A method for testing the wind resistance of unmanned aerial vehicles

Abstract

The application relates to the technical field of unmanned aerial vehicle testing, and discloses an unmanned aerial vehicle wind resistance grading test method; the method is used for solving the problem that the traditional method cannot accurately simulate the combined action of natural turbulence and multidirectional wind, thereby leading to a large deviation of grading results and failing to reflect the real stability of the unmanned aerial vehicle; the method firstly establishes a test session, defines a model state, a load state, a control state, a test boundary and grading conditions, and sets different grades; before starting each test stage, the actual action wind field in which the tested unmanned aerial vehicle is located is determined; in each stage, position keeping, attitude change, disturbance recovery and out-of-boundary state are continuously judged, and the wind resistance grading is determined in combination with direction coverage, fluctuation coverage, boundary rechecking and abnormality processing. The method is favorable for improving the accuracy and reviewability of wind resistance grading under a complex wind field.

Description

Technical Field

[0001] This invention relates to the field of unmanned aerial vehicle (UAV) testing technology, specifically to a method for classifying and testing the wind resistance capability of UAVs. Background Technology

[0002] The wind resistance rating test of unmanned aerial vehicles (UAVs) typically requires testing the flight status of the tested UAV under different wind speeds, directions, and attitudes in a controlled wind environment, and then determining its wind resistance rating based on the test results. Existing technologies already exist that utilize artificially constructed wind environments for UAV-related wind field testing. For example, the published invention patent application CN106226024B discloses a dual-rotor wind tunnel test platform, which enables simultaneous dual-rotor testing in a wind tunnel environment and reduces the impact of the test equipment on the flow field. Another example is the published invention patent application CN106768801B, which discloses a high-precision dual-axis sideslip angle adjustment device for wind tunnel testing. This device can adjust the sideslip angle during wind tunnel testing to meet the testing requirements under different attitude conditions. These existing technologies provide fundamental technical means for UAV wind field testing from the aspects of test platform construction and test attitude adjustment.

[0003] However, the aforementioned technologies address issues such as the establishment of wind tunnel test platforms, rotor drive control, or attitude angle adjustment, primarily improving the feasibility of test implementation and the controllability of test conditions. They cannot fundamentally solve the problem of inconsistency between the test wind field and the actual wind field in unmanned aerial vehicle (UAV) wind resistance rating tests. Existing test environments can only form relatively regular unidirectional wind fields or quasi-steady-state wind fields. Natural wind fields generally exhibit characteristics such as turbulence, gust disturbances, rapid changes in wind direction, and multi-directional airflow coupling. Because it is impossible to fully reproduce these complex wind field characteristics, the wind disturbance experienced by the tested UAV differs significantly from the disturbance experienced during actual flight. This leads to a certain degree of error between the wind resistance rating conclusions obtained from the test results and the actual wind resistance capability. Furthermore, without characterizing variations in turbulence intensity, multidirectional flow changes, and dynamic coupling disturbances, unmanned aerial vehicles (UAVs) of different models, structural forms, or loads cannot be compared under uniform conditions. This may lead to problems such as low accuracy in wind resistance classification, poor repeatability, and low reference value for engineering applications. Therefore, it is necessary to propose a classification test method for the wind resistance capability of UAVs, improve the simulation capability of the test wind field for complex natural wind fields, and establish the correspondence between wind field disturbance characteristics and wind resistance classification results. This would solve the technical problem that the current test environment cannot accurately simulate the combined effects of natural turbulence and multidirectional winds, resulting in large deviations in classification results and an inability to reflect the true stability of UAVs. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides a method for classifying and testing the wind resistance capability of unmanned aerial vehicles (UAVs). This method solves the problem that traditional methods cannot accurately simulate the combined effects of natural turbulence and multi-directional winds, resulting in large deviations in the classification results and failing to reflect the true stability of UAVs.

[0005] To achieve the above objectives, the present invention provides the following technical solution:

[0006] A method for classifying and testing the wind resistance capability of unmanned aerial vehicles, comprising:

[0007] S1: Establish a test session for the unmanned aerial vehicle under test, fix the aircraft type status, load status, control status and test boundaries, and set the wind resistance capability graded test conditions;

[0008] S2: Generate corresponding test phases according to the preset wind field level order, so that each test phase forms a test process with continuous changes in average wind speed, wind direction and wind field fluctuations. According to the wind direction coverage rules and phase generation rules, configure the target average wind speed range, target wind direction range, fluctuation amplitude range, vertical disturbance range and target duration for each level phase, so that the basic maintenance phase, direction change phase, fluctuation enhancement phase and recovery observation phase within the same level form a continuous test process according to the sequential relationship, and at least one of the average wind speed, fluctuation amplitude, direction change intensity or vertical disturbance intensity between adjacent levels increases. Generate a phase definition table and a phase execution sequence table, and determine the connection relationship of each phase according to the re-verification branch, downgrade branch and termination branch.

[0009] S3: Before each test phase, confirm the actual wind field around the unmanned aerial vehicle under test in the test space, and start the current test phase after confirming that it meets the requirements of the current test phase.

[0010] S4: During the current testing phase, continuously determine the position holding state, attitude change state, disturbance recovery state, and boundary crossing state of the tested unmanned aerial vehicle under the influence of the wind field, and enter the corresponding processing procedure based on the determination results;

[0011] S5: Based on the execution and judgment results of each test phase, determine the wind resistance rating of the tested unmanned aerial vehicle.

[0012] Preferably, a test session is established for the tested unmanned aerial vehicle, fixing the aircraft type, load, control status, and test boundaries, and setting graded test conditions for wind resistance capability, including:

[0013] The test session associates the session record table, field constraint table, and input data structure to perform consistency verification and locking on the airframe configuration, propulsion configuration, load level, center of gravity offset, control mode, control version, test space boundary, time boundary, and resource boundary;

[0014] Write the wind speed setting, wind direction setting, fluctuation setting, position deviation, attitude deviation, recovery time, and instantaneous out-of-bounds time according to the classification requirements to form a session initialization record.

[0015] Preferably, the corresponding test phases are generated according to the preset wind field level sequence, including:

[0016] Establish a phase record table based on the test session number, and configure the introduction phase, as well as the basic maintenance phase, direction change phase, and recovery observation phase corresponding to each level, in combination with the machine size, load level, control mode, test boundary, and classification conditions. Also configure the fluctuation enhancement phase and boundary verification phase as needed.

[0017] Perform parameter consistency verification, boundary verification, resource verification, and progression relationship verification on each stage. Stages that pass the verification are set to a locked state, while stages that fail the verification are set to an unexecutable state or deleted and regenerated.

[0018] Preferably, before each test phase, the actual wind field around the tested unmanned aerial vehicle in the test space is confirmed, including:

[0019] Before each test phase, key and auxiliary measuring points are set up around the body to establish a measuring point layout table, a real-time wind field record table, and a phase start-up condition table.

[0020] Based on the measuring points around the aircraft, the upstream reference measuring points, and the measuring points in the main wind-receiving area, the average wind speed, wind direction, wind field fluctuations, vertical disturbances, the number of effective measuring points, and the status of abnormal measuring points are determined.

[0021] The main wind-receiving area is determined based on the target wind direction range, the aircraft's orientation, and the distribution of the wind-receiving surface of the aircraft's shape.

[0022] Preferably, the current testing phase is initiated after confirming that the requirements of the current testing phase are met, including:

[0023] First, assess the validity of the data at each measuring point;

[0024] Then, verify the average wind speed, wind direction, wind field fluctuation, vertical disturbance, continuous satisfaction time and minimum number of effective measuring points in the order of partial confirmation, overall confirmation and establishment of timing. When the preset conditions are continuously met, form a stage confirmation record and enter the execution state.

[0025] When there are abnormalities in the main wind-receiving area measurement points, insufficient number of effective measurement points, excessive total confirmation time, or resource boundary triggering, the current testing phase will be recorded as a confirmation failure or an unformed phase, and will be handled according to the following methods: move forward, skip, downgrade, or terminate the relationship.

[0026] Preferably, during the current testing phase, the test unmanned aerial vehicle's position-holding state, attitude change state, disturbance recovery state, and boundary violation state under the influence of the wind field are continuously determined, including:

[0027] During the current testing phase, flight status records, boundary status records, and event trigger records are established based on the phase confirmation records. A fixed allowable position envelope is constructed using the phase start time and reference position. Combined with the allowable attitude boundary, recovery time, allowable instantaneous boundary crossing time, and attitude warning interval, a joint judgment is made on position offset, attitude change, disturbance recovery, and boundary crossing.

[0028] Recovery time is used to determine whether the tested unmanned aerial vehicle can return to the allowed position envelope and allowed attitude boundaries, while allowed instantaneous boundary crossing time is used to determine whether recovery observation is entered after the boundary crossing.

[0029] Preferably, based on the determination result, the process proceeds to the corresponding step, including:

[0030] The process flows through the following states: stable execution state, momentary out-of-bounds state, recovery observation state, boundary crossing state, failure state, re-verification state, degraded state, and termination state.

[0031] When the position, height, or orientation exceeds the allowed boundaries, record the type and duration of the boundary violation;

[0032] If the recovery is completed within the recovery time and the number of short-term boundary violations does not exceed the preset limit, the system will transition to a stable execution state or a boundary-passing state.

[0033] When data is missing, location jitter occurs, manual intervention is required, or resource boundaries are triggered, the process is handled according to the rules of pending verification, invalid interruption, failure, or termination, and a stage judgment record is generated.

[0034] Preferably, based on the execution status and judgment results of each test phase, including:

[0035] The system includes a call phase execution sequence table, a phase judgment record table, a re-verification record table, an exception event summary table, and a level judgment rule table. Necessary and optional phases are verified separately, and a summary judgment is made according to the rules of level progression, direction coverage, fluctuation coverage, boundary pass and positive conversion, exception removal, and supplementary testing. It also handles boundary pass, supplementary testing, postponement, skipping, and session failure.

[0036] Preferably, the wind resistance rating of the tested unmanned aerial vehicle is determined, including:

[0037] The process proceeds through the following stages: pending level determination, preliminary level establishment, pending level verification, level locked, level not established, and session end. Verification is performed level by level, starting from the lowest level.

[0038] When the current level meets the necessary stage conditions, coverage conditions, and re-verification conditions, it enters the level lock state;

[0039] When a critical phase fails, coverage is insufficient, retesting fails, or session conditions change, higher-level judgments are stopped, and a classification result record is generated.

[0040] Compared with the prior art, the present invention provides a method for classifying and testing the wind resistance capability of unmanned aerial vehicles, which has the following beneficial effects:

[0041] 1. This invention establishes reasonable constraints on the tested unmanned aerial vehicle (UAV) under the same aircraft type, load, control, and test boundaries. It organizes the stages of wind direction change, wind field change, recovery observation, and continuous testing in a graded manner, ensuring no difference when test conditions vary. In the initial stage, the actual wind field around the aircraft is used as a reference for continuous judgment of position maintenance, attitude change, disturbance recovery, and boundary crossing status during the subsequent stages. There is no judgment deviation due to inconsistency between equipment settings and actual wind conditions, nor is there any grade distortion caused by inconsistent test calibers under different aircraft types, structures, loads, etc., thereby achieving a graded classification of the true wind resistance stability of the UAV.

[0042] 2. This invention divides wind resistance testing into progressively graded testing stages, integrating directional coverage, fluctuation coverage, boundary verification, anomaly removal, and grade locking into unified rules. The wind resistance grade no longer depends on the results of each test flight or the temporary passage after a single wind speed, but is determined based on the results of comprehensive verification of multi-directional disturbances, continuous wind field changes, and the recovery process. The judgment results of each stage can be reviewed, supplemented, and graded according to unified rules, reducing the interference of critical state misjudgments and accidental passages on the final classification results, improving the stability and verifiability of the grade classification, and ultimately achieving a reliable determination of the wind resistance capability boundary of unmanned aerial vehicles. Attached Figure Description

[0043] Figure 1 This is a schematic diagram of the wind resistance rating test method for unmanned aerial vehicles according to the present invention. Detailed Implementation

[0044] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0045] Example 1: Figure 1 A method for classifying and testing the wind resistance capability of unmanned aerial vehicles is presented, including:

[0046] S1: Establish a test session for the tested unmanned aerial vehicle (UAV), fix the aircraft type, load, control status, and test boundaries, and set graded test conditions for wind resistance capability. The specific implementation is as follows:

[0047] Before testing, a unique test session is established for each individual aircraft under test. The aircraft status, load status, control status, test boundaries, and decision parameters are all written into the session initialization record. All test phases within the same test session should be conducted under the same conditions. If the object status, load, control parameters, or boundary conditions change during the test, it will no longer be used for formal classification to avoid losing the unified comparison basis for the final classification results and affecting the verifiability of the classification results.

[0048] Test session numbers are generated using traceable hierarchical coding, and at least include fields such as test date, model code, batch number, and session sequence number. Load level identifiers and control version identifiers can also be added. This coding method ensures that each test corresponds to a unique session, and data generated from different dates, batches, load states, and control versions are not bounded by boundaries, facilitating subsequent querying, verification, and comparison. When a test session is established, a session record table and a field constraint table are generated simultaneously. The session record table must at least contain fields such as fixed parameters, dynamic status, and abnormal events. The field constraint table must at least contain fields such as field name, unit of measurement, allowable range, filling definition, and update permission.

[0049] The aircraft status uses fixed status parameters, including at least the following fields: airframe configuration, number of rotors, rotor blade type, rotor blade installation direction, arm deployment status, landing gear status, battery type, mission payload type, payload installation location, center of gravity location, flight control version, inertial measurement calibration status, magnetic compass calibration status, positioning mode, and flight mode. Among these, airframe configuration, number of rotors, rotor blade type, and rotor blade installation direction represent a fixed propulsion layout and basic wind characteristics; payload installation location and center of gravity location represent a fixed inertia distribution; and flight control version, positioning mode, and flight mode represent fixed control response logic and navigation base. Accurate; The above-mentioned aircraft status is locked after being recorded in the session log table and cannot be modified before the end of the current test session; When battery replacement, propeller replacement, mission payload position adjustment, center of gravity rebalancing, flight control parameter rewriting, calibration result update or flight mode switching occurs during the test, the current test session is terminated, the obtained data is no longer used for formal classification, and a new test session is re-established; The reason for adopting this method is that the flight response of the unmanned aerial vehicle in the wind field will directly affect its stability in the wind field, and continuing to use the original test session will make it impossible to have a unified judgment standard for subsequent test results;

[0050] The load status uses fixed weight and center of gravity settings. The weight settings are set according to the relative proportion of the nominal maximum takeoff weight of the tested UAV, and can be selected as 60%, 80%, and 95%. Among them, 60% is used to characterize the wind resistance under light load conditions, 80% is used to characterize the wind resistance under common mission load conditions, and 95% is used to characterize the wind resistance boundary under conditions close to the rated upper limit. The above values ​​are based on the fact that when the load is below 60%, the power margin is usually large, which makes it difficult to reflect the wind resistance boundary under real mission conditions; when the load is above 95%, the load is close to the limit, and the test results are easily affected by loading errors and insufficient safety margins, which is not conducive to forming a stable level boundary. When determining the load, at least the fields of nominal maximum takeoff weight, rated mission load range, and actual mission load should be included, and the corresponding weight settings should be selected according to the normal light load conditions, standard mission conditions, and high load mission conditions. When the actual mission load is not completely consistent with the preset weight settings, the closest setting that is not lower than the actual mission condition and does not exceed the maximum takeoff weight can be selected.

[0051] The center of gravity (CG) is determined based on the aircraft's geometric center, recording the forward / backward and left / right offsets in millimeters. To ensure a clear basis for determining the CG offset values, the tested UAVs can be initially categorized according to their size or maximum takeoff weight. For example, small UAVs can be categorized by wheelbase no greater than 600 mm or maximum takeoff weight less than 7 kg; medium-sized UAVs can be categorized by wheelbase greater than 600 mm and less than 1500 mm, or maximum takeoff weight greater than 7 kg and less than 25 kg; larger aircraft can be further categorized based on these criteria. For small UAVs, the forward / backward and left / right offsets... For medium-sized UAVs, the offset can be within 20 mm; for medium-sized UAVs, the offset can be within 30 mm; for larger models, the offset can not exceed 3% of the wheelbase or main fuselage size. The above values ​​are based on the fact that small and medium-sized UAVs are more sensitive to changes in the center of gravity. When the center of gravity offset is too large, the attitude fluctuations and recovery delays that occur during the test reflect more the impact of abnormal loading than the actual wind resistance of the UAV itself. After the weight range and center of gravity range are determined, at least the fields such as load type, test weight, forward and backward offset, and left and right offset should be written and associated with the corresponding test session as the basis for subsequent test condition calls.

[0052] The control status adopts a unified control mode, unified control version, and unified control parameter configuration. The control status includes at least the following fields: flight control mode, flight control program version, limiting strategy version, control loop parameters, amplitude limiting parameters, filtering parameters, compensation parameters, and protection thresholds. The flight control mode remains consistent within the same test session, and can be selected as position hold mode, altitude hold mode, or orientation hold mode. The flight control program version and limiting strategy version remain consistent throughout the entire test session. The control loop parameters, amplitude limiting parameters, filtering parameters, compensation parameters, and protection thresholds remain unchanged at each level and stage, without temporary changes. The reason for adopting this approach is that wind resistance rating reflects the differences in flight response caused by different wind field conditions; otherwise, it would be difficult to switch control modes, change versions, or adjust control parameters. Wind resistance comparisons under different control strategies can be conducted through separate test sessions.

[0053] The test boundaries include spatial boundaries, temporal boundaries, and resource boundaries. The spatial boundaries establish a three-dimensional coordinate region based on the center point of the test site, including at least an effective test area, a buffer zone, and an emergency evacuation zone. The effective test area is used for wind resistance grading tests, the buffer zone is used to accommodate short-term offsets and recovery processes, and the emergency evacuation zone is used for exiting after triggering safety conditions. The horizontal dimension of the effective test area can be selected to be no less than 6 times the maximum diagonal dimension of the tested UAV and no less than 3 times the expected maximum recovery offset. For example, for small UAVs, the horizontal dimension can be 6 meters by 6 meters to 12 meters by 1... 2 meters; for medium-sized unmanned aerial vehicles (UAVs), the height can be appropriately increased based on wheelbase, inertia, and safety margin. The basis for adopting the above range is that the lower limit of the horizontal dimension needs to ensure that the UAV under test has the necessary maneuvering space during gust disturbances and recovery, so as to avoid normal recovery being judged as exceeding the limit. If the horizontal dimension is too large, it will increase the difficulty of wind field control and reduce the level differentiation. The height boundary can be selected from 3 meters to 15 meters. The basis for its selection is that the lower limit of the height should be more than 3 times the rotor diameter to reduce the impact of ground effect on position holding and attitude compensation. The upper limit of the height should take into account both wind field controllability and safety handling requirements.

[0054] The time boundary is used to constrain the execution duration of the test session and each test phase. For small and medium-sized electric unmanned aerial vehicles, the total duration of a single test session can be selected to be no more than 30 minutes, and the duration of a single test phase can be selected to be between 10 and 60 seconds. The reason for selecting a total duration of no more than 30 minutes for a single test session is that when the test duration is too long, the impact of battery degradation, power temperature rise, and environmental changes on the consistency of test results increases. The reason for selecting a duration of 10 to 60 seconds is that when the duration is less than 10 seconds, it is difficult to complete the effective disturbance establishment and recovery observation, and when the duration is more than 60 seconds, it will increase the impact of power and thermal state fluctuations on the test results.

[0055] Resource boundaries are used to constrain the battery status, temperature rise status, and human intervention status. The battery threshold can be set based on the stable discharge range of the test UAV's power system. For common small and medium-sized electric UAVs, the initial battery charge can be selected as no less than 85%, the charge before entering the formal grading test can be selected as no less than 70%, and the test termination battery threshold can be selected as 30%. The above values ​​are based on the fact that above 85% usually ensures relatively stable output capability at the start of the test, above 70% usually ensures basically stable power response during the formal grading test, and below 30% is prone to triggering low battery protection, thrust decay, or return-to-home control logic, thus affecting the objectivity of the test results. The temperature rise boundary can be set based on the allowable continuous operating temperature of the power system and the allowable discharge temperature of the battery. When the derating boundary is reached, higher-level testing will not proceed. The human intervention boundary includes at least manual takeover, manual correction, and manual cancellation of protection actions. Once any of these situations occur, the current test session is marked as a manual intervention session, and formal grading results will no longer be output.

[0056] After determining the above conditions, establish the test session input data structure. The input data structure is stored using a structured record table, with each field having a field name, field type, unit of measurement, allowable range, filling rules, and update status. The input data includes at least the following fields: session number, aircraft type identifier, rotor configuration, test weight, center of gravity offset, load type, control version, control mode, test boundary coordinates, initial power, ambient temperature, ambient humidity, site type, test date, and operator identifier. The unit of measurement for test weight can be kilograms, for center of gravity offset millimeters, for power, for spatial boundary coordinates, for ambient temperature, and for ambient humidity, it can be percentage. To facilitate subsequent retrieval, the input data can be divided into fields such as object information, control information, boundary information, environmental information, and constraint information according to the record content. Object information is used to identify the uniqueness of the test object; control information is used to verify whether the control mode and control version have changed; boundary information is used for position maintenance and boundary violation determination; environmental information is used to analyze wind field formation conditions; and constraint information is used for generation and status determination during the test phase.

[0057] Based on the input data structure, a wind resistance capability grading condition table is further established. The grading condition table, in order of grade, includes at least the following fields: average wind speed level, wind direction level, wind field fluctuation level, allowable position deviation, allowable attitude deviation, allowable recovery time, allowable instantaneous boundary crossing time, and retest triggering rules. Various thresholds are preferably set in gradations according to aircraft size, maximum takeoff weight, control mode, test altitude, and target wind speed level. When the preset values ​​are inconsistent with the parameters in the test UAV's manual or the actual mission conditions, values ​​no lower than those required by the actual mission conditions and without exceeding the test safety margin can be selected within the corresponding grading range. The closest value to the boundary; the average wind speed range can be, for example, 5 m / s to 7 m / s, 7 m / s to 9 m / s, or 9 m / s to 11 m / s, and preferably the average wind speed range of adjacent levels is increased by at least 2 m / s; the above values ​​are based on the fact that 5 m / s to 7 m / s can be used to characterize the basic wind resistance under low to moderate wind disturbance conditions, 7 m / s to 9 m / s can be used to characterize the stability maintenance capability under common complex wind field conditions, and 9 m / s to 11 m / s can be used to characterize the boundary maintenance capability under stronger wind disturbance conditions, and the incremental increase of no less than 2 m / s between adjacent levels is conducive to forming a clear level distinction;

[0058] The permissible positional deviation is preferably determined according to the size of the aircraft and the test mode. For small unmanned aerial vehicles (UAVs), the horizontal deviation can be, for example, 0.5 meters to 1 meter, and preferably not less than twice the diagonal size of the aircraft. The altitude deviation can be, for example, ±0.3 meters to ±0.8 meters. The reason is that small UAVs have high control precision, but are easily affected by instantaneous wind disturbances. Therefore, a reasonable judgment range needs to be left between normal control error and actual instability. The permissible attitude deviation can be, for example, the absolute values ​​of roll angle and pitch angle not exceeding 20° to 30°. The reason is that after exceeding this range, the aircraft usually enters a continuous high-compensation state and should no longer be regarded as stable. The allowable recovery time is preferably determined by combining the aircraft size, weight range, and target wind speed range, for example, it can be 1 to 5 seconds. The reason is that the larger the aircraft size, the higher the load, and the higher the target wind speed, the greater the aircraft inertia and recovery burden, and the longer the recovery time required. The allowable instantaneous boundary breach time is preferably determined by combining the flight control cycle, the state sampling cycle, and the expected gust peak duration, for example, it can be 0.3 to 1 second, and preferably no less than 3 consecutive control cycles and no more than 1 complete recovery judgment cycle. The reason is that gust peaks in complex wind fields may cause short-term boundary breaches, but when the duration exceeds this range, it is usually no longer a normal instantaneous disturbance response.

[0059] To ensure clear entry, exit, and exception handling conditions during the test session establishment process, the process is divided into four states: preparation, parameter locking, session activation, and exception blocking. The process begins in the preparation state, where at least the session number, aircraft type identifier, load information, control information, test boundaries, and classification conditions are entered, and consistency checks are performed. Once the basic fields are filled in, and the aircraft type, load, control, and test boundaries meet preset conditions, the process enters the parameter locking state. In this state, critical fields are read-only and cannot be modified. Finally, the site conditions, equipment conditions, and the current state of the tested UAV must pass the test. When the session is active, it can proceed to the subsequent testing phase. If any key fields are not met, there are conflicts between fields, the range is exceeded, the resource status is below the minimum requirement, or the key verification parameters have changed, the session will be in an abnormal blocking state and will not directly proceed to the subsequent process. At least one of the following actions will be taken: suspension, correction, or reconstruction. If the abnormality is due to a filling error or a unit error, it can be corrected and the test drone can be put back into the preparation state if the status remains unchanged. If the abnormality is due to a change in the object status, the battery can be replaced, the center of gravity can be re-balanced, or the control version can be changed. The current test session will then be stopped and renumbered.

[0060] The session initialization record includes at least the following fields: session number, fixed parameters, valid boundaries, hierarchical conditions, and status identifier. The status identifier is used to indicate whether the current test session meets the conditions for entering the subsequent test phase generation process.

[0061] S2: Generate corresponding test phases according to the preset wind field level order, so that each test phase forms a test process with continuous changes in average wind speed, wind direction, and wind field fluctuations. Based on the wind direction coverage rules and phase generation rules, configure the target average wind speed range, target wind direction range, fluctuation amplitude range, vertical disturbance range, and target duration for each level phase. Make the base maintenance phase, direction change phase, fluctuation enhancement phase, and recovery observation phase within the same level form a continuous test process according to their sequential relationship. Ensure that at least one of the average wind speed, fluctuation amplitude, direction change intensity, or vertical disturbance intensity increases between adjacent levels. Generate a phase definition table and a phase execution sequence table. Determine the connection relationship between each phase according to the re-verification branch, downgrade branch, and termination branch. The specific implementation is as follows:

[0062] After the session initialization record is generated, a test phase sequence corresponding to the current state of the unmanned aerial vehicle under test is generated according to the classification conditions in the record. Each test phase is generated according to the continuous change relationship of average wind speed, wind direction, wind field fluctuation, vertical disturbance, and recovery observation.

[0063] During the test phase, a phase record table is generated using the session number as an index, and the aircraft size, maximum takeoff weight, weight level, center of gravity level, control mode, test area size, power boundary, and classification conditions are read. The phase record table uses a structured field format to store at least one phase number, level, phase sequence, phase type, target average wind speed lower limit, target average wind speed upper limit, target wind direction range, target fluctuation range, target vertical disturbance range, target duration, start-up preconditions, completion conditions, recovery requirements, re-verification flag, downgrade flag, and status flag. The units for average wind speed, fluctuation range, and vertical disturbance are selected as meters per second, the wind direction is selected as an angle range, the duration is selected as seconds, and the status flag is selected as a preset status code. After this recording method is used, the subsequent confirmation of the actual wind field, phase execution judgment, and level judgment are all made using the same phase parameters.

[0064] The testing phase can be divided into an introduction phase and multiple graded testing phases. The introduction phase is used to determine the basic maintenance status of the tested unmanned aerial vehicle under low-intensity wind disturbance, as well as test sessions that cannot continue formal graded testing due to abnormal loading status, abnormal control status, or insufficient resource status. The graded testing phases increase the wind field complexity in order of grade. Within the same grade, the phases can be set in the order of basic maintenance, directional change, fluctuation enhancement, and recovery observation. If necessary, a boundary verification phase can also be set. Among them, basic maintenance determines the basic wind state under the current grade, directional change determines the changes in multi-directional incoming flow, fluctuation enhancement determines gusts and fluctuation disturbances, recovery observation determines the recovery after disturbance, and boundary verification determines the critical stage. Through these settings, all grades are composed of interrelated testing phases, and the grade determination is based on the comprehensive test results under continuously changing wind fields.

[0065] The stage parameters can be set according to the aircraft size, maximum takeoff weight, weight class, test altitude, site safety margin, and wind field construction capability. The average wind speed, fluctuation amplitude, vertical disturbance, and duration are within the range of values ​​for small and medium-sized electric unmanned aerial vehicles. If the parameters in the instruction manual, mission conditions, and site conditions of the unmanned aerial vehicle under test do not match the range of values, the nearest value that meets the actual test needs and does not exceed the safety boundary can be selected within the range.

[0066] The target average wind speed during the introduction phase can be selected as a low-disturbance range below the lower limit of the official minimum level, such as 1 to 5 meters per second; the fluctuation range can be selected as 5% to 10% of the average wind speed; the target duration can be selected as 15 to 20 seconds. The basis for adopting the above value range is that the introduction phase is mainly used to verify the basic maintenance status of the test unmanned aerial vehicle and whether the current test session has the conditions to continue execution. The wind disturbance intensity should not be too high, but it should be able to identify basic maintenance anomalies under mild disturbance conditions. When the wind disturbance intensity is below this range, the screening effect is insufficient; when the wind disturbance intensity is above this range, it is easy to trigger boundary instability before the official level determination.

[0067] The formal classification stages are set in a progressive manner; the target average wind speed for the first level can be selected from 5 to 7 meters per second, the target wind direction must cover at least three types of directions: frontal, crosswind, and diagonal, the fluctuation range can be selected from 10% to 20% of the average wind speed, the target duration can be selected from 20 to 30 seconds, and the vertical disturbance amplitude can be selected from 0.5 to 1 meter per second; the target average wind speed for the second level can be selected from 7 to 9 meters per second, the fluctuation range can be selected from 15% to 25%, the direction change angle can be selected from 45° to 90°, the vertical disturbance amplitude can be selected from 1 to 1.5 meters per second, and the target duration can be selected from 25 to 35 seconds; the target average wind speed for the third level can be selected from 9 to 11 meters per second, the fluctuation range can be... The parameters can be selected as 20% to 30%, the direction change can be selected as multiple switches, the vertical disturbance amplitude can be selected as 1.5 m / s to 2 m / s, and the target duration can be selected as 30 to 45 seconds. The above parameters are set progressively according to the complexity of the wind field. Among them, the average wind speed is used to characterize the intensity of the horizontal incoming flow, the fluctuation amplitude is used to characterize the degree of wind speed fluctuation, the vertical disturbance amplitude is used to characterize the additional disturbance in the direction of rise and fall, and the target duration is used to ensure that each level has a sufficient disturbance observation window. For models with larger loads or higher wind resistance levels, higher levels can be added on the basis of the third level. When the candidate level exceeds the wind resistance value specified in the model manual, the site wind field construction capability, or the resource boundary of the current test session, higher level stages will no longer be generated.

[0068] The above parameter ranges are based on the following criteria: Average wind speed is used to characterize horizontal flow conditions of different intensities; the target average wind speed for adjacent levels can be increased by at least 2 meters per second. When the difference in average wind speed is too small, the wind resistance differences between different levels are difficult to distinguish. Fluctuation amplitude is used to characterize the degree of wind speed fluctuation; the fluctuation amplitude for adjacent levels can be increased by at least 5 percentage points. When the fluctuation amplitude is below 10%, the characteristics of complex wind fields are not obvious; when the fluctuation amplitude is above 30%, the test results tend to overemphasize extreme peaks, reducing the distinguishability of sustained wind resistance. Vertical disturbance is used to characterize additional disturbances in the direction of uplift and subsidence; the vertical disturbance for adjacent levels can be selected as... The target wind speed should be increased by at least 0.5 m / s and kept lower than the average wind speed of the same level. When the vertical disturbance is less than 0.5 m / s, the heave compensation effect is not obvious. When the vertical disturbance is greater than 2 m / s, the test is likely to mainly reflect strong heave interference, reducing the level differentiation under horizontal flow conditions. The target duration should be appropriately extended as the level increases to ensure that there is a sufficient observation window in the high-level stage to cover the stability maintenance, disturbance response and recovery process. For small unmanned aerial vehicles under high load conditions, the upper limit of the target average wind speed can be determined in combination with the nominal wind resistance value in the model manual, the current weight level and safety margin. For example, it can be selected as no higher than 11 m / s.

[0069] Wind direction parameters are constrained using a wind direction coverage table. This table includes at least the fields for frontal wind, tailwind, left-side wind, right-side wind, front-left oblique wind, front-right oblique wind, rear-left oblique wind, and rear-right oblique wind. Each wind direction is represented by an angle range, with the range width selectable from 30° to 60°. The range width is chosen because a narrow range increases the difficulty of wind field construction, while a wide range reduces the distinction between different wind direction phases. Once established, the wind direction coverage table is bound to the session number and remains unchanged within the same test session. When generating test phases, different wind direction conditions are configured sequentially according to the level requirements and wind direction coverage rules. For example, at the same level, a basic frontal wind phase, a crosswind change phase, and an oblique wind recovery phase can be generated to ensure that the level determination is based on the comprehensive performance under multi-directional flow conditions.

[0070] The phase generation rules are used to determine the composition of the test phases corresponding to each level and the sequential relationships between the test phases. The phase generation rules include at least the minimum number of phases, necessary phase types, optional enhancement phases, directional coverage requirements, fluctuation increment requirements, vertical disturbance introduction conditions, recovery observation insertion conditions, re-verification insertion conditions, and degradation connection conditions. Typically, a formal level includes at least three necessary phases: a base maintenance phase, a directional change phase, and a recovery observation phase. When it is necessary to improve the accuracy of the judgment, a fluctuation enhancement phase can be inserted after the directional change phase, and a boundary re-verification phase can be inserted after the recovery observation phase. Through these rules, each test phase can have clear test content and sequential relationships, avoiding the mixing of different wind field conditions in the same time period, and facilitating subsequent corresponding analysis of the test results.

[0071] The generation of test stages can be achieved by combining sequential generation with constraint verification. First, candidate test stage sequences are generated in order of level. Then, executability verification is performed based on the boundary conditions and fixed parameters in the session initialization record. During generation, the completeness of test stages within the same level is checked first, followed by checking whether the parameters between adjacent levels maintain a progressive relationship. Test stages within the same level can include at least a basic maintenance stage, a direction change stage, and a recovery observation stage. The progressive relationship between parameters between adjacent levels can be expressed as follows: at least one of the following parameters—average wind speed, fluctuation amplitude, direction change intensity, or vertical disturbance intensity—is higher than the previous level, and not lower than the corresponding locked parameter of the previous level. If the lower limit of the target average wind speed of the next level is lower than the upper limit of the target average wind speed of the previous level, or if the fluctuation amplitude of the next level is not higher than the previous level, and the wind direction change intensity and vertical disturbance intensity have not increased, then the test stage definition for that level is not adopted. The reason for using this verification method is that if there is parameter overlap or inversion between levels, it will lead to unclear level boundaries and affect the stability of the final classification results.

[0072] The feasibility verification can be sequentially divided into wind speed limit verification, site size verification, and resource status verification. Wind speed limit verification determines whether the target average wind speed exceeds the test safety limit for the current model at the current weight level. This limit can be predetermined based on the model's nominal wind resistance, weight level, control mode, and safety margin. When the target average wind speed in a candidate stage exceeds the test safety limit, that stage and its higher-level stages are marked as unexecutable. Site size verification determines whether the test space is sufficient to support wind direction changes and recovery processes in the current stage. When the site's lateral dimensions are insufficient to support large-angle crosswind switching and recovery observation, stages containing large-angle directional changes are marked as site-restricted. The current test session does not include the phase; resource status verification is used to determine whether the remaining power and available time are sufficient to support the execution of the current phase sequence; when the remaining power does not match the total duration of the candidate phase, the duration of a single phase can be shortened first, wherein the duration of the basic hold phase and the direction change phase can be selected to be no less than 10 seconds, and the duration of the recovery observation phase can be selected to be no less than 5 seconds; the above values ​​are based on the fact that the basic hold phase and the direction change phase must cover at least one stable hold window and one obvious disturbance response window, and the recovery observation phase must cover one complete recovery process; if the minimum effective observation time is still not met after shortening, the corresponding level is marked as insufficient resources and will not enter the locked state;

[0073] The anomaly and boundary handling identification phase defines anomalies, phase duplication, and resource insufficiency. When a phase during generation has similar wind direction coverage to the previous phase, and the average wind speed, fluctuation amplitude, and vertical disturbance are not enhanced, this level is not considered a phase and is not required. Instead, an alternative phase is generated based on insufficient wind direction coverage or insufficient wind field enhancement. When a level lacks necessary phase types, such as only including the base maintenance phase and excluding the direction change phase, this level is not considered structurally incomplete and does not participate in subsequent actual wind field confirmation. When a level, after generation, conflicts with test session boundaries, such as height boundaries, time boundaries, or remaining resource boundaries, this level cannot be executed, while the phases already locked in the previous level are retained. Through the above processing, a closed-loop process of anomaly identification, anomaly marking, phase blocking, and regeneration can be formed.

[0074] To clarify the generation conditions and state transition relationships of the testing phase, the phase states are managed. Phase states include at least four states: pending generation, pending verification, locked, and unexecutable. Each candidate phase initially starts in the pending generation state, enters the pending verification state after phase parameters are written, and enters the locked state after passing parameter consistency verification, session boundary verification, and resource executability verification. If any critical verification fails, the phase enters the unexecutable state. Once in the locked state, the target average wind speed, wind direction range, fluctuation amplitude, vertical disturbance, and duration are no longer directly modified. If adjustments are necessary, the current phase is deleted and regenerated. The locked phase serves as the input condition for subsequent confirmation of the actual wind field and phase execution determination.

[0075] After the phase is generated, a phase definition table and a phase execution sequence table are generated. The phase definition table should include at least the target average wind speed range, target wind direction range, target fluctuation range, target vertical disturbance range, target duration, start conditions, completion conditions, and recovery requirements, which serve as the basis for confirming the actual wind field effect. The phase execution sequence table should include at least the phase before and after the phase, verification branches, downgrade branches, and termination branches, which serve as the basis for phase execution judgment and level closure. For example, if a phase with a change in direction at a certain level is judged to have passed the boundary, the phase execution sequence table will point to the boundary verification phase of the same level. If a phase with a basic maintenance at a certain level fails to be judged, the phase execution sequence table will point to the downgrade branch or the session termination branch, and it will be impossible to enter a higher level phase. After the phase definition table and the phase execution sequence table are generated, the current test session will enter the subsequent actual wind field confirmation process.

[0076] S3: Before each test phase, confirm the actual wind field around the unmanned aerial vehicle under test in the test space. After confirming that the requirements of the current test phase are met, start the current test phase. The specific implementation is as follows:

[0077] Using the phase definition table and phase execution sequence table as input, the actual wind field around the test UAV is confirmed before the phase start confirmation. The current phase is not started directly based on the setting parameters of the wind field generating equipment. The reason for adopting this method is that the equipment setting value cannot directly represent the actual wind conditions around the test UAV. When there is local backflow, boundary reflection or rotor-induced flow interference, there may be a deviation between the equipment setting value and the actual wind field around the aircraft.

[0078] Complete the pre-launch confirmation for the current testing phase, and establish a test space measurement point layout table, a real-time wind field record table, and a phase launch condition table. The test space measurement point layout table should at least include measurement point number, measurement point spatial location, measurement direction, installation height, and measurement point category. The real-time wind field record table should at least include recording time, measurement point number, average wind speed, instantaneous wind speed, wind direction angle, vertical velocity component, data validity identifier, and anomaly identifier fields. The phase launch condition table should at least include average wind speed compliance interval, wind direction compliance interval, wind field fluctuation compliance interval, vertical disturbance compliance interval, duration of compliance, main windward area condition, minimum number of valid measurement points condition, and number of abnormal measurement points condition fields. The above records are associated with the test session number and phase number, and are used to record the wind field formation process and launch confirmation status of the current testing phase.

[0079] The measurement points are arranged around the perimeter of the tested UAV, using a cluster of key measurement points instead of relying solely on single-point wind measurements upstream of the test site to determine whether the current test phase can commence. Key measurement points preferably cover the area in front of, behind, to the left, to the right, above, and upstream of the aircraft. When the current test phase involves large-angle directional changes or significant vertical disturbances, auxiliary measurement points can also be placed diagonally in front of, behind, or below the aircraft. For small UAVs, the radius of key measurement point placement can be selected from 0.5 to 2 meters, and the number of key measurement points can be selected from 5 to 9. For medium-sized UAVs, the radius of key measurement point placement can be selected from 1 to 3 meters, and the number of key measurement points can be selected from 7 to 11. The reason for using these ranges is that when measurement points are too close to the aircraft, they are easily affected by rotor-induced flow and airflow around the aircraft, resulting in the measured wind field reflecting more the propulsive airflow than the externally influenced wind field. When the measuring points are too far from the aircraft, it is difficult to accurately represent the actual wind conditions around the aircraft. The number of key measuring points is determined by the fact that wind field confirmation should at least cover the reference areas in front of, to the side of, above and upstream of the aircraft to ensure that the main incoming flow direction and key disturbance direction are observable. The installation height of each measuring point is preferably the same as or close to the current test height of the UAV under test, and the height deviation can be selected from 0.2 meters to 1 meter. The reason for using this range is that the lower limit is used to avoid installation errors and local measurement interference, and the upper limit is used to ensure that the measuring points are still within the main wind-affected height range of the UAV under test, thereby reducing the confirmation deviation caused by the difference in vertical wind speed distribution. When the test height is low, the height deviation is preferably smaller. When the test height is high and the site layout is limited, it can be appropriately widened without deviating from the main wind-affected height range. When the test height changes in stages, the installation height of the measuring points is adjusted synchronously.

[0080] The real-time wind field recording table uses a time-series structure to store measurement point data, including at least the fields of recording time, measurement point number, average wind speed, instantaneous wind speed, wind direction angle, and vertical velocity component. The recording period can be selected to be no more than 50 milliseconds, with a preferred value of 10 to 20 milliseconds. The rationale for using this range is that an excessively long recording period will lose the key transient processes of wind field fluctuations and wind direction switching, which is not conducive to accurately confirming the fluctuation amplitude and establishment time. Although an excessively short recording period can improve resolution, it will increase the amount of data and processing burden. Therefore, in most engineering test scenarios, 10 to 20 milliseconds can balance recording accuracy and processing efficiency. To ensure that the confirmation results of each test stage are compared under a unified standard, the average wind speed is taken as the interval average within a continuous time window, the instantaneous wind speed is taken as the value at the current sampling time, the wind direction angle is taken as the angle value at the current sampling time relative to the fixed direction of the test space or relative to the direction of the nacelle, and the vertical velocity component is taken as the vertical velocity value at the current sampling time. The units for all fields are uniformly set in meters per second and angle.

[0081] The phase initiation condition table is generated based on the current test phase definition and includes at least the following conditions: average wind speed compliance range, wind direction compliance range, wind field fluctuation compliance range, vertical disturbance compliance range, and duration of compliance. Taking a test phase with a target average wind speed of 7 to 9 meters per second as an example, the average wind speed compliance range condition requires that the average wind speed at effective measuring points around the aircraft be between 7 and 9 meters per second within a continuous confirmation window. The wind direction compliance range condition, for example, is based on the wind direction angle at the measuring point entering the corresponding angle range. The wind field fluctuation compliance range condition, for example, is based on the fluctuation amplitude being 15% to 25% of the average wind speed, using the ratio of the wind speed change amplitude to the average wind speed within a continuous window as the basis for judgment. The vertical disturbance compliance range condition... For example, the value can be set to 0.8 m / s to 1.5 m / s, with the vertical velocity component falling within the corresponding range as the criterion for judgment; the continuous satisfaction time requirement is that the above conditions must be met simultaneously within a continuous time period; the continuous satisfaction time can be selected from 3 seconds to 10 seconds, with 5 seconds being recommended; the basis for this value range is that if the establishment time is too short, it is easy to mistakenly judge the overlap of short-term sporadic conditions as the current test phase having been stabilized; if the establishment time is too long, it will occupy test session time and power resources, reducing the overall test efficiency; the establishment time can be configured according to the phase type and target wind speed setting, where a shorter establishment time can be selected for the basic maintenance phase, and a longer establishment time can be selected for the direction change phase and the fluctuation enhancement phase, so as to ensure that the complex wind field transformation is completed before entering the formal execution state;

[0082] The pre-launch confirmation process for the current testing phase follows this sequence: partial confirmation, then overall confirmation, followed by timing establishment, and finally, triggering the launch. During partial confirmation, data validity is assessed for each key measurement point. If a measurement point experiences continuous data loss exceeding one second, or if the measured value jumps beyond a preset physical reasonable range, or if the direction data is invalid for multiple consecutive sampling periods, that measurement point is temporarily marked as invalid and will not participate in the overall assessment for the current confirmation period. The one-second continuous data loss time threshold is chosen because it should be greater than the duration of short-term communication jitter and instantaneous sampling packet loss, but not so long as to affect the formation of the judgment in the current testing phase. An interruption can be determined by multiple consecutive sampling periods of abnormality that have exceeded the duration of short-term communication jitter. The preset reasonable physical range can be determined based on the upper limit wind speed of the current test session's model, the upper limit wind speed of the site, and the upper limit of the equipment's output. For example, if the instantaneous wind speed exceeds twice the maximum target wind speed of the current test session, and there is no corresponding change in the sampling values ​​before and after, it can be judged as an abnormal jump. The basis for setting this threshold to twice is that this value can reduce the risk of misjudging normal fluctuations as abnormal while retaining the ability to identify gust peaks. The specific value can be determined by combining the upper limit of the target wind speed of the current test level, the site's extreme wind field capability, and the noise level of the measuring point.

[0083] Overall confirmation refers to, after removing invalid measuring points, first determining whether the number of remaining valid measuring points reaches the minimum number of valid measuring points, and then determining whether the activation conditions such as average wind speed, wind direction, wind field fluctuations, and vertical disturbances are simultaneously met. For small unmanned aerial vehicles (UAVs), the minimum number of valid measuring points can be no less than 5; for medium-sized UAVs, it can be no less than 7. This value is based on the fact that the wind field confirmation before the start of the current testing phase should at least cover the main incoming flow direction, the lateral incoming flow direction, the direction of the overhead disturbance, and the upstream reference area. If the number of valid measuring points is insufficient, it will be difficult to confirm the validity of the test. To determine whether a complete wind field has formed around the aircraft, the current test phase remains in a pending confirmation state and is not initiated. The allowable number of abnormal measuring points can be set based on the total number of measuring points and the number of measuring points in the main wind-receiving area, preferably not exceeding one-third of the total number of measuring points, and preferably having no abnormal measuring points in the main wind-receiving area. When relaxed, the number of abnormal measuring points should not exceed one. The reason for adopting this range is that when a small number of abnormalities occur at the peripheral auxiliary measuring points, it is not advisable to directly terminate the confirmation of the current test phase; if there are too many abnormal measuring points in the main wind-receiving area, it will significantly weaken the authenticity and completeness of the wind field confirmation results.

[0084] Establishing a timer means that, provided the number of valid measurement points meets the requirement, further determining whether the average wind speed, wind direction, wind field fluctuation, and vertical disturbance within the current confirmation window simultaneously meet the start conditions for the current test phase; the timer is only started when all the above conditions are met simultaneously; if only some conditions are met, such as the average wind speed and wind direction being met but the fluctuation amplitude being insufficient, or the fluctuation amplitude being met but the vertical disturbance being insufficient, the current test phase remains in a partially satisfied state and does not enter the timer establishment phase; after the timer establishment phase is reached through continuous satisfaction, the current test phase switches from the pending start state to the execution state, and a phase confirmation record is generated; using this confirmation method can avoid prematurely starting the current test phase due to the satisfaction of a single condition or short-term overlap of conditions;

[0085] During pre-start confirmation, upstream reference measuring points around the aircraft are first checked. If the average wind speed, wind direction, and fluctuation of the upstream reference measuring points meet the conditions of the current test phase, but the measuring points in front of, to the side of, or above the aircraft do not meet the requirements, then the results of the actual measuring points around the aircraft shall prevail, and the current test phase shall not be started. The reason for adopting this rule is that the upstream reference measuring points can only represent the input airflow state and cannot guarantee that the target wind field has been formed around the aircraft. If the average wind speed meets the requirements but the wind direction deviation is too large, the wind speed condition shall not replace the overall condition. When the vertical disturbance is less than the corresponding condition, the start-up condition can be postponed for a period of time, but the conditions of the current test phase cannot be reduced, nor can the state of no vertical disturbance be considered as having met the start-up conditions.

[0086] To ensure that anomaly handling matches the current wind field characteristics, priority can be given to determining the main windward area. The main windward area can be determined based on the target wind direction range, aircraft orientation, and the distribution of the windward surface of the aircraft shape in the current test phase. That is, the measurement point area that is most affected by the wind load of the aircraft under the current incoming flow conditions. For example, under the condition of left front wind, the measurement points on the left front, front, and top of the aircraft are given priority as the main windward area; under the condition of left and right crosswind, the measurement points on the left and right sides of the aircraft are given priority as the main windward area. If key measurement points in the main windward area show continuous anomalies, long-term failures, or severe jumps, the current test phase will not be carried out and will be recorded as confirmation failure or partial invalidity. If the abnormal measurement point is not in the main windward area, and the number of remaining valid measurement points meets the minimum number of valid measurement points, wind field confirmation can begin after recording the anomaly mark. This can avoid the distortion of the main windward area being masked by the peripheral measurement points, and can also avoid unnecessary phase interruptions caused by uniformly applying the same anomaly threshold to all measurement points.

[0087] The time constraints employ a two-tiered control system: setup time constraint and total confirmation time constraint. The setup time constraint refers to maintaining a stable wind field for 3 to 10 seconds after all conditions are met before the current test phase enters the execution state. This range is based on the fact that a setup time that is too short may lead to misjudging short-term conditions as a stable wind field, while a setup time that is too long will increase power and time resource consumption. The total confirmation time constraint refers to the longest waiting time from the start of the current test phase until the conditions for activation are met or the conditions for activation are determined to be unavailable. The total time can be set to no more than 60 seconds. This upper limit is based on the fact that if a wind field meeting the definition of the current test phase cannot be formed after 60 seconds, it usually indicates that the current site conditions, equipment capabilities, or test session resources are insufficient to support continued waiting. If a wind field meeting the definition of the current test phase cannot be formed within 60 seconds, the test phase will proceed as planned. If the actual wind field defined in the previous testing phase is not formed, the current testing phase will be marked as an incomplete phase. Incomplete phases are not directly treated as failed phases, but are handled in conjunction with the remaining power, remaining time, priority of subsequent phases, and phase execution order table of the current testing session. If the current testing phase is a necessary phase of this level and the session resources are sufficient, it can be moved to the end of this level for retry, with the number of moves being no more than 1 or 2. If it still does not form after moving, it is treated as incomplete. If the current testing phase is an optional enhancement phase, it can be skipped directly and treated as insufficient coverage during level determination. If the remaining resources are insufficient to continue waiting, the current testing session enters a downgrade or termination branch. By distinguishing between incomplete phases and failed phases, the insufficiency of site wind field construction can be avoided as a misjudgment of insufficient wind resistance of the tested UAV.

[0088] The pre-start confirmation of the current testing phase is managed by a state machine. The states include at least the pending confirmation state, partially satisfied state, overall satisfied state, execution state, and confirmation failure state. The pending confirmation state indicates that the start conditions have not yet been met, or only the measurement point initialization has been completed. The partially satisfied state indicates that some conditions are met, but not all conditions. The overall satisfied state indicates that all conditions are met and the setup timer has started. The execution state indicates that the setup time requirement has been met, and the current testing phase has started. The confirmation failure state indicates that the start cannot be achieved due to prolonged failure to meet key conditions, failure of measurement points in the main windward area, insufficient minimum number of effective measurement points, resource boundary triggering, or session boundary conflict. State transitions are performed according to preset conditions. When the average wind speed, wind direction, wind field fluctuation, vertical disturbance, and minimum number of effective measurement points are simultaneously met, the system transitions from the partially satisfied state to the overall satisfied state. If a key condition is interrupted during the setup time, the system transitions from the overall satisfied state to the partially satisfied state or the pending confirmation state. If there are consecutive timeouts or insufficient resources, the system transitions from the pending confirmation state or the partially satisfied state to the confirmation failure state. The state machine converts key decisions during the pre-start confirmation process into objective branch conditions.

[0089] The phase confirmation record should include at least the phase number, valid status of the measuring point, status of the measuring point in the main windward area, start time of continuous satisfaction, end time of continuous satisfaction, establishment time length, start judgment conclusion, anomaly marker, and status transition path; the phase start record should include the establishment completion time and the time of entering the execution state, and the phase failure record should include the reason for failure; the phase confirmation record serves as the input for subsequent flight status judgment.

[0090] S4: During the current testing phase, continuously determine the position-holding state, attitude change state, disturbance recovery state, and boundary violation state of the tested unmanned aerial vehicle under the influence of the wind field, and enter the corresponding processing procedure based on the determination results. The specific implementation is as follows:

[0091] After the phase confirmation record is generated, the current test phase will be entered to perform the judgment work. The phase number, target wind field conditions, phase start time and location of the phase confirmation record will be used as the unified basis for continuous judgment. The judgment of the current test phase will be based on the flight response of the test unmanned aerial vehicle in the actual wind field.

[0092] During the current testing phase, three types of records are generated: flight status records, boundary status records, and event trigger records. Flight status records include at least the following fields: timestamp, spatial position coordinates, altitude, roll angle, pitch angle, yaw angle, horizontal speed, vertical speed, and current flight mode identifier. Boundary status records include at least the following fields: current position relative to the permissible spatial boundary, current attitude relative to the permissible attitude boundary, recovery timing, boundary violation start time, and boundary violation duration. Event trigger records include at least the following fields: protection action trigger, manual intervention trigger, phase interruption trigger, re-verification trigger, resource boundary trigger, and data anomaly trigger. The three types of records are associated with a unified time base, and position anomalies, attitude anomalies, recovery processes, and event triggers correspond to the same time interval.

[0093] At the start of the current testing phase, the start time recorded in the phase confirmation record and the position of the tested UAV at that time are used as reference benchmarks to establish an allowable position envelope. The allowable position envelope can be a three-dimensional envelope centered on the reference position. For small UAVs, the horizontal radius can be selected from 0.5 meters to 1 meter, and the allowable altitude deviation can be selected from ±0.3 meters to ±0.8 meters. For medium-sized UAVs, the horizontal radius can be selected from 1 meter to 3 meters, and the allowable altitude deviation can be selected from ±0.5 meters to ±1.5 meters. The above ranges are based on the fact that small UAVs have smaller mass and lower inertia, and generally higher position-keeping accuracy, but are more susceptible to [unspecified factors]. Due to the impact of transient wind disturbances, the envelope range should not be too tight to avoid misjudging normal short-term disturbances as failures, nor too wide to avoid weakening the class differentiation. Medium-sized unmanned aerial vehicles have larger inertia and slower recovery processes under the same wind field, so the envelope range can be appropriately widened in combination with the airframe size, control characteristics, and test targets. The position envelope is allowed to remain fixed within the current test phase and should not be re-sampling due to airframe offset. The specific value of the allowable position envelope can be set in combination with the airframe diagonal size, test altitude, control mode, and target wind speed settings. A smaller value can be selected for hovering tests, and the envelope range can be appropriately widened for test phases that include changes in direction and increased fluctuations, provided that the class differentiation boundaries are not exceeded.

[0094] Attitude change status is continuously determined using permissible attitude boundaries. Permissible attitude boundaries include at least parameters such as the permissible absolute values ​​of roll angle and pitch angle. For small and medium-sized unmanned aerial vehicles, the permissible absolute values ​​of roll angle and pitch angle can be selected from 20° to 30°. The basis for this range is that when the roll angle or pitch angle continuously approaches or exceeds this range, the aircraft is usually in a state of continuous high compensation, indicating that the control system needs to continuously output a large correction amount to maintain flight under the current wind field conditions. Therefore, even if the position has not obviously exceeded the boundary, it should not be considered as stable holding. The specific values ​​of permissible attitude boundaries can be determined in combination with the aircraft layout, load level, and control mode. The position holding mode can select a smaller value, while the strong disturbance stage or high load condition can select a higher value within the corresponding range. The yaw angle is not used as a separate failure criterion in the current determination, but it can be used as an auxiliary indicator for direction switching and recovery observation.

[0095] The disturbance recovery status is used to determine whether the tested unmanned aerial vehicle (UAV) can return to the allowed position envelope and allowed attitude boundaries within a limited time after a change in wind direction, an increase in wind field fluctuations, or the occurrence of vertical disturbances during the current test phase. The recovery time is set in different ranges according to the airframe size, load status, target wind speed setting, and phase type. For small UAVs, the recovery time can be selected from 1 to 3 seconds; for medium-sized UAVs, the recovery time can be selected from 2 to 5 seconds. The basis for these ranges is that small UAVs have smaller inertia and faster response, and can usually recover under low to medium wind disturbance conditions. Recovery can be completed in a relatively short time; medium-sized unmanned aerial vehicles require a longer recovery time due to increased inertia, heavier load, and differences in propulsion response characteristics; the recovery time is fixed by the phase definition table before the start of the current test phase and will not be adjusted temporarily during execution; if the position deviation or attitude exceeds the limit due to changes in wind field conditions during the current test phase, the recovery timer will be started; if the aircraft returns to the allowable position envelope and allowable attitude boundary simultaneously within the specified recovery time, it is considered a valid recovery; if the aircraft fails to return to the allowable range after the recovery time has expired, it is considered a recovery timeout, and the conclusion of the current test phase will be determined accordingly.

[0096] The boundary crossing status is used to determine whether the current position, altitude, and attitude values ​​exceed the allowable boundaries of the current testing phase, and to determine whether the duration of the boundary crossing exceeds the allowable instantaneous boundary crossing time. The allowable instantaneous boundary crossing time can be selected from 0.3 seconds to 1 second, with a recommended value of 0.5 seconds to 1 second. The basis for this value range is that short-term peak wind disturbances often occur in natural complex wind fields. A slight boundary crossing at the peak moment does not necessarily mean that the aircraft has lost its wind resistance capability. However, if the boundary crossing duration is too long, it usually indicates that the aircraft cannot be recovered in time under the current wind field conditions. The allowable instantaneous boundary crossing time is preferably not less than 3 consecutive state sampling cycles or 3 consecutive control cycles to avoid misjudging instantaneous sampling noise as a real boundary crossing. At the same time, it should not be longer than the duration of the short-term peak disturbance before a complete recovery judgment to avoid misjudging continuous instability as a recoverable deviation. The allowable instantaneous boundary crossing time and the recovery time are related but not the same. The former is used to determine whether a short-term boundary crossing has entered the recovery observation stage, and the latter is used to determine whether effective recovery can be completed after entering the recovery observation stage.

[0097] The current testing phase employs a combination of continuous judgment and branch processing. Upon phase startup, a stable execution state is entered. In this state, flight status and boundary status records are read. Based on a unified timeframe, it is determined whether the current position is within the allowed position envelope, whether the current attitude is within the allowed attitude boundary, whether there are disturbance events triggering recovery timers, and whether boundary violations have occurred. If the position, altitude, and attitude remain within the allowed range throughout the entire execution process, and no recovery timeout, protection action trigger, or manual intervention occurs, the test is considered successful. If a single or a few short-term boundary violations occur, and the duration of each violation does not exceed the allowed instantaneous boundary violation time, and the tester returns to the allowed position envelope and allowed attitude boundary within the recovery time, the test is considered successful. If continuous boundary violations occur, the altitude value exceeds the allowed altitude deviation and the duration of the exceedance exceeds the allowed instantaneous boundary violation time, the attitude continuously exceeds the allowed attitude boundary, recovery timeout occurs, protection action is triggered, or manual intervention is triggered, the test is considered a failure, and the process enters a downgrade or termination phase based on the current level position and session resource boundaries.

[0098] To ensure a clear state transition relationship in the judgment process, a state machine is used to manage the execution process of the current test phase. The state machine includes at least a stable execution state, a momentary out-of-bounds state, a recovery observation state, a boundary crossing state, a failure state, a re-verification state, a degradation state, and a termination state. The current test phase begins in the stable execution state. When any judgment variable (position, height, or attitude) exceeds the allowable boundary, it transitions to the momentary out-of-bounds state, recording the start time, type, and boundary category of the out-of-bounds error. If the position returns to the allowable range within the allowed momentary out-of-bounds time, it transitions to the recovery observation state and starts the recovery timer. If the out-of-bounds duration exceeds the allowed momentary out-of-bounds time, it transitions to the failure state. In the recovery observation state, if both position and attitude are recovered within the specified recovery time, and the number of short-term out-of-bounds errors does not exceed a preset limit, then the test phase is considered successful. Based on the cumulative number of boundary violations, the magnitude of boundary violations, and the number of recovery attempts, the system transitions to a stable execution state or a boundary-passing state. If the recovery time exceeds the limit, or if a more severe boundary violation occurs again during the recovery observation process, the system transitions to a failure state. The upper limit for the number of short-term boundary violations can be set in accordance with the boundary-passing judgment rule, and can be selected from 1 to 2 times. The value is based on the fact that 1 time can be used to control occasional short-term boundary violations under critical boundary conditions, and 2 times can be used to retain the judgment space for a small number of recoverable peak disturbances in complex wind fields. Exceeding this range usually indicates that the stability of the current test phase is insufficient. The boundary-passing state is used to indicate that the current test phase has not reached the immediate failure condition, but needs to enter the same level of re-verification. The failure state is used to indicate that the current test phase has not passed. After failure, the system transitions to a re-verification, downgrade, or termination state according to the phase execution sequence table and the resource boundary of the current test session.

[0099] To improve the objectivity of the judgment, a multi-dimensional state correlation judgment is adopted, and a single position deviation, a single attitude deviation, or a single event trigger is not used as an independent conclusion basis. If the spatial position deviates for a short time, but the roll angle, pitch angle, horizontal velocity, and vertical velocity are all within a reasonable range, and then return to the allowable position envelope, it is judged as a recoverable deviation caused by a short-term disturbance. If the position deviation occurs simultaneously with a sustained high-compensation attitude, abnormal velocity increase, or rapid altitude decrease, it is judged as boundary instability. If only the attitude is detected to be close to the allowable attitude boundary, while the position remains stable and no recovery timeout occurs, the judgment of the current test stage is maintained. If the attitude continuously reaches the preset warning interval of the allowable attitude boundary and is accompanied by continuous recovery timer triggers, it is judged that the wind resistance capability of the current test stage is close to the limit, and will be checked in detail in subsequent re-verification. The preset warning interval is set within the allowable attitude boundary and can be selected from 80% to 95% of the allowable attitude boundary. The basis for this range is that it is difficult to identify the critical compensation trend in time when it is below 80%, and it is too close to the formal attitude boundary when it is above 95%, resulting in insufficient warning lead time.

[0100] The anomaly handling rules primarily target data loss, positioning jitter, and manual intervention. If the continuous loss of flight status data does not exceed 100 milliseconds, and the preceding and following data are consistent and reasonable in terms of position, velocity, and attitude, the current judgment state can remain unchanged, and a verification mark will be added to the boundary state record. The 100-millisecond threshold is chosen because it is preferably no longer than the total duration of multiple consecutive sampling periods, and can be used to distinguish between short-term sampling interruptions and continuous data loss. If the continuous loss of flight status data exceeds 500 milliseconds, the current testing phase enters an invalid interruption state and is not used for formal judgment. The 500-millisecond threshold is chosen because exceeding this time... After a long period, the critical observation window required for boundary recognition and recovery timing has usually been crossed; if the tested UAV shows positional anomalies due to short-term jitter in positioning data, but the roll angle, pitch angle, horizontal speed, and vertical speed do not show corresponding instability characteristics, the current judgment state can be maintained and a verification mark can be added; if it is confirmed after verification that it is only a positioning drift, the original state is restored; if it is confirmed that there is a correlation anomaly between position, attitude, and speed, it is treated as a real offset; when a human safety officer intervenes in the control, the current test phase is immediately terminated and marked as a failure or session aborted, based on the fact that after human intervention, the flight response is no longer an autonomous performance under the current complex wind field conditions;

[0101] Time and resource constraints invoke resource boundaries already obtained in the test session; the execution time of the current test phase is the duration shown in the phase definition table, which cannot be extended during execution or recalculated due to the appearance of boundaries; the conclusion of the current test phase is based on the actual flight results within the duration; when the battery level drops to the minimum execution threshold during the execution of the current test phase, the current test phase will end, and the phase conclusion will retain the judgment information that has been completed, and subsequent levels will not continue to execute; when the aircraft temperature rises above the set derating boundary, the current judgment will be completed in the current state, and it will not re-enter a higher level; when the current state has entered the recovery observation state and the resource boundary is triggered at the same time, the current state will first perform the current recovery observation judgment, and then enter the termination state;

[0102] Upon completion of the phase determination, the phase determination record shall include at least the following fields: phase name, effective time, position maintenance conclusion, attitude change conclusion, disturbance recovery conclusion, boundary crossing conclusion, anomaly handling conclusion, trigger state transition path, and final phase conclusion, which shall serve as the basis for subsequent summarization.

[0103] S5: Based on the execution and judgment results of each test phase, determine the wind resistance capability classification of the tested unmanned aerial vehicle. The specific implementation is as follows:

[0104] Once the stage judgment records corresponding to each test stage are formed, the unified level judgment rules will be used to summarize the execution results, re-verification results, and abnormal results of each stage within the same test session, and obtain the wind resistance rating results of the tested unmanned aerial vehicle in the current test session according to the level progression relationship.

[0105] The input data required for level determination is the test session number master index. At a minimum, the level determination results must be present in the stage execution sequence table, stage determination record table, re-verification record table, exception event summary table, and level determination rule table. The stage execution sequence table records the order of each test stage under different levels, necessary and optional stages, re-verification branches, demotion branches, and termination branches. The stage determination record table records the final execution result of each stage, such as pass, boundary pass, failure, invalid interruption, not formed, and retest. The re-verification record table records the test results of same-level re-verification, retesting, or reconfirmation, such as whether the re-verification is approved, etc. Whether it is still in a boundary state or has ultimately failed; the abnormal event summary table records abnormal events that occurred in each stage, such as measurement point anomalies, data interruptions, manual intervention, resource boundary triggering, session condition changes, and site boundary changes; the level determination rule table records the conditions for each level to be established, including the minimum number of necessary stages, direction coverage requirements, wind field fluctuation coverage requirements, direction change requirements, recovery observation requirements, boundary pass-through conditions, allowed number of failed stages, and allowed invalid stage handling methods; by inputting data, the test information formed in each stage can be uniformly transformed into the structured judgment objects required for the final level determination;

[0106] The classification rules are progressive, coverage priority, boundary verification, and anomaly removal. Progressive means that a classification level must be valid before a higher level can be classified, and it is not permissible to classify a lower level as valid even if the previous level is invalid. Coverage priority means that both directional coverage and disturbance coverage must be satisfied when a classification level is valid, and the number of passed stages is not the sole criterion. Boundary verification means that after a boundary-passing stage occurs at a certain level, its validity is determined by reviewing the verification records. Anomaly removal means that invalid stages caused by abnormal measurement points, data interruption, manual intervention, or changes in session conditions are removed, and then supplementary measurement, postponement, or termination is carried out according to the classification rules.

[0107] In the classification coverage rules, it is preferable to cover at least three directions—frontal, crosswind, and oblique wind—with at least one directional change and one wind field fluctuation enhancement within the same classification level. Frontal wind reflects the basic maintenance capability under positive inflow conditions, crosswind reflects the resistance to lateral disturbances, and oblique wind reflects the positional and attitude coordinated stability capability under combined inflow conditions. Therefore, the three directions can serve as the basic directional coverage set under complex wind fields. If a classification level only passes under frontal conditions but not under crosswind or oblique wind conditions, then that classification level is invalid. Fluctuation coverage is preferably required to pass at least one basic maintenance phase and one fluctuation enhancement phase within the same classification level. In classification levels without a separate fluctuation enhancement phase, passing one directional change phase and one recovery observation phase can be used as alternative coverage conditions. The reason for adopting this rule is that the results of passing under low fluctuation conditions cannot directly prove that the organism can still maintain stability under continuous wind field fluctuations. Only by incorporating basic maintenance, directional change, fluctuation enhancement, and recovery observation into joint coverage can the wind resistance boundary under the current classification level be more accurately reflected.

[0108] The number of necessary stages can be determined based on the level settings. Preferably, each formal level should contain at least three necessary stages, corresponding to basic maintenance, direction change, and recovery observation, respectively. This setting is adopted because basic maintenance confirms the basic stability under the current level, direction change confirms the resilience under wind direction switching conditions, and recovery observation confirms the recovery capability after short-term instability. Without any one of these stages, the current level cannot simultaneously cover maintenance, adaptability, and recovery capabilities. The number of allowed failure stages can be determined by combining the division of necessary and optional stages. Necessary stages are preferably not allowed to fail. Optional enhancement stages can be allowed 0 to 1 failure stages without affecting direction and fluctuation coverage. If a failure of an optional enhancement stage leads to insufficient direction or fluctuation coverage, the failure stage will no longer be treated as an allowed failure stage, but as a coverage failure. This rule is adopted because necessary stages directly determine the basis for the level's establishment and should not fail. Optional enhancement stages are mainly used to improve resolution accuracy, and limited failures are allowed without compromising the integrity of the current level's coverage.

[0109] The grading system proceeds from lowest to highest grading level. Starting with the lowest grading level, the system checks if the stage conclusion in the stage grading record table is "passed" or if the boundary has passed and been corrected. A grading level is considered valid when all necessary stages are met and directional and fluctuation coverage meet the grading rules. If a necessary stage fails, directional coverage is insufficient, or the boundary fails and is not corrected, the grading level is considered invalid and marked as not reaching the minimum grading level. Once a grading level is deemed valid, the system continues to check the next grading level according to the same rules. A grading level is considered valid when the next grading level meets the necessary stage requirements, the coverage rules are met, and the re-verification loop is completed. A grading level is considered invalid if a critical stage fails, coverage is insufficient, re-verification fails, supplementary testing fails, or session resources are interrupted, and no further grading is performed. After a grading level is deemed invalid, the final grading level is set to the grading level that was previously valid, and no other passing stages from higher grading levels are used to replace the gaps in the grading level. This rule is used because grading levels have a progressive relationship; the validity of a higher grading level presupposes the stable validity of a lower grading level. Directly determining a grading level as valid at a non-valid grading level would disrupt the grading logic.

[0110] For boundary crossing scenarios, a positive conversion rule is applied. Boundary crossing indicates that although the current stage has not reached the level of direct failure, it has touched the boundary stability limit. Therefore, it is not directly counted as a formal pass, but rather a decision on whether to convert to positive status is made based on the re-verification record. Ideally, boundary crossing conversion should simultaneously meet the following conditions: the number of boundary crossings in a single stage does not exceed a preset upper limit, which can be selected as 1 to 2 times; the duration of each boundary crossing does not exceed the allowable instantaneous boundary crossing time for the current stage; each recovery process is completed within the specified recovery time; and the re-verification results do not show more severe issues such as continuous attitude exceeding limits, continuous altitude exceeding limits, recovery timeouts, or an increase in the number of consecutive boundary crossings. The boundary crossing limit is set at 1 to 2 times because occasional peak disturbances in complex wind fields may cause occasional short-term boundary crossings. However, if the number of boundary crossings is too high, it indicates that the current stage is in a clearly unstable boundary. Only when the above conditions are met simultaneously will the boundary crossing stage be converted to formal passing after re-verification. If the re-verification result is still boundary passing, or if new serious boundary behaviors occur, the stage will be treated as not being formally passed and will not be included in the formal passing stage required for the establishment of the grade. The boundary passing conversion rule is preferably consistent with the upper limit of the number of short-term boundary crossings in the stage execution judgment to ensure that the stage judgment and the grade summary use the same boundary caliber.

[0111] Abnormal events are handled using a combination of elimination and retesting in the final classification. If a stage is deemed invalid due to abnormal measurement points, interruption of critical data, manual intervention, changes in session conditions, or changes in site boundaries, that stage is neither counted as passed nor failed, and retesting is determined according to the stage execution sequence table. If the current stage is a necessary stage, it can be selected as a same-level retest; if it is an optional enhancement stage, it can be selected to be moved or skipped, but skipping it must not affect the necessary coverage requirements. The number of same-level retests for necessary stages is preferably no more than 1 to 2. The basis for adopting this range is that the purpose of retesting is to restore the coverage integrity that the current level should have, rather than to repeat the test indefinitely. If the number of retests is too high, the test conditions will be affected. Consistency and objectivity of results will be significantly reduced; if a valid stage cannot be formed after supplementary testing, the current level will be treated as incomplete and will not be directly recognized as established; for example, if the target wind field is not formed after multiple confirmations, or if manual intervention is triggered again during supplementary testing; if battery replacement, flight control version change, mission payload position change, or test site boundary readjustment occur during the session, it is considered that the session conditions have changed substantially, and the session will not output a formal level result, but only an invalid session mark, and a new session will be established; by removing invalid stages from the formal level determination and clarifying the processing boundaries of supplementary testing, postponement, skipping, and session failure, it can be ensured that the final level is established on the basis of valid, verifiable, and consistent stage records;

[0112] To ensure clear process constraints for the grading determination, a grading state machine is established. This state machine includes at least four states: Grade Pending Determination, Grade Initially Established, Grade Pending Verification, Grade Locked, Grade Invalid, and Session End. Each grade is initially in the Grade Pending Determination state before aggregation begins. When the necessary stage conclusions meet the minimum passing conditions and directional and fluctuation coverage requirements are met, the grade enters the Grade Initially Established state. When there is a boundary passing stage or a critical stage requiring supplementary testing and confirmation, the grade enters the Grade Pending Verification state. When verification or supplementary testing is completed and meets the conversion rules, the grade enters the Grade Locked state. When a critical stage fails, coverage is insufficient, verification fails, supplementary testing fails, or session conditions change, the grade enters the Grade Invalid state, and higher-level grading is halted. When all grading determinations are completed, resource boundary triggers termination, or the session is deemed invalid, the session ends. Once the grading state transition conditions are fixed, they serve as the basis for grading output and review judgment. Before a grade enters the Locked state, no formal grading result is formed; after a grade enters the Invalid state, it is no longer directly reclassified as established.

[0113] In time constraints, a single level determination should preferably be completed within 5 minutes after all stages of the current level have ended. This is based on the fact that level summarization is mainly based on the existing structured stage records and does not require lengthy manual analysis. If the summarization cannot be completed within 5 minutes, it usually indicates that the stage determination record table, re-verification record table, or abnormal event summary table has incomplete records, unclosed re-verification, incomplete supplementary testing, or incomplete abnormal handling. In this case, the completeness of the relevant records should be checked before level determination. When a session triggers a resource boundary during level determination, such as when the remaining battery power is lower than the session lower limit or the device temperature rises to the derating boundary, the current level has entered the state machine determination process, but it will not enter a higher level determination.

[0114] The classification result record should include at least the following fields: session number, final level, list of established levels, list of non-established levels, corresponding pass stage number for each level, boundary pass stage number, re-verification stage number, failure stage number, invalid stage number, direction coverage conclusion, fluctuation coverage conclusion, anomaly stage handling conclusion, and session validity conclusion. The classification result record should also include the level determination result, coverage determination result, anomaly handling result, and session validity result of the current test session, and generate a result summary as a test report, but this does not replace the classification result record. After the level summary is completed, the wind resistance classification result of the current test session should be displayed.

[0115] The technical solution of this embodiment takes a quadcopter unmanned aerial vehicle (UAV) for power line inspection as an example. First, a corresponding test session is established; the aircraft type, load status, control mode, site boundaries, and classification conditions are fixed. Based on the set levels, a test phase sequence is generated, including basic maintenance, direction change, fluctuation enhancement, and recovery observation. Before each test phase begins, the test does not start based on the set values ​​of the air supply equipment. Instead, it is determined through multiple measuring points around the aircraft that the wind field meets the target wind speed, target wind direction, fluctuation conditions, and vertical disturbance conditions before entering each test phase. During the test, the UAV's position, attitude, speed, recovery process, and boundary violations are recorded, judged by the allowed position envelope, allowed attitude boundaries, recovery time, and allowed instantaneous boundary violation time. After all tests are completed, based on the direction coverage, fluctuation coverage, boundary verification, and anomaly removal, the wind resistance capability classification result of the UAV under the current load and control conditions is determined sequentially according to the classification.

[0116] It should be noted that this invention can be deployed on the device itself to realize embedded applications, or it can run on a PC or other terminal with a user interface, thereby meeting various hardware environments and usage requirements.

[0117] The above embodiments can be implemented in whole or in part by software, hardware, firmware, or any other combination. When implemented in software, the above embodiments can be implemented in whole or in part by a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the processes or functions of the embodiments of this application are implemented in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted wirelessly or wiredly from one website, computer, server, or data center to another website, computer, server, or data center. Wired methods include optical fiber, twisted pair, coaxial cable, etc. Wireless methods include infrared, microwave, etc. Available media include any available media that can be accessed by a computer or data storage devices such as servers and data centers that contain one or more sets of available media. Available media can be magnetic media (floppy disks, hard disks, magnetic tapes), optical media (DVDs), or semiconductor media. Semiconductor media can be solid-state drives.

[0118] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0119] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for grading and testing the wind resistance capability of unmanned aerial vehicles, characterized in that, include: S1: Establish a test session for the unmanned aerial vehicle under test, fix the aircraft type status, load status, control status and test boundaries, and set the wind resistance capability graded test conditions; S2: Generate corresponding test phases according to the preset wind field level order, so that each test phase forms a test process with continuous changes in average wind speed, wind direction and wind field fluctuations. According to the wind direction coverage rules and phase generation rules, configure the target average wind speed range, target wind direction range, fluctuation amplitude range, vertical disturbance range and target duration for each level phase, so that the basic maintenance phase, direction change phase, fluctuation enhancement phase and recovery observation phase within the same level form a continuous test process according to the sequential relationship, and at least one of the average wind speed, fluctuation amplitude, direction change intensity or vertical disturbance intensity between adjacent levels increases. Generate a phase definition table and a phase execution sequence table, and determine the connection relationship of each phase according to the re-verification branch, downgrade branch and termination branch. S3: Before each test phase, confirm the actual wind field around the unmanned aerial vehicle under test in the test space, and start the current test phase after confirming that it meets the requirements of the current test phase. S4: During the current testing phase, continuously determine the position holding state, attitude change state, disturbance recovery state, and boundary crossing state of the tested unmanned aerial vehicle under the influence of the wind field, and enter the corresponding processing procedure based on the determination results; S5: Based on the execution and judgment results of each test phase, determine the wind resistance rating of the tested unmanned aerial vehicle.

2. The method for grading and testing the wind resistance capability of an unmanned aerial vehicle according to claim 1, characterized in that, Establish a test session for the tested unmanned aerial vehicle (UAV), fix the aircraft type, load, control status, and test boundaries, and set graded test conditions for wind resistance capability, including: The test session associates the session record table, field constraint table, and input data structure to perform consistency verification and locking on the airframe configuration, propulsion configuration, load level, center of gravity offset, control mode, control version, test space boundary, time boundary, and resource boundary; Write the wind speed setting, wind direction setting, fluctuation setting, position deviation, attitude deviation, recovery time, and instantaneous out-of-bounds time according to the classification requirements to form a session initialization record.

3. The method for grading and testing the wind resistance capability of an unmanned aerial vehicle according to claim 1, characterized in that, The test phases are generated according to the preset wind field level order, including: Establish a phase record table based on the test session number, and configure the introduction phase, as well as the basic maintenance phase, direction change phase, and recovery observation phase corresponding to each level, in combination with the machine size, load level, control mode, test boundary, and classification conditions. Also configure the fluctuation enhancement phase and boundary verification phase as needed. Perform parameter consistency verification, boundary verification, resource verification, and progression relationship verification on each stage. Stages that pass the verification are set to a locked state, while stages that fail the verification are set to an unexecutable state or deleted and regenerated.

4. The method for grading and testing the wind resistance capability of an unmanned aerial vehicle according to claim 1, characterized in that, Before each test phase, the actual wind field around the tested unmanned aerial vehicle in the test space is confirmed, including: Before each test phase, key and auxiliary measuring points are set up around the body to establish a measuring point layout table, a real-time wind field record table, and a phase start-up condition table. Based on the measuring points around the aircraft, the upstream reference measuring points, and the measuring points in the main wind-receiving area, the average wind speed, wind direction, wind field fluctuations, vertical disturbances, the number of effective measuring points, and the status of abnormal measuring points are determined. The main wind-receiving area is determined based on the target wind direction range, the aircraft's orientation, and the distribution of the wind-receiving surface of the aircraft's shape.

5. The method for grading and testing the wind resistance capability of an unmanned aerial vehicle according to claim 1, characterized in that, Once the requirements for the current testing phase are confirmed to be met, the current testing phase will begin, including: First, determine the validity of the data at each measuring point; Then, verify the average wind speed, wind direction, wind field fluctuation, vertical disturbance, continuous satisfaction time and minimum number of effective measuring points in the order of partial confirmation, overall confirmation and establishment of timing. When the preset conditions are continuously met, form a stage confirmation record and enter the execution state. When there are abnormalities in the main wind-receiving area measurement points, insufficient number of effective measurement points, excessive total confirmation time, or resource boundary triggering, the current testing phase will be recorded as a confirmation failure or an unformed phase, and will be handled according to the following methods: move forward, skip, downgrade, or terminate the relationship.

6. The method for grading and testing the wind resistance capability of an unmanned aerial vehicle according to claim 1, characterized in that, During the current testing phase, the test unmanned aerial vehicle's position-holding state, attitude change state, disturbance recovery state, and boundary violation state under wind conditions are continuously determined, including: During the current testing phase, flight status records, boundary status records, and event trigger records are established based on the phase confirmation records. A fixed allowable position envelope is constructed using the phase start time and reference position. Combined with the allowable attitude boundary, recovery time, allowable instantaneous boundary crossing time, and attitude warning interval, a joint judgment is made on position offset, attitude change, disturbance recovery, and boundary crossing. Recovery time is used to determine whether the tested unmanned aerial vehicle can return to the allowed position envelope and allowed attitude boundaries, while allowed instantaneous boundary crossing time is used to determine whether recovery observation should be entered after the boundary crossing.

7. The method for grading and testing the wind resistance capability of an unmanned aerial vehicle according to claim 1, characterized in that, Based on the judgment result, the corresponding processing flow will be initiated, including: The process flows through the following states: stable execution state, momentary out-of-bounds state, recovery observation state, boundary crossing state, failure state, re-verification state, degraded state, and termination state. When the position, height, or orientation exceeds the allowed boundaries, record the type and duration of the boundary violation; If the recovery is completed within the recovery time and the number of short-term boundary violations does not exceed the preset limit, the system will transition to a stable execution state or a boundary-passing state. When data is missing, location jitter occurs, manual intervention is required, or resource boundaries are triggered, the process is handled according to the rules of pending verification, invalid interruption, failure, or termination, and a stage judgment record is generated.

8. The method for classifying and testing the wind resistance capability of an unmanned aerial vehicle according to claim 1, characterized in that, Based on the execution and judgment results of each test phase, including: The system includes a call phase execution sequence table, a phase judgment record table, a re-verification record table, an exception event summary table, and a level judgment rule table. Necessary and optional phases are verified separately, and a summary judgment is made according to the rules of level progression, direction coverage, fluctuation coverage, boundary pass and positive conversion, exception removal, and supplementary testing. It also handles boundary pass, supplementary testing, postponement, skipping, and session failure.

9. The method for grading and testing the wind resistance capability of an unmanned aerial vehicle according to claim 1, characterized in that, Determine the wind resistance rating of the tested unmanned aerial vehicle, including: The process proceeds through the following stages: pending level determination, preliminary level establishment, pending level verification, level locked, level not established, and session end. Verification is performed level by level, starting from the lowest level. When the current level meets the necessary stage conditions, coverage conditions, and re-verification conditions, it enters the level lock state; When a critical phase fails, coverage is insufficient, retesting fails, or session conditions change, higher-level judgments are stopped, and a classification result record is generated.

Images