A multi-rotor unmanned aerial vehicle endurance test method and device

Abstract

The application relates to the technical field of multi-rotor unmanned aerial vehicles, in particular to a multi-rotor unmanned aerial vehicle endurance mileage test method and equipment. The method comprises the following steps: performing flight test of the multi-rotor unmanned aerial vehicle at an outfield site, recording a unit mileage energy consumption measured value and a first influence factor measured value; performing flight test of the multi-rotor unmanned aerial vehicle according to standard benchmark conditions and benchmark test working conditions, recording a unit mileage energy consumption true value; calculating a correction coefficient true value; training a neural network model; performing flight test of the multi-rotor unmanned aerial vehicle according to the benchmark test working conditions at the outfield site, and recording a second influence factor measured value and an endurance mileage measured value; inputting the second influence factor measured value into the neural network model to obtain a correction coefficient, and calculating a standard endurance mileage of the multi-rotor unmanned aerial vehicle under the standard benchmark conditions and the benchmark test working conditions. The application overcomes the defects of the prior art, such as strict environmental requirements, poor test operability and neglect of multi-factor nonlinear coupling.

Description

Technical Field

[0001] This application relates to the field of multi-rotor unmanned aerial vehicle (UAV) technology, and more specifically, to a method and device for testing the flight range of a multi-rotor UAV. Background Technology

[0002] Flight range is a core performance indicator for multi-rotor drones, directly determining their operational radius and mission capabilities. Currently, there are three major pain points in the industry's flight range testing and labeling: First, testing environmental conditions cannot be completely standardized. Test results from different institutions, at different times, and in different locations vary significantly, making direct benchmarking impossible. Drones are extremely sensitive to air density, wind resistance, and battery temperature characteristics; the flight range of the same model can vary by more than 50% under different environments. The constant temperature, calm wind, and sea-level conditions required by existing standards can only be achieved in laboratory environmental chambers, which cannot be met in field testing. Second, existing testing methods have extremely poor operability. Existing solutions require performing a complete flight range test until the drone's low battery protection activates, with a single test lasting tens of minutes. This not only results in significant battery depletion but also makes it difficult to maintain a stable field environment throughout the test, compromising consistency of operating conditions. The test data itself is subject to environmental interference, making engineering implementation extremely difficult. Third, existing correction methods ignore the nonlinear coupling effect between multiple factors and the interactive amplification effect between factors. For example, the power consumption increase caused by the increase in load under strong wind conditions is much greater than the power consumption change of the same load increase under windless conditions, resulting in serious insufficient correction accuracy and even reverse deviation.

[0003] Therefore, there is an urgent need for a standardized correction method for range that does not require strict control of the test environment, does not require flying the entire range, and can accurately fit the coupling effect of multiple factors, so as to achieve benchmarking, traceability and reproducibility of test results under different field conditions. Summary of the Invention

[0004] The purpose of this application is to provide a method and equipment for testing the endurance of multi-rotor drones, aiming to overcome the shortcomings of existing technologies, such as stringent environmental requirements, poor test operability, and neglect of multi-factor nonlinear coupling.

[0005] To achieve the above objectives, this application adopts the following technical solution: Firstly, this application provides a method for testing the flight range of a multi-rotor unmanned aerial vehicle (UAV), including: Define the standard baseline conditions and benchmark test conditions for testing the endurance of multi-rotor UAVs; the standard baseline conditions include environmental conditions and load baselines. Based on the aforementioned standard benchmark conditions, multiple sets of test cases are generated; For each set of test cases, flight tests of the multi-rotor UAV were conducted in the field according to the benchmark test conditions, and the measured values ​​of energy consumption per unit distance and the measured values ​​of the first influencing factor were recorded. Perform flight tests on the multi-rotor UAV according to standard benchmark conditions and benchmark test conditions, and record the true value of energy consumption per unit distance. Based on the measured value of energy consumption per unit mileage and the true value of energy consumption per unit mileage, calculate the true value of the correction coefficient for each test case. The measured value of the first influencing factor is used as the input training sample, and the true value of the correction coefficient is used as the label to train the neural network model. Flight tests of the multi-rotor UAV were conducted at the field site according to the benchmark test conditions, and the measured values ​​of the second influencing factor and the measured values ​​of the flight range were recorded. The measured value of the second influencing factor is input into the trained neural network model to obtain the correction coefficient; Based on the measured range and correction factor, the standard range of the multi-rotor UAV under standard reference conditions and benchmark test conditions is calculated.

[0006] Secondly, this application provides an electronic device, comprising: At least one processor, and a memory communicatively connected to at least one of the processors; The memory stores instructions that can be executed by at least one of the processors, which enable the at least one processor to perform a multi-rotor drone endurance testing method.

[0007] This application has the following substantial features and significant progress compared to the prior art.

[0008] 1. This application solves the industry pain point of poor operability of field testing: there is no need to fly the full range in a standard environment. Only 20 sets of short, lightweight tests of about 10 minutes each are needed to complete model training. A single test does not require depleting the battery, which greatly reduces the testing threshold, time cost and battery consumption. 2. The first application of GA-BP neural network to multi-factor coupling correction of UAV endurance: It automatically learns the nonlinear coupling relationship between multiple factors through data-driven approach, without the need for manual derivation of complex physical formulas, avoiding errors and omissions of coupling terms in manual modeling, and the correction accuracy is much higher than that of traditional solutions. 3. Full boundary coverage of samples is achieved through Latin hypercube sampling: the generalization ability of the model is guaranteed with the fewest number of tests, avoiding sample redundancy and coverage blind spots. The model can achieve high-precision correction under any combination of environments within the applicable boundary. 4. A complete closed-loop system of "sample design - data collection - model training - correction and application - verification and iteration" has been constructed, which can be directly applied to all scenarios such as drone product labeling, quality inspection and spot checks, and model benchmarking. Attached Figure Description

[0009] To more clearly illustrate the technical solutions in the specific embodiments of this application or the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0010] Figure 1 This is a flowchart of a method for testing the endurance of a multi-rotor unmanned aerial vehicle (UAV) according to an embodiment of this application; Figure 2 This is a schematic diagram of the structure of the electronic device provided in the embodiments of this application. Detailed Implementation

[0011] The following description, in conjunction with the accompanying drawings, illustrates exemplary embodiments of this application, including various details to aid understanding. These should be considered merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of this application. Similarly, for clarity and brevity, descriptions of well-known functions and structures are omitted in the following description.

[0012] This application aims to overcome the shortcomings of existing technologies, such as stringent environmental requirements, poor test operability, and neglect of multi-factor nonlinear coupling, by providing a standardized correction method for the range of multi-rotor UAVs based on a GA-BP neural network. The GA-BP neural network is a hybrid algorithm that combines the global search capability of a genetic algorithm (GA) with the local search and learning capability of a backpropagation (BP) neural network, aiming to solve the problem of traditional BP neural networks easily getting trapped in local optima. Through Latin hypercube sampling design for controlled variable testing, lightweight sample collection, and the construction of a multi-factor coupled correction model using a GA-BP neural network, only a limited number of short-duration tests are needed to accurately correct the measured range under any outdoor environment to the standard range under a unified standard benchmark, completely solving the industry problems of poor operability and insufficient correction accuracy in outdoor testing.

[0013] Figure 1 This is a flowchart illustrating a method for testing the endurance of a multi-rotor drone, provided in this embodiment. It is applicable to field endurance testing and standardization benchmarking of electric multi-rotor drones with a maximum takeoff weight not exceeding 25 kg. See also... Figure 1 The method provided in this embodiment includes: S110 defines the standard baseline conditions and baseline test conditions for testing the endurance of multi-rotor UAVs; the standard baseline conditions include environmental conditions and load baselines.

[0014] Referring to standard atmospheric specifications, standard baseline conditions are determined as the benchmark for correcting all test results, specifically including: Environmental conditions: Altitude 0m, standard atmospheric pressure 101.325kPa, ambient temperature 25℃, relative humidity 55%, windless environment (wind speed 0m / s). Payload baseline: Rated takeoff weight of the UAV, including standard onboard equipment and rated mission payload.

[0015] Benchmark test conditions (with a one-to-one correspondence with the neural network model): These are fixed and unique test conditions, including: hovering condition, constant speed cruise condition, and designated mission profile condition. The hovering condition for multi-rotor UAVs refers to the UAV maintaining a stationary position, altitude, and heading in the air. In this condition, the total lift generated by the rotors is precisely equal to its own weight, making it the fundamental condition for measuring the stability of the flight control system. The constant speed cruise condition refers to the UAV maintaining a stable state of constant speed in level flight. This is typically aimed at optimizing endurance, so the flight speed is set near the "economic speed" that minimizes energy consumption. The designated mission profile condition is a complex sequence of actions performed by the UAV to complete a specific operational objective. It organically combines basic flight maneuvers such as hovering, climbing, descending, turning, and constant speed cruise according to mission requirements. For example, a typical reconnaissance or logistics mission profile might include a complete sequence of steps such as takeoff, climb, cruise to the target area, hovering operation, return cruise, and descent landing. In actual testing, it is sufficient to select one test condition as the benchmark. Subsequent test cases and tests on multi-rotor drones will all use the benchmark test condition.

[0016] S120. Generate multiple sets of test cases based on the aforementioned standard benchmark conditions.

[0017] Based on the aforementioned standard baseline conditions, multiple test cases are generated using the Latin hypercube orthogonal sampling method.

[0018] This step, targeting the S110 benchmark test condition, employs the Latin hypercube orthogonal sampling method to design a control variable test scheme, covering all factor combinations within the applicable correction boundary with the fewest possible test runs. The core influencing factors and their level ranges are determined as follows: Factor 1: Ambient temperature, within the rated operating temperature range of lithium batteries (-10℃~45℃). Factor 2: Ambient wind speed, horizontal range from 0 to 70% of the tested drone's nominal maximum wind resistance rating; Factor 3: Altitude, horizontal range of 0~3000m; Factor 4: Takeoff weight, ranging from empty to maximum rated takeoff weight; Each factor was uniformly stratified into five levels. Twenty unique factor combinations were generated using Latin hypercube sampling, serving as test cases for the control variable test. This ensured uniform sample coverage of the entire applicable boundary with no blind spots. The influencing factors for each test case included ambient temperature, ambient wind speed, altitude, and load; the value ranges of these influencing factors varied between different test cases.

[0019] S130. For each set of test cases, perform flight tests of the multi-rotor UAV in the field according to the benchmark test conditions, and record the measured value of energy consumption per unit distance and the measured value of the first influencing factor.

[0020] Perform short-duration lightweight testing at a compliant outdoor site. Full-range flight is not required; only stable flight testing under baseline conditions is needed to collect training sample data. Specific requirements are as follows: Each test flight lasted 10-15 minutes, and the flight process strictly followed the benchmark test conditions to maintain stable flight status. Three types of core data were collected simultaneously: Flight energy consumption data: Energy consumption per unit distance, power consumption per unit time, flight speed and attitude angle sequence for this test; The first influencing factor measured values: average ambient temperature, average ambient wind speed and direction, altitude, atmospheric pressure, average relative humidity, etc. throughout the test, as well as the actual takeoff weight of the UAV (i.e., payload).

[0021] S140. Perform flight tests on the multi-rotor UAV according to standard reference conditions and benchmark test conditions, and record the true value of energy consumption per unit distance.

[0022] Design standard benchmark conditions and control the multi-rotor UAV to conduct flight tests according to benchmark test conditions; during the flight, collect the true value of energy consumption per unit distance as the benchmark true value for model calibration, and then obtain a complete model training sample set.

[0023] S150. Based on the measured value of energy consumption per unit mileage and the true value of energy consumption per unit mileage, calculate the true value of the correction coefficient for each test case.

[0024] Optionally, prior to S150, the recorded values ​​from the flight tests need to be validated and normalized. Validation includes: 1) Flight condition compliance verification: The deviation between the actual flight speed and altitude of the multi-rotor UAV and the speed and altitude corresponding to the benchmark test conditions is within the allowable range; 2) Environmental stability verification: During the test, the fluctuation of environmental parameters shall not exceed ±10%.

[0025] 3) Equipment status compliance verification: There are no abnormal alarms from the battery, motor, and flight controller, and there are no abnormal situations such as loss of control or forced landing during flight.

[0026] The recorded values ​​(mainly the measured values ​​of influencing factors) are normalized and mapped to the [0,1] interval to eliminate the influence of the difference in the dimensions of different factors.

[0027] Optionally, the true value of energy consumption per unit mileage can be divided by the measured value of energy consumption per unit mileage to obtain the true value of the correction coefficient for each test case.

[0028] S160. Use the measured value of the first influencing factor as the input training sample and the true value of the correction coefficient as the label to train the neural network model.

[0029] Optionally, the neural network model is a BP network model, and a genetic algorithm is used to optimize the parameters in the neural network model, ultimately constructing a multi-factor coupled correction model. The BP neural network structure is designed as follows: Input layer: 4 neurons, corresponding to the 4 core influencing factors of ambient temperature, ambient wind speed, altitude and load; Hidden layer: adopts a single hidden layer structure, the number of neurons is determined to be 12 through trial and error, and the activation function is the tansig function; Output layer: 1 neuron, corresponding to the correction coefficient of the driving range under the environmental combination and benchmark test conditions.

[0030] The optimization process of a genetic algorithm includes: Real-number encoding is used, with the encoding length equal to the total number of weights and thresholds in the BP network. The fitness function is the reciprocal of the mean square error between the predicted value and the true value of the correction coefficients in the BP network output; the smaller the mean square error, the higher the fitness. The population size is set to 20, the number of generations to 100, the crossover probability to 0.7, and the mutation probability to 0.01. Optimal initial weights and thresholds are obtained through selection, crossover, and mutation operations.

[0031] The preprocessed sample set was divided into a training set and a test set in an 8:2 ratio. The optimal weights and thresholds optimized by GA were assigned to the BP network. The trainlm algorithm was used to train the network, with the training objective error set to 0.001 and the maximum number of iterations to 1000. Leave-one-out cross-validation and test set validation were used to ensure the generalization ability of the model. The model was considered qualified if the prediction accuracy on the test set was ≥95%.

[0032] S170. Conduct flight tests of the multi-rotor UAV at the outdoor site according to the benchmark test conditions, and record the measured values ​​of the second influencing factor and the measured values ​​of the flight range.

[0033] S180. Input the measured value of the second influencing factor into the trained neural network model to obtain the correction coefficient.

[0034] S190. Calculate the standard range of the multi-rotor UAV under standard reference conditions and reference test conditions based on the measured range and correction coefficient.

[0035] The following steps, based on the measured flight range of a multi-rotor drone in any compliant outdoor environment, utilize a trained BP neural network model to perform standardized corrections. This process includes: Field endurance tests were conducted, strictly adhering to the benchmark test conditions specified in S110. Measured endurance values ​​were recorded, and measured values ​​of secondary influencing factors were simultaneously collected throughout the test, including average ambient temperature, wind speed, altitude, takeoff weight, and relative humidity. The collected measured values ​​of these secondary influencing factors were input into a trained BP neural network model. The model automatically outputs the correction coefficient K under the benchmark test conditions and influencing factors. The standard endurance of the multi-rotor UAV under standard benchmark conditions and benchmark test conditions was calculated using the following formula: Standard driving range = measured driving range × correction factor K; Optionally, multiple sets (more than 3 sets) of parallel tests can be conducted under different environmental conditions, and the corrected standard driving range can be calculated for each set. If the relative deviation of the standard driving range of each set is not greater than the set threshold (5%), the correction is deemed effective.

[0036] When the hardware parameters of a multi-rotor drone change, or the test environment exceeds the applicable boundaries of the original model, supplement the corresponding sample data, retrain the model, and complete iterative optimization. The applicable boundaries of the model are clearly defined. Test data exceeding the following ranges are not applicable to this application: average wind speed exceeding 70% of the nominal maximum wind resistance level of the tested UAV; ambient temperature exceeding the rated operating temperature range of the lithium battery (-10℃~45℃); takeoff weight exceeding the maximum rated takeoff weight of the UAV; altitude exceeding 3000m.

[0037] The following describes in detail one embodiment of this application, taking the 8m / s constant speed cruise condition of a civilian multi-rotor mapping UAV as an example.

[0038] 1. Scene recognition.

[0039] The multi-rotor UAV under test is a quadcopter mapping UAV with a rated takeoff weight of 5kg, a nominal rated payload of 1kg, and a nominal maximum wind resistance of 8m / s. A standardized correction model of the flight range of this model under the constant speed cruise condition of 8m / s needs to be constructed to achieve unified benchmarking of the field test results.

[0040] 2. Definition of standard reference conditions.

[0041] The correction baseline for this model is as follows: altitude 0m, atmospheric pressure 101.325kPa, temperature 25℃, relative humidity 55%, wind speed 0m / s, rated load (takeoff weight) 5kg. The baseline test condition is a constant speed cruise condition of 8m / s, and the model is bound to this condition.

[0042] 3. Test design with controlled variables.

[0043] Determine the level ranges for the four influencing factors: Ambient temperature: -10℃~45℃, 5 levels: -10℃, 5℃, 20℃, 35℃, 45℃; Ambient wind speed: 0~6m / s (70% of the maximum wind resistance level), 5 levels: 0m / s, 1m / s, 2m / s, 4m / s, 6m / s; Altitude: 0~3000m, 5 levels: 0m, 750m, 1500m, 2250m, 3000m; Load capacity: 250g~25kg, 5 levels: 250g, 1kg, 7kg, 15kg, 25kg; Twenty test cases were generated using Latin hypercube sampling, covering all applicable boundaries.

[0044] 4. Lightweight sample collection.

[0045] For 20 test cases, a 12-minute constant speed cruise test at 8m / s was performed. The measured values ​​of energy consumption per unit mileage and the measured values ​​of the first influencing factor were collected for each test. At the same time, the true values ​​of energy consumption per unit mileage under standard baseline conditions and baseline test conditions were collected as the baseline true values.

[0046] 5. Data preprocessing and model training.

[0047] After validating the sample data, normalization was performed. Using the true energy consumption per unit mileage under standard baseline conditions and benchmark test conditions as the benchmark, an energy consumption correction coefficient was calculated for each sample group as the label. A GA-BP neural network model was constructed: 4 neurons in the input layer, 12 neurons in the hidden layer, and 1 neuron in the output layer. The GA population size was set to 20, the number of generations to 100, the crossover probability to 0.7, and the mutation probability to 0.01. After optimizing the weight thresholds, the BP network was trained. After training, the prediction accuracy on the test set was 96.2%, meeting the accuracy requirement of ≥95%, and the model was deemed qualified.

[0048] 6. Field measurement correction.

[0049] Range tests were conducted under typical field testing conditions, with simultaneous recording of ambient temperature, wind speed, altitude, load, and measured range. This set of field test data was then input into a pre-trained GA-BP correction model, which automatically outputs correction coefficients under those conditions. Based on these correction coefficients, the measured range was converted to standard baseline conditions to obtain the standard range.

[0050] 7. Verification.

[0051] The corrected standard driving range was compared with the driving range tested under standard reference conditions. The deviation was within the allowable range, verifying the effectiveness and engineering applicability of the correction method in this application.

[0052] For example, tests were conducted at an outdoor site at an altitude of 1000m, with an ambient temperature of 28℃, an average headwind speed of 3m / s, a relative humidity of 40%, and an actual takeoff weight of 5kg. A complete range test was performed, achieving a measured range of 24.2km. The environmental and payload parameters were input into a trained BP neural network model, which output a correction coefficient K=1.35. The calculated standard range was approximately 24.2km × 1.35 ≈ 32.7km, a deviation of 4.8% from the environmental chamber test result of 31.2km, which is within the acceptable range, indicating the correction is effective.

[0053] The embodiments of this application have the following technical effects: 1. This application solves the industry pain point of poor operability of field testing: there is no need to fly the full range in a standard environment. Only 20 sets of short, lightweight tests of about 10 minutes each are needed to complete model training. A single test does not require depleting the battery, which greatly reduces the testing threshold, time cost and battery consumption. 2. The first application of GA-BP neural network to multi-factor coupling correction of UAV endurance: It automatically learns the nonlinear coupling relationship between multiple factors through data-driven approach, without the need for manual derivation of complex physical formulas, avoiding errors and omissions of coupling terms in manual modeling, and the correction accuracy is much higher than that of traditional solutions. 3. Full boundary coverage of samples is achieved through Latin hypercube sampling: the generalization ability of the model is guaranteed with the fewest number of tests, avoiding sample redundancy and coverage blind spots. The model can achieve high-precision correction under any combination of environments within the applicable boundary. 4. A complete closed-loop system of "sample design - data collection - model training - correction and application - verification and iteration" has been constructed, which can be directly applied to all scenarios such as drone product labeling, quality inspection and spot checks, and model benchmarking.

[0054] This embodiment provides an electronic device, see [link / reference] Figure 2 It includes at least one processor 301 and a memory 302 communicatively connected to at least one of the processors 301; The memory 302 stores instructions that can be executed by at least one of the processors 301, which enable at least one of the processors 301 to perform the above-described multi-rotor drone endurance testing method, thus having at least the same advantages as the above-described method.

[0055] Optionally, the electronic device also includes interfaces for connecting the various components, including high-speed interfaces and low-speed interfaces. The components are interconnected using different buses and can be mounted on a common motherboard or otherwise installed as needed. The processor can process instructions executed within the electronic device, including instructions stored in or on memory to display graphical information of a GUI (Graphical User Interface) on an external input / output device (such as a display device coupled to the interface). In other embodiments, multiple processors can be used with multiple memories, and / or multiple buses can be used with multiple memories, if desired. Similarly, multiple electronic devices (e.g., as a server array, a group of blade servers, or a multiprocessor system) can be connected, each providing some of the necessary operations.

[0056] The memory 302, as a computer-readable storage medium, can be used to store software programs, computer-executable programs, and modules, such as the program instructions / modules corresponding to the multi-rotor drone endurance testing method in this embodiment. The processor 301 executes various functional applications and data processing of the device by running the software programs, instructions, and modules stored in the memory 302, thereby realizing the aforementioned multi-rotor drone endurance testing method.

[0057] The memory 302 may primarily include a program storage area and a data storage area. The program storage area may store the operating system and at least one application program required for a given function; the data storage area may store data created based on terminal usage. Furthermore, the memory 302 may include high-speed random access memory and non-volatile memory, such as at least one disk storage device, flash memory, or other non-volatile solid-state storage device. In some instances, the memory 302 may further include memory remotely configured relative to the processor, which can be connected to the device via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

[0058] The electronic device may also include an input device 303 and an output device 304. The processor 301, memory 302, input device 303, and output device 304 may be connected via a bus or other means.

[0059] Input device 303 can receive input digital or character information, and output device 304 may include a display device, an auxiliary lighting device (e.g., an LED), and a haptic feedback device (e.g., a vibration motor). The display device may include, but is not limited to, a liquid crystal display (LCD), a light-emitting diode (LED) display, and a plasma display. In some embodiments, the display device may be a touchscreen.

[0060] This embodiment provides a medium storing computer instructions for instructing a computer to perform the methods described above. The computer instructions on this medium, used to instruct the computer to perform the methods described above, thus possess at least the same advantages as the methods described above.

[0061] The medium in this application may be any combination of one or more computer-readable media. The medium may be a computer-readable signal medium or a computer-readable storage medium. The medium may be, for example,, but not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatuses, or devices, or any combination thereof. More specific examples of the medium (a non-exhaustive list) include: electrical connections having one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this document, the medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.

[0062] Computer-readable signal media may include data signals propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. Computer-readable signal media may also be any computer-readable medium other than computer-readable storage media, capable of sending, propagating, or transmitting programs for use by or in connection with an instruction execution system, apparatus, or device.

[0063] Program code contained on a computer-readable medium may be transmitted using any suitable medium, including but not limited to wireless, wire, optical fiber, RF (Radio Frequency), or any suitable combination thereof.

[0064] Computer program code for performing the operations of this invention can be written in one or more programming languages ​​or a combination thereof. Programming languages ​​include object-oriented programming languages ​​such as Java, Smalltalk, and C++, as well as conventional procedural programming languages ​​such as C or similar languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0065] It should be understood that the various forms of processes shown above can be used to rearrange, add, or delete steps. For example, the steps described in this application can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution disclosed in this application can be achieved, and this is not limited herein.

[0066] The specific embodiments described above do not constitute a limitation on the scope of protection of this application. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A method for testing the flight range of a multi-rotor unmanned aerial vehicle (UAV), characterized in that, include: Define the standard benchmark conditions and benchmark test conditions for testing the endurance of multi-rotor UAVs; Standard reference conditions include environmental conditions and load references; Based on the aforementioned standard benchmark conditions, multiple sets of test cases are generated; For each set of test cases, flight tests of the multi-rotor UAV were performed in the field according to the benchmark test conditions. It was not necessary to fly the full range. The measured values ​​of energy consumption per unit distance and the measured values ​​of the first influencing factor were recorded. Perform flight tests on the multi-rotor UAV according to standard benchmark conditions and benchmark test conditions, and record the true value of energy consumption per unit distance. Based on the measured value of energy consumption per unit mileage and the true value of energy consumption per unit mileage, calculate the true value of the correction coefficient for each test case. The measured value of the first influencing factor is used as the input training sample, and the true value of the correction coefficient is used as the label to train the neural network model. Flight tests of the multi-rotor UAV were conducted at the field site according to the benchmark test conditions, and the measured values ​​of the second influencing factor and the measured values ​​of the flight range were recorded. The measured value of the second influencing factor is input into the trained neural network model to obtain the correction coefficient; Based on the measured range and correction factor, calculate the standard range of the multi-rotor UAV under standard reference conditions and benchmark test conditions. The influencing factors for each test case include ambient temperature, ambient wind speed, altitude, and load; the range of values ​​for the influencing factors differs between different test cases.

2. The method according to claim 1, characterized in that, Based on the aforementioned standard baseline conditions, multiple sets of test cases are generated, including: Based on the aforementioned standard baseline conditions, multiple test cases are generated using the Latin hypercube orthogonal sampling method.

3. The method according to claim 2, characterized in that, Before calculating the true value of the correction coefficient for each test case based on the measured energy consumption per unit mileage and the true value of energy consumption per unit mileage, the following steps are also included: The recorded values ​​from the flight tests were validated and normalized. The validity verification includes flight condition compliance verification, environmental stability verification, and equipment status compliance verification.

4. The method according to claim 3, characterized in that, The neural network model is a BP network model.

5. The method according to claim 4, characterized in that, Based on the measured range and correction factor, after calculating the standard range of the multi-rotor UAV under standard reference conditions and benchmark test conditions, the following steps are also included: Multiple parallel tests were conducted under different environmental conditions, and the corrected standard driving range was calculated for each test. If the relative deviation of the standard driving range of each group is not greater than the set threshold, the correction is deemed effective.

6. The method according to claim 5, characterized in that, The benchmark test conditions include: hovering condition, constant speed cruise condition, and specified mission profile condition.

7. The method according to claim 6, characterized in that, Training the neural network model includes: Genetic algorithms are used to optimize the parameters in a neural network model.

8. An electronic device, characterized in that, include: At least one processor, and a memory communicatively connected to at least one of the processors; The memory stores instructions that can be executed by at least one of the processors, which are executed by at least one of the processors to enable the at least one of the processors to perform the multi-rotor drone endurance testing method according to any one of claims 1-7.

Images