A method and system for autonomous flight control of unmanned aerial vehicles

Abstract

This application relates to an autonomous flight control method and system for unmanned aerial vehicles (UAVs). The UAV autonomous flight control method includes: applying a small rotational speed disturbance of a preset amplitude to the drive motor during UAV flight, wherein the small rotational speed disturbance is configured to excite vibration in the UAV's airframe structure without changing the UAV's macroscopic flight attitude; collecting vibration response data from the inertial measurement unit (IMU) in response to the vibration; performing real-time spectrum analysis on the vibration response data to identify characteristic frequencies with abnormally increased vibration amplitudes, and determining these characteristic frequencies as the current inherent structural resonance frequencies of the UAV's airframe structure; and dynamically adjusting a vibration suppression strategy based on the current inherent structural resonance frequencies to reduce vibration noise in the IMU data and optimize the position estimation logic of multi-sensor data fusion. This invention effectively improves the stability and operational accuracy of autonomous UAV flight.

Description

Technical Field

[0001] This application relates to the field of unmanned aerial vehicle (UAV) control technology, and in particular to a method and system for autonomous flight control of UAVs. Background Technology

[0002] With the rapid development of industrial inspection technology, drones have gradually shifted from conventional large-scale macroscopic monitoring (such as external factory inspections and power line inspections) to precision operations in confined spaces (such as 3D modeling of high-value equipment and optical detection of micro-cracks). In conventional inspection scenarios, the attitude control loop parameters of drones (such as PID gain) are usually calibrated based on stable flight conditions, which can effectively counteract common environmental disturbances such as light winds and maintain a stable flight path. However, when facing precision operation tasks, drones need to frequently perform rapid attitude adjustments and complex trajectory tracking in confined spaces, which places extremely high demands on the dynamic response speed of the flight control system.

[0003] To meet the demands of these highly dynamic missions, existing technologies typically employ strategies that increase the proportional and derivative gains of the attitude control loop to enhance the system's real-time response and control accuracy. However, this aggressive control strategy triggers a chain of negative effects: high-frequency, large-amplitude attitude commands force drastic fluctuations in motor speed, leading to a significant broadening of the vibration spectrum. When the excitation frequency of the motor at a specific speed couples with the natural frequency of the UAV's arm or fuselage structure, it induces strong structural resonance. At this point, the resulting vibration amplitude often far exceeds the design suppression range of traditional damping pads, causing the physical vibration isolation mechanism to fail, and high-amplitude vibrations are directly transmitted to the flight control motherboard.

[0004] The inertial measurement unit (IMU), the core of attitude calculation, suffers severe interference in this environment. Strong resonance causes the raw data output by the gyroscope and accelerometer to be overwhelmed by a large amount of high-frequency noise, resulting in a sharp drop in the signal-to-noise ratio. Although existing multi-sensor fusion algorithms (such as Kalman filtering) attempt to incorporate data sources such as vision and barometers for correction, they struggle to effectively remove noise when faced with IMU input errors that far exceed design margins. This directly leads to significant drift and high-frequency jitter in the position and velocity estimates output by the position estimation logic; that is, when the UAV is actually stationary and hovering, the system misjudges it as exhibiting random motion.

[0005] For precision operations requiring centimeter-level accuracy in confined spaces, the aforementioned distortion in position estimation constitutes a fatal flaw. Navigation programs rely on accurate position estimation to determine waypoint arrival status and trigger subsequent actions. When the estimated position fluctuates repeatedly within and outside the distance threshold due to noise, the UAV cannot stably lock onto the target, and may even malfunction, leading to mission failure.

[0006] In summary, existing technologies face a technical dilemma that cannot be resolved through simple parameter adjustments: adjusting the control gain to improve dynamic response can induce structural resonance in the airframe, thereby contaminating core sensor data and ultimately causing multi-sensor fusion positioning to fail. Currently, there is a lack of a comprehensive control scheme that can simultaneously meet the requirements of high dynamic response and anti-structural resonance capabilities, making it impossible to ensure truly reliable centimeter-level precise hovering and slow movement of UAVs in complex industrial scenarios. Therefore, there is an urgent need to develop a new autonomous flight control method and system for UAVs to solve the vibration coupling and positioning drift problems caused by the aforementioned control strategies. Summary of the Invention

[0007] This application provides an autonomous flight control method and system for unmanned aerial vehicles (UAVs) to at least solve the problem that adjusting the control gain to improve dynamic response in the prior art can induce resonance in the fuselage structure, thereby contaminating the core sensing data and ultimately causing the failure of multi-sensor fusion positioning.

[0008] Firstly, this application provides an autonomous flight control method for an unmanned aerial vehicle (UAV), comprising the following steps: During the flight of the UAV, a small speed disturbance of a preset amplitude is applied to the drive motor, wherein the small speed disturbance is configured to excite the UAV body structure to vibrate without changing the macroscopic flight attitude of the UAV. The vibration response data of the inertial measurement unit in response to the vibration output is collected; Real-time spectrum analysis was performed on the vibration response data to identify the characteristic frequencies of abnormally increased vibration amplitude, and the characteristic frequencies were determined as the current inherent structural resonance frequencies of the UAV body structure. Based on the current inherent structural resonance frequency, the vibration suppression strategy is dynamically adjusted to reduce vibration noise in the inertial measurement unit data and optimize the position estimation logic of multi-sensor data fusion.

[0009] Optionally, the step of performing real-time spectrum analysis on the vibration response data to identify characteristic frequencies with abnormally increased vibration amplitude includes: Acquire local ambient temperature data for key structural components of the drone; Based on the preset temperature-resonance frequency mapping relationship, combined with the local ambient temperature data, the expected resonance frequency range and drift trend under the current operating conditions are predicted. Based on the expected resonant frequency range and drift trend, the parameters of the spectrum analysis are adaptively adjusted, wherein the parameters include the data acquisition window length, the target search frequency range, and the peak identification threshold; Under the adjusted parameters, the frequency point with the highest vibration amplitude is retrieved within the target search frequency range, and the characteristic frequency is confirmed by the peak stability criterion.

[0010] Optionally, the step of dynamically adjusting the vibration suppression strategy based on the current inherent structural resonant frequency includes: Obtain the mission attributes of the current flight mission, wherein the mission attributes include priority requirements for response speed, thrust output capability and energy efficiency; Configure the execution parameters of the vibration suppression strategy according to the task attributes: If the task priority is high response speed, then reduce the bandwidth of the notch filter and appropriately reduce the filtering depth, while maintaining a high attitude control loop gain. If the task priority is high positioning accuracy, the bandwidth of the notch filter is increased and the filtering depth is increased. At the same time, the weight of the inertial measurement unit data is reduced and the weight of the external sensor data is increased in the multi-sensor data fusion algorithm. If the task priority is low energy consumption, then the resonant frequency should be avoided by fine-tuning the motor speed, thereby reducing the computational energy consumption caused by filtering operations.

[0011] Optionally, when multiple characteristic frequencies with similar frequencies are identified, the dynamic adjustment vibration suppression strategy includes: A multi-notch filter bank is generated, wherein the multi-notch filter bank comprises multiple independent notch filters, and the center frequency of each notch filter corresponds to a characteristic frequency. The multi-notch filter bank is applied to the preprocessing stage of inertial measurement unit data; Synchronous adjustment of motor speed command: If multiple characteristic frequencies are distributed in a narrow speed range, the motor is controlled to quickly pass through the speed range; if multiple characteristic frequencies are distributed in a wide speed range, the motor operating speed is finely adjusted to the non-resonance zone, and the speeds of other motors are adjusted in coordination to maintain the overall thrust balance, thereby maintaining the macroscopic flight attitude stability of the UAV while avoiding resonant frequencies.

[0012] Optionally, the method further includes a closed-loop feedback optimization step: Real-time monitoring of external environmental parameters and internal state parameters of the UAV, wherein the external environmental parameters include airflow disturbance intensity, and the internal state parameters include load change and energy consumption rate; Based on the external environmental parameters, internal state parameters, and the real-time noise level of the inertial measurement unit, the effectiveness score of the current vibration suppression strategy is calculated. When the performance score is lower than a preset threshold, a dynamic parameter adjustment mechanism is triggered. If the intensity of airflow disturbance increases, reduce the depth of the notch filter and increase the gain of the attitude control loop to enhance the wind disturbance resistance. If the load change causes the characteristic frequency to shift, the real-time spectrum analysis step is re-executed to update the current inherent structural resonant frequency and the notch filter center frequency is updated synchronously. If the energy consumption rate exceeds the preset limit, the system switches to a low-power suppression mode while meeting the positioning accuracy constraints. The low-power suppression mode prioritizes a motor speed fine-tuning strategy and simplifies the complexity of the filtering algorithm.

[0013] Optionally, the step of calculating the effectiveness score of the current vibration suppression strategy includes: Time synchronization and calibration of monitoring data; Select a combination of performance indicators that match the current flight phase and mission type. These performance indicators include attitude stability variance, position estimate drift, and signal-to-noise ratio. Set the weighting coefficients and judgment thresholds for each performance indicator based on the current task type; The effect score is generated based on the comparison between the weighted performance indicators and the judgment threshold.

[0014] Optionally, the dynamic adjustment mechanism for the trigger parameters further includes: Identify the set of parameters to be adjusted and their interdependencies, and construct a parameter adjustment priority sequence; Based on the aforementioned priority sequence, the notch filter parameters, attitude control loop gain, and motor speed command are adjusted collaboratively. After the adjustment is implemented, the improvement of the vibration suppression effect is monitored in real time. If the target is not met, iterative optimization is carried out until the effect score meets the requirements.

[0015] Secondly, this application provides an autonomous flight control system for unmanned aerial vehicles (UAVs), the system comprising: The disturbance application module is used to apply a small speed disturbance of a preset amplitude to the drive motor during the flight of the UAV, wherein the small speed disturbance is configured to excite the UAV body structure to vibrate without changing the macroscopic flight attitude of the UAV. The data acquisition module is used to acquire vibration response data of the inertial measurement unit in response to the vibration output; The spectrum analysis module is used to perform real-time spectrum analysis on the vibration response data, identify the characteristic frequencies of abnormally increased vibration amplitude, and determine the characteristic frequencies as the current inherent structural resonance frequencies of the UAV body structure. The strategy adjustment module is used to dynamically adjust the vibration suppression strategy based on the current inherent structural resonance frequency in order to reduce vibration noise in the inertial measurement unit data and optimize the position estimation logic of multi-sensor data fusion.

[0016] Thirdly, this application provides a computer device, including a memory and a processor, wherein the memory stores a computer program, characterized in that the processor executes the computer program to implement the steps of the method provided in the first aspect.

[0017] Fourthly, this application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method provided in the first aspect.

[0018] Compared with related technologies, the autonomous flight control method and system for unmanned aerial vehicles provided in this application have at least the following technical advantages: By applying a small, preset-amplitude rotational speed disturbance to the drive motor during UAV flight, this disturbance excites vibration in the UAV's airframe structure without affecting its macroscopic flight attitude. Subsequently, vibration response data from the inertial measurement unit (IMU) in response to this vibration is collected, and real-time spectral analysis is performed to identify the characteristic frequencies with abnormally increased vibration amplitudes, which are then determined as the current inherent structural resonance frequencies of the UAV's airframe. Based on the identified current inherent structural resonance frequencies, vibration suppression strategies are dynamically adjusted to effectively reduce vibration noise in the IMU data and optimize the position estimation logic based on multi-sensor data fusion.

[0019] Through the above technical solution, this application enables UAVs to stably maintain their position within the distance threshold of the waypoint determined in the navigation program when performing high-precision tasks such as centimeter-level precise hovering and slow movement in confined spaces. This effectively solves the dilemma that the positioning accuracy in the prior art cannot meet the task requirements, and improves the stability and operational accuracy of UAV autonomous flight.

[0020] Details of one or more embodiments of this application are set forth in the following drawings and description to make other features, objects and advantages of this application more readily apparent. Attached Figure Description

[0021] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings: Figure 1 This is a flowchart illustrating an autonomous flight control method for unmanned aerial vehicles (UAVs) according to an exemplary embodiment.

[0022] Figure 2 This is a partial flowchart illustrating step S3 according to an exemplary embodiment.

[0023] Figure 3 This is a partial flowchart illustrating step S4 according to an exemplary embodiment.

[0024] Figure 4 This is a block diagram illustrating an autonomous flight control system for an unmanned aerial vehicle (UAV) according to an exemplary embodiment. Detailed Implementation

[0025] To make the objectives, technical solutions, and advantages of this application clearer, the application is described and illustrated below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application. All other embodiments obtained by those skilled in the art based on the embodiments provided in this application without inventive effort are within the scope of protection of this application.

[0026] Obviously, the accompanying drawings described below are merely some examples or embodiments of this application. Those skilled in the art can apply this application to other similar scenarios based on these drawings without any inventive effort. Furthermore, it is understood that although the efforts made in this development process may be complex and lengthy, for those skilled in the art related to the content disclosed in this application, any changes to design, manufacturing, or production based on the technical content disclosed in this application are merely conventional technical means and should not be construed as insufficient disclosure of the content of this application.

[0027] In this application, the reference to "embodiment" means that a specific feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places in the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment that is mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described in this application may be combined with other embodiments without conflict.

[0028] Unless otherwise defined, the technical or scientific terms used in this application shall have the ordinary meaning understood by one of ordinary skill in the art to which this application pertains. The terms “a,” “an,” “an,” “the,” and similar words used in this application do not indicate quantity limitation and may represent singular or plural. The terms “comprising,” “including,” “having,” and any variations thereof used in this application are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that includes a series of steps or modules (units) is not limited to the listed steps or units, but may also include steps or units not listed, or may include other steps or units inherent to such processes, methods, products, or apparatus.

[0029] Example 1

[0030] This invention provides a method for autonomous flight control of unmanned aerial vehicles (UAVs). Figure 1 This is a flowchart illustrating an autonomous flight control method for a drone according to an exemplary embodiment. Figure 1 As shown, the method includes the following steps: S1. During the flight of the UAV, a small speed disturbance of a preset amplitude is applied to the drive motor, wherein the small speed disturbance is configured to excite the UAV body structure to vibrate without changing the macroscopic flight attitude of the UAV. In this embodiment, one way to implement a small speed perturbation is to superimpose a low-amplitude sine or square wave signal into the motor control command. The frequency of this signal can be preset to cover the frequency range where the UAV may resonate, or it can be gradually excited using a frequency sweep method. Another approach is to send brief, small speed increment or decrement commands to one or more drive motors via the flight control system. For example, applying a very short-duration speed pulse with an amplitude only a few percent of the normal flight speed. Although this perturbation is small, it is sufficient to generate vibrations in the UAV's airframe structure that can be detected by the inertial measurement unit.

[0031] S2. Collect vibration response data of the inertial measurement unit in response to the vibration output; In this embodiment, the inertial measurement unit (IMU) is typically mounted on the UAV flight control motherboard. Its internal accelerometer and gyroscope continuously read raw data at a high sampling rate. This raw data includes not only the UAV's attitude and motion information but also high-frequency noise components caused by structural vibrations. For example, the accelerometer outputs high-frequency acceleration changes during vibrations, while the gyroscope outputs high-frequency angular velocity changes. This vibration response data is transmitted in real-time to the flight control system for further processing.

[0032] S3. Perform real-time spectrum analysis on the vibration response data, identify the characteristic frequency of abnormally increased vibration amplitude, and determine the characteristic frequency as the current inherent structural resonance frequency of the UAV body structure; In this embodiment, real-time spectrum analysis is implemented using algorithms such as Fast Fourier Transform. For example, the flight control system periodically captures a segment of vibration response data from the inertial measurement unit and performs a Fourier transform on this data to obtain its spectrum. In the spectrum, frequency points with vibration amplitudes significantly higher than surrounding frequency points can be identified by setting a fixed amplitude threshold and determined as characteristic frequencies. However, this fixed-parameter spectrum analysis method may not be able to adapt to resonant frequency drift caused by changes in ambient temperature, and may also misjudge due to simple peak identification, thus affecting the accuracy and robustness of resonant frequency identification.

[0033] S4. Based on the current inherent structural resonance frequency, dynamically adjust the vibration suppression strategy to reduce vibration noise in the inertial measurement unit data and optimize the position estimation logic of multi-sensor data fusion. In this embodiment, based on the above steps, once the current inherent structural resonance frequency is identified, a notch filter is configured with its center frequency precisely set at this resonance frequency to filter out vibration noise. The bandwidth and depth of the notch filter can be adjusted according to the magnitude of the resonance amplitude. In the position estimation logic of multi-sensor data fusion, when the vibration noise in the inertial measurement unit data is effectively reduced, its signal-to-noise ratio is improved, and the fusion algorithm can more accurately use the inertial measurement unit data for position estimation.

[0034] The technical solution of the above embodiments constructs a closed-loop vibration sensing, identification, and dynamic suppression system. By actively applying minute rotational speed disturbances during UAV flight and collecting vibration response data from the inertial measurement unit in real time for spectral analysis, the current inherent structural resonant frequency of the UAV's airframe can be dynamically identified. This active excitation and real-time identification mechanism enables the system to accurately capture real-time changes in the resonant frequency. Based on this real-time identification result, the vibration suppression strategy is dynamically adjusted, for example, by precisely adjusting the center frequency and bandwidth of the notch filter, or intelligently fine-tuning the motor speed to avoid the resonant range. Therefore, this application ensures that the UAV maintains high-precision positioning and stable autonomous flight in complex environments, especially in detailed surveying tasks requiring centimeter-level precise hovering and slow movement, greatly improving the UAV's mission execution capability and reliability.

[0035] In one possible design, Figure 2 This is a partial flowchart illustrating step S3 according to an exemplary embodiment. (Refer to the attached diagram.) Figure 2 In step S3, the step of performing real-time spectrum analysis on the vibration response data to identify the characteristic frequencies of abnormally increased vibration amplitude includes: S31. Obtain local ambient temperature data for key structural components of the UAV; In this embodiment, temperature sensors are deployed inside the drone's fuselage or near key structural components to monitor and acquire temperature information in these areas in real time. Because physical parameters such as the elastic modulus and density of a material directly affect the structure's natural vibration frequency, obtaining temperature information from these areas is crucial for understanding how material properties change with temperature.

[0036] S32. Based on the preset temperature-resonance frequency mapping relationship and combined with the local ambient temperature data, predict the expected resonance frequency range and drift trend under the current operating conditions; In this embodiment, during the UAV design and testing phase, the variation law of the inherent resonant frequency of the UAV's airframe structure under different temperatures is established through experiments or simulations. This mapping relationship can be a lookup table, empirical formula, or physical model. Once real-time local ambient temperature data is obtained, the system can accurately predict the possible range of the resonant frequency under the current temperature conditions, as well as its drift direction and amplitude relative to the standard temperature, based on this mapping relationship.

[0037] The following further explains the temperature-resonance frequency mapping relationship and prediction logic. The natural frequencies of the UAV's airframe structure mainly depend on the material's stiffness and mass. Since mass changes very little with temperature, the main variable is the material's stiffness (elastic modulus). According to the principles of materials mechanics, the natural frequency (… Stiffness is proportional to the square root of the material's stiffness. For most engineering materials, stiffness exhibits an approximately linear negative correlation with temperature (i.e., as temperature decreases, stiffness increases, and frequency increases).

[0038] Therefore, in this embodiment, the mapping relationship formula system uses the following linear mapping model for prediction: In the formula, For the current temperature at Current predicted resonant frequency (Hz) As a reference temperature The calibrated resonant frequency (Hz) at a temperature (usually 25℃). The temperature drift coefficient (Hz / ℃) is obtained through environmental stress screening tests before leaving the factory. This represents the local ambient temperature (°C) of the current critical component.

[0039] In one example, assuming the drone in this embodiment was calibrated at the factory, the following measurements were taken: At 25℃, the first-order modal resonance frequency of its arm It is 150.0 Hz.

[0040] Based on historical big data analysis, the temperature drift coefficient of the carbon fiber arm of this model It is approximately +0.2 Hz / ^\circ C$ (that is, for every 1°C decrease in temperature, the resonant frequency increases by 0.2 Hz).

[0041] When the drone flies into a low-temperature environment to perform a mission, step S31 obtains the local ambient temperature. At 5℃: Based on this calculation, the system predicts that the current resonant frequency will drift to 146.0 Hz. In step S33, the spectrum analysis module will no longer scan the entire 0-500Hz frequency band, but will adaptively adjust the target search frequency range to 142Hz-150Hz, and dynamically adjust the peak identification threshold to 3dB higher than the normal environment. This allows it to quickly and accurately lock the actual 146.2Hz resonant peak within 20ms, avoiding the waste of computing power and delay caused by full-band scanning.

[0042] S33. Based on the expected resonance frequency range and drift trend, adaptively adjust the parameters of the spectrum analysis, wherein the parameters include the data acquisition window length, the target search frequency range, and the peak identification threshold. In this embodiment, adjusting the length of the data acquisition window can balance frequency resolution and real-time requirements; the target search frequency range can be dynamically narrowed according to the predicted resonant frequency range to avoid invalid searches in irrelevant frequency regions; and the peak identification threshold can be adjusted according to the expected vibration intensity and environmental noise level to improve the sensitivity of resonant peak identification and reduce false alarms.

[0043] S34. Under the adjusted parameters, retrieve the frequency point with the highest vibration amplitude within the target search frequency range, and confirm the characteristic frequency through the peak stability criterion. In this embodiment, the peak stability criterion refers to continuously monitoring the identified high-amplitude frequency points over a period of time to determine whether their frequency and amplitude remain relatively stable, thereby eliminating false peaks caused by transient noise or non-structural vibrations. For example, it can be required that the frequency point exhibits significant vibration amplitude within multiple consecutive time windows, and that its frequency fluctuations are within a preset allowable range.

[0044] The technical solutions described above improve the accuracy and robustness of UAVs in identifying the inherent resonant frequencies of their structures under complex and variable environmental conditions. Specifically, by introducing local ambient temperature data and a temperature-resonant frequency mapping relationship, the UAV can perceive and predict potential drifts in the inherent resonant frequencies of its structure in real time. By adaptively adjusting spectrum analysis parameters such as the data acquisition window length, target search frequency range, and peak identification threshold, the system can locate the true resonant frequency with higher efficiency and accuracy, effectively reducing the computational resources and time required for spectrum analysis and improving the system's real-time response capability. Furthermore, by incorporating peak stability criteria, it is further ensured that the identified characteristic frequencies are stable and reliable structural resonances, rather than transient interference or noise.

[0045] In one example, suppose a drone is performing a high-altitude inspection mission, its flight altitude rising from the ground to several thousand meters, causing the ambient temperature to drop from 25°C to 5°C. Before the drone leaves the factory, experiments have determined that the main resonant frequency of its airframe structure at 25°C is 150Hz, and a temperature-resonant frequency mapping relationship exists, indicating that for every 1°C decrease in temperature, the resonant frequency shifts to a higher frequency direction by approximately 0.2Hz. When the drone ascends to an environment of 5°C, the onboard system first obtains the local ambient temperature as 5°C. Based on the preset mapping relationship, the system predicts that the expected resonant frequency range under the current operating condition may be around 150Hz + (25-5) * 0.2Hz = 154Hz, for example, between 153Hz and 155Hz.

[0046] Based on this prediction, the spectrum analysis module adaptively adjusts its parameters. Specifically, the data acquisition window length is optimized to quickly respond to frequency changes while ensuring sufficient frequency resolution; the target search frequency range is precisely set to 150Hz to 160Hz, rather than the broad 0Hz to 500Hz, thus significantly reducing computational load; and the peak identification threshold is dynamically adjusted based on the current ambient noise level and the expected resonance intensity. Under the adjusted parameters, the system quickly retrieves the frequency point with the highest vibration amplitude within the 150Hz to 160Hz range and, combined with peak stability criteria (such as requiring the amplitude of this frequency point to be higher than a preset threshold and the frequency fluctuation to be less than 0.1Hz for 500 consecutive milliseconds), accurately confirms the current inherent structural resonance frequency as 154.2Hz. This precisely identified frequency is then used to dynamically adjust vibration suppression strategies such as notch filters to ensure effective suppression of vibration noise under temperature changes.

[0047] In one possible design, Figure 3 This is a partial flowchart illustrating step S4 according to an exemplary embodiment. (Refer to the attached diagram.) Figure 3 In step S4, the step of dynamically adjusting the vibration suppression strategy based on the current inherent structural resonant frequency includes: S41. Obtain the mission attributes of the current flight mission, wherein the mission attributes include priority requirements for response speed, thrust output capability and energy consumption efficiency. In this embodiment, mission attributes refer to the performance preferences and constraints related to the flight mission currently being performed by the UAV. These attributes are preset by the mission planning system or dynamically updated according to real-time mission instructions. For example, in high-speed cruise or emergency response missions, response speed is a high priority; in precision mapping or hovering missions, positioning accuracy is a high priority; and in long-endurance reconnaissance or logistics delivery missions, energy efficiency is a more prominent priority.

[0048] S42. Configure the execution parameters of the vibration suppression strategy according to the task attributes: S421. If the task priority is high response speed, reduce the bandwidth of the notch filter and appropriately reduce the filtering depth, while maintaining a high attitude control loop gain. S422. If the task priority is high positioning accuracy, then expand the bandwidth of the notch filter and increase the filtering depth. At the same time, in the multi-sensor data fusion algorithm, reduce the weight of the inertial measurement unit data and increase the weight of the external sensor data. S423. If the task priority is low energy consumption, then the resonant frequency should be avoided by fine-tuning the motor speed to reduce the computational energy consumption caused by filtering operations.

[0049] In this embodiment, the execution parameters for configuring the vibration suppression strategy are customized based on the task attributes in step S41 to create a specific suppression scheme. When the task priority is high response speed, the bandwidth of the notch filter is reduced and the filtering depth is appropriately decreased to ensure that the UAV can respond quickly to control commands. This is because excessively wide bandwidth and excessively deep filtering depth may introduce additional latency, affecting the real-time response capability of the system. At the same time, to compensate for the control performance loss that filtering may cause, the attitude control loop gain is maintained at a high level to enhance the ability to quickly correct attitude changes.

[0050] When the task priority is high positioning accuracy, the bandwidth of the notch filter is expanded and the filtering depth is increased to minimize the impact of vibration noise on position estimation, thereby more thoroughly filtering out noise near the resonant frequency. Furthermore, in multi-sensor data fusion algorithms, the weight of inertial measurement unit (IMU) data is reduced due to its susceptibility to vibration noise, while the weight of data from external sensors (such as GPS and vision sensors) is increased to utilize their more stable positioning information and improve overall positioning accuracy.

[0051] When the task priority is low energy consumption, the system will preferentially use fine-tuning of motor speed to avoid resonant frequencies. By changing the frequency of the excitation source to avoid coinciding with the natural frequency of the airframe structure, vibration is fundamentally reduced, thus reducing the need for filtering operations. Compared to complex filtering algorithms, fine-tuning of motor speed typically has lower computational energy consumption, helping to extend the drone's endurance.

[0052] To more clearly illustrate the technical solution of this application, this embodiment provides the specific adjustment parameter range of the notch filter and configuration examples under different task modes.

[0053] First, the parameters of the notch filter are defined. In this embodiment, the system uses a second-order IIR notch filter, whose transfer function is: In the formula, the core adjustable parameters include: Center frequency ( ): , ( The sampling frequency is typically 1kHz-8kHz. The adjustment range is 50 Hz - 400 Hz. Quality factor (Q): determines bandwidth, calculated using the following formula: BW = In this embodiment, the adjustment range of Q is 2.0-50.0; Filter depth (D): in dB, adjustable from 0dB (no filtering) to -30dB (depth filtering).

[0054] Assuming the system identifies the current resonant frequency Based on the task attributes obtained in step S41, the specific parameter configurations are as follows: Task Scenario 1: High Response Speed ​​Mode (e.g., Emergency Tracking) Task requirement: Sacrifice some filtering effect in exchange for control bandwidth.

[0055] The configuration parameters are as follows: Center frequency Locked to 145 Hz Quality factor Q: set to 40 (extremely narrow bandwidth, BW (≈3.6 Hz, reducing attenuation of useful surrounding signals); Filter depth D: Set to -6 dB (shallow filtering, only removing spikes); Attitude loop gain: Maintain a high gain of 0.85.

[0056] Task Scenario 2: High Positioning Accuracy Mode (e.g., Power Line Inspection) Task requirement: Completely eliminate noise to ensure accurate location estimation.

[0057] The configuration parameters are as follows: Center frequency Locked to 145 Hz Quality factor Q: Set to 10 (for wider bandwidth, BW ≈14.5 Hz, covering frequency fluctuations); Filter depth D: Set to -24 dB (depth filtering, completely eliminating noise bands); External sensor weighting: The weighting of visual / RTK data is increased to 0.7.

[0058] Scenario 3: Low-energy mode (e.g., long-endurance monitoring) Task requirement: Reduce CPU computing power consumption.

[0059] The configuration parameters are as follows: Notch filter: Off (depth set to 0 dB); Motor speed fine-tuning: Fine-tune the average motor speed by +50 RPM to avoid the speed range corresponding to 145Hz; Algorithm complexity: Switch to first-order high-pass filtering instead of second-order notch filtering.

[0060] In the technical solution of the above embodiments, by introducing task attributes as the basis for dynamically adjusting the vibration suppression strategy, the vibration suppression strategy can be intelligently adjusted according to the specific needs of the current flight mission, thereby overcoming the problem of insufficient adaptability of traditional single or general suppression strategies in multi-task scenarios. This allows the UAV to better balance key performance indicators such as response speed, positioning accuracy, and energy efficiency when performing different tasks. For example, in missions requiring rapid maneuvering, it can ensure that the UAV has agile response capabilities; in missions requiring precise hovering or mapping, it can significantly improve positioning accuracy; and in missions requiring long-term flight, it can effectively reduce system energy consumption and extend endurance. Thus, this application significantly improves the overall performance and mission success rate of the UAV in complex and variable mission environments, achieving a high degree of synergy between vibration suppression and mission objectives.

[0061] In one example, suppose a drone is deployed to perform two different types of missions: one is a high-speed reconnaissance mission, which requires the drone to quickly traverse a designated area and transmit data in real time; the other is a precision agriculture spraying mission, which requires the drone to perform precise hovering and path tracking within a specific area.

[0062] For high-speed reconnaissance missions, the system will identify the mission attribute as "high response speed" priority. In this case, the execution parameters of the vibration suppression strategy will be configured as follows: the bandwidth of the notch filter is reduced, and the filtering depth is moderately decreased to reduce data processing latency and ensure that the control system can quickly respond to commands from the pilot or autonomous navigation system. Simultaneously, the attitude control loop gain is maintained at a high level to enhance the attitude stability of the UAV during high-speed flight and rapid maneuvers.

[0063] For precision agriculture spraying tasks, the system identifies the task attribute as "high positioning accuracy" priority. In this case, the execution parameters of the vibration suppression strategy will be adjusted: the bandwidth of the notch filter is expanded, and the filtering depth is increased to more thoroughly filter out vibration noise in the inertial measurement unit (IMU) data. Furthermore, in the multi-sensor data fusion algorithm, the weight of IMU data will be reduced, while the weight of external sensor data from high-precision GPS or RTK modules will be increased, thereby ensuring that the UAV achieves centimeter-level positioning accuracy during spraying and avoiding spraying deviations.

[0064] By adjusting strategies under the two different mission scenarios described above, the solution proposed in this application enables UAVs to effectively suppress the adverse effects of vibration while meeting the core requirements of their respective missions, thereby significantly improving the efficiency and quality of mission execution.

[0065] In one possible design, in step S4, when multiple characteristic frequencies with similar frequencies are identified, the dynamic adjustment vibration suppression strategy includes: S431. Generate a multi-notch filter bank, wherein the multi-notch filter bank includes multiple independent notch filters, and the center frequency of each notch filter corresponds to a characteristic frequency. In this embodiment, when the spectrum analysis module identifies two or more frequency points with abnormally increased vibration amplitudes, and these frequency points are relatively close to each other, the system generates a multi-notch filter bank. This filter bank consists of multiple independent notch filters, each precisely configured so that its center frequency corresponds to an identified characteristic frequency. For example, if frequencies f1 and f2 are identified, two notch filters will be generated, with f1 and f2 as their center frequencies, respectively.

[0066] S432. The multi-notch filter bank is applied to the preprocessing stage of the inertial measurement unit data; In this embodiment, the notch filter generated in step S431 is applied to the raw vibration response data output by the inertial measurement unit to effectively filter out vibration noise of specific frequencies before the data enters subsequent processing (such as attitude calculation and position estimation).

[0067] S433, Synchronous adjustment of motor speed command: If multiple characteristic frequencies are distributed in a narrow speed range, the motor is controlled to quickly pass through the speed range; if multiple characteristic frequencies are distributed in a wide speed range, the motor operating speed is finely adjusted to the non-resonance zone, and the speeds of other motors are coordinated to maintain the overall thrust balance, thereby maintaining the macroscopic flight attitude stability of the UAV while avoiding the resonance frequency. In this embodiment, if multiple characteristic frequencies are concentrated within a relatively narrow motor speed range, the control system will instruct the motor to quickly traverse this speed range to minimize the cumulative resonance effect and prevent the UAV from operating within this resonance range for an extended period. Conversely, if multiple characteristic frequencies are distributed within a wider speed range, making it difficult to avoid them through rapid traversal, the system will fine-tune the motor's operating speed to avoid all identified resonance frequency regions and enter a non-resonant operating area. During this process, to ensure the overall thrust balance and flight attitude stability of the UAV, the speeds of other motors will also be adjusted in a coordinated manner.

[0068] In the technical solution of the above embodiments, a dual suppression mechanism is formed by introducing a multi-notch filter bank and a refined motor speed synchronization adjustment mechanism. This mechanism performs filtering at the data level and source control at the physical level, thereby more comprehensively and accurately suppressing vibration noise caused by multiple resonant frequencies that may occur in the UAV under complex operating conditions. Ultimately, this improves the purity of the inertial measurement unit data, providing more reliable input for subsequent attitude calculation and position estimation, thus enhancing the overall flight control accuracy and stability of the UAV. Specifically, the multi-notch filter bank can precisely filter each identified characteristic frequency, avoiding over-filtering or under-filtering that may occur with traditional single broadband notch filters. Simultaneously, the motor speed synchronization adjustment strategy, whether for quickly traversing narrow resonance ranges or fine-tuning to avoid wide resonance ranges, aims to reduce or avoid resonance at its source, thereby reducing vibration noise.

[0069] In one example, suppose a drone is performing a high-precision mapping task. Its inertial measurement unit (IMU) data is identified by a real-time spectrum analysis module at two characteristic frequencies with abnormally high vibration amplitudes: 120 Hz and 125 Hz. These two frequencies are determined to be close. At this point, the system generates a multi-notch filter bank containing two independent notch filters, one with a center frequency set to 120 Hz and the other to 125 Hz. This filter bank is immediately applied to the IMU's raw output data stream to precisely filter out vibration noise at these two frequencies.

[0070] Simultaneously, the control system evaluates the motor speed ranges that generate these two characteristic frequencies. If the motor speed range corresponding to 120 Hz and 125 Hz is found to be very narrow, such as only around a specific speed point, the system will instruct all drive motors to quickly pass through this speed range to avoid lingering at the resonance point for an extended period. If the speed range corresponding to these two frequencies is wider, such as covering a large portion of the motor's normal operating range, the system will fine-tune the motor's current operating speed to avoid the resonance speeds corresponding to 120 Hz and 125 Hz, and coordinately adjust the speeds of other motors to ensure that the UAV maintains overall thrust balance while flying stably in the non-resonance zone. In this way, the UAV can effectively cope with multiple resonance challenges and ensure the accuracy of mapping data.

[0071] In one possible design, the method further includes step S5, closed-loop feedback optimization, which provides adaptive and self-optimizing capabilities for the autonomous flight control of the UAV. The steps are as follows: S51. Real-time monitoring of external environmental parameters and UAV internal state parameters, wherein the external environmental parameters include airflow disturbance intensity, and the internal state parameters include load change and energy consumption rate; In this embodiment, S52. Based on the external environmental parameters, internal state parameters, and the real-time noise level of the inertial measurement unit, calculate the effectiveness score of the current vibration suppression strategy; In this embodiment, external environmental parameters, such as the intensity of airflow disturbance, directly affect the vibration characteristics and attitude stability of the UAV body; internal state parameters, such as load changes, alter the overall mass distribution and structural stiffness of the UAV, thereby affecting the inherent structural resonant frequency; energy consumption rate reflects the current energy consumption status of the UAV and is an important indicator for evaluating the efficiency of the suppression strategy. The real-time noise level of the inertial measurement unit is a key indicator for directly measuring the vibration suppression effect, and its level directly reflects the effectiveness of the current strategy.

[0072] The effectiveness of the vibration suppression strategy will be further quantified below. This embodiment provides a weighted normalization-based scoring formula for explanation.

[0073] Among them, the effect score The calculation model is as follows: For this embodiment, three core indicators are selected for weighting: In the formula, These are the weighting coefficients for signal-to-noise ratio, attitude stability, and position estimation accuracy, respectively. ; The real-time signal-to-noise ratio (dB) of the current IMU data; The signal-to-noise ratio under standard non-resonance conditions (set to 60dB). Let be the variance of the current attitude angle (°). 2 ); Maximum allowable attitude variance (set to 0.5 (°)) 2 ); The cumulative drift (m) estimated for the current position; The maximum allowable drift threshold for triggering relocation (set to 0.5m); Assuming the drone is currently in "precision mapping" mission mode, the system sets the weights as follows: (Signal-to-noise ratio) (attitude), (Location).

[0074] The monitoring data is as follows: Real-time signal-to-noise ratio 48 dB Attitude stability variance : 0.1 (°) 2 Position estimation drift 0.05 m Substitute into the formula to calculate: 1. Signal-to-noise ratio score: 0.3 × (48 / 60) = 0.3 × 0.8 = 0.24 2. Posture score: 0.3 × (1 - 0.1 / 0.5) = 0.3 × 0.8 = 0.24 3. Position score: 0.4 × (1 - 0.05 / 0.5) = 0.4 × 0.9 = 0.36 The final result score is: The system's preset passing threshold is 0.70. Calculate the score. The system determines that the current vibration suppression strategy (i.e., the current notch filter parameters) is working well and there is no need to trigger the parameter adjustment mechanism in step S53.

[0075] S53. When the effect score is lower than the preset threshold, the parameter dynamic adjustment mechanism is triggered: In this embodiment, based on this real-time monitoring data, the system calculates an effectiveness score for the current vibration suppression strategy. This score is a comprehensive indicator used to quantify the performance of the current strategy in suppressing vibration, maintaining attitude stability, and optimizing energy consumption.

[0076] S531. If the intensity of airflow disturbance increases, reduce the depth of the notch filter and increase the gain of the attitude control loop to enhance the wind disturbance resistance. S532. If the load change causes the characteristic frequency to shift, the real-time spectrum analysis step is re-executed to update the current inherent structural resonance frequency and the notch filter center frequency is updated synchronously. S533. If the energy consumption rate exceeds the preset limit, then under the premise of meeting the positioning accuracy constraint, switch to low power consumption suppression mode. The low power consumption suppression mode preferentially adopts the motor speed fine-tuning strategy and simplifies the complexity of the filtering algorithm. In this embodiment, when the performance score is lower than a preset threshold, it indicates that the performance of the current strategy can no longer meet the requirements, and a dynamic parameter adjustment mechanism will be triggered. This mechanism adjusts the strategy parameters in a targeted manner based on specific monitoring data. For example, when the intensity of airflow disturbance increases, in order to enhance the UAV's resistance to wind disturbance, the system will reduce the depth of the notch filter to reduce the delay of the control signal and increase the gain of the attitude control loop to enhance the attitude response speed.

[0077] When load changes cause a shift in the characteristic frequency, the system restarts the real-time spectrum analysis process to accurately identify the new inherent structural resonance frequency and update the center frequency of the notch filter accordingly, ensuring that the filter can accurately suppress the new resonance point.

[0078] In addition, if the energy consumption rate exceeds the preset limit, in order to extend the battery life, the system will switch to a low power suppression mode while ensuring basic positioning accuracy. This mode prioritizes the use of motor speed fine-tuning strategy, which has less impact on energy consumption, and simplifies the complexity of the filtering algorithm to reduce the computational load.

[0079] In the technical solution of the above embodiments, by introducing a closed-loop feedback optimization mechanism, real-time and adaptive optimization of the vibration suppression strategy is achieved, significantly improving the flight stability, positioning accuracy, and energy efficiency of the UAV in complex and variable environments. Specifically, by monitoring the external environment and internal state in real time, and combining the noise level of the inertial measurement unit, the system can continuously evaluate the actual effect of the current vibration suppression strategy. When the effect is unsatisfactory, the system can intelligently trigger and execute targeted parameter adjustments according to specific operating conditions, such as adjusting the notch filter parameters, attitude control loop gain, or switching to a low-power mode. This adaptive adjustment capability enables the UAV to actively cope with the impact of airflow disturbances, load changes, and other factors on vibration characteristics, thereby avoiding performance degradation caused by environmental changes, ensuring the continuous effectiveness of the vibration suppression strategy, and ultimately enhancing the autonomous flight capability of the UAV.

[0080] In one example, suppose a drone is performing a long-distance inspection mission. Initially, the flight environment is relatively stable, the vibration suppression strategy works well, and the performance score is high. However, during flight, the drone suddenly encounters a strong lateral airflow disturbance, causing increased body vibration and a significant increase in the noise level output by the inertial measurement unit, resulting in a performance score below a preset threshold. At this point, the closed-loop feedback optimization mechanism is triggered. The system detects the increased intensity of the airflow disturbance and immediately reduces the depth of the notch filter to minimize delays in the attitude control signal. Simultaneously, it increases the gain of the attitude control loop, enhancing the drone's rapid response to external disturbances, thereby effectively suppressing vibrations caused by the airflow disturbance and quickly restoring attitude stability.

[0081] In another scenario, after completing an inspection, a drone needs to return carrying a small payload. After acquiring the payload, the drone's total mass and mass distribution change, which may cause a shift in the inherent resonant frequencies of its airframe structure. The system monitors the load change in real time and, combined with the real-time noise level of the inertial measurement unit, detects a decrease in performance score. At this point, the system re-executes the real-time spectrum analysis step to accurately identify the new inherent structural resonant frequencies caused by the load change. Once the new characteristic frequencies are determined, the system synchronously updates the center frequency of the notch filter, enabling it to accurately suppress the new resonant frequencies and thus avoid vibration problems caused by load changes.

[0082] For example, when a drone performs a long-duration reconnaissance mission, the battery level gradually decreases as flight time increases, and the energy consumption rate may exceed a preset limit. To extend mission duration, the system triggers a low-power suppression mode while ensuring that positioning accuracy meets minimum requirements. In this mode, the system prioritizes a motor speed fine-tuning strategy with minimal impact on energy consumption to avoid resonant frequencies and simplifies the complexity of some filtering algorithms to reduce the processor's computational load, thereby effectively reducing overall energy consumption and extending the drone's endurance.

[0083] In one possible design, step S52, the step of calculating the effectiveness score of the current vibration suppression strategy, includes: S521. Synchronize and calibrate the monitoring data in time; In this embodiment, data from different sensors (such as inertial measurement units, GPS, barometers, etc.) and the system's internal status monitoring module are aligned according to a unified time reference, and any potential sensor drift, bias, or delay is corrected. This ensures that all data used for performance scoring calculations are consistent in time, avoiding evaluation biases caused by data asynchrony or inaccuracy.

[0084] S522. Select a combination of performance indicators that match the current flight phase and mission type. The performance indicators include attitude stability variance, position estimate drift, and signal-to-noise ratio. In this embodiment, the most representative indicator of the vibration suppression strategy is dynamically selected based on the specific flight state of the UAV (e.g., takeoff, cruise, hovering, landing) and the nature of the mission being performed (e.g., reconnaissance, mapping, logistics, performance). For example, in mapping missions requiring high-precision positioning, position estimation drift is given greater attention; while during high-speed cruise, attitude stability variance may become a more critical evaluation indicator. Attitude stability variance measures the stability of the UAV's attitude during flight, reflecting the impact of vibration on attitude control; position estimation drift assesses the accuracy of position estimation after multi-sensor data fusion, directly reflecting the interference of vibration noise on navigation accuracy; and the signal-to-noise ratio directly quantifies the ratio of effective signal to vibration noise in the inertial measurement unit data, serving as a direct indicator of vibration suppression effectiveness.

[0085] S523. Set the weight coefficients and judgment thresholds for each performance indicator according to the current task type. In this embodiment, for tasks requiring high positioning accuracy, the weighting coefficient of the position estimation drift is significantly increased, and its judgment threshold is set more strictly; while for tasks requiring high response speed, the weighting coefficient of attitude stability variance may be higher. This allows the performance scoring to flexibly adapt to the performance requirements of different task scenarios, ensuring the relevance and effectiveness of the evaluation results.

[0086] S524. Based on the comparison results between the weighted performance indicators and the judgment threshold, the effect score is generated; In this embodiment, each weighted performance index is compared with its corresponding judgment threshold. For example, if an index exceeds the threshold, a corresponding score is deducted; if it is within the judgment threshold, a corresponding score is awarded. Finally, by summarizing the scores of all indices or using a comprehensive evaluation model, a quantitative performance score is generated. This score can intuitively reflect the overall performance of the current vibration suppression strategy.

[0087] In the technical solution of the above embodiments, by synchronizing and calibrating the monitoring data in time, the accuracy and consistency of the evaluation data are ensured, avoiding evaluation deviations caused by data quality issues. This allows the effect score to truly reflect the actual performance of the vibration suppression strategy under different flight stages and mission types. Specifically, by selecting a combination of performance indicators that matches the current flight stage and mission type, and setting the weight coefficients and judgment thresholds for each performance indicator according to the current mission type, the calculation of the effect score can fully consider the complexity and diversity of the actual operation of the UAV. For example, in missions with extremely high positioning accuracy requirements, the position estimation drift is given a higher weight and a stricter threshold, so that the effect score can more sensitively reflect the impact of vibration on positioning accuracy. Thus, an effect score is generated based on the comparison results of the weighted performance indicators and the judgment thresholds. This score can not only quantify the current performance of the vibration suppression strategy, but also adaptively adjust according to different flight scenarios and mission requirements, thereby providing a more accurate and reliable evaluation basis for subsequent closed-loop feedback optimization steps.

[0088] In one possible design, in step S53, the dynamic adjustment mechanism for the trigger parameters further includes: S531. Identify the set of parameters to be adjusted and their interdependencies, and construct a parameter adjustment priority sequence; In this embodiment, identifying the set of parameters to be adjusted refers to determining all relevant control parameters that need to be adjusted under the current operating conditions, such as the center frequency, bandwidth, and depth of the notch filter, the proportional gain, integral gain, and derivative gain of the attitude control loop, and the speed commands of each drive motor. Interdependence refers to the inherent correlation between these parameters in affecting the UAV's vibration suppression and flight performance. For example, increasing the bandwidth of the notch filter may reduce the response speed of the control system, while increasing the gain of the attitude control loop may enhance the response speed but may also amplify high-frequency noise. Constructing a parameter adjustment priority sequence means determining the order and importance of parameter adjustments based on the priority requirements of the current flight mission (such as high response speed, high positioning accuracy, or low energy consumption) and the sensitivity and impact range of the parameters. For example, when a rapid response to airflow disturbances is required, adjusting the attitude control loop gain may have a higher priority.

[0089] S532. Based on the priority sequence, coordinately adjust the notch filter parameters, attitude control loop gain, and motor speed command. In this embodiment, parameter adjustments are no longer made by modifying individual parameters in isolation, but rather by comprehensively considering the interaction of all relevant parameters for holistic optimization. For example, when it is necessary to reduce vibration noise, not only the depth and bandwidth of the notch filter may be adjusted, but the motor speed may also be fine-tuned to avoid resonant frequencies, and the attitude control loop gain may be moderately adjusted to maintain flight stability. This coordinated adjustment aims to ensure that the modifications to various parameters work together to achieve the best vibration suppression and flight control effects.

[0090] S533. After the adjustment is implemented, monitor the improvement of the vibration suppression effect in real time. If the standard is not met, iterative optimization is carried out until the effect score meets the requirements. In this embodiment, after completing one parameter adjustment, the system continuously monitors the real-time noise level, attitude stability variance, and position estimation drift of the inertial measurement unit, and recalculates the effectiveness score of the vibration suppression strategy. If the effectiveness score fails to reach the preset optimization target or remains below the safety threshold, the system restarts the parameter adjustment process, further fine-tuning and optimizing the parameters based on the latest monitoring data and effectiveness score until the vibration suppression effect reaches the expected standard. Through this iterative optimization mechanism, the system is ensured to continuously adapt to environmental changes and gradually converge to the optimal control state.

[0091] In the technical solutions of the above embodiments, by identifying the set of parameters to be adjusted and their interdependencies, and constructing a parameter adjustment priority sequence, a comprehensive understanding of the control dimensions that need to be optimized can be achieved, avoiding negative impacts or conflicts caused by isolated adjustments, and improving the adaptability and robustness of the UAV autonomous flight control method in complex environments. Specifically, by coordinating the adjustment of notch filter parameters, attitude control loop gain, and motor speed commands according to the priority sequence, this application can achieve refined and holistic control of the UAV vibration suppression strategy, avoiding the problem of overall performance degradation caused by local optimization due to single parameter adjustments. In addition, by executing a real-time monitoring and iterative optimization mechanism after adjustment, this application can form a closed-loop adaptive adjustment process, ensuring that the vibration suppression strategy can continuously converge to the optimal state in a dynamically changing flight environment, thereby effectively solving the problems of insufficient adjustment or adjustment conflicts that may exist in the basic solution.

[0092] In summary, the UAV autonomous flight control method proposed in this application applies a small, preset-amplitude speed disturbance to the drive motor during UAV flight. This disturbance excites the UAV's structure to vibrate without affecting its macroscopic flight attitude. Subsequently, vibration response data from the inertial measurement unit (IMU) in response to this vibration is collected, and real-time spectral analysis is performed on the data to identify the characteristic frequencies with abnormally increased vibration amplitudes, which are then determined as the current inherent structural resonance frequencies of the UAV's structure. Based on the identified current inherent structural resonance frequencies, the vibration suppression strategy is dynamically adjusted to effectively reduce vibration noise in the IMU data and optimize the position estimation logic of multi-sensor data fusion.

[0093] Example 2

[0094] Embodiment 2 of the present invention provides an autonomous flight control system for unmanned aerial vehicles. Figure 4 This is a block diagram illustrating an autonomous flight control system for a drone according to an exemplary embodiment. (Refer to the attached diagram.) Figure 4 The system includes: The disturbance application module 01 is used to apply a small speed disturbance of a preset amplitude to the drive motor during the flight of the UAV, wherein the small speed disturbance is configured to excite the UAV body structure to vibrate without changing the macroscopic flight attitude of the UAV. Data acquisition module 02 is used to acquire vibration response data of the inertial measurement unit in response to the vibration output; Spectrum analysis module 03 is used to perform real-time spectrum analysis on the vibration response data, identify the characteristic frequency of abnormally increased vibration amplitude, and determine the characteristic frequency as the current inherent structural resonance frequency of the UAV body structure. The strategy adjustment module 04 is used to dynamically adjust the vibration suppression strategy based on the current inherent structural resonance frequency in order to reduce vibration noise in the inertial measurement unit data and optimize the position estimation logic of multi-sensor data fusion.

[0095] In summary, the UAV autonomous flight control method and system provided in this invention apply a small, preset-amplitude rotational speed disturbance to the drive motor during UAV flight. This disturbance excites the UAV's structure to vibrate without affecting its macroscopic flight attitude. Subsequently, vibration response data from the inertial measurement unit (IMU) in response to this vibration is collected, and real-time spectral analysis is performed on the data to identify the characteristic frequency with abnormally increased vibration amplitude, which is then determined as the current inherent structural resonance frequency of the UAV's structure. Based on the identified current inherent structural resonance frequency, the vibration suppression strategy is dynamically adjusted to effectively reduce vibration noise in the IMU data and optimize the position estimation logic of multi-sensor data fusion. Through the above technical solution, this application enables the UAV to stably maintain its position within the distance threshold for reaching the waypoint as determined by the navigation program when performing high-precision tasks such as centimeter-level precise hovering and slow movement in confined spaces. This effectively solves the dilemma in the prior art where the positioning accuracy cannot meet the task requirements, and improves the stability and operational accuracy of the UAV's autonomous flight.

[0096] Example 3

[0097] Embodiment 3 of the present invention provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the method described in Embodiment 1 above.

[0098] Furthermore, in conjunction with the UAV autonomous flight control method in the above embodiments, this application embodiment can provide a computer-readable storage medium for implementation. This computer-readable storage medium stores computer program instructions; when executed by a processor, these computer program instructions implement the steps of the method in Embodiment 1 above.

[0099] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0100] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this patent application should be determined by the appended claims.

Claims

1. A method for autonomous flight control of an unmanned aerial vehicle (UAV), characterized in that, Includes the following steps: During the flight of the UAV, a small speed disturbance of a preset amplitude is applied to the drive motor, wherein the small speed disturbance is configured to excite the UAV body structure to vibrate without changing the macroscopic flight attitude of the UAV. The vibration response data of the inertial measurement unit in response to the vibration output is collected; Real-time spectrum analysis was performed on the vibration response data to identify the characteristic frequencies of abnormally increased vibration amplitude, and the characteristic frequencies were determined as the current inherent structural resonance frequencies of the UAV body structure. Based on the current inherent structural resonance frequency, the vibration suppression strategy is dynamically adjusted to reduce vibration noise in the inertial measurement unit data and optimize the position estimation logic of multi-sensor data fusion.

2. The method according to claim 1, characterized in that, The step of performing real-time spectrum analysis on the vibration response data to identify the characteristic frequencies of abnormally increased vibration amplitude includes: Acquire local ambient temperature data for key structural components of the drone; Based on the preset temperature-resonance frequency mapping relationship, combined with the local ambient temperature data, the expected resonance frequency range and drift trend under the current operating conditions are predicted. Based on the expected resonant frequency range and drift trend, the parameters of the spectrum analysis are adaptively adjusted, wherein the parameters include the data acquisition window length, the target search frequency range, and the peak identification threshold; Under the adjusted parameters, the frequency point with the highest vibration amplitude is retrieved within the target search frequency range, and the characteristic frequency is confirmed by the peak stability criterion.

3. The method according to claim 1, characterized in that, The step of dynamically adjusting the vibration suppression strategy based on the current inherent structural resonant frequency includes: Obtain the mission attributes of the current flight mission, wherein the mission attributes include priority requirements for response speed, thrust output capability and energy efficiency; Configure the execution parameters of the vibration suppression strategy according to the task attributes: If the task priority is high response speed, then reduce the bandwidth of the notch filter and appropriately reduce the filtering depth, while maintaining a high attitude control loop gain. If the task priority is high positioning accuracy, the bandwidth of the notch filter is increased and the filtering depth is increased. At the same time, the weight of the inertial measurement unit data is reduced and the weight of the external sensor data is increased in the multi-sensor data fusion algorithm. If the task priority is low energy consumption, then the resonant frequency should be avoided by fine-tuning the motor speed, thereby reducing the computational energy consumption caused by filtering operations.

4. The autonomous flight control method for unmanned aerial vehicles according to claim 1, characterized in that, When multiple characteristic frequencies with similar frequencies are identified, the dynamic vibration suppression strategy includes: A multi-notch filter bank is generated, wherein the multi-notch filter bank comprises multiple independent notch filters, and the center frequency of each notch filter corresponds to a characteristic frequency. The multi-notch filter bank is applied to the preprocessing stage of inertial measurement unit data; Synchronous adjustment of motor speed command: If multiple characteristic frequencies are distributed in a narrow speed range, the motor is controlled to quickly pass through the speed range; if multiple characteristic frequencies are distributed in a wide speed range, the motor operating speed is finely adjusted to the non-resonance zone, and the speeds of other motors are adjusted in coordination to maintain the overall thrust balance, thereby maintaining the macroscopic flight attitude stability of the UAV while avoiding resonant frequencies.

5. The method according to any one of claims 1 to 4, characterized in that, It also includes closed-loop feedback optimization steps: Real-time monitoring of external environmental parameters and internal state parameters of the UAV, wherein the external environmental parameters include airflow disturbance intensity, and the internal state parameters include load change and energy consumption rate; Based on the external environmental parameters, internal state parameters, and the real-time noise level of the inertial measurement unit, the effectiveness score of the current vibration suppression strategy is calculated. When the performance score is lower than a preset threshold, a dynamic parameter adjustment mechanism is triggered. If the intensity of airflow disturbance increases, reduce the depth of the notch filter and increase the gain of the attitude control loop to enhance the wind disturbance resistance. If the load change causes the characteristic frequency to shift, the real-time spectrum analysis step is re-executed to update the current inherent structural resonant frequency and the notch filter center frequency is updated synchronously. If the energy consumption rate exceeds the preset limit, the system switches to a low-power suppression mode while meeting the positioning accuracy constraints. The low-power suppression mode prioritizes a motor speed fine-tuning strategy and simplifies the complexity of the filtering algorithm.

6. The method according to claim 5, characterized in that, The steps for calculating the effectiveness score of the current vibration suppression strategy include: Time synchronization and calibration of monitoring data; Select a combination of performance indicators that match the current flight phase and mission type. These performance indicators include attitude stability variance, position estimate drift, and signal-to-noise ratio. Set the weighting coefficients and judgment thresholds for each performance indicator based on the current task type; The effect score is generated based on the comparison between the weighted performance indicators and the judgment threshold.

7. The method according to claim 5, characterized in that, The dynamic adjustment mechanism for trigger parameters also includes: Identify the set of parameters to be adjusted and their interdependencies, and construct a parameter adjustment priority sequence; Based on the aforementioned priority sequence, the notch filter parameters, attitude control loop gain, and motor speed command are adjusted collaboratively. After the adjustment is implemented, the improvement of the vibration suppression effect is monitored in real time. If the target is not met, iterative optimization is carried out until the effect score meets the requirements.

8. An autonomous flight control system for unmanned aerial vehicles (UAVs), characterized in that, The system includes: The disturbance application module is used to apply a small speed disturbance of a preset amplitude to the drive motor during the flight of the UAV, wherein the small speed disturbance is configured to excite the UAV body structure to vibrate without changing the macroscopic flight attitude of the UAV. The data acquisition module is used to acquire vibration response data of the inertial measurement unit in response to the vibration output; The spectrum analysis module is used to perform real-time spectrum analysis on the vibration response data, identify the characteristic frequencies of abnormally increased vibration amplitude, and determine the characteristic frequencies as the current inherent structural resonance frequencies of the UAV body structure. The strategy adjustment module is used to dynamically adjust the vibration suppression strategy based on the current inherent structural resonance frequency in order to reduce vibration noise in the inertial measurement unit data and optimize the position estimation logic of multi-sensor data fusion.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1-8.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1-8.

Images