Multi-parameter coupling oriented unmanned aerial vehicle metrological test requirement analysis method and system

Abstract

The application discloses a method and system for unmanned aerial vehicle metrological test requirement analysis facing multi-parameter coupling, relates to the technical field of metrological test, and comprises the following steps: acquiring real environment interference parameters when a logistics unmanned aerial vehicle executes a delivery task; constructing a multi-environment coupling simulation test platform to simulate a real flight environment; controlling the logistics unmanned aerial vehicle to execute a predetermined flight action, collecting dynamic change data of key performance parameters of the logistics unmanned aerial vehicle; analyzing a coupling relationship between the real environment interference parameters and the key performance parameters, and establishing a parameter coupling model; and performing test items, test conditions and test indexes in the real flight environment to generate a metrological test requirement analysis report. The application solves the technical problem that, in the prior art, the dynamic change law of key performance parameters under the coupling action of multi-environment parameters cannot be explored, and as a result, the flight safety and performance reliability of an unmanned aerial vehicle in a real environment cannot be guaranteed.

Description

Technical Field

[0001] This application relates to the field of metrology and testing technology, specifically to a method and system for analyzing the metrology and testing requirements of unmanned aerial vehicles (UAVs) with multi-parameter coupling. Background Technology

[0002] With the rapid development of drone technology, delivery missions using drones in complex environments such as high-altitude areas are gradually increasing in the logistics and distribution sector. When logistics drones perform missions in high-altitude areas, they face the combined effects of multiple complex environmental factors. These parameters do not exist independently but are coupled together, jointly affecting the drone's power system, flight performance, and other aspects.

[0003] In existing technologies, the metrological testing requirements analysis for UAVs mainly focuses on the study of single environmental parameters or single performance parameters, lacking in-depth research on the dynamic changes of key performance parameters of UAVs under the coupling effect of multiple environmental parameters. As a result, the test items, test conditions and test indicators formulated based on incomplete test requirements are difficult to effectively guarantee the flight safety and performance reliability of UAVs in real environments. Summary of the Invention

[0004] This application provides a method and system for UAV metrology and testing requirements analysis oriented towards multi-parameter coupling. This solves the technical problem that existing UAV metrology and testing requirements analysis lacks exploration of the dynamic changes of key performance parameters under the coupling of multiple environmental parameters, which leads to the inability to guarantee the flight safety and performance reliability of UAVs in real environments.

[0005] The technical solution to the above-mentioned technical problems in this application is as follows: Firstly, this application provides a method for analyzing the metrological testing requirements of unmanned aerial vehicles (UAVs) with multi-parameter coupling, the method comprising: The low-temperature environment parameters, low-pressure environment parameters, and strong wind environment parameters of logistics drones performing delivery tasks in plateau areas are obtained as real environmental interference parameters. A multi-environment coupling simulation test platform is constructed, wherein the multi-environment coupling simulation test platform is used to simultaneously apply the real environment interference parameters to simulate the real flight environment in which the logistics drone performs delivery tasks; In the multi-environment coupling simulation test platform, the logistics drone is controlled to perform predetermined flight actions, and dynamic change data of multiple key performance parameters of the logistics drone are collected simultaneously. The multiple key performance parameters include tension parameters, rotational speed parameters, current parameters, and voltage parameters. Based on the dynamic change data, the coupling relationship between the real environmental interference parameters and the multiple key performance parameters is analyzed, and a parameter coupling model describing the coupling relationship is established. The parameter coupling model is used to determine the test items, test conditions, and test indicators required for metrological testing of the logistics drone in the real flight environment, and a metrological testing requirements analysis report is generated.

[0006] Secondly, this application provides a UAV metrology and testing requirements analysis system for multi-parameter coupling, including: The platform construction module is used to build a multi-environment coupling simulation test platform, wherein the multi-environment coupling simulation test platform is used to simultaneously apply the real environment interference parameters to simulate the real flight environment in which the logistics drone performs delivery tasks; The flight test module is used to control the logistics drone to perform predetermined flight actions in the multi-environment coupling simulation test platform, and to synchronously collect dynamic change data of multiple key performance parameters of the logistics drone, including tension parameters, rotational speed parameters, current parameters and voltage parameters. The model training module is used to analyze the coupling relationship between the real environmental interference parameters and the multiple key performance parameters based on the dynamically changing data, and to establish a parameter coupling model describing the coupling relationship. The results generation module is used to determine the test items, test conditions, and test indicators required for the metrological testing of the logistics drone in the real flight environment using the parameter coupling model, and to generate a metrological testing requirements analysis report.

[0007] This application provides one or more technical solutions, which have at least the following technical effects or advantages: This application provides a method and system for analyzing the metrological testing requirements of unmanned aerial vehicles (UAVs) oriented towards multi-parameter coupling. First, it acquires low-temperature, low-pressure, and high-wind environmental parameters of a logistics UAV performing delivery missions in high-altitude areas, collectively forming a complex real-world flight environment. Second, it constructs a multi-environment coupling simulation test platform capable of simultaneously applying real-world environmental disturbance parameters, thereby reproducing the real-world flight environment encountered by the logistics UAV during delivery missions in a laboratory setting. Third, within this multi-environment coupling simulation test platform, the logistics UAV is controlled to perform predetermined flight maneuvers, simulating the most severe power output conditions the UAV might encounter in high-altitude areas. Simultaneously, dynamic change data of the UAV's thrust, rotational speed, current, and voltage parameters are collected, serving as the primary basis for analyzing the coupling relationship between environmental and performance parameters. Then, based on the collected dynamic change data, parameter pairs with strong coupling relationships between real-world environmental disturbance parameters and multiple key performance parameters are identified. A parameter coupling model is constructed to describe the nonlinear response laws of key performance parameters under different combinations of environmental parameters. Finally, the metrological testing requirements analysis report for logistics drones in real flight environments was determined, which effectively ensured the scientific nature and relevance of test items, test conditions and test indicators, and improved the ability to guarantee the flight safety and performance reliability of drones in complex real environments such as plateaus.

[0008] Through the above technical solution, this application breaks through the limitations of traditional single-parameter testing by dynamically coupling and analyzing real environmental interference parameters with key performance parameters of UAVs. It provides a comprehensive demand analysis framework for metrological testing of logistics UAVs in complex environments, which helps to promote the development of UAV metrological testing technology towards multi-parameter collaboration and real environment simulation. Attached Figure Description

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

[0010] Figure 1 This is a flowchart illustrating the UAV metrology and testing requirements analysis method for multi-parameter coupling provided in the embodiments of this application; Figure 2 This is a schematic diagram of the structure of the UAV metrology and testing requirements analysis system for multi-parameter coupling provided in the embodiments of this application.

[0011] The components represented by each number in the attached diagram are explained below: Parameter acquisition module 11, platform construction module 12, flight test module 13, model training module 14, and result generation module 15. Detailed Implementation

[0012] This application provides a method and system for UAV metrology and testing requirements analysis oriented towards multi-parameter coupling. This is intended to address the technical problem that existing UAV metrology and testing requirements analysis lacks exploration of the dynamic changes in key performance parameters under the coupling of multiple environmental parameters, which leads to the inability to guarantee the flight safety and performance reliability of UAVs in real-world environments.

[0013] Example 1, as Figure 1 As shown in the embodiments of this application, a method for analyzing the metrological testing requirements of unmanned aerial vehicles (UAVs) oriented towards multi-parameter coupling is provided, including: S10: Acquire low-temperature environmental parameters, low-pressure environmental parameters, and strong-wind environmental parameters when the logistics drone performs delivery tasks in plateau areas, as real environmental interference parameters; In this embodiment, real-world environmental interference parameters are first obtained by collecting real-time environmental data through meteorological monitoring stations deployed along typical flight routes in plateau regions. Simultaneously, airborne environmental sensor data recorded during historical flight missions of the logistics drone are collected, including real-time temperature, air pressure, and encountered gusts of wind. This data is then filtered and cleaned to remove outliers and invalid data, ultimately forming a dataset of real-world environmental interference parameters containing characteristics of low temperature, low air pressure, and strong winds.

[0014] Specifically, step S10 in the method includes: Meteorological observation data of the logistics drone at multiple typical flight sites in the plateau region were collected over multiple historical time periods. The meteorological observation data included at least temperature, air pressure and wind speed observations. Statistical analysis is performed on the meteorological observation data to determine the typical value ranges of the low temperature environmental parameters, the low pressure environmental parameters, and the strong wind environmental parameters. The typical value ranges include the statistical lower limit, statistical upper limit, and statistical average of the meteorological observation data. Based on the flight profile of the logistics drone performing delivery missions, at least one set of environmental parameter combinations that have the greatest impact on the logistics drone's power system during the takeoff, climb, cruise, and landing phases are identified, and the parameter values ​​corresponding to the environmental parameter combinations are used as the real environmental interference parameters.

[0015] In this embodiment of the application, firstly, historical meteorological data of different altitudes and seasons are retrieved from the meteorological database of the plateau region, including indicators such as temperature, air pressure, and wind speed, to ensure that the time span of the data is not less than 3 years, so as to reflect the environmental characteristics under different climatic conditions. The collected meteorological observation data are preprocessed to remove outliers in the temperature, air pressure, and wind speed data. For example, data that exceeds the mean ± 3 times the standard deviation are regarded as outliers and excluded to ensure the reliability of the data.

[0016] Then, the statistical lower limit, statistical upper limit, and statistical average values ​​of environmental parameters for low temperature, low pressure, and high wind are calculated separately to form the typical value range of each parameter. The statistical lower limit includes the lowest temperature, lowest air pressure, and highest wind speed; the statistical upper limit includes the highest temperature, highest air pressure, and lowest wind speed. For example, the typical value range for low temperature environmental parameters might be -25℃ to 5℃, for low pressure environmental parameters it might be 50kPa to 80kPa, and for high wind environmental parameters it might be 8m / s to 20m / s.

[0017] Subsequently, based on the flight profile of the logistics drone, the impact of different combinations of environmental parameters on the power system during the takeoff, climb, cruise, and landing phases was analyzed. Through simulation, the fluctuations in power system output power and changes in battery endurance caused by changes in environmental parameters during each phase were calculated, and the environmental parameter combinations with the greatest impact on the power system were identified.

[0018] Specifically, during takeoff, as the drone rapidly ascends from the ground, low air pressure reduces the air density of the rotor, causing a significant drop in thrust parameters. To maintain the lift required for takeoff, the motors need to increase their rotational speed, leading to a momentary surge in current and voltage parameters. During cruise, the continuous strong winds subject the drone to unstable lateral forces. To maintain flight attitude, the motor speed and current fluctuate at high frequencies, while the voltage slowly decreases due to continuous high power output.

[0019] For example, during the climb phase, the combination of low pressure (60 kPa) and strong crosswind (15 m / s) may significantly increase the load on the power system. The parameter value corresponding to this combination is determined as the real environmental disturbance parameter and used for subsequent test platform construction and model training.

[0020] S20: Construct a multi-environment coupling simulation test platform, wherein the multi-environment coupling simulation test platform is used to simultaneously apply the real environment interference parameters to simulate the real flight environment in which the logistics drone performs delivery tasks; In this embodiment, a multi-environment coupling simulation test platform is first constructed. This simulation test platform adopts a modular design and mainly consists of an environmental simulation chamber, a power loading system, a parameter acquisition system, and a centralized control system. Furthermore, the real environmental interference parameters acquired above are applied to the environmental simulation chamber to reproduce the complex environment of a logistics drone performing delivery tasks in a plateau region.

[0021] The construction of a multi-environment coupled simulation test platform includes: A comprehensive environmental simulation test chamber is established, wherein the comprehensive environmental simulation test chamber integrates a temperature regulation subsystem, an air pressure regulation subsystem, and a wind speed simulation subsystem; The low temperature environment parameters, low air pressure environment parameters, and strong wind environment parameters are simultaneously input into the centralized control system of the comprehensive environment simulation test chamber. The centralized control system adjusts the actual temperature, air pressure, and wind speed inside the comprehensive environment simulation test chamber in real time according to the input parameters. Throughout the entire process of the logistics drone performing the predetermined flight maneuvers, the actual values ​​of temperature, air pressure, and wind speed are continuously monitored and recorded, and it is ensured that the actual values ​​of temperature, air pressure, and wind speed remain stable within the allowable fluctuation ranges set by the low temperature environment parameter, low air pressure environment parameter, and high wind environment parameter, respectively.

[0022] In this embodiment, firstly, the temperature regulation subsystem of the integrated environmental simulation test chamber adopts dual-compressor cascade refrigeration technology, achieving a temperature regulation range of -40℃ to 60℃ with a temperature control accuracy of ±0.5℃. Multiple temperature sensors arranged within the chamber provide real-time feedback on the temperature field distribution, ensuring temperature uniformity within the chamber is within ±2℃. The air pressure regulation subsystem, through the coordinated operation of a vacuum pump and an air replenishment device, simulates an air pressure environment at altitudes of 0 to 10,000 meters, with a pressure regulation range of 30 kPa to 101 kPa and a control accuracy of ±0.5 kPa. It also possesses rapid depressurization and depressurization capabilities to simulate air pressure changes during rapid ascent or descent of a UAV. The wind speed simulation subsystem employs a high-power axial flow fan and air guide channel design, generating a stable wind speed of 0 to 30 m / s. Continuous wind speed adjustment is achieved through a frequency converter. Rotatable nozzles are also included to simulate gusts from different directions, with a wind direction adjustment range of 0° to 360° and a wind speed control accuracy of ±0.3 m / s.

[0023] Secondly, the centralized control system uses an industrial-grade PLC as the core controller. It communicates with the temperature, air pressure, and wind speed regulation subsystems via Ethernet to achieve real-time monitoring and closed-loop control of various environmental parameters. Based on the input parameters, it adjusts the actual temperature, air pressure, and wind speed values ​​inside the integrated environmental simulation test chamber in real time.

[0024] Furthermore, the centralized control system is equipped with a human-machine interface that can intuitively display current cabin environmental parameters, equipment operating status, and historical data curves. It also supports user-defined environmental parameter change curves, such as simulating the gradual change in air pressure and temperature as a drone gradually climbs from a plain to a plateau cruising altitude. To ensure the stability of environmental parameters during testing, a multi-level alarm mechanism is set up. When a parameter exceeds the set allowable fluctuation range, such as temperature fluctuation exceeding ±1℃, air pressure fluctuation exceeding ±1kPa, or wind speed fluctuation exceeding ±0.5m / s, an audible and visual alarm is triggered, and the adjustment and compensation program is automatically started. If the abnormality continues, the test is automatically interrupted and the fault information is recorded.

[0025] For example, when the low-temperature environment parameter is set to -15℃, the temperature regulation subsystem controls the operating power of the refrigeration compressor through a PID algorithm to ensure that the cabin temperature reaches the set value within 30 minutes and remains stable, with fluctuations controlled within ±0.5℃. For the low-pressure environment parameter of 60kPa, the pressure regulation subsystem first starts the vacuum pump to reduce the cabin pressure to 55kPa, and then slowly raises it to 60kPa through the air replenishment device to avoid sudden pressure changes from impacting the UAV structure. The stable pressure fluctuation does not exceed ±0.5kPa. The crosswind parameter of 15m / s is achieved through the axial flow fan and rotatable nozzle of the wind speed simulation subsystem. The fan frequency is adjusted to the corresponding speed, the nozzle is rotated to the set wind direction angle, and the wind speed is fed back in real time and finely adjusted through the frequency converter to ensure that the wind speed is stable at 15±0.3m / s.

[0026] S30: In the multi-environment coupling simulation test platform, control the logistics drone to perform a predetermined flight action, and synchronously collect dynamic change data of multiple key performance parameters of the logistics drone, wherein the multiple key performance parameters include tension parameters, rotational speed parameters, current parameters and voltage parameters. In this embodiment, after the multi-environment coupling simulation test platform is constructed and running stably, the logistics drone is fixedly installed and debugged. The drone is rigidly fixed to the center of the test platform inside the integrated environment simulation test chamber using a high-strength alloy bracket, ensuring the stability of the drone's fuselage attitude during simulated flight, while avoiding significant interference of the bracket with the airflow field.

[0027] Secondly, the logistics drone is controlled to perform predetermined flight maneuvers. These maneuvers are designed based on typical mission scenarios of logistics drones in high-altitude areas and include five consecutive phases: takeoff acceleration, constant speed climb, cruise hovering, wind-resistant maneuvering, and landing deceleration.

[0028] Furthermore, a high-precision tension sensor is installed below the drone's rotor. This sensor uses a strain gauge measurement principle, with a range of 0-50 kg, an accuracy class of 0.1, and a sampling frequency of up to 1000 Hz, to capture the dynamic changes in tension generated by rotor rotation. A non-contact speed encoder is installed at the connection between the motor output shaft and the rotor. It acquires the motor speed through electromagnetic induction, with a resolution of 1 rpm and a sampling frequency of 500 Hz, ensuring accurate recording of instantaneous fluctuations in speed. A high-precision current sensor and a voltage sensor are connected in series at the drone's battery output. The current sensor uses the Hall effect principle, with a range of 0-50 A and an accuracy of ±0.5% FS. The voltage sensor is a voltage divider design, with a range of 0-50 V and an accuracy of ±0.2% FS. Both sensors synchronously acquire current and voltage data at a sampling frequency of 1000 Hz. All sensor signals are converted into digital signals by a data acquisition card and then transmitted to the centralized control system for storage and analysis.

[0029] Specifically, step S30 in the method includes: The predetermined flight maneuver is set as a sudden climb maneuver under full load conditions, wherein the sudden climb maneuver includes the process of switching from hovering state to maximum power climb state, in order to simulate the most severe power output conditions of the logistics drone in high-altitude areas. In the multi-environment coupling simulation test platform, the logistics drone is controlled to switch from hovering state to maximum power climb state through preset flight control commands, and the maximum power climb state is maintained for a preset period of time. During the process of the logistics drone performing the sudden climb, the real-time change curves of the tension parameter, rotation speed parameter, current parameter, and voltage parameter are simultaneously collected at a rate not lower than the preset sampling frequency, as the dynamic change data.

[0030] In this embodiment, firstly, considering the heavy load on the power system in high-altitude areas, the predetermined flight maneuver is set as a sudden climb under full load conditions to simulate the most severe power output conditions that logistics drones may encounter in high-altitude environments. Full load refers to the drone carrying its rated maximum payload, typically determined by its design standards. For example, if a certain model of logistics drone has a rated payload of 15kg, then a 15kg standard counterweight will be used during the test. The sudden climb maneuver is the process of rapidly switching from a hovering state to a maximum power climb state, with a switching response time required to be no more than 0.5 seconds, to reproduce extreme operational scenarios when the drone is urgently avoiding obstacles or rapidly reaching the target altitude.

[0031] Secondly, in the multi-environment coupling simulation test platform, the flight control command module integrated into the centralized control system sends preset control commands to the flight control system of the logistics drone. Specifically, after the environmental parameters in the integrated environmental simulation test chamber stabilize to the set real environmental interference parameters, such as low temperature -15℃, low air pressure 60kPa, and crosswind of 15m / s, the flight control command module first sends a hovering command, causing the drone to hover stably at the set altitude for 30 seconds to ensure consistent initial state. Subsequently, a maximum power climb command is sent to control the drone to switch to maximum power output mode. At this time, the motor throttle opening rapidly increases to 100%, and this maximum power climb state is maintained for a preset time, usually set to 60 seconds, to fully capture the complete dynamic process of the power system from startup, loading to stable output and then to possible attenuation.

[0032] Furthermore, during the sudden climb maneuver performed by the logistics drone, the parameter acquisition system synchronously collects various key performance parameters at a rate no less than the preset sampling frequency. The preset sampling frequency is set differently according to the dynamic characteristics of the parameters. The sampling frequency for tension parameters and current and voltage parameters is set to 1000Hz to capture instantaneous peaks and fluctuations; the sampling frequency for rotational speed parameters is set to 500Hz, which can meet the recording requirements for rapid changes in rotational speed while avoiding data redundancy.

[0033] Specifically, the tension sensor monitors the tension generated by the rotor in real time, outputting a 4-20mA analog signal. After being converted into a digital signal by the data acquisition card, a real-time change curve of the tension parameter is generated. This curve reflects the dynamic trend of tension changes with altitude and environmental resistance during the climb, from the stable tension during hovering to the surge in tension during the climb. The speed encoder records the motor's rotational speed in real time, generating a real-time change curve of the speed parameter. It can observe the process of rapidly climbing from the lower stable speed during hovering to the maximum speed after receiving the climb command, as well as the high-frequency oscillation of speed that may occur under strong wind interference. The current and voltage sensors collect the real-time current and voltage supplied by the battery to the motor, respectively, generating corresponding real-time change curves. The current curve shows a large jump during the climb, such as jumping from 15A during hovering to 35A, while the voltage curve may show a brief voltage drop due to large current discharge, such as dropping from the nominal voltage of 22.2V to 20.5V. Finally, the above dynamic change data is stored in real time in the database of the centralized control system.

[0034] S40: Based on the dynamic change data, analyze the coupling relationship between the real environmental interference parameters and the multiple key performance parameters, and establish a parameter coupling model describing the coupling relationship; In this embodiment, based on the collected dynamic change data, time-domain features are extracted from the dynamic change curves of each key performance parameter. Combined with the specific values ​​of real-world environmental interference parameters, an input feature set and an output feature set for the parameter coupling model are constructed. The input feature set includes the specific values ​​of real-world environmental interference parameters, namely, low-temperature environmental parameters, low-pressure environmental parameters, and strong-wind environmental parameters. The output feature set consists of time-domain features extracted from the dynamic change data, such as the peak value, rise time, and steady-state value of the tension parameter; the maximum speed and fluctuation amplitude of the rotational speed parameter; the peak current and average current of the current parameter; and the minimum voltage and voltage drop rate of the voltage parameter.

[0035] Specifically, step S40 in the method includes: The dynamically changing data is preprocessed, including data cleaning, outlier removal, and data normalization, to obtain a standardized multidimensional time series dataset. The Pearson correlation coefficient between the real environmental interference parameters and the multiple key performance parameters is calculated using the correlation analysis method. Based on the magnitude of the correlation coefficient, parameter pairs with strong coupling relationships are identified, resulting in multiple strongly coupled parameter pairs. Based on the aforementioned multiple strongly coupled parameter pairs, a parameter coupling model based on a feedforward neural network is constructed.

[0036] In this embodiment, the dynamic data is first preprocessed. The data cleaning operation is mainly aimed at noise interference that may occur during the sensor acquisition process. A 5th-order Butterworth low-pass filter is used to smooth the original signal. The cutoff frequency is set according to the highest frequency component of each parameter. For example, the cutoff frequency of the tension and current signals is set to 100Hz, and the speed signal is set to 50Hz to effectively filter out high-frequency noise.

[0037] Specifically, outlier removal is based on the 3σ criterion. For each parameter's time-series data, the mean μ and standard deviation σ are calculated. Data points exceeding the range [μ-3σ, μ+3σ] are identified as outliers and replaced using linear interpolation with adjacent normal data points to ensure the continuity of the data sequence. Data normalization employs the min-max normalization method, mapping the time-domain characteristic values ​​of each parameter to the [0,1] interval. The specific formula is: x normalized value = (xx) min ) / (x max -x min ), where x min and x max The minimum and maximum values ​​of this feature in the historical sample data are respectively used to eliminate the influence of different units on subsequent analysis, and finally a standardized multidimensional time series dataset is obtained.

[0038] Secondly, correlation analysis was used to calculate the Pearson correlation coefficients between real environmental disturbance parameters and multiple key performance parameters. Low-temperature environmental parameter T, low-pressure environmental parameter P, and strong-wind environmental parameter V were used as environmental variables, while time-domain characteristics such as peak tensile force F, speed fluctuation amplitude N, peak current I, and minimum voltage U were used as performance variables. A 3×n correlation coefficient matrix was constructed, where n is the number of performance variables. The formula for calculating the Pearson correlation coefficient r is r=cov(X,Y) / (σX×σY), where cov(X,Y) is the covariance of variables X and Y, and σX and σY are the standard deviations of X and Y, respectively. By calculating the absolute values ​​of the correlation coefficients between each environmental variable and the performance variable, parameter pairs with absolute values ​​greater than 0.7 are identified as having a strong coupling relationship. For example, the correlation coefficient between the low air pressure parameter P and the peak tension F is -0.82, which is a negative correlation. The decrease in air pressure leads to a decrease in the peak tension. The correlation coefficient between the strong wind environmental parameter V and the speed fluctuation amplitude N is 0.85, which is a positive correlation. The increase in wind speed leads to an increase in speed fluctuation. Thus, multiple strongly coupled parameter pairs are obtained.

[0039] Finally, based on multiple strongly coupled parameter pairs, a parameter coupling model based on a feedforward neural network is constructed. This neural network adopts a 3-layer structure. The number of neurons in the input layer corresponds to the environmental interference parameter dimension in the strongly coupled parameter pair, such as 3 environmental parameters. There are 2 hidden layers. The first layer contains 32 neurons and uses the ReLU activation function, and the second layer contains 16 neurons and also uses the ReLU activation function. The number of neurons in the output layer corresponds to the key performance parameter dimension in the strongly coupled parameter pair, such as 4 performance parameters with temporal features. A linear activation function is used to output continuous values.

[0040] Furthermore, the network's training dataset is derived from dynamically changing data under multiple combinations of different environmental parameters, and is divided into training and validation sets. The loss function is mean squared error, the optimizer is the Adam optimizer, and the initial learning rate is set to 0.001, dynamically adjusted during training using a learning rate decay strategy. During training, early stopping is used to prevent overfitting; training stops when the validation set loss no longer decreases after 10 consecutive training epochs, resulting in a parameter coupling model that accurately describes the nonlinear coupling relationship between real-world environmental disturbance parameters and key performance parameters.

[0041] Specifically, based on the multiple strongly coupled parameter pairs, a parameter coupling model based on a feedforward neural network is constructed, including: Based on the multiple strongly coupled parameter pairs, the input layer nodes and output layer nodes of the feedforward neural network are determined, wherein the real environmental interference parameters in the multiple strongly coupled parameter pairs are used as the input variables of the input layer nodes, and the key performance parameters in the multiple strongly coupled parameter pairs corresponding to the real environmental interference parameters are used as the output variables of the output layer nodes. The hidden layer structure of the feedforward neural network is configured such that the hidden layer structure includes at least one hidden layer, each hidden layer contains multiple neuron nodes, and the number of hidden layers and the number of neuron nodes are optimized according to the dimensions of the input layer nodes and the output layer nodes and the amount of training sample data. The feedforward neural network is trained using a standardized multidimensional time series dataset obtained after data preprocessing. During the training process, the backpropagation algorithm is used to continuously adjust the weight parameters and bias parameters of the feedforward neural network, so that the loss function value between the predicted output value of the feedforward neural network and the actual collected dynamic data gradually decreases and converges to a preset threshold range. After the feedforward neural network is trained, the trained parameter coupling model is validated using an independent test dataset. The root mean square error is calculated to evaluate the fitting accuracy of the parameter coupling model to the nonlinear mapping relationship between the strongly coupled parameter pairs. If the fitting accuracy does not meet the preset requirements, the training parameters are reset and iterative training is performed until the fitting accuracy meets the requirements.

[0042] In this embodiment, firstly, the input layer nodes and output layer nodes of the feedforward neural network are defined based on multiple strongly coupled parameter pairs. For example, if the strongly coupled parameter pairs include low air pressure and peak tension, strong wind and speed fluctuation amplitude, low temperature and peak current, and low temperature and minimum voltage, then the input variables of the input layer nodes are the real environmental disturbance parameters, including low temperature environmental parameters, low air pressure environmental parameters, and strong wind environmental parameters, for a total of 3 input nodes; the output variables of the output layer nodes are the corresponding 4 output nodes. The number of nodes in the input layer and output layer is directly determined by the number of types of environmental parameters and performance parameters in the strongly coupled parameter pairs.

[0043] Secondly, the hidden layer structure of the feedforward neural network is set, and the hidden layer typically contains at least one hidden layer. In this embodiment, considering that the input layer is 3-dimensional, the output layer is 4-dimensional, and the amount of training sample data is relatively sufficient, such as hundreds of sets of dynamic change data collected under different combinations of temperature, air pressure, and wind speed through a multi-environment coupling simulation test platform, the initial setting is 2 hidden layers. The first hidden layer contains 32 neurons and uses the ReLU activation function, which can effectively solve the gradient vanishing problem, accelerate network training, and increase the non-linear expressive ability of the model. The second hidden layer contains 16 neurons and also uses the ReLU activation function. By reducing the number of neurons layer by layer, the dimensionality reduction and abstraction of features are achieved. The number of hidden layers and the number of neurons are not fixed and can be optimized according to the actual training effect to balance the fitting ability and generalization ability of the feedforward neural network. For example, if the model underfits, the number of hidden layers or the number of neurons per layer can be increased; if overfitting occurs and the validation set loss is high, the number of neurons can be appropriately reduced.

[0044] Next, the feedforward neural network is trained using the standardized multidimensional time series dataset obtained after data preprocessing. The preprocessed dataset is divided into training and validation sets in a predetermined ratio of 7:3. During training, the backpropagation algorithm is used to continuously adjust the weight and bias parameters of each layer of the network based on the input-output pairs of the training set.

[0045] Specifically, the input variables are first passed from the input layer to the hidden layer. After being processed by the weighted summation and activation function of the neurons in the hidden layer, they are passed to the output layer to obtain the predicted output value. Then, the loss function value between the predicted output value and the actual value of the corresponding performance parameter in the actually collected dynamic data is calculated. In this embodiment, the mean squared error is used as the loss function, and its formula is Mean Squared Error MSE = (1 / n). Where yi is the actual value, i represents the predicted value, and n represents the number of samples. The Adam optimizer is used, with an initial learning rate of 0.001. A learning rate decay strategy is employed, for example, multiplying the learning rate by 0.9 every 10 training epochs to fine-tune the parameters in the later stages of training. Through continuous iteration, the loss function value gradually decreases and converges to a preset threshold range, for example, a range less than 0.001.

[0046] Finally, after the feedforward neural network is trained, the trained parameter-coupled model is validated using an independent test dataset—data not used in training and validation. The model's accuracy in fitting the nonlinear mapping relationship between strongly coupled parameter pairs is evaluated by calculating the root mean square error (RMSE) between the predicted and actual values ​​on the test set. The formula is: RMSE = If the calculated RMSE is less than the preset accuracy requirement, the model fitting accuracy is considered to meet the requirement; if it does not meet the preset requirement, the training parameters need to be reset, such as adjusting the learning rate, hidden layer structure, number of iterations, or increasing the number of training samples, improving data preprocessing methods, etc., and iterative training should be carried out until the model's fitting accuracy on the test set meets the preset requirement, thereby ensuring that the parameter coupling model can accurately reflect the dynamic coupling law between the real environmental interference parameters and the key performance parameters of the logistics drone in the complex plateau environment.

[0047] S50: Using the parameter coupling model, determine the test items, test conditions, and test indicators required for the metrological testing of the logistics drone in the real flight environment, and generate a metrological testing requirements analysis report.

[0048] In this embodiment of the application, based on the constructed parameter coupling model, the changing trend of key performance parameters of logistics drones can be predicted by inputting interference parameter combinations under different real flight environments. Then, the test items, test conditions and test indicators required for metrology testing can be deduced, and a metrology testing requirements analysis report can be generated.

[0049] Specifically, the parameter coupling model is used to determine the test items, test conditions, and test indicators required for metrological testing of the logistics drone in the actual flight environment, including: Based on the parameter coupling model, at least one sensitive parameter among the multiple key performance parameters whose dynamic fluctuation amplitude exceeds a preset safety threshold under the simultaneous action of the real environmental interference parameters is identified, and the sensitive parameter is determined as the core test item required for the metrological test. Based on the parameter coupling model, the dynamic response characteristics of the core test item under different combinations of real-world environmental interference parameters are simulated and calculated. Based on the dynamic response characteristics, test conditions corresponding to the core test item are formulated. The test conditions include at least the specific numerical range of the real-world environmental interference parameters and the precise execution sequence of the predetermined flight maneuvers. Based on the parameter coupling model and the design safety boundary of the logistics drone, the maximum permissible error range of the core test item in the real flight environment is calculated, and the maximum permissible error range is determined as the test index, which is used to determine whether the performance of the logistics drone is qualified in subsequent metrological tests.

[0050] In this embodiment of the application, firstly, based on the parameter coupling model, sensitive parameters whose dynamic fluctuation amplitude exceeds a preset safety threshold among multiple key performance parameters are identified and used as core test items required for metrological testing.

[0051] Specifically, the preset safety thresholds are set according to the design standards and safe operation requirements of logistics drones. For example, the safety threshold for peak tensile force is set at ±10% of its rated tensile force, the safety threshold for speed fluctuation is set at ±5%, the safety threshold for peak current is set at 1.2 times the rated current, and the safety threshold for minimum voltage is set at 80% of the battery's nominal voltage. By comparing the dynamic fluctuation amplitude of each key performance parameter predicted by the parameter coupling model under the action of real environmental interference parameter combinations with the preset safety thresholds, such as the real environmental interference parameter combination being low temperature -20℃, low air pressure 50kPa, and strong wind 12m / s, if the fluctuation amplitude of a certain parameter exceeds the threshold, it is identified as a sensitive parameter.

[0052] For example, when low pressure and strong winds act simultaneously, the model predicts that the peak tension fluctuation reaches 15%, exceeding the safety threshold of 10%, and the speed fluctuation reaches 8%, exceeding the safety threshold of 5%. At this point, the peak tension and speed fluctuation are determined as core test items.

[0053] Secondly, based on the parameter coupling model, the dynamic response characteristics of the core test items under different combinations of real-world environmental disturbance parameters are simulated and calculated. Taking the peak thrust and rotational speed fluctuation amplitude of the core test items as examples, by inputting different combinations of low temperature, low air pressure, and high wind speed values ​​into the parameter coupling model, the curves of the change of peak thrust and rotational speed fluctuation amplitude over time are simulated when the logistics drone performs predetermined flight actions under this environmental combination, such as takeoff and climb, hovering, cruise, and landing.

[0054] For example, under the combined conditions of -25°C, 45 kPa low air pressure, and 14 m / s strong wind, the instantaneous maximum value of the thrust peak, the time point at which the maximum value is reached, and the periodicity and decay characteristics of the fluctuations are simulated during the acceleration of a UAV from a standstill to cruising speed. Simultaneously, the amplitude range and frequency of the speed fluctuations, as well as the differences in fluctuations during different flight phases, such as acceleration, constant speed, and deceleration, are recorded. Through the analysis of dynamic response characteristics, the degree and pattern of influence of different combinations of environmental parameters on the core test items are clarified, thereby enabling the formulation of test conditions corresponding to the core test items.

[0055] Furthermore, the specific numerical ranges of the real-world environmental interference parameters in the test conditions must cover the extreme environmental ranges that the UAV may actually encounter, as well as the sensitive ranges that significantly affect performance. For example, the low-temperature range can be set to -35℃ to 5℃, the low-pressure range to 40kPa to 101kPa, and the high-wind-speed range to 0m / s to 20m / s. The precise execution sequence of the predetermined flight maneuvers must specify the start time, duration, and rate of change of speed for each maneuver. For example, the takeoff phase requires acceleration from a standstill to a cruising speed of 10m / s within 3 seconds, the hovering phase lasts for 60 seconds, and the cruising phase maintains a stable speed within ±0.5m / s.

[0056] Furthermore, the method also includes the step of establishing an field data feedback calibration mechanism: At least one representative typical flight station is selected in the plateau region, and the logistics drone is organized to perform the same real flight mission as the predetermined flight maneuver at the typical flight station, and real field flight data of the multiple key performance parameters are collected simultaneously during the real flight. The real field flight data and the dynamic change data collected in the multi-environment coupling simulation test platform are compared and analyzed in multiple dimensions to calculate the deviation values ​​between the two in the time domain and frequency domain. The construction parameters of the multi-environment coupling simulation test platform are adjusted based on the deviation value, and the model parameters of the parameter coupling model are corrected and calibrated so that the environment simulated by the multi-environment coupling simulation test platform is closer to the real flight environment, and the predicted output value of the parameter coupling model is closer to the real field flight data.

[0057] In this embodiment, firstly, considering factors such as altitude, climate characteristics, and topography, at least one representative typical flight station is selected in the plateau region. For example, mountainous stations above 3500 meters in altitude with frequent strong winds and drastic temperature changes, and plateau basin stations above 4500 meters in altitude with extremely low air pressure are selected. After selecting the stations, the logistics drones are organized to perform real flight tasks at the typical flight stations that are exactly the same as the predetermined flight actions preset in the multi-environment coupling simulation test platform, such as standard take-off, climb, hovering, cruise, turning, and landing sequences. During the real flight, real field flight data of multiple key performance parameters are simultaneously collected through the drone's onboard three-axis accelerometer, gyroscope, tension sensor, speed sensor, current and voltage sensor, and ground monitoring system. The sampling frequency is no less than the sampling frequency of the simulation test platform to ensure consistent data temporal resolution.

[0058] Secondly, a multi-dimensional comparative analysis was conducted between the collected real field flight data and the dynamic change data collected in the multi-environment coupled simulation test platform. In the time domain, the curve shapes of key performance parameters changing over time were compared, and the numerical deviation, peak deviation, valley deviation, and fluctuation period deviation at the same moment were calculated. In the frequency domain, the time domain signal was converted into a frequency domain signal through Fourier transform, and the spectral characteristics of the two were compared to analyze the amplitude differences and phase differences of the main frequency components, identifying the similarities and differences between the simulated data and the real data in frequency response.

[0059] For example, by comparing the dominant frequency amplitude of the thrust fluctuation during the hovering phase, if the dominant frequency amplitude of the simulated data is 80% of that of the real data, then its deviation value is calculated to be 20%. Through multi-dimensional comparison, the differences between the simulation test platform and the real flight environment are comprehensively evaluated.

[0060] Finally, the construction parameters of the multi-environment coupling simulation test platform are adjusted based on the calculated deviation values, and the model parameters of the parameter coupling model are corrected and calibrated. For the simulation test platform, if real field data shows that the actual peak thrust of the UAV under specific low-temperature conditions is 10% lower than the simulated value, the temperature control accuracy of the low-temperature environment simulation module of the platform is adjusted, or the coupling coefficient between airflow velocity and pressure field in the wind field simulation system is corrected to reproduce the interaction between low temperature and wind field in the real environment.

[0061] For parameter-coupled models, if the predicted rotational speed fluctuations under high wind conditions deviate significantly from the actual field data, the actual field data is used as new training samples for incremental training of the model. Alternatively, the model's hyperparameters, such as the number of hidden layer neurons and the learning rate, are adjusted, and the model is retrained to reduce prediction bias. Through a field data feedback calibration mechanism, the simulation test platform and parameter-coupled model are continuously iteratively optimized, making the simulation environment closer to the real flight environment and the model's predicted output values ​​closer to the actual field flight data.

[0062] Furthermore, a metrology and testing requirements analysis report is generated, including: The test items, test conditions, and test indicators shall be written into the main body of the metrological test requirements analysis report in the form of a structured table; The metrology and testing requirements analysis report displays the coupling relationship between the real environmental interference parameters and the multiple key performance parameters in the form of a visual curve or scatter plot, and marks the key coupling nodes and the dynamic change trend of the parameters over time in the coupling relationship graph.

[0063] In this embodiment of the application, the test items, test conditions and test indicators are first written into the main body of the metrological test requirements analysis report in the form of a structured table. The column headings of the table can be set as "core test items", "test environment parameter range", "predetermined flight actions and sequence" and "maximum permissible error range".

[0064] Secondly, the metrology and testing requirements analysis report should display the coupling relationship between real-world environmental interference parameters and multiple key performance parameters using visual graphs or scatter plots. For example, a 3D surface plot can be drawn to show the trend of peak tension changing with wind speed under different combinations of temperature and air pressure. The color depth of the surface can represent the magnitude of the peak tension, thus intuitively reflecting the coupling effect between the three. For a single key performance parameter, such as the fluctuation range of rotational speed, a graph showing its change over time should be drawn, and the corresponding curves of environmental parameters such as temperature, air pressure, and wind speed should be overlaid on the graph to show how the dynamic changes of environmental parameters affect the fluctuation of performance parameters.

[0065] Simultaneously, key coupling nodes are marked in the coupling relationship diagram, such as the point where the peak thrust shows a significant jump when the wind speed reaches 12 m / s and the air pressure is below 50 kPa, or the critical point where the speed fluctuation exceeds the safety threshold. Furthermore, the dynamic change trend of parameters over time is clearly marked in the diagram, such as the rapid rise of the peak thrust to its maximum value during takeoff acceleration, followed by stabilization during cruise, and periodic fluctuations under strong wind interference.

[0066] In summary, compared to existing technologies, this application constructs a multi-environment coupling simulation test platform capable of simultaneously applying multiple real-world environmental disturbance parameters. This platform controls a logistics drone to perform predetermined flight maneuvers under conditions such as sudden climbs under full load, while simultaneously collecting dynamic change data of key performance parameters. Correlation analysis identifies parameter pairs with strong coupling relationships, and a parameter coupling model is constructed based on a feedforward neural network. This model reveals the nonlinear mapping law between environmental disturbance parameters and performance parameters, thereby determining the test items, test conditions, and test indicators required for metrological testing of the logistics drone in a real flight environment. The resulting metrological testing requirements analysis report improves the consistency between metrological test results and real flight conditions, effectively ensuring the flight safety and performance reliability of the logistics drone when performing delivery missions in high-altitude areas.

[0067] In summary, the embodiments of this application have at least the following technical effects: This application provides a method for analyzing the metrological testing requirements of unmanned aerial vehicles (UAVs) oriented towards multi-parameter coupling. First, it acquires low-temperature, low-pressure, and high-wind environmental parameters of a logistics UAV performing delivery missions in high-altitude areas, collectively forming a complex real-world flight environment. Second, it constructs a multi-environment coupling simulation test platform capable of simultaneously applying real-world environmental disturbance parameters, thereby reproducing the real-world flight environment encountered by the logistics UAV during delivery missions in a laboratory setting. Third, within this multi-environment coupling simulation test platform, the logistics UAV is controlled to perform predetermined flight maneuvers, simulating the most severe power output conditions the UAV might encounter in high-altitude areas. Simultaneously, dynamic change data of the UAV's thrust, rotational speed, current, and voltage parameters are collected, serving as the primary basis for analyzing the coupling relationship between environmental and performance parameters. Then, based on the collected dynamic change data, parameter pairs with strong coupling relationships between real-world environmental disturbance parameters and multiple key performance parameters are identified. A parameter coupling model is constructed to describe the nonlinear response laws of key performance parameters under different combinations of environmental parameters. Finally, the metrological testing requirements analysis report for logistics drones in real flight environments was determined, which effectively ensured the scientific nature and relevance of test items, test conditions and test indicators, and improved the ability to guarantee the flight safety and performance reliability of drones in complex real environments such as plateaus.

[0068] Through the above technical solution, this application breaks through the limitations of traditional single-parameter testing by dynamically coupling and analyzing real environmental interference parameters with key performance parameters of UAVs. It provides a comprehensive demand analysis framework for metrological testing of logistics UAVs in complex environments, which helps to promote the development of UAV metrological testing technology towards multi-parameter collaboration and real environment simulation.

[0069] Example 2, as Figure 2 As shown, based on the same inventive concept as the UAV metrology and testing requirements analysis method for multi-parameter coupling provided in Embodiment 1, this application also provides a UAV metrology and testing requirements analysis system for multi-parameter coupling, including: The parameter acquisition module 11 is used to acquire low temperature environmental parameters, low air pressure environmental parameters and strong wind environmental parameters when the logistics drone performs delivery tasks in the plateau area, as real environmental interference parameters; Platform construction module 12 is used to construct a multi-environment coupling simulation test platform, wherein the multi-environment coupling simulation test platform is used to simultaneously apply the real environment interference parameters to simulate the real flight environment in which the logistics drone performs delivery tasks; The flight test module 13 is used to control the logistics drone to perform predetermined flight actions in the multi-environment coupling simulation test platform, and to synchronously collect dynamic change data of multiple key performance parameters of the logistics drone, wherein the multiple key performance parameters include tension parameters, rotational speed parameters, current parameters and voltage parameters. The model training module 14 is used to analyze the coupling relationship between the real environment interference parameters and the multiple key performance parameters based on the dynamically changing data, and to establish a parameter coupling model describing the coupling relationship. The result generation module 15 is used to determine the test items, test conditions and test indicators required for the metrological testing of the logistics drone in the real flight environment using the parameter coupling model, and to generate a metrological testing requirements analysis report.

[0070] In one embodiment, the parameter acquisition module 11 is specifically used for: Meteorological observation data of the logistics drone at multiple typical flight sites in the plateau region were collected over multiple historical time periods. The meteorological observation data included at least temperature, air pressure and wind speed observations. Statistical analysis is performed on the meteorological observation data to determine the typical value ranges of the low temperature environmental parameters, the low pressure environmental parameters, and the strong wind environmental parameters. The typical value ranges include the statistical lower limit, statistical upper limit, and statistical average of the meteorological observation data. Based on the flight profile of the logistics drone performing delivery missions, at least one set of environmental parameter combinations that have the greatest impact on the logistics drone's power system during the takeoff, climb, cruise, and landing phases are identified, and the parameter values ​​corresponding to the environmental parameter combinations are used as the real environmental interference parameters.

[0071] Furthermore, in one embodiment of the application, a multi-environment coupling simulation test platform is constructed, comprising: A comprehensive environmental simulation test chamber is established, wherein the comprehensive environmental simulation test chamber integrates a temperature regulation subsystem, an air pressure regulation subsystem, and a wind speed simulation subsystem; The low temperature environment parameters, low air pressure environment parameters, and strong wind environment parameters are simultaneously input into the centralized control system of the comprehensive environment simulation test chamber. The centralized control system adjusts the actual temperature, air pressure, and wind speed inside the comprehensive environment simulation test chamber in real time according to the input parameters. Throughout the entire process of the logistics drone performing the predetermined flight maneuvers, the actual values ​​of temperature, air pressure, and wind speed are continuously monitored and recorded, and it is ensured that the actual values ​​of temperature, air pressure, and wind speed remain stable within the allowable fluctuation ranges set by the low temperature environment parameter, low air pressure environment parameter, and high wind environment parameter, respectively.

[0072] In one embodiment, the flight test module 13 is specifically used for: The predetermined flight maneuver is set as a sudden climb maneuver under full load conditions, wherein the sudden climb maneuver includes the process of switching from hovering state to maximum power climb state, in order to simulate the most severe power output conditions of the logistics drone in high-altitude areas. In the multi-environment coupling simulation test platform, the logistics drone is controlled to switch from hovering state to maximum power climb state through preset flight control commands, and the maximum power climb state is maintained for a preset period of time. During the process of the logistics drone performing the sudden climb, the real-time change curves of the tension parameter, rotation speed parameter, current parameter, and voltage parameter are simultaneously collected at a rate not lower than the preset sampling frequency, as the dynamic change data.

[0073] In one embodiment, the model training module 14 is specifically used for: The dynamically changing data is preprocessed, including data cleaning, outlier removal, and data normalization, to obtain a standardized multidimensional time series dataset. The Pearson correlation coefficient between the real environmental interference parameters and the multiple key performance parameters is calculated using the correlation analysis method. Based on the magnitude of the correlation coefficient, parameter pairs with strong coupling relationships are identified, resulting in multiple strongly coupled parameter pairs. Based on the aforementioned multiple strongly coupled parameter pairs, a parameter coupling model based on a feedforward neural network is constructed.

[0074] Furthermore, in one embodiment of the application, a parameter coupling model based on a feedforward neural network is constructed based on the plurality of strongly coupled parameter pairs, including: Based on the multiple strongly coupled parameter pairs, the input layer nodes and output layer nodes of the feedforward neural network are determined, wherein the real environmental interference parameters in the multiple strongly coupled parameter pairs are used as the input variables of the input layer nodes, and the key performance parameters in the multiple strongly coupled parameter pairs corresponding to the real environmental interference parameters are used as the output variables of the output layer nodes. The hidden layer structure of the feedforward neural network is configured such that the hidden layer structure includes at least one hidden layer, each hidden layer contains multiple neuron nodes, and the number of hidden layers and the number of neuron nodes are optimized according to the dimensions of the input layer nodes and the output layer nodes and the amount of training sample data. The feedforward neural network is trained using a standardized multidimensional time series dataset obtained after data preprocessing. During the training process, the backpropagation algorithm is used to continuously adjust the weight parameters and bias parameters of the feedforward neural network, so that the loss function value between the predicted output value of the feedforward neural network and the actual collected dynamic data gradually decreases and converges to a preset threshold range. After the feedforward neural network is trained, the trained parameter coupling model is validated using an independent test dataset. The root mean square error is calculated to evaluate the fitting accuracy of the parameter coupling model to the nonlinear mapping relationship between the strongly coupled parameter pairs. If the fitting accuracy does not meet the preset requirements, the training parameters are reset and iterative training is performed until the fitting accuracy meets the requirements.

[0075] Furthermore, the parameter coupling model is used to determine the test items, test conditions, and test indicators required for metrological testing of the logistics drone in the actual flight environment, including: Based on the parameter coupling model, at least one sensitive parameter among the multiple key performance parameters whose dynamic fluctuation amplitude exceeds a preset safety threshold under the simultaneous action of the real environmental interference parameters is identified, and the sensitive parameter is determined as the core test item required for the metrological test. Based on the parameter coupling model, the dynamic response characteristics of the core test item under different combinations of real-world environmental interference parameters are simulated and calculated. Based on the dynamic response characteristics, test conditions corresponding to the core test item are formulated. The test conditions include at least the specific numerical range of the real-world environmental interference parameters and the precise execution sequence of the predetermined flight maneuvers. Based on the parameter coupling model and the design safety boundary of the logistics drone, the maximum permissible error range of the core test item in the real flight environment is calculated, and the maximum permissible error range is determined as the test index, which is used to determine whether the performance of the logistics drone is qualified in subsequent metrological tests.

[0076] Furthermore, the method also includes the step of establishing an field data feedback calibration mechanism: At least one representative typical flight station is selected in the plateau region, and the logistics drone is organized to perform the same real flight mission as the predetermined flight maneuver at the typical flight station, and real field flight data of the multiple key performance parameters are collected simultaneously during the real flight. The real field flight data and the dynamic change data collected in the multi-environment coupling simulation test platform are compared and analyzed in multiple dimensions to calculate the deviation values ​​between the two in the time domain and frequency domain. The construction parameters of the multi-environment coupling simulation test platform are adjusted based on the deviation value, and the model parameters of the parameter coupling model are corrected and calibrated so that the environment simulated by the multi-environment coupling simulation test platform is closer to the real flight environment, and the predicted output value of the parameter coupling model is closer to the real field flight data.

[0077] Furthermore, in one embodiment, generating a metrology and testing requirements analysis report includes: The test items, test conditions, and test indicators shall be written into the main body of the metrological test requirements analysis report in the form of a structured table; The metrology and testing requirements analysis report displays the coupling relationship between the real environmental interference parameters and the multiple key performance parameters in the form of a visual curve or scatter plot, and marks the key coupling nodes and the dynamic change trend of the parameters over time in the coupling relationship graph.

Claims

1. A method for analyzing the metrological testing requirements of unmanned aerial vehicles (UAVs) with multi-parameter coupling, characterized in that: The methods include: The low-temperature environment parameters, low-pressure environment parameters, and strong wind environment parameters of logistics drones performing delivery tasks in plateau areas are obtained as real environmental interference parameters. A multi-environment coupling simulation test platform is constructed, wherein the multi-environment coupling simulation test platform is used to simultaneously apply the real environment interference parameters to simulate the real flight environment in which the logistics drone performs delivery tasks; In the multi-environment coupling simulation test platform, the logistics drone is controlled to perform predetermined flight actions, and dynamic change data of multiple key performance parameters of the logistics drone are collected simultaneously. The multiple key performance parameters include tension parameters, rotational speed parameters, current parameters, and voltage parameters. Based on the dynamic change data, the coupling relationship between the real environmental interference parameters and the multiple key performance parameters is analyzed, and a parameter coupling model describing the coupling relationship is established. The parameter coupling model is used to determine the test items, test conditions, and test indicators required for metrological testing of the logistics drone in the real flight environment, and a metrological testing requirements analysis report is generated.

2. The method for analyzing UAV metrological testing requirements based on multi-parameter coupling as described in claim 1, characterized in that, The parameters of low-temperature environment, low-pressure environment, and high-wind environment when logistics drones perform delivery missions in plateau areas are obtained as real-world environmental interference parameters, including: Meteorological observation data of the logistics drone at multiple typical flight sites in the plateau region were collected over multiple historical time periods. The meteorological observation data included at least temperature, air pressure and wind speed observations. Statistical analysis is performed on the meteorological observation data to determine the typical value ranges of the low temperature environmental parameters, the low pressure environmental parameters, and the strong wind environmental parameters. The typical value ranges include the statistical lower limit, statistical upper limit, and statistical average of the meteorological observation data. Based on the flight profile of the logistics drone performing delivery missions, at least one set of environmental parameter combinations that have the greatest impact on the logistics drone's power system during the takeoff, climb, cruise, and landing phases are identified, and the parameter values ​​corresponding to the environmental parameter combinations are used as the real environmental interference parameters.

3. The method for analyzing UAV metrological testing requirements based on multi-parameter coupling as described in claim 1, characterized in that, Construct a multi-environment coupled simulation test platform, including: A comprehensive environmental simulation test chamber is established, wherein the comprehensive environmental simulation test chamber integrates a temperature regulation subsystem, an air pressure regulation subsystem, and a wind speed simulation subsystem; The low temperature environment parameters, low air pressure environment parameters, and strong wind environment parameters are simultaneously input into the centralized control system of the comprehensive environment simulation test chamber. The centralized control system adjusts the actual temperature, air pressure, and wind speed inside the comprehensive environment simulation test chamber in real time according to the input parameters. Throughout the entire process of the logistics drone performing the predetermined flight maneuvers, the actual values ​​of temperature, air pressure, and wind speed are continuously monitored and recorded, and it is ensured that the actual values ​​of temperature, air pressure, and wind speed remain stable within the allowable fluctuation ranges set by the low temperature environment parameter, low air pressure environment parameter, and high wind environment parameter, respectively.

4. The method for analyzing UAV metrological testing requirements based on multi-parameter coupling according to claim 1, characterized in that, In the multi-environment coupling simulation test platform, the logistics drone is controlled to perform predetermined flight maneuvers, and dynamic change data of multiple key performance parameters of the logistics drone are collected simultaneously, including: The predetermined flight maneuver is set as a sudden climb maneuver under full load conditions, wherein the sudden climb maneuver includes the process of switching from hovering state to maximum power climb state, in order to simulate the most severe power output conditions of the logistics drone in high-altitude areas. In the multi-environment coupling simulation test platform, the logistics drone is controlled to switch from hovering state to maximum power climb state through preset flight control commands, and the maximum power climb state is maintained for a preset period of time. During the process of the logistics drone performing the sudden climb, the real-time change curves of the tension parameter, rotation speed parameter, current parameter, and voltage parameter are simultaneously collected at a rate not lower than the preset sampling frequency, as the dynamic change data.

5. The method for analyzing UAV metrological testing requirements based on multi-parameter coupling as described in claim 1, characterized in that, Based on the dynamically changing data, the coupling relationship between the real environmental disturbance parameters and the multiple key performance parameters is analyzed, and a parameter coupling model describing the coupling relationship is established, including: The dynamically changing data is preprocessed, including data cleaning, outlier removal, and data normalization, to obtain a standardized multidimensional time series dataset. The Pearson correlation coefficient between the real environmental interference parameters and the multiple key performance parameters is calculated using the correlation analysis method. Based on the magnitude of the correlation coefficient, parameter pairs with strong coupling relationships are identified, resulting in multiple strongly coupled parameter pairs. Based on the aforementioned multiple strongly coupled parameter pairs, a parameter coupling model based on a feedforward neural network is constructed.

6. The method for analyzing UAV metrological testing requirements based on multi-parameter coupling as described in claim 5, characterized in that, Based on the aforementioned multiple strongly coupled parameter pairs, a parameter coupling model based on a feedforward neural network is constructed, including: Based on the multiple strongly coupled parameter pairs, the input layer nodes and output layer nodes of the feedforward neural network are determined, wherein the real environmental interference parameters in the multiple strongly coupled parameter pairs are used as the input variables of the input layer nodes, and the key performance parameters in the multiple strongly coupled parameter pairs corresponding to the real environmental interference parameters are used as the output variables of the output layer nodes. The hidden layer structure of the feedforward neural network is configured such that the hidden layer structure includes at least one hidden layer, each hidden layer contains multiple neuron nodes, and the number of hidden layers and the number of neuron nodes are optimized according to the dimensions of the input layer nodes and the output layer nodes and the amount of training sample data. The feedforward neural network is trained using a standardized multidimensional time series dataset obtained after data preprocessing. During the training process, the backpropagation algorithm is used to continuously adjust the weight parameters and bias parameters of the feedforward neural network, so that the loss function value between the predicted output value of the feedforward neural network and the actual collected dynamic data gradually decreases and converges to a preset threshold range. After the feedforward neural network is trained, the trained parameter coupling model is validated using an independent test dataset. The root mean square error is calculated to evaluate the fitting accuracy of the parameter coupling model to the nonlinear mapping relationship between the strongly coupled parameter pairs. If the fitting accuracy does not meet the preset requirements, the training parameters are reset and iterative training is performed until the fitting accuracy meets the requirements.

7. The method for analyzing UAV metrological testing requirements based on multi-parameter coupling as described in claim 1, characterized in that, The parameter coupling model is used to determine the test items, test conditions, and test indicators required for metrological testing of the logistics drone in the actual flight environment, including: Based on the parameter coupling model, at least one sensitive parameter among the multiple key performance parameters whose dynamic fluctuation amplitude exceeds a preset safety threshold under the simultaneous action of the real environmental interference parameters is identified, and the sensitive parameter is determined as the core test item required for the metrological test. Based on the parameter coupling model, the dynamic response characteristics of the core test item under different combinations of real-world environmental interference parameters are simulated and calculated. Based on the dynamic response characteristics, test conditions corresponding to the core test item are formulated. The test conditions include at least the specific numerical range of the real-world environmental interference parameters and the precise execution sequence of the predetermined flight maneuvers. Based on the parameter coupling model and the design safety boundary of the logistics drone, the maximum permissible error range of the core test item in the real flight environment is calculated, and the maximum permissible error range is determined as the test index, which is used to determine whether the performance of the logistics drone is qualified in subsequent metrological tests.

8. The method for analyzing UAV metrological testing requirements based on multi-parameter coupling according to claim 1, characterized in that, It also includes the step of establishing an field data feedback calibration mechanism: At least one representative typical flight station is selected in the plateau region, and the logistics drone is organized to perform the same real flight mission as the predetermined flight maneuver at the typical flight station, and real field flight data of the multiple key performance parameters are collected simultaneously during the real flight. The real field flight data and the dynamic change data collected in the multi-environment coupling simulation test platform are compared and analyzed in multiple dimensions to calculate the deviation values ​​between the two in the time domain and frequency domain. The construction parameters of the multi-environment coupling simulation test platform are adjusted based on the deviation value, and the model parameters of the parameter coupling model are corrected and calibrated so that the environment simulated by the multi-environment coupling simulation test platform is closer to the real flight environment, and the predicted output value of the parameter coupling model is closer to the real field flight data.

9. The method for analyzing UAV metrological testing requirements based on multi-parameter coupling according to claim 1, characterized in that, Generate a metrology and testing requirements analysis report, including: The test items, test conditions, and test indicators shall be written into the main body of the metrological test requirements analysis report in the form of a structured table; The metrology and testing requirements analysis report displays the coupling relationship between the real environmental interference parameters and the multiple key performance parameters in the form of a visual curve or scatter plot, and marks the key coupling nodes and the dynamic change trend of the parameters over time in the coupling relationship graph.

10. A UAV metrology and testing requirements analysis system for multi-parameter coupling, characterized in that, The method for performing UAV metrology and testing requirements analysis oriented towards multi-parameter coupling as described in any one of claims 1-9 includes: The parameter acquisition module is used to acquire low-temperature environmental parameters, low-pressure environmental parameters, and strong wind environmental parameters when the logistics drone performs delivery tasks in plateau areas, as real environmental interference parameters. The platform construction module is used to build a multi-environment coupling simulation test platform, wherein the multi-environment coupling simulation test platform is used to simultaneously apply the real environment interference parameters to simulate the real flight environment in which the logistics drone performs delivery tasks; The flight test module is used to control the logistics drone to perform predetermined flight actions in the multi-environment coupling simulation test platform, and to synchronously collect dynamic change data of multiple key performance parameters of the logistics drone, including tension parameters, rotational speed parameters, current parameters and voltage parameters. The model training module is used to analyze the coupling relationship between the real environmental interference parameters and the multiple key performance parameters based on the dynamically changing data, and to establish a parameter coupling model describing the coupling relationship. The results generation module is used to determine the test items, test conditions, and test indicators required for the metrological testing of the logistics drone in the real flight environment using the parameter coupling model, and to generate a metrological testing requirements analysis report.

Images