An unmanned helicopter rotor aerodynamic performance calculation method based on edge computing

Abstract

The application relates to the technical field of edge computing and flight data processing, and discloses an unmanned helicopter rotor aerodynamic performance calculation method based on edge computing, which comprises the following steps: collecting multi-source operation data in the flight process of an unmanned helicopter; generating rotor cycle state data according to rotor rotating speed data in the rotor aerodynamic calculation data flow; identifying the current rotor operation working condition of the unmanned helicopter according to the rotor cycle state data; operating a lightweight aerodynamic basic calculation core according to the rotor cycle state data and the rotor working condition state unit; calculating an aerodynamic consistency deviation group; performing source decomposition on the aerodynamic consistency deviation group; updating an edge end aerodynamic residual error cluster library and generating an aerodynamic model correction parameter according to the aerodynamic residual error decomposition result; and outputting a rotor aerodynamic performance calculation result according to the aerodynamic model correction parameter. The application realizes real-time calculation of rotor aerodynamic performance by adopting edge residual error correction.

Description

Technical Field

[0001] This invention relates to the field of edge computing and flight data processing technology, and in particular to a method for calculating the aerodynamic performance of an unmanned helicopter rotor based on edge computing. Background Technology

[0002] Current calculations of the aerodynamic performance of unmanned helicopter rotors are typically based on flight test data, ground station data analysis, or offline aerodynamic models. These methods collect data on rotor speed, blade pitch angle, flight speed, altitude, attitude angle, motor power, and environmental parameters to calculate rotor lift, drag, torque, power consumption, and rotor efficiency. Some methods also incorporate momentum theory, blade element calculations, and empirical aerodynamic parameters to estimate rotor aerodynamic performance under different flight conditions.

[0003] The methods described above still suffer from insufficient data processing and computational adaptability during actual flight. Multi-source operational data is typically processed according to fixed time windows, making it difficult to correlate with the actual rotor rotation cycle; there is a lack of consistency deviation analysis between the basic aerodynamic calculation results and the flight dynamic response data, making it difficult to distinguish the sources of calculation deviations; when environmental disturbances, load changes, rotor efficiency decay, or sensor synchronization errors affect the calculation results, existing methods struggle to perform real-time corrections to the aerodynamic model at the airborne edge computing end, thus affecting the real-time performance and accuracy of the rotor aerodynamic performance calculation results.

[0004] Therefore, how to provide a method for calculating the aerodynamic performance of unmanned helicopter rotors based on edge computing is a problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0005] One objective of this invention is to propose a method for calculating the aerodynamic performance of an unmanned helicopter rotor based on edge computing. This invention fully utilizes airborne edge computing, multi-source operational data fusion, rotor periodic segment division, aerodynamic consistency deviation analysis, and aerodynamic residual decomposition and correction techniques. It details the processing flow for real-time calculation of rotor aerodynamic performance during unmanned helicopter flight. By collecting rotor speed data, blade pitch angle data, flight speed data, altitude data, attitude angle data, motor power data, environmental data, and airframe vibration data, a rotor aerodynamic calculation data stream is generated. Rotor periodic segments are formed according to the rotor rotation cycle. Then, by combining rotor operating condition state units, basic rotor aerodynamic performance results, aerodynamic consistency deviation groups, and edge-end aerodynamic residual cluster libraries, the basic aerodynamic calculation results are dynamically corrected, and the rotor aerodynamic performance calculation results are output. This method has the advantages of high real-time performance, strong computational adaptability, distinguishable deviation sources, and high accuracy in aerodynamic performance calculation.

[0006] A method for calculating the aerodynamic performance of an unmanned helicopter rotor based on edge computing according to an embodiment of the present invention includes the following steps:

[0007] Collect multi-source operational data during the flight of the unmanned helicopter and generate rotor aerodynamic calculation data stream at the airborne edge computing terminal;

[0008] Rotor cycle segments are generated based on rotor speed data in rotor aerodynamic calculation data stream, and multi-source operating data are mapped to rotor cycle segments to generate rotor cycle state data.

[0009] Identify the current rotor operating condition of the unmanned helicopter based on rotor cycle state data, and generate rotor operating condition state unit;

[0010] Based on the rotor cycle state data and rotor operating condition state unit, a lightweight aerodynamic foundation calculation kernel is run to obtain the rotor foundation aerodynamic performance results;

[0011] Calculate the aerodynamic consistency deviation group based on the rotor's basic aerodynamic performance results and flight dynamic response data;

[0012] The sources of aerodynamic consistency deviation groups are decomposed to generate aerodynamic residual decomposition results.

[0013] Update the edge-end aerodynamic residual cluster library based on the aerodynamic residual decomposition results, and generate aerodynamic model correction parameters;

[0014] The basic aerodynamic performance results of the rotor are corrected based on the aerodynamic model correction parameters, and the rotor aerodynamic performance calculation results are output.

[0015] Optionally, the airborne edge computing terminal adds a unified time stamp to multi-source operating data from different sources, and establishes a data access queue according to data source, sampling frequency and data update interval to generate rotor aerodynamic computing data stream.

[0016] Optionally, the step of generating rotor cycle segments based on rotor speed data in the rotor aerodynamic calculation data stream, and mapping multi-source operating data to rotor cycle segments to generate rotor cycle state data specifically includes:

[0017] Generate a time-ordered sequence of rotor angular velocities based on rotor speed data and a unified time stamp from the rotor aerodynamic calculation data stream;

[0018] The data boundary corresponding to the completion of a single rotor rotation is determined based on the rotor angular velocity sequence and a unified time stamp.

[0019] Generate rotor cycle segments based on the fluctuation state of multiple consecutive data boundaries and rotor speed data;

[0020] Based on the start and end data boundaries of the rotor cycle segment, the multi-source operating data in the rotor aerodynamic calculation data stream is mapped to the corresponding rotor cycle segment.

[0021] Generate intra-segment alignment data based on the rotor cycle segment after writing multi-source operational data;

[0022] Generate rotor periodic state data containing average change, maximum change, change direction, and fluctuation concentration location based on the intra-segment alignment data.

[0023] Optionally, the step of identifying the current rotor operating condition of the unmanned helicopter based on rotor cycle state data and generating a rotor operating condition state unit specifically includes:

[0024] Generate operating condition identification input groups based on rotor cycle state data;

[0025] The ascent and descent status, forward flight status, and attitude adjustment status are determined based on the working condition recognition input group.

[0026] The rotor load change state and disturbance state are determined based on the operating condition identification input group;

[0027] The rotor operating condition state unit is generated based on the lift-down state, forward flight state, attitude adjustment state, rotor load change state, and disturbance state.

[0028] Optionally, the step of running a lightweight aerodynamic foundation calculation kernel based on rotor cycle state data and rotor operating condition state unit to obtain rotor foundation aerodynamic performance results specifically includes:

[0029] The calculation inputs for the lightweight aerodynamic basic calculation core are generated based on the rotor cycle state data and rotor operating condition state units.

[0030] Calculate air density and attitude change based on ambient temperature data, air pressure data, and attitude angle data in the calculation input;

[0031] Based on the calculation input, air density, and attitude change, a simplified momentum calculation unit is run to obtain the momentum lift calculation and induced velocity calculation.

[0032] Based on the calculation input, air density and rotor angular velocity, the blade element calculation unit is run to obtain the blade element lift calculation, blade element drag calculation, and blade element torque calculation.

[0033] The basic aerodynamic performance results of the rotor are generated based on the momentum lift calculation, the blade element lift calculation, the blade element drag calculation, the blade element torque calculation, and the rotor operating condition unit.

[0034] Optionally, the calculation of the aerodynamic consistency deviation group based on the rotor's basic aerodynamic performance results and flight dynamic response data specifically includes:

[0035] Flight dynamic response data is generated based on rotor cycle segments and intra-segment alignment data;

[0036] The power consistency deviation is calculated based on the basic power consumption results in the rotor basic aerodynamic performance results and the actual motor power in the flight dynamic response data.

[0037] The lift response deviation is calculated based on the basic lift results in the rotor's basic aerodynamic performance results and the average value of the vertical acceleration data and the altitude change in the flight dynamic response data.

[0038] Based on the basic torque and drag results in the basic aerodynamic performance results of the rotor, as well as the attitude angular velocity and flight speed changes in the flight dynamic response data, the torque response deviation and drag response deviation are calculated respectively.

[0039] The efficiency response deviation is calculated based on the basic rotor efficiency results in the basic rotor aerodynamic performance results, as well as the altitude change and flight speed change corresponding to unit power.

[0040] Aerodynamic consistency deviation group is generated based on power consistency deviation, lift response deviation, torque response deviation, drag response deviation, and efficiency response deviation.

[0041] Optionally, the step of performing source decomposition on the aerodynamic consistency deviation group to generate aerodynamic residual decomposition results specifically includes:

[0042] Generate residual decomposition input groups based on aerodynamic consistency deviation groups and corresponding rotor cycle segments;

[0043] Calculate the environmental disturbance residuals based on the drag response bias, wind speed estimation data variation, and flight speed variation in the residual decomposition input group;

[0044] The load variation residual is calculated based on the lift response deviation, torque response deviation, pitch angle data change, motor power data change, and vertical acceleration data average value in the residual decomposition input group.

[0045] Calculate the rotor efficiency attenuation residual based on the power consistency deviation, efficiency response deviation and maximum value of airframe vibration data in the residual decomposition input group;

[0046] Calculate the sensor synchronization residual based on the jump position of the aerodynamic consistency deviation group and the time stamp misalignment of multi-source operating data;

[0047] The aerodynamic residual decomposition results are generated based on the environmental disturbance residual, load change residual, rotor efficiency decay residual, and sensor synchronization residual.

[0048] Optionally, updating the edge-end aerodynamic residual cluster library based on the aerodynamic residual decomposition results and generating aerodynamic model correction parameters specifically includes:

[0049] Based on the aerodynamic residual decomposition results, the operating condition residual is generated and written into the record;

[0050] Write the operating condition residuals into the corresponding operating condition residual clusters according to the rotor operating condition status unit;

[0051] Update the occurrence frequency, duration, direction of change, average magnitude, and corresponding flight dynamic response of various residuals based on the written residual clusters;

[0052] When a new rotor cycle segment enters the calculation, similar operating condition residual clusters are matched from the edge aerodynamic residual cluster library based on the current rotor operating condition state unit and the current aerodynamic residual decomposition results.

[0053] Residual correction factors are generated based on the average amplitude, duration, and direction of change of various residuals in the residual cluster of similar working conditions.

[0054] Based on the residual correction factor set, the lift correction parameters, drag correction parameters, torque correction parameters, induced velocity correction parameters, power consumption correction parameters, and rotor efficiency correction parameters are calculated to form the aerodynamic model correction parameters.

[0055] Optionally, when a new rotor cycle segment enters the calculation, matching similar operating condition residual clusters from the edge-end aerodynamic residual cluster library based on the current rotor operating condition state unit and the current aerodynamic residual decomposition result specifically includes:

[0056] Generate the current operating condition index based on the current rotor operating condition state unit corresponding to the new rotor cycle segment;

[0057] Generate the current residual proportion vector based on the current aerodynamic residual decomposition result corresponding to the new rotor cycle segment;

[0058] Generate candidate operating condition residual clusters and average residual proportion vectors based on the edge end aerodynamic residual cluster library;

[0059] Calculate the condition matching value based on the current condition index and the condition index of the candidate condition residual cluster;

[0060] Calculate the residual proportion matching value based on the current residual proportion vector and the average residual proportion vector of the candidate working condition residual cluster;

[0061] Similar operating condition residual clusters are determined based on the operating condition matching value and the residual proportion matching value. When there are multiple candidate operating condition residual clusters with the highest similarity, the candidate operating condition residual cluster with the most recent written operating condition residual record is determined as the similar operating condition residual cluster.

[0062] Optionally, the step of correcting the rotor's basic aerodynamic performance results based on the aerodynamic model correction parameters and outputting the rotor aerodynamic performance calculation results specifically includes:

[0063] The correction calculation input is generated based on the aerodynamic model correction parameters and the basic aerodynamic performance results of the rotor.

[0064] The base lift, base drag, and base torque results are corrected based on the correction calculation input to obtain the corrected lift, base drag, and base torque results.

[0065] Write the corrected lift, drag, and torque results into the correction calculation record corresponding to the current rotor cycle segment;

[0066] Based on the corrected calculation input, the calculation quantities related to the rotor flow field, the basic power consumption results, and the basic rotor efficiency results are corrected to obtain the corrected induced velocity calculation quantities, the corrected power consumption results, and the corrected rotor efficiency results.

[0067] Write the corrected induced velocity calculation, corrected power consumption result, and corrected rotor efficiency result into the corrected calculation record corresponding to the current rotor cycle segment;

[0068] The aerodynamic stability results are calculated based on the corrected lift, drag, torque, power consumption, rotor efficiency, and aerodynamic residual decomposition results.

[0069] The rotor aerodynamic performance calculation results are generated based on the corrected lift, drag, torque, power consumption, rotor efficiency, and aerodynamic stability results.

[0070] The beneficial effects of this invention are:

[0071] This invention generates rotor cycle segments based on rotor speed data and maps multi-source operational data to the corresponding rotor cycle segments, ensuring that the data organization for rotor aerodynamic performance calculations corresponds to the actual rotor rotation cycle. Compared to fixed-time-window processing, this method reduces data mixing issues caused by rotor speed variations, aligning and statistically analyzing pitch angle data, flight speed data, attitude angle data, motor power data, environmental data, and airframe vibration data within the same rotor cycle segment. This improves the responsiveness of rotor cycle state data to the current rotor operating state.

[0072] This invention generates an aerodynamic consistency deviation set, including power consistency deviation, lift response deviation, torque response deviation, drag response deviation, and efficiency response deviation, by comparing the consistency between the basic aerodynamic performance results of the rotor and the flight dynamic response data. This allows the basic aerodynamic calculation results to be verified against the actual flight response of the unmanned helicopter. This processing method can reduce the influence of deviations caused by directly outputting results from a fixed aerodynamic model and improve the matching degree between the rotor aerodynamic performance calculation results and the actual flight state.

[0073] This invention decomposes the aerodynamic consistency deviation group into source residuals, generating residuals for environmental disturbances, load variations, rotor efficiency decay, and sensor synchronization. This enables the airborne edge computing end to distinguish the sources of influence of different factors on the rotor aerodynamic performance calculation results. Through this residual decomposition process, aerodynamic model correction no longer relies solely on uniform compensation based on overall deviations, but can instead make corresponding corrections to lift, drag, torque, induced velocity, power consumption, and rotor efficiency based on the source of the residuals, thereby improving the specificity of aerodynamic model correction.

[0074] This invention updates the edge-end aerodynamic residual cluster library based on the aerodynamic residual decomposition results, and matches similar operating condition residual clusters when a new rotor cycle segment enters the calculation, generating aerodynamic model correction parameters. This enables the airborne edge computing end to dynamically correct the rotor's basic aerodynamic performance results using historical residual patterns under similar rotor operating conditions. This approach reduces reliance on offline analysis at ground stations and centralized cloud computing, improving the real-time performance, continuity, and adaptability of rotor aerodynamic performance calculations during unmanned helicopter flight. Attached Figure Description

[0075] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings:

[0076] Figure 1 This is a flowchart of a method for calculating the aerodynamic performance of an unmanned helicopter rotor based on edge computing, as proposed in this invention.

[0077] Figure 2 This is a flowchart of the rotor periodic segment generation process in the edge computing-based unmanned helicopter rotor aerodynamic performance calculation method proposed in this invention.

[0078] Figure 3 This is a flowchart of aerodynamic residual decomposition in an edge computing-based method for calculating the aerodynamic performance of an unmanned helicopter rotor proposed in this invention.

[0079] Figure 4 This is a flowchart illustrating the generation of aerodynamic model correction parameters in an edge computing-based method for calculating the aerodynamic performance of an unmanned helicopter rotor proposed in this invention. Detailed Implementation

[0080] The present invention will now be described in further detail with reference to the accompanying drawings. These drawings are simplified schematic diagrams, illustrating only the basic structure of the invention, and therefore only show the components relevant to the invention.

[0081] refer to Figures 1-4 A method for calculating the aerodynamic performance of an unmanned helicopter rotor based on edge computing includes the following steps:

[0082] Collect multi-source operational data during the flight of the unmanned helicopter and generate rotor aerodynamic calculation data stream at the airborne edge computing terminal;

[0083] Rotor cycle segments are generated based on rotor speed data in rotor aerodynamic calculation data stream, and multi-source operating data are mapped to rotor cycle segments to generate rotor cycle state data.

[0084] Identify the current rotor operating condition of the unmanned helicopter based on rotor cycle state data, and generate rotor operating condition state unit;

[0085] Based on the rotor cycle state data and rotor operating condition state unit, a lightweight aerodynamic foundation calculation kernel is run to obtain the rotor foundation aerodynamic performance results;

[0086] Calculate the aerodynamic consistency deviation group based on the rotor's basic aerodynamic performance results and flight dynamic response data;

[0087] The sources of aerodynamic consistency deviation groups are decomposed to generate aerodynamic residual decomposition results.

[0088] Update the edge-end aerodynamic residual cluster library based on the aerodynamic residual decomposition results, and generate aerodynamic model correction parameters;

[0089] The basic aerodynamic performance results of the rotor are corrected based on the aerodynamic model correction parameters, and the rotor aerodynamic performance calculation results are output.

[0090] In this embodiment, collecting multi-source operational data during the flight of the unmanned helicopter and generating a rotor aerodynamic calculation data stream at the onboard edge computing terminal specifically includes:

[0091] Establish a data source mapping relationship for multi-source operation data, and connect the rotor speed sensor, pitch angle acquisition unit, flight speed acquisition unit, altitude acquisition unit, inertial measurement unit, motor controller, environmental parameter acquisition unit, wind speed estimation unit, and airframe vibration acquisition unit to the airborne edge computing terminal respectively;

[0092] Establish a one-to-one correspondence between each data source and rotor speed data, blade pitch angle data, flight speed data, altitude data, attitude angle data, angular velocity data, vertical acceleration data, motor current data, motor voltage data, ambient temperature data, air pressure data, wind speed estimation data, and airframe vibration data, and generate data source numbers for multi-source operation data;

[0093] The original sampling time of each type of multi-source operational data is obtained, and a unified time stamp is generated for each sample value according to the local clock of the airborne edge computing terminal. The unified time stamp is obtained by adding the original sampling time of the corresponding sample value to the time correction amount of the multi-source operational data relative to the local clock of the airborne edge computing terminal, so that multi-source operational data obtained from different acquisition sources can be classified into the same time reference.

[0094] Calculate the sampling frequency and data update interval for each type of multi-source operational data. Subtract one from the number of samples of the same type of multi-source operational data in the current statistical interval, and divide by the time difference between the unified time stamp corresponding to the last sample value and the unified time stamp corresponding to the first sample value of the multi-source operational data to obtain the sampling frequency of the multi-source operational data.

[0095] Subtract the unified timestamp corresponding to the previous sample value from the unified timestamp corresponding to the current sample value of this type of multi-source operational data to obtain the data update interval of this type of multi-source operational data.

[0096] A data access queue is established according to the data source number, sampling frequency and data update interval. Rotor speed data, pitch angle data, motor current data and motor voltage data are classified into the rotor power related queue, and flight speed data, altitude data, attitude angle data, angular velocity data and vertical acceleration data are classified into the flight response related queue.

[0097] Ambient temperature data, air pressure data, and wind speed estimation data are categorized into an environmental impact-related queue, while airframe vibration data are categorized into a rotor vibration-related queue. The sampled values ​​of each data entry queue are then arranged from first to last according to a unified time stamp.

[0098] Perform edge data integrity sorting on the sampled values ​​in each data access queue, group multi-source running data with the same time mark or time interval falling within the same acquisition alignment range into the same data frame, and record the data positions in the same data frame that have not received sampled values ​​as missing markers;

[0099] Record sampling values ​​that exceed the normal sampling range of the corresponding data source as an anomaly marker, and retain the original sampling values ​​of rotor speed data, blade pitch angle data, motor current data, motor voltage data, attitude angle data, and vertical acceleration data;

[0100] Based on the order of completion and the integrity of the edge data, each data is entered into the queue to generate a rotor aerodynamic calculation data stream;

[0101] The rotor aerodynamic calculation data stream includes a unified time stamp, data source number, rotor speed data, blade pitch angle data, flight speed data, altitude data, attitude angle data, angular velocity data, vertical acceleration data, motor current data, motor voltage data, ambient temperature data, air pressure data, wind speed estimation data, airframe vibration data, missing data markers, and anomaly markers.

[0102] Input the rotor aerodynamic calculation data stream into the edge buffer used to generate rotor periodic segments.

[0103] In this embodiment, generating rotor cycle segments based on rotor speed data in the rotor aerodynamic calculation data stream, and mapping multi-source operating data to rotor cycle segments to generate rotor cycle state data specifically includes:

[0104] Based on the rotor speed data and unified time stamp in the rotor aerodynamic calculation data stream, a rotor angular velocity sequence arranged by time is generated. The airborne edge computing terminal extracts the rotor speed data and the corresponding unified time stamp from the rotor aerodynamic calculation data stream and establishes a correspondence between each rotor speed data and the corresponding unified time stamp.

[0105] Based on the conversion relationship between the units of rotational speed and angular velocity, the rotational speed data of each rotor are converted into the rotor angular velocity under the corresponding unified time mark, and the rotor angular velocities under the unified time mark are arranged in chronological order to form a rotor angular velocity sequence.

[0106] Based on the rotor angular velocity sequence and unified time markers, the data boundary corresponding to the rotor completing a single rotation is determined. The airborne edge computing terminal continuously accumulates the rotation angle formed by the rotor between adjacent sampling times, according to the unified time markers in the order from first to last, combined with the time interval between adjacent unified time markers and the corresponding rotor angular velocity.

[0107] When the accumulated rotation angle reaches the angle corresponding to the rotor completing one revolution, the current unified time mark is determined as a data boundary, and the rotation angle for the next revolution is re-accumulated from the data boundary.

[0108] Based on the fluctuation state of multiple consecutive data boundaries and rotor speed data, a rotor cycle segment is generated. The airborne edge computing terminal divides the data between two adjacent data boundaries into a single rotation data segment, and calculates the speed fluctuation amplitude and average rotor speed based on the rotor speed data currently entering the edge buffer area.

[0109] The rotational speed fluctuation range is obtained from the difference between the maximum and minimum values ​​of the rotor rotational speed data currently entering the edge buffer zone, and the average rotor rotational speed is obtained by averaging the rotational speed data of each rotor currently entering the edge buffer zone;

[0110] The larger the proportion of the speed fluctuation amplitude to the average rotor speed, the fewer the number of single rotation data segments contained in a rotor cycle segment; the smaller the proportion of the speed fluctuation amplitude to the average rotor speed, the more the number of single rotation data segments contained in a rotor cycle segment. The corresponding number of consecutive single rotation data segments are combined into a rotor cycle segment.

[0111] Based on the start and end data boundaries of the rotor cycle segment, the multi-source operation data in the rotor aerodynamic calculation data stream is mapped to the corresponding rotor cycle segment. The airborne edge computing terminal uses the start and end data boundaries of each rotor cycle segment as the segment time range.

[0112] Extract the rotor pitch angle data, flight speed data, altitude data, attitude angle data, angular velocity data, vertical acceleration data, motor current data, motor voltage data, ambient temperature data, air pressure data, wind speed estimation data, and airframe vibration data from the rotor aerodynamic calculation data stream, and write the extracted multi-source operating data into the corresponding rotor cycle segment.

[0113] Based on the rotor cycle segment after writing multi-source operation data, the segment alignment data is generated with the unified time mark of the rotor speed data as the reference. The airborne edge computing terminal uses the unified time mark corresponding to the rotor speed data as the segment alignment time point.

[0114] The propeller pitch angle data, flight speed data, altitude data, attitude angle data, angular velocity data, vertical acceleration data, ambient temperature data, air pressure data, wind speed estimation data, and airframe vibration data are assigned to the corresponding intra-segment alignment time points according to the principle of minimizing the time interval between them and the intra-segment alignment time points. The motor voltage data and motor current data at the same intra-segment alignment time point are multiplied to obtain the motor power data corresponding to that intra-segment alignment time point. The intra-segment alignment data is formed by the multi-source operating data and motor power data after alignment.

[0115] Based on the aligned data within the segment, rotor cycle state data including average change, maximum change, change direction, and fluctuation concentration location are generated. The airborne edge computing terminal statistically analyzes the changes of each type of multi-source operation data within the same rotor cycle segment.

[0116] The average change is obtained by averaging the changes in sampled values ​​between aligned time points in adjacent segments within a segment. The maximum change is determined by the maximum absolute value of the changes in sampled values ​​between aligned time points in adjacent segments within a segment. The direction of change is determined by the increase or decrease relationship between the sampled value corresponding to the data boundary at the end of the segment and the sampled value corresponding to the data boundary at the beginning of the segment. The location of the fluctuation concentration is determined by the aligned time point within the segment corresponding to the maximum change. The rotor cycle state data is generated from the average change, maximum change, direction of change, and location of the fluctuation concentration.

[0117] In this embodiment, identifying the current rotor operating condition of the unmanned helicopter based on rotor cycle state data and generating a rotor operating condition state unit specifically includes:

[0118] Based on the rotor cycle state data, a working condition identification input group is generated. The airborne edge computing terminal extracts the average change, maximum change, change direction, and fluctuation concentration location of altitude data, vertical acceleration data, flight speed data, attitude angle data, angular velocity data, pitch angle data, motor power data, wind speed estimation data, and airframe vibration data from the rotor cycle state data corresponding to the same rotor cycle segment.

[0119] For any type of data involved in working condition identification, the ratio of the absolute value of the average change of the data type to the absolute value of the maximum change is calculated to obtain the normalized change of the data type. When the maximum change of the data type is zero, the normalized change of the data type is determined to be zero. The working condition identification input group is generated from the normalized change, change direction and fluctuation concentration location of each type of data.

[0120] The ascent and descent state, forward flight state, and attitude adjustment state are determined based on the working condition identification input group. The airborne edge computing terminal determines the climb state, descent state, or altitude holding state based on the normalized change of altitude data, the normalized change of vertical acceleration data, and the direction of change of both. The ascent and descent state is formed by the climb state, descent state, or altitude holding state.

[0121] Based on the normalized change and direction of the flight speed data, determine the acceleration forward flight state, deceleration forward flight state, or steady speed forward flight state, and form the forward flight state from the acceleration forward flight state, deceleration forward flight state, or steady speed forward flight state.

[0122] Based on the corresponding data with the largest normalized change in attitude angle data and angular velocity data, determine the main attitude adjustment direction and the main character velocity adjustment direction, and form the attitude adjustment state by the main attitude adjustment direction and the main character velocity adjustment direction.

[0123] The rotor load change state and disturbance state are determined based on the working condition identification input group. The airborne edge computing end determines the load increase state, load decrease state or load stable state based on the normalized change amount and direction of the pitch angle data, the normalized change amount and direction of the motor power data, and the increase and decrease relationship of the motor power data formed by the motor current data and the motor voltage data within the same rotor cycle segment. The rotor load change state is formed by the load increase state, load decrease state or load stable state.

[0124] Based on the relationship between the normalized change of wind speed estimation data and the normalized change of body vibration data, the wind speed disturbance state, vibration disturbance state, or combined disturbance state is determined, and the disturbance state is formed by the wind speed disturbance state, vibration disturbance state, or combined disturbance state.

[0125] The rotor operating condition state unit is generated based on the lift-down state, forward flight state, attitude adjustment state, rotor load change state and disturbance state. The airborne edge computing terminal classifies the lift-down state, forward flight state, attitude adjustment state, rotor load change state and disturbance state into the same rotor cycle segment.

[0126] The normalized change, change direction, fluctuation concentration location, and the start and end data boundaries of the corresponding rotor cycle segment for each state are written into the same operating condition record to generate a rotor operating condition state unit.

[0127] In this embodiment, the rotor basic aerodynamic performance results obtained by running the lightweight aerodynamic foundation calculation kernel based on the rotor periodic state data and rotor operating condition state unit specifically include:

[0128] The calculation inputs for the lightweight aerodynamic basic calculation core are generated based on the rotor cycle state data and rotor operating condition state units.

[0129] The airborne edge computing terminal reads the rotor cycle state data and rotor operating condition state unit corresponding to the same rotor cycle segment, and extracts the average value of rotor speed data, average value of pitch angle data, average value of flight speed data, average value of altitude data, average value of motor power data, average value of wind speed estimation data, average value of ambient temperature data, average value of air pressure data, and average change of attitude angle data from the rotor cycle state data.

[0130] The average values ​​of rotor speed data, pitch angle data, flight speed data, altitude data, motor power data, wind speed estimation data, ambient temperature data, air pressure data, attitude angle data, and rotor operating condition units are combined as calculation inputs.

[0131] Based on the ambient temperature data, air pressure data and attitude angle data in the calculation input, the air density and attitude change are calculated. The airborne edge computing terminal converts the average ambient temperature data within the current rotor cycle segment into absolute temperature.

[0132] The air density is obtained by dividing the average air pressure data by the product of the dry air gas constant and the absolute temperature.

[0133] The airborne edge computing terminal obtains the absolute values ​​of the average change in roll angle, pitch angle, and yaw angle within the current rotor cycle segment, and adds the three together to obtain the attitude change.

[0134] Based on the calculation input, air density, and attitude change, a simplified momentum calculation unit is run to obtain the momentum lift calculation and induced velocity calculation.

[0135] The airborne edge computing terminal reads the rotor radius from the rotor structure parameter table of the unmanned helicopter and calculates the rotor disk area based on the rotor radius;

[0136] The airborne edge computing terminal converts the average value of the rotor speed data into the rotor angular velocity corresponding to the current rotor cycle segment according to the conversion relationship between the units of rotational speed and angular velocity.

[0137] The airborne edge computing terminal multiplies the average value of the motor power data with the motor power conversion coefficient to obtain the effective input power. It then calculates the induced velocity calculation based on the effective input power, air density, and rotor disk area. Finally, it calculates the momentum lift calculation based on the air density, rotor disk area, and induced velocity calculation.

[0138] Based on the calculation input, air density and rotor angular velocity, the blade element calculation unit is run to obtain the blade element lift calculation, blade element drag calculation, and blade element torque calculation.

[0139] The airborne edge computing terminal calculates the forward ratio based on the average value of flight speed data, the average value of wind speed estimation data, rotor angular velocity, and rotor radius;

[0140] The airborne edge computing terminal reads the basic lift coefficient, pitch lift coefficient, advance ratio lift coefficient, basic drag coefficient, pitch drag coefficient, and advance ratio drag coefficient from the empirical aerodynamic parameter calculation unit, and calculates the lift coefficient and drag coefficient based on the average value of the pitch angle data and the advance ratio.

[0141] The airborne edge computing terminal calculates the lift of the blade element based on air density, rotor angular velocity, rotor radius, rotor disk area and lift coefficient, and calculates the drag of the blade element based on air density, rotor angular velocity, rotor radius, rotor disk area and drag coefficient. The blade element drag is then multiplied by the effective lever arm in the unmanned helicopter rotor structure parameter table to obtain the blade element torque.

[0142] The basic aerodynamic performance results of the rotor are generated based on the momentum lift calculation, the blade element lift calculation, the blade element drag calculation, the blade element torque calculation, and the rotor operating condition unit.

[0143] The airborne edge computing terminal reads the momentum calculation weight and blade element calculation weight corresponding to the current rotor operating condition from the empirical aerodynamic parameter calculation unit according to the rotor operating condition state unit, and makes the sum of the momentum calculation weight and the blade element calculation weight one.

[0144] The airborne edge computing terminal multiplies the momentum lift calculation quantity with the momentum calculation weight to obtain the momentum lift weight value, multiplies the blade element lift calculation quantity with the blade element calculation weight to obtain the blade element lift weight value, and adds the momentum lift weight value with the blade element lift weight value to obtain the basic lift result.

[0145] The airborne edge computing terminal corrects the calculated amount of blade element drag based on the advance ratio, attitude change, advance ratio drag correction coefficient, and attitude change drag correction coefficient to obtain the basic drag result.

[0146] The torque calculation of the blade elements is corrected based on the attitude change amount and the attitude change torque correction coefficient to obtain the basic torque result;

[0147] The basic power consumption is calculated based on the basic torque result, rotor angular velocity, and motor power conversion coefficient.

[0148] When the base power consumption result is not zero, divide the base lift result by the base power consumption result to obtain the base rotor efficiency result;

[0149] The basic aerodynamic performance results of the rotor are generated from the basic lift results, basic drag results, basic torque results, basic power consumption results, and basic rotor efficiency results.

[0150] In this embodiment, the calculation of the aerodynamic consistency deviation group based on the rotor's basic aerodynamic performance results and flight dynamic response data specifically includes:

[0151] Flight dynamic response data is generated based on rotor cycle segments and intra-segment alignment data. The airborne edge computing terminal reads the intra-segment alignment data corresponding to the current rotor cycle segment, multiplies the motor voltage data and motor current data at the same intra-segment alignment time point, and obtains the instantaneous motor power corresponding to the intra-segment alignment time point.

[0152] The actual motor power is obtained by averaging the instantaneous motor power within the current rotor cycle segment.

[0153] The change in altitude is obtained by subtracting the starting value of the altitude data from the ending value of the altitude data within the current rotor cycle segment.

[0154] The change in flight speed is obtained by subtracting the initial value of the flight speed data from the final value of the flight speed data within the current rotor cycle segment.

[0155] The change in attitude angular velocity is obtained by subtracting the initial value of the attitude angular velocity data from the final value of the attitude angular velocity data within the current rotor cycle segment.

[0156] The average vertical acceleration data within the current rotor cycle segment is obtained by averaging the vertical acceleration data.

[0157] Flight dynamic response data is generated from the actual power of the motor, the average value of vertical acceleration data, the change in altitude, the change in attitude angular velocity, and the change in flight speed;

[0158] The power consistency deviation is calculated based on the basic power consumption result in the rotor basic aerodynamic performance results and the actual motor power in the flight dynamic response data. The airborne edge computing terminal reads the basic power consumption result and the actual motor power corresponding to the current rotor cycle segment, subtracts the basic power consumption result from the actual motor power and takes the absolute value to obtain the power difference.

[0159] Add the absolute value of the actual power of the motor to the power stability value to obtain the power normalization reference; divide the power difference by the power normalization reference to obtain the power consistency deviation.

[0160] Among them, the power stability is determined by the minimum effective resolution of the motor power data stored at the airborne edge computing terminal;

[0161] The lift response deviation is calculated based on the basic lift results in the rotor's basic aerodynamic performance results and the average value of the vertical acceleration data and the altitude change in the flight dynamic response data.

[0162] The airborne edge computing terminal reads the mass and gravitational acceleration of the unmanned helicopter from the rotor structure parameter table, and reads the duration of the current rotor cycle segment;

[0163] Multiply the altitude change by two and divide by the square of the duration of the current rotor cycle segment to obtain the equivalent vertical acceleration corresponding to the altitude change.

[0164] Based on the effective number of vertical acceleration data samples and the effective number of altitude data samples in the current rotor cycle segment, the vertical acceleration response weight and altitude change response weight are determined respectively, and the sum of the vertical acceleration response weight and the altitude change response weight is made equal to one.

[0165] The vertical acceleration response components are obtained by multiplying the average value of the vertical acceleration data with the weight of the vertical acceleration response.

[0166] Multiply the equivalent vertical acceleration corresponding to the height change by the height change response weight to obtain the height change response component;

[0167] The lift response is obtained by adding the gravitational acceleration, vertical acceleration response components and altitude change response components together and then multiplying them by the mass of the unmanned helicopter.

[0168] Subtract the base lift result from the lift response and take the absolute value to obtain the lift difference; add the absolute value of the lift response to the lift stability to obtain the lift normalization reference; divide the lift difference by the lift normalization reference to obtain the lift response deviation.

[0169] Among them, the lift stability is obtained by multiplying the mass of the unmanned helicopter by the minimum effective resolution of the vertical acceleration data;

[0170] Based on the basic torque and drag results in the basic aerodynamic performance results of the rotor, as well as the attitude angular velocity and flight speed changes in the flight dynamic response data, the torque response deviation and drag response deviation are calculated respectively.

[0171] The airborne edge computing terminal reads the yaw moment of inertia and equivalent translational mass from the rotor structure parameter table of the unmanned helicopter, and reads the duration of the current rotor cycle segment;

[0172] Multiply the yaw moment of inertia by the change in attitude angular velocity, and then divide by the duration of the current rotor cycle segment to obtain the torque response.

[0173] Subtract the base torque result from the torque response and take the absolute value to obtain the torque difference; add the absolute value of the torque response to the torque stability to obtain the torque normalization benchmark.

[0174] Divide the torque difference by the torque normalization reference to obtain the torque response deviation;

[0175] Multiply the equivalent translational mass by the absolute value of the change in flight speed, and then divide by the duration of the current rotor cycle segment to obtain the drag response.

[0176] Subtract the basic resistance result from the resistance response and take the absolute value to obtain the resistance difference;

[0177] The absolute value of the resistance response is added to the resistance stability value to obtain the resistance normalization benchmark.

[0178] Divide the resistance difference by the resistance normalization benchmark to obtain the resistance response deviation.

[0179] The efficiency response deviation is calculated based on the basic rotor efficiency results in the basic rotor aerodynamic performance results, as well as the altitude change and flight speed change corresponding to unit power.

[0180] The airborne edge computing terminal reads the basic rotor efficiency results, actual motor power, altitude change, and flight speed change corresponding to the current rotor cycle segment;

[0181] When the actual power of the motor is not zero, the absolute value of the altitude change is added to the absolute value of the flight speed change to obtain the flight change composite quantity, and the flight change composite quantity is divided by the actual power of the motor to obtain the actual efficiency response quantity.

[0182] The efficiency difference is obtained by subtracting the basic rotor efficiency result from the actual efficiency response and taking the absolute value.

[0183] Add the absolute value of the actual efficiency response to the efficiency stability value to obtain the efficiency normalization benchmark; divide the efficiency difference by the efficiency normalization benchmark to obtain the efficiency response deviation.

[0184] When the actual power of the motor is zero, the efficiency response deviation is marked as invalid, and an efficiency invalidation flag is generated.

[0185] Aerodynamic consistency deviation group is generated based on power consistency deviation, lift response deviation, torque response deviation, drag response deviation, and efficiency response deviation;

[0186] The airborne edge computing terminal writes the current rotor cycle segment number, rotor operating condition unit, power consistency deviation, lift response deviation, torque response deviation, drag response deviation, efficiency response deviation, efficiency invalidation flag, and the start and end data boundaries of the corresponding rotor cycle segment into the same deviation record, and uses the deviation record as the aerodynamic consistency deviation group corresponding to the current rotor cycle segment.

[0187] In this embodiment, the source decomposition of the aerodynamic consistency deviation group and the generation of aerodynamic residual decomposition results specifically include:

[0188] Generate residual decomposition input groups based on aerodynamic consistency deviation groups and corresponding rotor cycle segments;

[0189] The airborne edge computing terminal reads the aerodynamic consistency deviation group corresponding to the current rotor cycle segment and extracts the power consistency deviation, lift response deviation, torque response deviation, drag response deviation, and efficiency response deviation.

[0190] Extract the wind speed estimation data change, flight speed change, blade pitch angle data change, motor power data change, vertical acceleration data average, airframe vibration data maximum, and unified time stamp of multi-source operation data from the intra-segment alignment data corresponding to the current rotor cycle segment;

[0191] Among them, the change in wind speed estimation data is obtained by subtracting the initial value from the final value of wind speed estimation data within the current rotor cycle segment;

[0192] The change in pitch angle data is obtained by subtracting the initial value from the final value of the pitch angle data within the current rotor cycle segment;

[0193] The change in motor power data is obtained by subtracting the starting value from the ending value of the motor power data within the current rotor cycle segment;

[0194] The residual decomposition input group is generated from power consistency deviation, lift response deviation, torque response deviation, drag response deviation, efficiency response deviation, wind speed estimation data variation, flight speed variation, propeller pitch angle data variation, motor power data variation, average vertical acceleration data, maximum airframe vibration data, and uniform time stamp.

[0195] Calculate the environmental disturbance residuals based on the drag response bias, wind speed estimation data variation, and flight speed variation in the residual decomposition input group;

[0196] The airborne edge computing terminal divides the absolute value of the change in wind speed estimation data by the sum of the absolute value of the change in wind speed estimation data and the minimum effective resolution of wind speed estimation data to obtain the wind speed disturbance intensity.

[0197] When the change in wind speed estimation data and the change in flight speed change in the same direction, the wind speed-velocity correlation value is set to one; when the change in wind speed estimation data and the change in flight speed change in opposite directions or when either change is zero, the wind speed-velocity correlation value is set to zero.

[0198] Multiply the drag response deviation, wind speed disturbance intensity, and wind speed correlation to obtain the environmental disturbance residual;

[0199] The load variation residual is calculated based on the lift response deviation, torque response deviation, pitch angle data change, motor power data change, and vertical acceleration data average value in the residual decomposition input group.

[0200] The airborne edge computing terminal divides the absolute value of the pitch angle data change by the sum of the absolute value of the pitch angle data change and the minimum effective resolution of the pitch angle data to obtain the pitch change intensity.

[0201] The power change intensity is obtained by dividing the absolute value of the change in motor power data by the sum of the absolute value of the change in motor power data and the minimum effective resolution of the motor power data.

[0202] The vertical acceleration response intensity is obtained by dividing the absolute value of the average vertical acceleration data by the sum of the absolute value of the average vertical acceleration data and the minimum effective resolution of the vertical acceleration data.

[0203] The load variation data are obtained by averaging the pitch variation intensity, power variation intensity, and vertical acceleration response intensity.

[0204] Based on the absolute values ​​of the lift response deviation and the torque response deviation, determine the weights of the lift response deviation and the torque response deviation respectively, and make the sum of the weights of the lift response deviation and the torque response deviation equal to one.

[0205] The lift response deviation is multiplied by the lift response deviation weight to obtain the lift deviation component, and the torque response deviation is multiplied by the torque response deviation weight to obtain the torque deviation component. The lift deviation component and the torque deviation component are added together and then multiplied by the load change data to obtain the load change residual.

[0206] Calculate the rotor efficiency attenuation residual based on the power consistency deviation, efficiency response deviation and maximum value of airframe vibration data in the residual decomposition input group;

[0207] The airborne edge computing terminal reads the power consistency deviation and efficiency response deviation corresponding to the current rotor cycle segment and the previous consecutive rotor cycle segments, counts the number of segments in which the power consistency deviation and efficiency response deviation both increase or decrease in the same direction between adjacent rotor cycle segments, and divides the number of segments by the total number of rotor cycle segments included in the statistics to obtain the same-direction persistence coefficient.

[0208] The vibration influence coefficient is obtained by dividing the maximum value of the body vibration data by the sum of the maximum value of the body vibration data and the minimum effective resolution of the body vibration data.

[0209] The efficiency deviation is obtained by averaging the power consistency deviation and the efficiency response deviation. The efficiency deviation is then multiplied by the efficiency deviation, the same direction duration coefficient, and the non-vibration influence ratio obtained from the vibration influence coefficient to obtain the rotor efficiency attenuation residual.

[0210] When an efficiency invalid mark is recorded in the aerodynamic consistency deviation group, the rotor efficiency decay residual is determined based on the continuous change of the power consistency deviation in the current rotor cycle segment and the previous consecutive rotor cycle segments.

[0211] Calculate the sensor synchronization residual based on the jump position of the aerodynamic consistency deviation group and the time stamp misalignment of multi-source operating data;

[0212] The airborne edge computing terminal calculates the difference between the unified time stamp of each type of multi-source operation data in the current rotor cycle segment and the unified time stamp corresponding to the rotor speed data. The absolute values ​​of each type of difference are averaged and then divided by the duration of the current rotor cycle segment to obtain the time stamp misalignment.

[0213] The power consistency deviation, lift response deviation, torque response deviation, drag response deviation and efficiency response deviation corresponding to the current rotor cycle segment are calculated with the same type of deviation corresponding to the previous rotor cycle segment. The absolute values ​​of each difference are averaged to obtain the deviation jump variable.

[0214] Multiply the time stamp misalignment by the deviation jump variable to obtain the sensor synchronization residual;

[0215] The aerodynamic residual decomposition results are generated based on the environmental disturbance residual, load change residual, rotor efficiency decay residual, and sensor synchronization residual.

[0216] The airborne edge computing terminal adds up the environmental disturbance residual, load change residual, rotor efficiency decay residual, and sensor synchronization residual to obtain the total residual;

[0217] When the total residual is not zero, the environmental disturbance residual, load change residual, rotor efficiency decay residual and sensor synchronization residual are divided by the total residual to obtain the proportion of environmental disturbance residual, load change residual, rotor efficiency decay residual and sensor synchronization residual respectively.

[0218] When the total residual is zero, the percentage of each residual is recorded as zero. The current rotor cycle segment number, rotor operating condition unit, environmental disturbance residual, load change residual, rotor efficiency decay residual, sensor synchronization residual, percentage of each residual, and the starting and ending data boundaries of the corresponding rotor cycle segment are written into the same residual decomposition record to generate aerodynamic residual decomposition results.

[0219] In this embodiment, updating the edge-end aerodynamic residual cluster library based on the aerodynamic residual decomposition results and generating aerodynamic model correction parameters specifically includes:

[0220] Based on the aerodynamic residual decomposition results, the operating condition residual is written into the record. The airborne edge computing terminal reads the aerodynamic residual decomposition results corresponding to the current rotor cycle segment and extracts the environmental disturbance residual, load change residual, rotor efficiency decay residual, sensor synchronization residual, and the corresponding proportions of environmental disturbance residual, load change residual, rotor efficiency decay residual, and sensor synchronization residual.

[0221] It reads the rotor operating condition status unit, the duration of the current rotor cycle segment, and the flight dynamic response data, and combines the environmental disturbance residual, load change residual, rotor efficiency decay residual, sensor synchronization residual, the proportion of each residual, the rotor operating condition status unit, the duration of the current rotor cycle segment, and the flight dynamic response data into the operating condition residual and writes it into the record.

[0222] Write the operating condition residuals into the corresponding operating condition residual clusters according to the rotor operating condition status unit;

[0223] The airborne edge computing terminal combines the lift, forward flight, attitude adjustment, rotor load change, and disturbance states in the rotor condition state unit into a condition index, and searches for a condition residual cluster that matches the condition index in the edge aerodynamic residual cluster library; if a matching condition residual cluster exists, the condition residual is written into the record and appended to that condition residual cluster.

[0224] If no consistent operating condition residual cluster exists, a new operating condition residual cluster is created using the operating condition index, and the operating condition residual write records are written to the newly created operating condition residual cluster.

[0225] Update the occurrence frequency, duration, direction of change, average magnitude, and corresponding flight dynamic response of various residuals based on the written residual clusters;

[0226] The airborne edge computing terminal performs statistical analysis on the environmental disturbance residuals, load change residuals, rotor efficiency attenuation residuals, and sensor synchronization residuals in the residual cluster under the same operating condition.

[0227] The number of records with residual values ​​greater than zero is taken as the occurrence count of the corresponding residual. The duration of the rotor cycle segment corresponding to the records with residual values ​​greater than zero is added together to obtain the duration of the corresponding residual. The residual value of the current condition residual written into the record is compared with the same type of residual value of the previous condition residual written into the record.

[0228] And determine the direction of change of the corresponding residual as increasing, stable, or decreasing based on whether the difference is positive, zero, or negative;

[0229] The average magnitude of the corresponding residual is obtained by averaging the residual values ​​of the same type of residual in the same operating condition residual cluster, and the corresponding flight dynamic response data is written into the operating condition residual cluster.

[0230] When a new rotor cycle segment enters the calculation, similar operating condition residual clusters are matched from the edge aerodynamic residual cluster library based on the current rotor operating condition state unit and the current aerodynamic residual decomposition results.

[0231] The airborne edge computing terminal combines the current rotor operating condition state unit corresponding to the new rotor cycle segment into the current operating condition index, and combines the proportion of environmental disturbance residual, load change residual, rotor efficiency decay residual and sensor synchronization residual in the current aerodynamic residual decomposition result into the current residual proportion vector.

[0232] The airborne edge computing terminal reads the working condition index and average residual ratio vector of each candidate working condition residual cluster, and divides the number of the same state in the current working condition index and the working condition index of the candidate working condition residual cluster by the total number of states to obtain the working condition matching value.

[0233] Take the absolute value of the difference between the current residual proportion vector and the average residual proportion vector of the candidate working condition residual cluster, and add the absolute values ​​together and take half of them to obtain the residual proportion difference; subtract the residual proportion difference from the first value to obtain the residual proportion matching value.

[0234] The working condition matching value and the residual proportion matching value are weighted and added together according to the working condition matching weight to obtain the similarity of the candidate working condition residual clusters, and the candidate working condition residual cluster with the highest similarity is determined as the similar working condition residual cluster.

[0235] Residual correction factors are generated based on the average amplitude, duration, and direction of change of various residuals in the residual cluster of similar working conditions.

[0236] The airborne edge computing terminal reads the average amplitude, duration and direction of change of environmental disturbance residual, load change residual, rotor efficiency attenuation residual and sensor synchronization residual in the residual cluster of similar working conditions;

[0237] For any type of residual, divide the duration of the residual by the sum of the duration of the residual and the duration stability to obtain the duration influence ratio, and convert the change direction of the residual into the change direction adjustment amount.

[0238] Among them, the direction of increase corresponds to positive adjustment, the direction of stability corresponds to zero adjustment, and the direction of decrease corresponds to negative adjustment;

[0239] Multiply the average amplitude and duration of the residuals by the direction adjustment result obtained from the change direction adjustment to obtain the residual correction factor corresponding to the residuals. The residual correction factor group is composed of the environmental disturbance residual correction factor, the load change residual correction factor, the rotor efficiency decay residual correction factor, and the sensor synchronization residual correction factor.

[0240] Based on the residual correction factor group, aerodynamic model correction parameters are generated. The airborne edge computing terminal reads the mapping coefficients of various residual correction factors on lift, drag, torque, induced velocity, power consumption and rotor efficiency from the correction parameter mapping table.

[0241] When generating the lift correction parameters, the various residual correction factors are multiplied by their corresponding lift mapping coefficients and summed. One is then added to the summation result to obtain the lift correction parameters.

[0242] When generating the resistance correction parameters, the various residual correction factors are multiplied by their corresponding resistance mapping coefficients and summed. One is then added to the summation result to obtain the resistance correction parameters.

[0243] When generating torque correction parameters, the various residual correction factors are multiplied by their corresponding torque mapping coefficients and summed, and then one is added to the summation result to obtain the torque correction parameters.

[0244] When generating the induced velocity correction parameters, the various residual correction factors are multiplied by their corresponding induced velocity mapping coefficients and summed. One is then added to the summation result to obtain the induced velocity correction parameters.

[0245] When generating power consumption correction parameters, the various residual correction factors are multiplied by their corresponding power consumption mapping coefficients and summed. One is then added to the summation result to obtain the power consumption correction parameters.

[0246] When generating rotor efficiency correction parameters, the various residual correction factors are multiplied by their corresponding rotor efficiency mapping coefficients and summed. One is then added to the summation result to obtain the rotor efficiency correction parameters.

[0247] The aerodynamic model correction parameters consist of lift correction parameters, drag correction parameters, torque correction parameters, induced velocity correction parameters, power consumption correction parameters, and rotor efficiency correction parameters.

[0248] In this embodiment, when a new rotor cycle segment enters the calculation, matching similar operating condition residual clusters from the edge-end aerodynamic residual cluster library based on the current rotor operating condition state unit and the current aerodynamic residual decomposition result specifically includes:

[0249] Generate the current operating condition index based on the current rotor operating condition state unit corresponding to the new rotor cycle segment;

[0250] The airborne edge computing terminal reads the lift-up status, forward flight status, attitude adjustment status, rotor load change status and disturbance status from the current rotor operating condition status unit, and writes each status into the same operating condition index record according to the arrangement order of the above statuses in the rotor operating condition status unit to generate the current operating condition index.

[0251] Generate the current residual proportion vector based on the current aerodynamic residual decomposition result corresponding to the new rotor cycle segment;

[0252] The airborne edge computing terminal reads the proportions of environmental disturbance residuals, load change residuals, rotor efficiency decay residuals, and sensor synchronization residuals from the current aerodynamic residual decomposition results. Then, in a fixed order of environmental disturbance residuals, load change residuals, rotor efficiency decay residuals, and sensor synchronization residuals, it writes each residual proportion into the same residual proportion record to generate the current residual proportion vector.

[0253] Generate candidate operating condition residual clusters and their average residual proportion vectors based on the edge-end aerodynamic residual cluster library;

[0254] The airborne edge computing terminal reads the operating condition residual cluster containing the operating condition residual write record from the edge aerodynamic residual cluster library as a candidate operating condition residual cluster;

[0255] The average residual proportions of environmental disturbance residuals, load change residuals, rotor efficiency decay residuals, and sensor synchronization residuals in each candidate operating condition residual cluster are calculated separately. The average residual proportion vector corresponding to each candidate operating condition residual cluster is generated according to the residual type order of the current residual proportion vector.

[0256] The operating condition matching value is calculated based on the current operating condition index and the operating condition index of the candidate operating condition residual cluster. The airborne edge computing terminal will calculate the lift-down state, forward flight state, attitude adjustment state, rotor load change state and disturbance state in the current operating condition index.

[0257] Each candidate working condition residual cluster is compared with the same type of state in the working condition index. The number of consistent states is divided by the total number of states participating in the comparison to obtain the working condition matching value corresponding to the candidate working condition residual cluster.

[0258] Calculate the residual proportion matching value based on the current residual proportion vector and the average residual proportion vector of the candidate working condition residual cluster;

[0259] The airborne edge computing terminal subtracts the residual proportions of environmental disturbance, load change, rotor efficiency decay, and sensor synchronization from the residual proportions in the current residual proportion vector with the residual proportions of the same type in the average residual proportion vector of the candidate operating condition residual cluster, and takes the absolute value. The absolute values ​​are then added together to obtain the residual proportion difference. The residual proportion matching value corresponding to the candidate operating condition residual cluster is determined based on the difference between the residual proportion difference and the complete residual proportion.

[0260] Based on the working condition matching value and the residual proportion matching value, a cluster of similar working condition residuals is determined, and the airborne edge computing terminal reads the working condition matching weight and the residual proportion matching weight.

[0261] Multiply the working condition matching value corresponding to the candidate working condition residual cluster by the working condition matching weight to obtain the working condition matching weighted value. Multiply the residual proportion matching value corresponding to the candidate working condition residual cluster by the residual proportion matching weight to obtain the residual proportion matching weighted value. Add the working condition matching weighted value and the residual proportion matching weighted value to obtain the similarity of the candidate working condition residual cluster. The candidate working condition residual cluster with the highest similarity is determined as the similar working condition residual cluster.

[0262] When there are multiple candidate operating condition residual clusters with the highest similarity, the candidate operating condition residual cluster with the most recent written operating condition residual record is determined as the similar operating condition residual cluster.

[0263] In this embodiment, the basic aerodynamic performance results of the rotor are corrected according to the aerodynamic model correction parameters, and the output rotor aerodynamic performance calculation results specifically include:

[0264] The correction calculation input is generated based on the aerodynamic model correction parameters and the basic aerodynamic performance results of the rotor.

[0265] The airborne edge computing terminal reads the aerodynamic model correction parameters and the basic aerodynamic performance results of the rotor corresponding to the current rotor cycle segment, and extracts lift correction parameters, drag correction parameters, torque correction parameters, induced velocity correction parameters, power consumption correction parameters and rotor efficiency correction parameters from the aerodynamic model correction parameters;

[0266] Extract the basic lift, drag, torque, power consumption, and rotor efficiency results from the basic rotor aerodynamic performance results. Incorporate the aerodynamic model correction parameters and the basic rotor aerodynamic performance results into the same current rotor cycle segment to generate correction calculation inputs.

[0267] The basic lift, drag, and torque results are corrected based on the correction calculation input to obtain the corrected lift, drag, and torque results. The airborne edge computing terminal multiplies the basic lift result with the lift correction parameter to obtain the corrected lift result.

[0268] Multiply the base resistance result by the resistance correction parameter to obtain the corrected resistance result;

[0269] Multiply the base torque result by the torque correction parameter to obtain the corrected torque result, and write the corrected lift result, corrected drag result, and corrected torque result into the correction calculation record corresponding to the current rotor cycle segment;

[0270] Based on the corrected calculation input, the calculation quantities related to the rotor flow field, the basic power consumption results, and the basic rotor efficiency results are corrected to obtain the corrected induced velocity calculation quantities, the corrected power consumption results, and the corrected rotor efficiency results.

[0271] The airborne edge computing terminal reads the induced velocity calculation amount obtained from the simplified momentum calculation unit, and multiplies the induced velocity calculation amount with the induced velocity correction parameter to obtain the corrected induced velocity calculation amount;

[0272] Multiply the base power consumption result by the power consumption correction parameter to obtain the corrected power consumption result;

[0273] Multiply the basic rotor efficiency result by the rotor efficiency correction parameter to obtain the corrected rotor efficiency result, and write the corrected induced velocity calculation, the corrected power consumption result, and the corrected rotor efficiency result into the correction calculation record corresponding to the current rotor cycle segment.

[0274] The aerodynamic stability results are calculated based on the corrected lift, drag, torque, power consumption, rotor efficiency, and aerodynamic residual decomposition results.

[0275] The airborne edge computing terminal reads the environmental disturbance residual ratio, load change residual ratio, rotor efficiency decay residual ratio, and sensor synchronization residual ratio corresponding to the current rotor cycle segment, and selects the largest residual ratio from the four types of residual ratios.

[0276] The airborne edge computing terminal reads the power consistency deviation, lift response deviation, torque response deviation, drag response deviation, and efficiency response deviation corresponding to the current rotor cycle segment, and averages the five types of deviations to obtain the comprehensive consistency deviation.

[0277] Multiply the maximum residual percentage by the overall consistency deviation to obtain the stability deduction; subtract the stability deduction from one to obtain the initial aerodynamic stability result.

[0278] When the initial aerodynamic stability result is less than zero, the aerodynamic stability result is set to zero.

[0279] When the initial aerodynamic stability result is greater than one, the aerodynamic stability result is set to one.

[0280] When the initial aerodynamic stability result is between zero and one, the initial aerodynamic stability result is determined as the aerodynamic stability result.

[0281] The rotor aerodynamic performance calculation results are generated based on the corrected lift, drag, torque, power consumption, rotor efficiency, and aerodynamic stability results.

[0282] The airborne edge computing terminal writes the current rotor cycle segment number, rotor operating condition unit, corrected lift result, corrected drag result, corrected torque result, corrected induced velocity calculation, corrected power consumption result, corrected rotor efficiency result, aerodynamic stability result, and the start and end data boundaries of the corresponding rotor cycle segment into the same result record, and outputs the result record as the rotor aerodynamic performance calculation result.

[0283] Example 1: To verify the feasibility of this invention in practice, it was applied to a flight test scenario of the rotor aerodynamic performance of an unmanned helicopter. In this scenario, the unmanned helicopter was equipped with an onboard edge computing terminal, flight control data interface, motor controller data interface, inertial measurement unit, barometric altitude acquisition unit, environmental parameter acquisition unit, and fuselage vibration acquisition unit. It completed hovering, climbing, descending, forward flight, turning, and load-changing flight within a closed test airspace. During the test, the rotor speed of the unmanned helicopter was maintained between 1450 and 1680 revolutions per minute, the flight altitude between 18 and 52 meters, and the flight speed between 0 and 13.6 meters per second. Each flight lasted 18 minutes, and data from five consecutive flights were collected. Due to crosswind disturbances, load changes, and inconsistent sensor sampling frequencies during flight, using a fixed time window for rotor aerodynamic performance calculations easily leads to the problem of multiple rotor rotation states being mixed within the same calculation window. This results in discrepancies between the lift, power consumption, and rotor efficiency results output by the basic aerodynamic model and the actual flight dynamic response.

[0284] In this scenario, the airborne edge computing terminal first collects rotor speed, pitch angle, flight speed, altitude, attitude angle, angular velocity, vertical acceleration, motor current, motor voltage, ambient temperature, air pressure, wind speed estimation, and airframe vibration data during the unmanned helicopter's flight. It then adds a unified time stamp to the multi-source operational data from different sources, generating a rotor aerodynamic calculation data stream. Based on the rotor speed data, the airborne edge computing terminal determines the data boundary corresponding to a single rotor rotation, combines multiple consecutive single-rotation data segments into rotor cycle segments, and maps multi-source operational data within the same time range to the corresponding rotor cycle segments, generating rotor cycle state data. Subsequently, the airborne edge computing terminal identifies the rotor operating conditions based on the rotor cycle state data, generates rotor operating condition state units, and inputs the rotor cycle state data and rotor operating condition state units into a lightweight aerodynamic foundation calculation kernel to obtain foundation lift, foundation drag, foundation torque, foundation power consumption, and foundation rotor efficiency results.

[0285] After obtaining the basic aerodynamic performance results of the rotor, the airborne edge computing terminal compares the consistency of the basic power consumption results with the actual motor power, the basic lift results with the flight response corresponding to vertical acceleration and altitude changes, the basic torque results with the attitude angular velocity changes, the basic drag results with the flight speed changes, and the basic rotor efficiency results with the altitude and flight speed changes corresponding to unit power, generating an aerodynamic consistency deviation group. For rotor cycle segments with concentrated deviations, the airborne edge computing terminal further decomposes the aerodynamic consistency deviation group into environmental disturbance residuals, load change residuals, rotor efficiency attenuation residuals, and sensor synchronization residuals based on the time stamp misalignment of wind speed estimation data, load change data, airframe vibration data, and multi-source operation data, generating aerodynamic residual decomposition results. Subsequently, the airborne edge computing terminal writes the aerodynamic residual decomposition results into the edge aerodynamic residual cluster library, and matches the residual cluster of similar operating conditions when a new rotor cycle segment enters the calculation, generates aerodynamic model correction parameters, and finally corrects the basic aerodynamic performance results of the rotor, outputting the rotor aerodynamic performance calculation results.

[0286] In five consecutive flight sorties, when comparing fixed-time-window and ground-station batch calculation methods, the average delay per calculation window was 1.86 seconds, the average rotor efficiency calculation error was 9.7%, and during climb and load change phases, the average deviation between the base power consumption result and the actual motor power reached 12.4%. After adopting this invention, the airborne edge computing terminal organizes data in rotor cycle segments, with an average processing time of 0.18 seconds per rotor cycle segment, reducing the average rotor efficiency calculation error to 4.1%, and the average deviation between the base power consumption result and the actual motor power to 5.3%. In flight segments with significant crosswind disturbances, this invention can primarily attribute drag response deviations to environmental disturbance residuals; in flight segments with load changes, this invention can primarily attribute lift response deviations and torque response deviations to load change residuals; in short-term sensor misalignment segments, this invention can generate sensor synchronization residuals based on the deviation jump position and time marker misalignment position, thus providing a corresponding data source for subsequent aerodynamic model correction parameters.

[0287] Table 1 Comparison of the Implementation Results of Rotor Aerodynamic Performance Calculation

[0288] Test Project Fixed time window calculation method Calculation method of this invention Improvement Results Average processing time per computation segment 1.86 seconds 0.18 seconds Reduced by 90.3% Average error in rotor efficiency calculation 9.7% 4.1% Decreased by 5.6 percentage points Average deviation of power consumption 12.4% 5.3% Decreased by 7.1 percentage points Lift response mean deviation 10.8% 4.9% Decreased by 5.9 percentage points Average deviation of resistance response 13.1% 6.2% Decreased by 6.9 percentage points Torque response average deviation 11.6% 5.7% Decreased by 5.9 percentage points Crosswind disturbance segment residual identification accuracy 72.5% 89.6% An increase of 17.1 percentage points Accuracy of residual identification for load variation segments 70.8% 88.3% An increase of 17.5 percentage points Sensor synchronization anomaly segment recognition rate 68.4% 86.9% An increase of 18.5 percentage points Effective aerodynamic output rate per flight 91.2% 98.1% An increase of 6.9 percentage points

[0289] As can be seen from the table above, this invention significantly improves the real-time performance, accuracy, and residual identification capabilities of unmanned helicopter rotor aerodynamic performance calculations compared to the fixed-time-window calculation method. Regarding computational efficiency, the average processing time for a single computation segment using the fixed-time-window method is 1.86 seconds, while this invention directly generates the rotor aerodynamic calculation data stream through an airborne edge computing terminal and organizes multi-source operational data according to rotor cycle segments, reducing the average processing time for a single rotor cycle segment to 0.18 seconds, a reduction of 90.3%. This result demonstrates that this invention reduces reliance on batch calculations at ground stations, making rotor aerodynamic performance calculations more suitable for real-time processing requirements during flight.

[0290] Regarding calculation accuracy, this invention reduces the calculation deviations for rotor efficiency, power consumption, lift response, drag response, and torque response. Specifically, the average error in rotor efficiency calculation is reduced from 9.7% to 4.1%, the average deviation in power consumption from 12.4% to 5.3%, the average deviation in lift response from 10.8% to 4.9%, the average deviation in drag response from 13.1% to 6.2%, and the average deviation in torque response from 11.6% to 5.7%. These data demonstrate that this invention, by comparing the consistency between the basic rotor aerodynamic performance results and the flight dynamic response data, can promptly detect deviations between the basic aerodynamic calculation results and the actual flight response. Furthermore, by correcting the aerodynamic model correction parameters, the basic lift, drag, torque, power consumption, and rotor efficiency results are adjusted, thereby improving the matching degree between the rotor aerodynamic performance calculation results and the actual flight conditions.

[0291] In terms of residual source identification, this invention outperforms the fixed-time-window calculation method in identifying crosswind disturbance segments, load change segments, and sensor synchronization anomaly segments. The accuracy rate of residual identification for crosswind disturbance segments increased from 72.5% to 89.6%, for load change segments from 70.8% to 88.3%, and for sensor synchronization anomaly segments from 68.4% to 86.9%. This result demonstrates that by decomposing the aerodynamic consistency deviation group, this invention can generate corresponding residuals for environmental disturbances, load changes, rotor efficiency decay, and sensor synchronization problems. Furthermore, by utilizing the edge-end aerodynamic residual cluster library to match residual clusters of similar operating conditions, the aerodynamic model correction has a more explicit data source. The effective aerodynamic result output rate per flight increased from 91.2% to 98.1%, indicating that this invention can improve the continuous output capability of rotor aerodynamic performance calculation results during flight.

[0292] The above are merely preferred embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. An edge-computing-based unmanned helicopter rotor aerodynamic performance calculation method, characterized in that, Includes the following steps: Collect multi-source operational data during the flight of the unmanned helicopter and generate rotor aerodynamic calculation data stream at the airborne edge computing terminal; Rotor cycle segments are generated based on rotor speed data in rotor aerodynamic calculation data stream, and multi-source operating data are mapped to rotor cycle segments to generate rotor cycle state data. Identify the current rotor operating condition of the unmanned helicopter based on rotor cycle state data, and generate rotor operating condition state unit; Based on the rotor cycle state data and rotor operating condition state unit, the lightweight aerodynamic foundation calculation kernel is run to obtain the rotor foundation aerodynamic performance results; Calculate the aerodynamic consistency deviation group based on the rotor's basic aerodynamic performance results and flight dynamic response data; The sources of aerodynamic consistency deviation groups are decomposed to generate aerodynamic residual decomposition results. Update the edge-end aerodynamic residual cluster library based on the aerodynamic residual decomposition results, and generate aerodynamic model correction parameters; The basic aerodynamic performance results of the rotor are corrected based on the aerodynamic model correction parameters, and the rotor aerodynamic performance calculation results are output.

2. The method for calculating the aerodynamic performance of an unmanned helicopter rotor based on edge computing according to claim 1, characterized in that, The airborne edge computing terminal adds a unified time stamp to multi-source operating data from different sources, and establishes a data access queue according to data source, sampling frequency and data update interval to generate rotor aerodynamic computing data stream.

3. The method for calculating the aerodynamic performance of an unmanned helicopter rotor based on edge computing according to claim 1, characterized in that, The process of generating rotor cycle segments based on rotor speed data in the rotor aerodynamic calculation data stream, and mapping multi-source operating data to rotor cycle segments to generate rotor cycle state data specifically includes: Generate a time-ordered sequence of rotor angular velocities based on rotor speed data and a unified time stamp from the rotor aerodynamic calculation data stream; The data boundary corresponding to the completion of a single rotor rotation is determined based on the rotor angular velocity sequence and a unified time stamp. Generate rotor cycle segments based on the fluctuation state of multiple consecutive data boundaries and rotor speed data; Based on the start and end data boundaries of the rotor cycle segment, the multi-source operating data in the rotor aerodynamic calculation data stream is mapped to the corresponding rotor cycle segment. Generate intra-segment alignment data based on the rotor cycle segment after writing multi-source operational data; Generate rotor periodic state data containing average change, maximum change, change direction, and fluctuation concentration location based on the intra-segment alignment data.

4. The method for calculating the aerodynamic performance of an unmanned helicopter rotor based on edge computing according to claim 1, characterized in that, The step of identifying the current rotor operating condition of the unmanned helicopter based on rotor cycle state data and generating a rotor operating condition state unit specifically includes: Generate operating condition identification input groups based on rotor cycle status data; The ascent and descent status, forward flight status, and attitude adjustment status are determined based on the working condition recognition input group. The rotor load change state and disturbance state are determined based on the operating condition identification input group; The rotor operating condition state unit is generated based on the lift-down state, forward flight state, attitude adjustment state, rotor load change state, and disturbance state.

5. The method for calculating the aerodynamic performance of an unmanned helicopter rotor based on edge computing according to claim 1, characterized in that, The process of running a lightweight aerodynamic foundation calculation kernel based on rotor cycle state data and rotor operating condition state unit to obtain the rotor foundation aerodynamic performance results specifically includes: The calculation inputs for the lightweight aerodynamic basic calculation core are generated based on the rotor cycle state data and rotor operating condition state units. Calculate air density and attitude change based on ambient temperature data, air pressure data, and attitude angle data in the calculation input; Based on the calculation input, air density, and attitude change, a simplified momentum calculation unit is run to obtain the momentum lift calculation and induced velocity calculation. Based on the calculation input, air density and rotor angular velocity, the blade element calculation unit is run to obtain the blade element lift calculation, blade element drag calculation, and blade element torque calculation. The basic aerodynamic performance results of the rotor are generated based on the momentum lift calculation, the blade element lift calculation, the blade element drag calculation, the blade element torque calculation, and the rotor operating condition unit.

6. The method for calculating the aerodynamic performance of an unmanned helicopter rotor based on edge computing according to claim 1, characterized in that, The calculation of the aerodynamic consistency deviation group based on the rotor's basic aerodynamic performance results and flight dynamic response data specifically includes: Flight dynamic response data is generated based on rotor cycle segments and intra-segment alignment data; The power consistency deviation is calculated based on the basic power consumption results in the rotor basic aerodynamic performance results and the actual motor power in the flight dynamic response data. The lift response deviation is calculated based on the basic lift results in the rotor's basic aerodynamic performance results and the average value of vertical acceleration data and altitude change in the flight dynamic response data. Based on the basic torque and drag results in the basic aerodynamic performance results of the rotor, as well as the attitude angular velocity and flight speed changes in the flight dynamic response data, the torque response deviation and drag response deviation are calculated respectively. The efficiency response deviation is calculated based on the basic rotor efficiency results in the basic rotor aerodynamic performance results, as well as the altitude change and flight speed change corresponding to unit power. Aerodynamic consistency deviation group is generated based on power consistency deviation, lift response deviation, torque response deviation, drag response deviation, and efficiency response deviation.

7. The method for calculating the aerodynamic performance of an unmanned helicopter rotor based on edge computing according to claim 1, characterized in that, The specific steps for decomposing the aerodynamic consistency deviation group to generate aerodynamic residual decomposition results include: Generate residual decomposition input groups based on aerodynamic consistency deviation groups and corresponding rotor cycle segments; Calculate the environmental disturbance residuals based on the drag response bias, wind speed estimation data variation, and flight speed variation in the residual decomposition input group; The load variation residual is calculated based on the lift response deviation, torque response deviation, pitch angle data change, motor power data change, and vertical acceleration data average value in the residual decomposition input group. Calculate the rotor efficiency attenuation residual based on the power consistency deviation, efficiency response deviation and maximum value of airframe vibration data in the residual decomposition input group; Calculate the sensor synchronization residual based on the jump position of the aerodynamic consistency deviation group and the time stamp misalignment of multi-source operating data; The aerodynamic residual decomposition results are generated based on the environmental disturbance residual, load change residual, rotor efficiency decay residual, and sensor synchronization residual.

8. The method for calculating the aerodynamic performance of an unmanned helicopter rotor based on edge computing according to claim 1, characterized in that, The step of updating the edge-end aerodynamic residual cluster library based on the aerodynamic residual decomposition results and generating aerodynamic model correction parameters specifically includes: Based on the aerodynamic residual decomposition results, the operating condition residual is generated and written into the record; Write the operating condition residual into the corresponding operating condition residual cluster according to the rotor operating condition status unit; Update the occurrence frequency, duration, direction of change, average magnitude, and corresponding flight dynamic response of various residuals based on the written residual clusters; When a new rotor cycle segment enters the calculation, similar operating condition residual clusters are matched from the edge aerodynamic residual cluster library based on the current rotor operating condition state unit and the current aerodynamic residual decomposition results. Residual correction factors are generated based on the average amplitude, duration, and direction of change of various residuals in the residual cluster of similar working conditions. Based on the residual correction factor set, the lift correction parameters, drag correction parameters, torque correction parameters, induced velocity correction parameters, power consumption correction parameters, and rotor efficiency correction parameters are calculated to form the aerodynamic model correction parameters.

9. The method for calculating the aerodynamic performance of an unmanned helicopter rotor based on edge computing according to claim 8, characterized in that, When a new rotor cycle segment enters the calculation, matching similar operating condition residual clusters from the edge-end aerodynamic residual cluster library based on the current rotor operating condition state unit and the current aerodynamic residual decomposition result specifically includes: Generate the current operating condition index based on the current rotor operating condition state unit corresponding to the new rotor cycle segment; Generate the current residual proportion vector based on the current aerodynamic residual decomposition result corresponding to the new rotor cycle segment; Generate candidate operating condition residual clusters and average residual proportion vectors based on the edge end aerodynamic residual cluster library; Calculate the condition matching value based on the current condition index and the condition index of the candidate condition residual cluster; Calculate the residual proportion matching value based on the current residual proportion vector and the average residual proportion vector of the candidate working condition residual cluster; Similar operating condition residual clusters are determined based on the operating condition matching value and the residual proportion matching value. When there are multiple candidate operating condition residual clusters with the highest similarity, the candidate operating condition residual cluster with the most recent written operating condition residual record is determined as the similar operating condition residual cluster.

10. The method for calculating the aerodynamic performance of an unmanned helicopter rotor based on edge computing according to claim 1, characterized in that, The process of correcting the rotor's basic aerodynamic performance results based on aerodynamic model correction parameters and outputting the rotor aerodynamic performance calculation results specifically includes: The correction calculation input is generated based on the aerodynamic model correction parameters and the basic aerodynamic performance results of the rotor. The base lift, base drag, and base torque results are corrected based on the correction calculation input to obtain the corrected lift, base drag, and base torque results. Write the corrected lift, drag, and torque results into the correction calculation record corresponding to the current rotor cycle segment; Based on the corrected calculation input, the calculation quantities related to the rotor flow field, the basic power consumption results, and the basic rotor efficiency results are corrected to obtain the corrected induced velocity calculation quantities, the corrected power consumption results, and the corrected rotor efficiency results. Write the corrected induced velocity calculation, corrected power consumption result, and corrected rotor efficiency result into the corrected calculation record corresponding to the current rotor cycle segment; The aerodynamic stability results are calculated based on the corrected lift, drag, torque, power consumption, rotor efficiency, and aerodynamic residual decomposition results. The rotor aerodynamic performance calculation results are generated based on the corrected lift, drag, torque, power consumption, rotor efficiency, and aerodynamic stability results.

Images