates:

[0069]

[0070] In the formula, φ is the geographic latitude, λ is the geographic longitude, E is the plane eastward coordinate, N is the plane northward coordinate, k 0 is the projection scale factor, which is taken as 0.9996 in this example, f is the radius of curvature, T is the latitude function, C is the ellipsoid parameter, A is the longitude difference parameter; E 0 is the eastward offset; M is the meridian arc length; M 0 is the meridian arc length of the central meridian

[0071] According to the waypoint sequence {p i =(x i , y i , z i )} n i＝1 , the flight path curve can be obtained according to the following formula:

[0072] S(t)=a i (t - t i ) 3 + b i (t - t i ) 2 + c i (t - t i ) + d i , t ∈ [t i , t i+1

[0073] In the formula, t is the normalized path parameter, t ∈ [0, 1]; a i , b i , c i , d i are the cubic polynomial coefficient matrices;

[0074] According to:

[0075]

[0076] Perform sectional speed planning; where m is the mass of the UAV, g is the acceleration due to gravity, γ is the climb angle, ρ is the air density, A is the frontal area of the UAV, C rr is the rolling resistance coefficient; C D is the aerodynamic drag coefficient; v climb is the speed of the UAV during the climb phase; v cruise is the speed of the UAV during the cruise phase; v descent is the speed of the UAV during the descent phase; v 0 is the initial speed of the UAV during descent; a safe is the safe acceleration of the UAV during descent; z is the initial height of the UAV during descent; z target is the target height of the UAV during descent, μ is the equivalent rolling resistance coefficient; v is the airspeed of the UAV; P max is the maximum available power of the UAV power system.

[0077] According to:

[0078]

[0079] Perform heading angle calculation; where ψ is the heading angle, δ ψ is the magnetic declination correction term, x k 、x k+1 、y k 、y k+1 are the coordinate positions of the UAV.

[0080] According to:

[0081]

[0082] Perform roll angle calculation; where φ is the roll angle, R is the instantaneous turning radius; Δψ is the change in direction angle between adjacent flight segments;

[0083] According to:

[0084]

[0085] Perform pitch angle calculation; where θ is the pitch angle, θ trim is the trim angle correction.

[0086] Integrate the waypoint coordinate data, flight speed control data, and flight attitude angle control data into test instructions; these instructions include detailed information such as flight paths, speed curves, and attitude adjustment strategies; send the test instructions to the unmanned aerial vehicle via wireless communication technology. After receiving the instructions, the unmanned aerial vehicle will execute the flight mission according to the instruction requirements.

[0087] In an optional embodiment proposed by the present invention, step 12 may include:

[0088] Step 121, obtain at least one of the attitude data, altitude data, speed data, and battery status data of the unmanned aerial vehicle according to multiple sensors installed on the unmanned aerial vehicle;

[0089] Step 122, perform anomaly detection processing on at least one of the attitude data, altitude data, speed data, and battery status data to obtain multiple detection data;

[0090] Step 123, merge and process the multiple detection data according to a preset format to obtain the flight status data of the unmanned aerial vehicle.

[0091] In this embodiment, the unmanned aerial vehicle uses a variety of high-precision sensors installed thereon to capture key flight information in real time. These sensors include: attitude sensors, such as an inertial measurement unit composed of a combination of a gyroscope, an accelerometer, a magnetometer, etc., which is used to measure attitude parameters such as the pitch angle, yaw angle, and roll angle of the unmanned aerial vehicle, reflecting the spatial orientation and dynamic stability of the aircraft; altitude sensors, such as a barometer, lidar, or ultrasonic sensor, which are used to accurately measure the vertical height of the aircraft relative to the ground to ensure flight safety and avoid collisions; speed sensors, such as a pitot tube, a GPS receiver combined with algorithm calculations, etc., which provide instant speed information of the aircraft, including ground speed, airspeed, etc., and are helpful for flight control and navigation; battery status monitoring sensors, which are used to monitor key parameters such as battery voltage, current, remaining power, and temperature to ensure the stability and safety of power supply during flight.

[0092] After obtaining the flight status data of the above-mentioned unmanned aerial vehicle, compare the real-time data with a preset safe range or threshold, and the data outside the range is regarded as abnormal; after excluding the abnormal values, data that can truly reflect the current flight status of the unmanned aerial vehicle can be obtained.

[0093] Integrate all the data detected by anomaly detection to form a set of structured flight status data reports. Specifically, organize the detected data in a unified format, such as JSON, XML, etc., for subsequent data storage, transmission, and analysis; at the same time, add accurate time stamps to each piece of data to ensure the timeliness and traceability of the data; pack the formatted data to form a data matrix, which consists of multiple vectors with the same number of dimensions, and the elements in each vector include the attitude data, altitude data, speed data, battery status data, time stamp data, etc. of the unmanned aerial vehicle obtained.

[0094] In an optional embodiment proposed by the present invention, step 13 may include:

[0095] Step 131, obtain the endurance time and climb rate according to the battery status data and the speed data;

[0096] Step 132, obtain the maximum range according to the endurance time data and the speed data;

[0097] Step 133, obtain the maximum flight altitude according to the altitude data, the battery status data, and the speed data;

[0098] Step 134, obtain the maximum flight speed according to the battery status data and the attitude data.

[0099] In this embodiment, the endurance time of the unmanned aerial vehicle is related to the flight speed and the battery status; at different flight speeds, the energy consumption of the aircraft is different. Generally speaking, when flying at a high speed, the motor needs to consume more electrical energy to overcome air resistance, and the energy consumption is relatively high; while when flying at a low speed, the energy consumption is relatively low. For example, when the aircraft flies at a speed of 10 m / s, the motor power is 50 W; while when flying at a speed of 20 m / s, the motor power may increase to 150 W. Specifically, for a fixed-wing unmanned aerial vehicle, it can be obtained by:

[0100]

[0101] Get the endurance time, where T is the endurance time, E 1 is the total battery energy, E 1 = Uq, q is the battery capacity, U is the voltage, P 1 is the cruise power, m is the mass of the unmanned aerial vehicle, g is the acceleration due to gravity, v is the cruise speed of the unmanned aerial vehicle, k is the lift-to-drag ratio of the unmanned aerial vehicle, and η is the motor efficiency.

[0102] Thus, the maximum range of the unmanned aerial vehicle can be obtained:

[0103] S = v·T

[0104] Among them, S is the maximum range of the unmanned aerial vehicle.

[0105] The formula for the climbing rate of the unmanned aerial vehicle is:

[0106]

[0107] Among them, Λ is the climbing rate of the unmanned aerial vehicle, P 2 is the remaining power of the battery, P is the total power of the battery, and I is the working current of the motor.

[0108] The formula for the maximum flight altitude is:

[0109]

[0110] Among them, ρ is the air density, and A is the windward area of the unmanned aerial vehicle.

[0111] The formula for the maximum flight speed is:

[0112]

[0113] Among them, C D is the drag coefficient, which is related to the aerodynamic design of the fuselage.

[0114] In an optional embodiment proposed by the present invention, step 14 may include:

[0115] Step 141, process the flight state data and the performance data to obtain an input vector;

[0116] Step 142, input the input vector into a flight state test model for processing to obtain test data; the flight state test model processes the input vector through a node function to obtain an intermediate result, and corrects the deviation of the intermediate result to obtain test data.

[0117] In this embodiment, the flight state data obtained after anomaly detection and the performance data obtained through calculation, a total of 13 types of key feature data, so the dimension of the input vector is 13; arrange the 13 types of key feature data in a certain order to form an input vector; input the input vector into the flight state test model for processing; among them, the flight state test model is a model constructed based on mathematical algorithms and machine learning techniques, which can process the input vector, simulate the flight state of the unmanned aerial vehicle, and output corresponding test data. This model is trained with a large amount of training data to ensure that it can accurately reflect the actual flight situation of the aircraft; the node function is one of the core parts of the flight state test model, and according to each feature in the input vector, it is calculated according to certain rules and algorithms to obtain an intermediate result; the design of the node function depends on the purpose and requirements of the test. For example, if the focus of the test is the stability of the aircraft, the node function may pay more attention to attitude data and acceleration data; if the focus of the test is the endurance of the aircraft, the node function may focus more on battery state data and endurance time data. Due to reasons such as model simplification, data incompleteness, and environmental factor interference, the intermediate result obtained by the node function may have deviations; for example, the model may not consider some complex aerodynamic effects, or there are certain errors in the data collection and transmission process, which will cause a certain deviation between the intermediate result and the actual situation; in order to obtain more accurate test data, it is necessary to correct the deviation of the intermediate result to obtain the test data.

[0118] In an alternative embodiment proposed by the present invention, the flight state test model is trained through the following process:

[0119] Step 151, obtain historical flight state data, historical performance data, and historical fault type data;

[0120] Step 152, perform induction processing on the historical flight state data, the historical performance data, and the historical fault type data to obtain sample data;

[0121] Step 153, perform random sampling processing on the sample data to obtain a training data set and a test sample data set;

[0122] Step 154, input the training data set into the flight state test model for processing to obtain a training result data set;

[0123] Step 155, compare the training result data set with the test sample data set to obtain a comparison result;

[0124] Step 156, according to the comparison result, adjust the node function parameters of the flight state test model to obtain the trained flight state test model.

[0125] In this embodiment, historical flight state data, historical performance data, and historical fault type data are first obtained. Among them, the historical flight state data records various state information of the unmanned aerial vehicle during past flights. For example, the variation curve of flight altitude over time can reflect the vertical position of the aircraft at different flight stages; the flight speed data, including horizontal speed and vertical speed, can reflect the movement speed and direction change of the aircraft; the attitude angle data, such as pitch angle, roll angle, and yaw angle, can describe the spatial attitude of the aircraft in the air. The historical performance data relates to various performance indicators of the unmanned aerial vehicle. For instance, the power output data of the engine reflects the working ability of the engine under different working conditions; the power consumption of the battery can help understand the endurance ability of the aircraft; the maximum climb rate, maximum level flight speed, etc. of the aircraft directly reflect the performance limits of the aircraft. The historical fault type data records various fault situations that occurred during the past flights of the unmanned aerial vehicle. For example, engine faults may manifest as abnormal power decline, excessive vibration, etc.; sensor faults may lead to inaccurate data acquisition; communication faults will affect the information transmission between the aircraft and the ground control station. Each fault type has corresponding fault feature descriptions, such as the time of fault occurrence, fault phenomena, fault codes, etc. These data are helpful for analyzing the causes and laws of fault occurrence.

[0126] After obtaining the historical flight state data, historical performance data, and historical fault type data, it is necessary to perform inductive processing on the historical flight state data, the historical performance data, and the historical fault type data to obtain sample data.

[0127] Check the acquired historical data and remove the noise data, error data, and missing values. For example, due to sensor failures or signal interference, some abnormal data points may occur, which will affect the training effect of the model and need to be removed. For missing values, interpolation methods, mean filling, etc. can be used for processing. Associate and integrate different types of data. For example, align the flight state data, performance data, and fault type data according to the timestamp so that each set of data can correspond to the complete information of the aircraft at a certain moment. In this way, the internal connection between the data can be established. Extract the features that are of great significance for flight state testing and fault diagnosis from the integrated data. For example, calculate the change rate of the flight state data (such as the change rate of speed, the change rate of attitude angle) to reflect the dynamic characteristics of the aircraft; calculate the ratio of some performance indicators (such as the ratio of power to weight) according to the performance data to evaluate the performance efficiency of the aircraft. These features will be used as the input of the model for training and testing the flight state testing model. According to the extracted features, divide the data into independent samples. Each sample contains a set of input features (such as flight state features, performance features) and the corresponding output label (such as fault type or no-fault identification).

[0128] To obtain the training data set and the test sample data set, the sample data can be randomly sampled; random sampling is to ensure that the training data set and the test sample data set are representative and can reflect the distribution of the entire sample data. In this way, the problems of overfitting or underfitting of the model during the training process can be avoided, and the generalization ability of the model can be improved. Usually, the sample data will be divided into the training data set and the test sample data set according to a certain ratio, and the common division ratios are 7:3, 8:2, etc. For example, if there are 1000 samples and they are divided according to the ratio of 7:3, the training data set contains 700 samples, and the test sample data set contains 300 samples.

[0129] Specifically, the flight state data and performance data of the unmanned aircraft at the i-th moment can be represented by the vector X i , X i = {x i1 , x i2 , …, x ij} = {v i , ψ i , φ i , θ i , h i , U i , I i , T i , S i , Λ i , h maxi , v maxi , t i}; The test result at this moment can be expressed as Y i ={y i1 , y i2 , …, y ij}, then {X i , Y i} forms a sample;

[0130] After obtaining the training data set and the test sample data set, the training data set is input into the flight state test model for processing to obtain a training result data set. Among them, the flight state test model is a machine learning-based model, including an input layer, a hidden layer, and an output layer. The input layer receives the sample features in the training data set, the hidden layer performs non-linear transformation and feature extraction on the input features, and the output layer gives the prediction result of the model. The model processing process includes: sequentially inputting each sample in the training data set into the flight state test model. The model calculates and processes the features through the node function according to the input features, transmits information layer by layer, and finally obtains the prediction result of each sample in the output layer. These prediction results constitute the training result data set. The node function parameters determine the processing method and mapping relationship of the model to the input features. Different parameter values will cause the model to produce different outputs for the same input. During the training process, these parameters need to be continuously adjusted to make the prediction result of the model as close as possible to the true label.

[0131] Specifically, first perform normalization processing on the input vector X i :

[0132]

[0133] In the formula, X max is the maximum value in a certain sample index; X min is the minimum value in a certain sample index.

[0134] The input layer of the flight state test model passes through the first node function:

[0135]

[0136] to obtain the output result of the input layer and input it to the hidden layer; the hidden layer passes through the second node function:

[0137]

[0138] to obtain the output result of the hidden layer and input it to the output layer; the output layer passes through the third node function:

[0139]

[0140] Obtain the output result of the flight state test model. The output result of this model is the test data, which is also a multi-dimensional vector, and the elements therein respectively represent the time of fault occurrence, fault phenomenon, fault code, fault type, etc.; in the formula, ω j and V j are preset weights, and r j and λ j are preset thresholds.

[0141] The correction error of the output layer is:

[0142] d i =(o 2j -y j )y j (1 - y j )

[0143] The correction error of the hidden layer is:

[0144]

[0145] Update the preset weights and preset thresholds according to the correction error, and obtain:

[0146] V j+1 =V j +ad i y j

[0147] r j+1 =r j +ad i

[0148] ω j+1 =ω j +be j x i

[0149] λ j+1 =λ j +be j

[0150] In the formula: a and b are the corresponding learning rates, 0 < a < 1, 0 < b < 1.

[0151] Compare the predicted category of each sample in the training result dataset with the corresponding true category in the test sample dataset, and calculate metrics such as classification accuracy, recall rate, and F1 value. Train j learning samples. When the global error of the network is less than the limit value, the output of the network approximates the historical fault type data, and the training ends to obtain the trained flight state test model.

[0152] A specific embodiment of the unmanned aerial vehicle test method provided by the embodiment of the present invention is:

[0153] Step 1: According to the preset route data, convert the longitude and latitude coordinates into plane coordinates to obtain waypoint coordinate data; according to the waypoint coordinate data, obtain flight speed control data and flight attitude angle control data; according to the waypoint coordinate data, the flight speed control data and the flight attitude angle control data, generate a test instruction and send it to the unmanned aerial vehicle.

[0154] Step 2: According to multiple sensors installed on the unmanned aerial vehicle, obtain the attitude data, altitude data, speed data and battery status data of the unmanned aerial vehicle; according to multiple sensors installed on the unmanned aerial vehicle, obtain at least one of the attitude data, altitude data, speed data and battery status data of the unmanned aerial vehicle; according to multiple sensors installed on the unmanned aerial vehicle, obtain at least one of the attitude data, altitude data, speed data and battery status data of the unmanned aerial vehicle.

[0155] Step 3: According to the battery status data and the speed data, obtain the endurance time and the climb rate; according to the endurance time data and the speed data, obtain the maximum range; according to the altitude data, the battery status data and the speed data, obtain the maximum flight altitude; according to the battery status data and the attitude data, obtain the maximum flight speed.

[0156] Step 4: Process the flight state data and the performance data to obtain an input vector, where the i-th input vector can be expressed as: X i ={x i1 , x i2 , …, x ij}={v i , ψ i , φ i , θ i , h i , U i , I i , T i , S i , Λ i , h maxi , v maxi , t i}; Input the input vector into the flight state test model for processing to obtain test data, where the flight state test model is a neural network including an input layer, a hidden layer and an output layer, as Figure 2 shown; The specific process of data processing includes:

[0157] First, perform normalization processing on the input vector X i :

[0158]

[0159] In the formula, Xmax is the maximum value in a certain sample index; X min is the minimum value in a certain sample index.

[0160] The input layer of the flight state test model passes through the first node function:

[0161]

[0162] obtains the output result of the input layer and inputs it to the hidden layer; the hidden layer passes through the second node function:

[0163]

[0164] obtains the output result of the hidden layer and inputs it to the output layer; the output layer passes through the third node function:

[0165]

[0166] obtains the output result of the flight state test model, and the output result of this model is the test data, and the test data is also a multi-dimensional vector, and the elements in it respectively represent the time of fault occurrence, fault phenomenon, fault code, fault type, etc.; in the formula, ω j , V j are preset weights, and r j , λ j are preset thresholds.

[0167] Through the above technical solution proposed by the present invention, by generating test instructions and sending them to the unmanned aerial vehicle, the flight state data thereof is obtained in real time, and then these data are processed to obtain at least one of the key performance data including endurance time, climb rate, maximum range, maximum flight altitude, maximum flight speed, etc., realizing a comprehensive quantitative evaluation of the performance of the unmanned aerial vehicle. In addition, the present invention also inputs the flight state data and performance data into the flight state test model for processing, and utilizes the comprehensive analysis ability of this model to obtain more accurate and comprehensive test data. This technical solution not only improves the automation degree of the test process, reduces human intervention, but also provides strong support for the performance optimization and improvement of the unmanned aerial vehicle in a data-driven manner, helping to improve the overall performance and reliability of the unmanned aerial vehicle.

[0168] As Figure 3 shown, the embodiment of the present invention also provides a test device 30 for an unmanned aerial vehicle, including:

[0169] A generation module 31 for generating test instructions and sending them to the unmanned aerial vehicle;

[0170] An acquisition module 32 for acquiring the flight state data of the unmanned aerial vehicle;

[0171] The processing module 33 is configured to process the flight status data to obtain performance data; input the flight status data and the performance data into a flight status test model for processing to obtain test data.

[0172] Optionally, the generating module 31 is specifically configured to:

[0173] Obtain waypoint coordinate data according to preset route data;

[0174] Obtain flight speed control data and flight attitude angle control data according to the waypoint coordinate data;

[0175] Generate a test instruction according to the waypoint coordinate data, the flight speed control data, and the flight attitude angle control data, and send it to the unmanned aerial vehicle.

[0176] Optionally, the obtaining module 32 is specifically configured to:

[0177] Obtain at least one of the attitude data, altitude data, speed data, and battery status data of the unmanned aerial vehicle according to a plurality of sensors installed on the unmanned aerial vehicle;

[0178] Perform anomaly detection processing on at least one of the attitude data, altitude data, speed data, and battery status data to obtain a plurality of detection data;

[0179] Merge and process the plurality of detection data according to a preset format to obtain the flight status data of the unmanned aerial vehicle.

[0180] Optionally, the processing module 33 is specifically configured to:

[0181] Obtain the endurance time and the climb rate according to the battery status data and the speed data;

[0182] Obtain the maximum range according to the endurance time data and the speed data;

[0183] Obtain the maximum flight altitude according to the altitude data, the battery status data, and the speed data;

[0184] Obtain the maximum flight speed according to the battery status data and the attitude data.

[0185] Optionally, the processing module 33 is further specifically configured to:

[0186] Process the flight status data and the performance data to obtain an input vector;

[0187] Input the input vector into the flight state test model for processing to obtain test data; the flight state test model processes the input vector through a node function to obtain an intermediate result, and performs deviation correction on the intermediate result to obtain test data.

[0188] Optionally, the flight state test model is trained through the following process:

[0189] Obtain historical flight state data, historical performance data, and historical fault type data;

[0190] Perform induction processing on the historical flight state data, the historical performance data, and the historical fault type data to obtain sample data;

[0191] Perform random sampling processing on the sample data to obtain a training data set and a test sample data set;

[0192] Input the training data set into the flight state test model for processing to obtain a training result data set;

[0193] Compare the training result data set with the test sample data set to obtain a comparison result;

[0194] Adjust the node function parameters of the flight state test model according to the comparison result to obtain a trained flight state test model.

[0195] Optionally, the device 30 further includes:

[0196] An output module 34 for outputting a test report of the unmanned aerial vehicle, where the test report of the unmanned aerial vehicle includes: performance data and test data of the unmanned aerial vehicle.

[0197] It should be noted that this device corresponds to the above method, and all implementation manners in the above method embodiments are applicable to this embodiment and can achieve the same technical effects.

[0198] As Figure 4 shown, an embodiment of the present invention further provides a computing device 40, including a processor 41, a memory 42, a program or instruction stored on the memory 42 and executable on the processor 41. When the program or instruction is executed by the processor 41, each process of the above-mentioned unmanned aerial vehicle test method embodiment is implemented and can achieve the same technical effects. To avoid repetition, it will not be elaborated here. It should be noted that the computing device in the embodiment of the present invention includes the above-mentioned mobile electronic device and non-mobile electronic device.

[0199] Those of ordinary skill in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are executed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of the present invention.

[0200] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the systems, devices, and units described above can refer to the corresponding processes in the foregoing method embodiments and will not be elaborated herein.

[0201] In the embodiments provided by the present invention, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed coupling or direct coupling or communication connection between each other can be through some interfaces. The indirect coupling or communication connection of devices or units can be in electrical, mechanical, or other forms.

[0202] The units described as separate components may or may not be physically separated. The components displayed as units may or may not be physical units, that is, they can be located in one place, or distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0203] In addition, the functional units in each embodiment of the present invention can be integrated into one processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit.

[0204] If the described 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, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods in each embodiment of the present invention. The foregoing storage medium includes: various media such as USB flash drives, mobile hard disks, ROM, RAM, magnetic disks, or optical discs that can store program codes.

[0205] In addition, it should be noted that in the device and method of the present invention, obviously, each component or each step can be decomposed and / or recombined. These decompositions and / or recombinations shall be regarded as equivalent solutions of the present invention. Moreover, the steps of performing the above series of processes can naturally be executed in chronological order according to the described order, but it is not necessary to be executed in chronological order. Some steps can be executed in parallel or independently of each other. For those of ordinary skill in the art, it is understandable that all or any steps or components of the method and device of the present invention can be implemented in any computing device (including processors, storage media, etc.) or a network of computing devices in the form of hardware, firmware, software, or a combination thereof, which can be achieved by those of ordinary skill in the art using their basic programming skills after reading the description of the present invention.

[0206] Therefore, the object of the present invention can also be achieved by running a program or a set of programs on any computing device. The computing device can be a well-known general-purpose device. Therefore, the object of the present invention can also be achieved only by providing a program product containing program code for implementing the method or device. That is to say, such a program product also constitutes the present invention, and a storage medium storing such a program product also constitutes the present invention. Obviously, the storage medium can be any well-known storage medium or any storage medium developed in the future. It should also be noted that in the device and method of the present invention, obviously, each component or each step can be decomposed and / or recombined. These decompositions and / or recombinations shall be regarded as equivalent solutions of the present invention. Moreover, the steps of performing the above series of processes can naturally be executed in chronological order according to the described order, but it is not necessary to be executed in chronological order. Some steps can be executed in parallel or independently of each other.

[0207] The above are the preferred embodiments of the present invention. It should be pointed out that for those of ordinary skill in the art, without departing from the principle of the present invention, several improvements and refinements can be made, and these improvements and refinements should also be regarded as the protection scope of the present invention.

Claims

1. A method for testing an unmanned aerial vehicle, characterized in that: include: Generate test instructions and send them to the UAV; Obtain flight status data of unmanned aerial vehicles; Processing the flight status data to obtain performance data, wherein the performance data includes at least one of endurance time, climb rate, maximum range, maximum flight altitude, and maximum flight speed; The flight status data and the performance data are input into a flight status test model for processing to obtain test data.

2. The unmanned aerial vehicle testing method according to claim 1, characterized in that: The step of generating a test instruction and sending the instruction to the unmanned aerial vehicle comprises: According to the preset route data, the waypoint coordinate data is obtained; According to the waypoint coordinate data, flight speed control data and flight attitude angle control data are obtained; A test instruction is generated according to the waypoint coordinate data, the flight speed control data and the flight attitude angle control data and sent to the unmanned aerial vehicle.

3. The unmanned aerial vehicle testing method according to claim 1, characterized in that: The step of obtaining the flight status data of the unmanned aerial vehicle includes: Acquire at least one of attitude data, altitude data, speed data, and battery status data of the unmanned aerial vehicle according to a plurality of sensors installed on the unmanned aerial vehicle; Performing abnormality detection processing on at least one of the posture data, the height data, the speed data, and the battery status data to obtain a plurality of detection data; The multiple detection data are combined and processed according to a preset format to obtain the flight status data of the unmanned aerial vehicle.

4. The unmanned aerial vehicle testing method according to claim 3, characterized in that: The flight status data is processed to obtain performance data, including: Obtaining the endurance time and climbing rate according to the battery status data and the speed data; Obtaining a maximum range according to the endurance time data and the speed data; Obtaining a maximum flight altitude according to the altitude data, the battery status data and the speed data; A maximum flight speed is obtained according to the battery status data and the attitude data.

5. The unmanned aerial vehicle testing method according to claim 1, characterized in that: The flight status data and the performance data are input into a flight status test model for processing to obtain test data, including: Processing the flight status data and the performance data to obtain an input vector; The input vector is input into the flight status test model for processing to obtain test data; the flight status test model processes the input vector through a node function to obtain an intermediate result, and performs deviation correction on the intermediate result to obtain test data.

6. The unmanned aerial vehicle testing method according to claim 1, characterized in that: The flight state test model is trained through the following process: Obtain historical flight status data, historical performance data, and historical fault type data; Summarizing the historical flight status data, the historical performance data, and the historical fault type data to obtain sample data; Randomly extract the sample data to obtain a training data set and a test sample data set; Inputting the training data set into the flight state test model for processing to obtain a training result data set; Comparing the training result data set with the test sample data set to obtain a comparison result; According to the comparison result, the node function parameters of the flight status test model are adjusted to obtain a trained flight status test model.

7. The unmanned aerial vehicle testing method according to claim 1, characterized in that: The method further comprises: Outputting an unmanned aerial vehicle test report, wherein the unmanned aerial vehicle test report includes: performance data and test data of the unmanned aerial vehicle.

8. An unmanned aerial vehicle testing device, characterized in that: include: A generation module, used for generating test instructions and sending them to the unmanned aerial vehicle; An acquisition module, used to obtain flight status data of the unmanned aerial vehicle; The processing module is used to process the flight status data to obtain performance data; the flight status data and the performance data are input into a flight status test model for processing to obtain test data.

9. A computing device, characterized in that include: A processor and a memory storing a computer program, wherein when the computer program is executed by the processor, the method according to any one of claims 1 to 7 is performed.

10. A computer-readable storage medium, characterized in that: Instructions are stored, and when the instructions are executed on a computer, the computer is caused to execute the method according to any one of claims 1 to 7.

Images