Unmanned aerial vehicle flight attitude self-adaptive imbalance control method, device and equipment

Abstract

The invention provides an unmanned aerial vehicle flight attitude self-adaptive imbalance control method, device and equipment, and aims to accurately obtain an ideal control track and recognize a beneficial imbalance mode to generate a controllable imbalance strategy library by constructing a reverse track analysis and controllable imbalance strategy generation mechanism. A dynamic balance reference establishment and track reconstruction technology is adopted to construct an unbalance-recovery control track, and dynamic response characteristics are accurately obtained; an unbalance recovery space-time manifold construction and topology analysis technology is introduced, a core unbalance control feature set is extracted through space-time folding transformation, and unbalance control key parameters are determined; establishing an unbalance-recovery mapping rule and self-adaptive parameter fusion mechanism, balancing the control effect and the safety requirement, and generating a disturbance control parameter and a self-adaptive mapping parameter; a real-time control output regulation mechanism is constructed to generate a flight control signal in combination with reverse unbalance combined control and a multi-axis cooperative control mechanism; finally, a self-adaptive unbalance control system is formed, and precise unbalance control over the attitude of the unmanned aerial vehicle is achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of unmanned aerial vehicle (UAV) control, and particularly to a method, device and equipment for adaptively controlling the flight attitude of a UAV in case of imbalance. Background Art

[0002] As an important application of modern aviation technology, UAVs are increasingly widely used in various fields. However, traditional UAV attitude control technologies mainly adopt an error correction control idea, regarding the attitude deviation and imbalance phenomena occurring during flight as interference factors to be eliminated, and pulling the attitude back to the predetermined trajectory through feedback control. Although this control method can maintain flight stability, it cannot fully utilize the maneuvering potential contained in the imbalance phenomenon, restricting the maneuvering performance and environmental adaptability of UAVs.

[0003] The existing technologies mainly have the following problems: lack of in-depth analysis and effective utilization of the imbalance phenomenon, unable to identify beneficial imbalance patterns and convert them into control resources; lack of an inverse design method from the ideal trajectory to the actual control, making it difficult to establish an accurate mapping relationship between the imbalance phenomenon and the control effect; lack of a dynamic balance benchmark and an adaptive control mechanism, unable to flexibly adjust the imbalance control strategy according to changes in the flight state; lack of an imbalance control system for multi-axis coordination, making it difficult to achieve unified coordination and optimal cooperation of the imbalance control for each axis. Therefore, there is an urgent need to invent a new UAV attitude control method that can actively identify and utilize the imbalance phenomenon, establish an adaptive imbalance control strategy, and solve the technical problems such as insufficient utilization of imbalance, low control efficiency, and poor adaptability in the existing technologies. Summary of the Invention

[0004] The present invention provides a method, device and equipment for adaptively controlling the flight attitude of a UAV in case of imbalance, aiming to achieve the active utilization and intelligent control of the imbalance phenomenon during the flight of the UAV, integrating key technologies such as inverse trajectory analysis, imbalance strategy generation, dynamic balance benchmark, construction of an imbalance recovery space-time manifold, topological feature extraction, mapping rule design, and adaptive control, performing full-process intelligent processing on the flight attitude imbalance phenomenon, and realizing imbalance identification, intelligent generation of control strategies, and multi-axis coordinated linkage through an innovative analysis method in the space-time dimension, forming a UAV imbalance control system with prediction ability, rapid response, path optimization, effective coordination, and adaptive adjustment.

[0005] In the first aspect of the present invention, a method for adaptively controlling the flight attitude of a UAV in case of imbalance is proposed, including the following steps: Obtain the current attitude state and the target attitude state of the UAV, and perform inverse trajectory analysis on the current attitude state and the target attitude state to generate an ideal control trajectory sequence; Obtain the imbalance points of the ideal control trajectory sequence, determine the disturbance injection positions according to the imbalance points, and generate a controllable imbalance strategy library according to the disturbance injection positions; Determine the imbalance control interval according to the ideal control trajectory sequence and the controllable imbalance strategy library, generate a disturbance injection sequence according to the imbalance control interval, and establish a dynamic balance benchmark according to the disturbance injection sequence; Perform trajectory reconstruction based on the dynamic balance benchmark to obtain an imbalance-recovery control trajectory, and obtain the corresponding dynamic response feature set of the imbalance-recovery control trajectory; Obtain the balance recovery index corresponding to the dynamic response feature set, construct an imbalance recovery spatio-temporal manifold based on the balance recovery index, and perform feature mapping on the imbalance recovery spatio-temporal manifold to obtain a core imbalance control feature set; Establish an imbalance-recovery mapping rule according to the core imbalance control feature set, generate disturbance control parameters according to the imbalance-recovery mapping rule, and establish an adaptive mapping parameter based on the disturbance control parameters; Generate an inverse imbalance joint control according to the adaptive mapping parameter, output a flight control signal according to the inverse imbalance joint control, and realize the adaptive imbalance control of the UAV flight attitude.

[0006] The second aspect of the present invention proposes a UAV flight attitude adaptive imbalance control device, including: A trajectory analysis module, configured to obtain the current attitude state and the target attitude state of the UAV, and perform reverse trajectory analysis on the current attitude state and the target attitude state to generate an ideal control trajectory sequence; An imbalance acquisition module, configured to obtain the imbalance points of the ideal control trajectory sequence, determine the disturbance injection positions according to the imbalance points, and generate a controllable imbalance strategy library according to the disturbance injection positions; A benchmark establishment module, configured to determine the imbalance control interval according to the ideal control trajectory sequence and the controllable imbalance strategy library, generate a disturbance injection sequence according to the imbalance control interval, and establish a dynamic balance benchmark according to the disturbance injection sequence; A trajectory reconstruction module, configured to perform trajectory reconstruction based on the dynamic balance benchmark to obtain an imbalance-recovery control trajectory, and obtain the corresponding dynamic response feature set of the imbalance-recovery control trajectory; A feature extraction module, configured to obtain the balance recovery index corresponding to the dynamic response feature set, construct an imbalance recovery spatio-temporal manifold based on the balance recovery index, and perform feature mapping on the imbalance recovery spatio-temporal manifold to obtain a core imbalance control feature set; A mapping generation module, configured to establish an imbalance-recovery mapping rule according to the core imbalance control feature set, generate disturbance control parameters according to the imbalance-recovery mapping rule, and establish an adaptive mapping parameter based on the disturbance control parameters; A control execution module is used to generate a reverse imbalance joint control according to the adaptive mapping parameter, and output a flight control signal according to the reverse imbalance joint control, so as to realize the adaptive imbalance control of the UAV flight attitude.

[0007] A third aspect of the present invention provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, the steps of an UAV flight attitude adaptive imbalance control method disclosed in the first aspect are implemented.

[0008] The beneficial effects of the present invention are reflected in the following aspects: 1. By using the reverse trajectory analysis technology to reverse from the target attitude state to the current attitude state, the imbalance points in the trajectory can be accurately identified and their controllability can be evaluated, and a strategy library containing multiple imbalance control modes is established, converting the traditional passive error-correcting imbalance phenomenon into an actively utilized control resource, improving the maneuverability and control efficiency of the UAV. 2. Mapping the imbalance recovery process to a four-dimensional spacetime manifold, identifying the geometric characteristics of the manifold through topological analysis, using spacetime folding transformation to compress the long recovery path into an efficient control trajectory, and extracting the folding node set to form the core imbalance control characteristics, improving the control efficiency and response speed, and solving the problem of long recovery paths in traditional methods. 3. By establishing an imbalance-recovery mapping rule and an adaptive parameter fusion technology, the imbalance control strategy can be automatically adjusted according to the flight state and environmental changes, realizing the coordinated imbalance control and unified output of the pitch, roll, and yaw axes, solving the problems of poor adaptability and difficult multi-axis coordination in traditional control, and forming a complete adaptive imbalance control system.

[0009] It should be understood that the above general description and the following detailed description are only exemplary and explanatory, and cannot limit the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] The drawings here show specific examples of the technical solutions of the present invention and form a part of the description of the specific implementation manners, and are used to explain the technical solutions, principles, and effects of the present invention.

[0011] Unless otherwise specified or defined, in different drawings, the same reference numerals represent the same or similar technical features, and for the same or similar technical features, different reference numerals may also be used to represent them.

[0012] Figure 1 It is a schematic flowchart of an UAV flight attitude adaptive imbalance control method of the present invention.

[0013] Figure 2 It is a structural block diagram of an UAV flight attitude adaptive imbalance control device of the present invention.

[0014] Figure 3 It is a schematic structural diagram of a computer device according to the present invention. Specific embodiments

[0015] In the following description, specific details such as specific system architectures, technologies, etc. are presented for the purpose of illustration rather than limitation, so as to thoroughly understand the embodiments of the present application. However, those skilled in the art should clearly understand that the present application can also be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to avoid unnecessary details from interfering with the description of the present application.

[0016] It should be understood that when used in the specification and appended claims of the present application, the term "comprising" indicates the presence of the described features, wholes, steps, operations, elements, and / or components, but does not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components, and / or their combinations.

[0017] It should also be understood that the term "and / or" used in the specification and appended claims of the present application refers to any combination and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0018] As used in the specification and appended claims of the present application, the term "if" can be interpreted as "when", "once", "in response to determining", or "in response to detecting" according to the context. Similarly, the phrase "if determined" or "if [the described condition or event] is detected" can be interpreted as meaning "once determined", "in response to determining", "once [the described condition or event] is detected", or "in response to detecting [the described condition or event]" according to the context.

[0019] In addition, in the description of the specification and appended claims of the present application, the terms "first", "second", "third", etc. are only used for distinguishing descriptions and cannot be understood as indicating or implying relative importance.

[0020] The reference to "one embodiment" or "some embodiments" etc. described in the specification of the present application means that a specific feature, structure, or characteristic described in conjunction with the embodiment is included in one or more embodiments of the present application. Thus, statements such as "in one embodiment", "in some embodiments", "in other some embodiments", "in still other embodiments", etc. that appear in different places in this specification do not necessarily all refer to the same embodiment, but mean "one or more but not all embodiments", unless otherwise specifically emphasized in another way. The terms "comprising", "including", "having", and their variants all mean "including but not limited to", unless otherwise specifically emphasized in another way.

[0021] The technical solutions of the embodiments of the present application will be introduced below.

[0022] As Figure 1 shown, the embodiment of the present invention provides a method for adaptive imbalance control of the flight attitude of an unmanned aerial vehicle, including the following steps S110-S170: Step S110, obtain the current attitude state and the target attitude state of the unmanned aerial vehicle, and perform reverse trajectory analysis on the current attitude state and the target attitude state to generate an ideal control trajectory sequence.

[0023] Specifically, first obtain the current attitude state of the unmanned aerial vehicle, and collect attitude information in real time through the inertial measurement unit (IMU) of the flight control system. The IMU integrates a three-axis gyroscope, a three-axis accelerometer, and a three-axis magnetometer, and the sampling frequency is set to 1000 Hz to ensure high-precision update of attitude data. The gyroscope measures the angular velocity of the unmanned aerial vehicle around the X, Y, and Z axes, with a range of ±2000° / s and an accuracy of 0.01° / s; the accelerometer detects the three-axis linear acceleration, with a range of ±16g and a resolution of 0.001g; the magnetometer measures the geomagnetic field intensity to obtain the heading angle information, with an accuracy of ±0.5°. The data of the three types of sensors are fused through the extended Kalman filter algorithm to calculate the real-time attitude angles of the unmanned aerial vehicle, including the pitch angle θ, the roll angle φ, and the yaw angle ψ. At the same time, collect the angular velocity and angular acceleration information of each axis. The GPS module provides three-dimensional position coordinates with an accuracy of up to centimeter level, the optical flow sensor assists in measuring the horizontal velocity vector, and the ultrasonic altimeter measures the height from the ground. Through multi-sensor data fusion, the current attitude state is obtained. During a certain flight, the current attitude state shows a pitch angle of 15.2°, a roll angle of -3.8°, and a yaw angle of 127.5°, and the corresponding angular velocities are 0.8° / s, -0.2° / s, and 1.5° / s respectively. Then obtain the target attitude state of the unmanned aerial vehicle, and obtain the expected attitude parameters through the flight mission planning module and control instruction parsing. The mission planning module calculates the expected attitude angles and change rates at each moment according to the preset flight waypoints, maneuver actions, and mission requirements. For cruise flight, the target attitude state usually requires the attitude angles to be stable, the pitch angle to be maintained at the horizontal flight angle, the roll angle to be close to zero, and the yaw angle to be set according to the heading requirement. For maneuvering flight, the target attitude state needs to precisely define the target values and time constraints of attitude conversion. When performing a specific maneuvering action, such as a quick turn maneuver, the target attitude state is set to keep the pitch angle at 15°±2° to maintain altitude, the roll angle is adjusted to 25° to generate a turning moment, and the yaw angle changes from the current 127.5° to the target 217.5° to complete a 90° turn, and the entire attitude conversion process is required to be completed within 3.5 seconds. Through real-time sensor data collection and mission instruction parsing, the current attitude state and the target attitude state are obtained.

[0024] In some embodiments, the reverse trajectory analysis of the current attitude state and the target attitude state to generate an ideal control trajectory sequence includes: performing a reverse path planning from the target attitude state to the current attitude state; determining intermediate control nodes according to the reverse path planning, and performing a timing mark on the intermediate control nodes; obtaining the attitude change gradient of the intermediate control nodes; and generating an ideal control trajectory sequence according to the timing mark and the attitude change gradient.

[0025] First, perform a reverse path planning from the target attitude state to the current attitude state. Use a reverse search algorithm, with the obtained target attitude state as the starting point and the obtained current attitude state as the ending point, to construct a reverse kinematics solution process. Taking a turning maneuver as an example, perform a reverse path search from the target attitude state (pitch angle 15°, roll angle 25°, yaw angle 217.5°) to the current attitude state (pitch angle 15.2°, roll angle -3.8°, yaw angle 127.5°). Compared with traditional forward planning, the advantage of reverse planning is that it can start from the desired final state to ensure the accuracy and reachability of the trajectory end point. The reverse path planning uses an optimization search method in the three-dimensional attitude space to find a feasible path from the target state to the current state under the premise of satisfying the UAV dynamics constraints and actuator limitations. Considering the maximum angular velocity limit of the UAV (pitch angular velocity ≤ 30° / s, roll angular velocity ≤ 45° / s, yaw angular velocity ≤ 25° / s) and angular acceleration constraints, ensure the physical realizability of the planned path. Traverse the attitude space through the reverse search algorithm, calculate the time cost and energy cost of the angular change in each axis, and select the path with the minimum overall cost as the optimal solution. In addition, the reverse path planning also needs to consider the attitude coupling effect. The change in the roll angle will affect the yaw control efficiency, and the yaw action will generate an additional roll moment. Therefore, the planning process needs to comprehensively analyze the mutual influence relationship of the three-axis attitudes. Through reverse search and constraint optimization, generate a reverse path planning from the target attitude state to the current attitude state.

[0026] Next, determine the intermediate control nodes according to the reverse path planning and perform timing marking on the intermediate control nodes. Based on the three-dimensional attitude change trajectory of the reverse path planning, use the equal time interval segmentation method to determine the intermediate control nodes. Divide the total time of 3.5 seconds in the reverse path planning into 7 time periods of 0.5 seconds each, and extract the corresponding attitude states at each time node as the intermediate control nodes. For example, the first intermediate control node corresponds to the moment of t = 0.5s, with the attitude angles of pitch angle 15.0°, roll angle 21.2°, and yaw angle 205.1°; the second node corresponds to the moment of t = 1.0s, with the attitude angles of pitch angle 15.0°, roll angle 17.8°, and yaw angle 192.8°; and so on until the seventh node. In addition to the attitude angle information, each intermediate control node also records the corresponding angular velocity and angular acceleration values. Subsequently, perform timing marking on the determined intermediate control nodes, and use the decreasing timestamp marking method to mark them sequentially from the target state to the current state direction. The target attitude state is marked as T0 = 3.5s, the first intermediate control node is marked as T1 = 3.0s, the second node is marked as T2 = 2.5s, and so on, and the current attitude state is marked as T7 = 0s. The timing marking not only records the time information but also marks the priority and criticality of each node. Among them, the nodes with larger attitude change amplitudes are marked as key control points and require precise control; the nodes with gentle attitude changes are marked as transition points and allow a certain control error. Through equal time segmentation and decreasing marking, obtain the intermediate control nodes with timing marking.

[0027] Then, obtain the attitude change gradient of the intermediate control nodes. Based on the intermediate control nodes with timing marking, calculate the attitude change rate and change acceleration between each node. Use the attitude angle difference between adjacent intermediate control nodes and the time interval of the timing marking to calculate the attitude change gradient in each axis. The pitch angle change gradient is obtained by dividing the pitch angle difference between adjacent nodes by the time interval. From the T1 to T2 nodes, the pitch angle remains unchanged at 15.0°, and the change gradient is 0° / s; the roll angle changes from 21.2° to 17.8°, the time interval is 0.5s, and the change gradient is -6.8° / s; the yaw angle changes from 205.1° to 192.8°, and the change gradient is -24.6° / s. Further calculate the second-order gradient, that is, the attitude change acceleration, which is obtained by the difference between the change gradients of adjacent time periods. For example, the change gradient of the roll angle in the T0 - T1 section is -7.6° / s, and in the T1 - T2 section is -6.8° / s, and the second-order gradient is 1.6° / s², indicating that the change rate of the roll angle is slowing down. The calculation of the second-order gradient of the yaw angle shows that it maintains a relatively stable change rate during the maneuver. In addition, the attitude change gradient also includes coupled gradient analysis, the influence coefficient of roll angle change on yaw control, and the disturbance degree of yaw movement on roll stability. Through multi-order gradient calculation and coupled analysis, obtain the attitude change gradient of the intermediate control nodes.

[0028] Finally, an ideal control trajectory sequence is generated based on the timing markers and the attitude change gradient. By integrating the time constraint information of the timing markers and the dynamic characteristics of the attitude change gradient, a complete ideal control trajectory is constructed. Using the time nodes T0 to T7 in the timing markers, a time reference for the trajectory sequence is established to ensure the precise definition of each control moment. At the same time, by combining the first-order and second-order derivative information in the attitude change gradient, the control instruction intensity and change trend for each time period are designed. In the time period T1 - T2, according to the change gradient of the roll angle of -6.8° / s and the second-order gradient of 1.6° / s², the generated roll control instruction should start with medium intensity and gradually weaken; in the same time period, the stable change gradient of the yaw angle of -24.6° / s requires that the yaw control instruction be output with a constant intensity. Through the priority information of the timing markers, the trajectory density is increased at key control points to improve the control accuracy; interpolation smoothing is performed on transition points to ensure the continuity of the trajectory. The ideal control trajectory sequence contains the precise attitude target values and the corresponding control instruction amplitudes at each 0.1 s moment from t = 0 to t = 3.5 s. The trajectory sequence also integrates feedforward control information to predict the control requirements at the next moment based on the attitude change gradient and adjust the control output in advance. An ideal control trajectory sequence is generated through the time synchronization of the timing markers and the dynamic matching of the attitude change gradient.

[0029] Step S120: Obtain the imbalance points of the ideal control trajectory sequence, determine the disturbance injection positions according to the imbalance points, and generate a controllable imbalance strategy library according to the disturbance injection positions.

[0030] Specifically, the purpose of this step is to convert the natural imbalance phenomenon in the ideal control trajectory sequence into a controllable control resource. By identifying the key nodes in the trajectory sequence that are prone to imbalance, analyzing the imbalance characteristics and controllability of these nodes, determining the optimal disturbance injection timing and method, and finally establishing a strategy library containing various imbalance control modes, the control idea is transformed from passive error correction to active utilization of imbalance.

[0031] In some embodiments, the obtaining of the imbalance points of the ideal control trajectory sequence includes: analyzing the attitude conversion key points in the ideal control trajectory sequence; determining the available natural imbalance directions at the attitude conversion key points; evaluating the controllability of the natural imbalance directions to generate a controllability evaluation result; and obtaining the imbalance points according to the natural imbalance directions and the controllability evaluation result.

[0032] First, based on the generated ideal control trajectory sequence, deeply analyze the exact attitude target values at each 0.1 s moment from t = 0 to t = 3.5 s contained therein, and identify the key transition points where the attitude change amplitude exceeds the preset threshold. By calculating the attitude angle change rate ω = (θ(t+Δt) - θ(t)) / Δt between adjacent time points, where θ represents the pitch angle, roll angle φ or yaw angle ψ, and Δt = 0.1 s, when the pitch angle change rate |ωθ| > 5° / 0.1 s, roll angle change rate |ωφ| > 8° / 0.1 s or yaw angle change rate |ωψ| > 10° / 0.1 s, mark this time point as the attitude transition key point. In the ideal control trajectory sequence of the turning maneuver, identify the main attitude transition key points: the roll angle starts to increase rapidly at t = 0.3 s, the roll angle reaches the maximum value of 25° at t = 1.2 s, the yaw angle change rate reaches the peak at t = 2.1 s, the roll angle starts to decrease at t = 2.8 s, and the attitude approaches the target state at t = 3.2 s. Each key point corresponds to a significant change in the dynamic characteristics of the UAV and is the moment most prone to control deviation and imbalance phenomena. In high-speed reconnaissance missions, the key points of the turning maneuver usually concentrate on the roll axis change stage, while in the precise hovering mission, the key points are mainly distributed at the fine-tuning moments of each axis. Considering the particularity of different maneuver types at the same time, the key points of the climbing maneuver mainly concentrate on the moments when the pitch axis changes violently, and the key points of the rapid evasion maneuver are manifested as rapid changes in multiple axes simultaneously. Obtain the main attitude transition key points through the time series analysis and gradient calculation of the ideal control trajectory sequence.

[0033] Next, determine the available natural imbalance directions at the key points of attitude conversion. Based on the identified main attitude conversion key points, deeply analyze the dynamic characteristics and imbalance tendencies of the UAV at these moments. At the roll start key point at t = 0.3 s, the UAV starts to establish a roll angle from the balanced state. At this time, the airframe is most sensitive to disturbances around the roll axis. The natural imbalance directions mainly manifest as overshoot or deficiency of the roll angle. The available imbalance directions are the positive and negative offsets of the roll axis. At the roll peak key point at t = 1.2 s, the airframe is in the state of the maximum roll angle. At this time, the coupling effect between the yaw axis and the pitch axis is the strongest. The natural imbalance directions include the premature change of the yaw angle and the additional fluctuation of the pitch angle. At the yaw change key point at t = 2.1 s, the yaw angular velocity of the airframe reaches the maximum. At this time, the main natural imbalance directions are the angular velocity fluctuation of the yaw axis and the secondary oscillation of the roll angle. For the payload delivery mission, there will be an obvious imbalance tendency around the pitch axis at the moment of delivery; for the formation flight mission, the airflow interference between adjacent UAVs will cause the coupling imbalance of the roll axis and the yaw axis. In different flight phases, the manifestation forms of the imbalance directions are also significantly different. The imbalance in the rapid maneuvering phase is mainly reflected in the sudden change of the angular velocity, while the imbalance in the steady flight phase is manifested as the slow offset of the angle. By analyzing the dynamic characteristics and coupling relationships of each key point, multiple main natural imbalance directions can be identified at each conversion point, and multiple available natural imbalance directions can be obtained.

[0034] Then, evaluate the controllability of the natural imbalance direction to generate a controllability evaluation result. For each identified natural imbalance direction, a linearization analysis method is used to evaluate the controllability of each imbalance direction. The controllability evaluation mainly considers three aspects: the predictability of the imbalance amplitude, the controllability of the recovery time, and the degree of influence on the overall trajectory. Taking the positive imbalance of the roll axis at the key point of t = 0.3s as an example, through simulation analysis, it is found that the 5° positive roll imbalance can be completely restored within 0.8 seconds through control instructions, and the influence time on the subsequent trajectory does not exceed 1.5 seconds, and the controllability rating is high. In contrast, although the yaw axis imbalance at the key point of t = 2.1s has a smaller amplitude, due to being in the high-speed yaw stage, the recovery time is 1.2 seconds, and it will affect the accuracy of the second half of the entire turn, and the controllability rating is medium. Under complex meteorological conditions, such as when performing tasks in a gusty wind environment, the difficulty of imbalance recovery will increase significantly, and the controllability rating needs to be adjusted accordingly. By establishing a three-dimensional evaluation model of the imbalance amplitude A, the recovery time T, and the influence range R, the controllability index C = w1·f1(A) + w2·f2(T) + w3·f3(R), where f1(A) = e^(-A / A0), f2(T) = e^(-T / T0), f3(R) = e^(-R / R0), w1, w2, and w3 are weight coefficients and w1 + w2 + w3 = 1, A0, T0, and R0 are normalization parameters, calculate the controllability index for each natural imbalance direction respectively. The index range is 0 - 1, where above 0.8 is high controllability, 0.5 - 0.8 is medium controllability, and below 0.5 is low controllability. The evaluation results show that most imbalance directions have high controllability, a few have medium controllability, and a few are low controllability, generating a complete controllability evaluation result.

[0035] Finally, obtain the imbalance points based on the natural imbalance direction and the controllability assessment results. By synthesizing each natural imbalance direction and the corresponding controllability assessment results, screen out the imbalance points suitable as control strategies. Preferentially select the imbalance directions with high-level controllability assessment results as the main imbalance points, which have the characteristics of controllable imbalance amplitude, short recovery time, and limited influence range. The imbalance directions with medium-level controllability are used as alternative imbalance points and can be used under specific conditions, but additional safety margins need to be added. The imbalance directions with low controllability are not included in the candidate imbalance points temporarily to avoid control risks. For the emergency obstacle avoidance scenario, some of the imbalance points with medium-level controllability can be temporarily enabled to achieve rapid maneuvering. Through the screening of the imbalance directions, multiple effective imbalance points are obtained, and each imbalance point has been strictly verified for safety and feasibility. For example, the imbalance point FP1 is located at the key roll start point at t = 0.3 s, the imbalance direction is the positive direction of the roll axis, the expected imbalance amplitude is 3 - 7°, and the recovery time is 0.6 - 1.0 seconds; the imbalance point FP2 is located at the same key point but the imbalance direction is the negative direction of the roll axis, with similar parameters but opposite directions. There is also a timing correlation between different imbalance points, and the triggering of some imbalance points will affect the availability of subsequent imbalance points. Therefore, it is necessary to establish the constraint relationship between the imbalance points. Through the analysis of the natural imbalance direction and the controllability screening, obtain the verified imbalance points.

[0036] Determine the perturbation injection position based on the imbalance points. Based on the multiple effective imbalance points obtained, analyze the time positions and axial characteristics of each imbalance point to determine the optimal perturbation injection timing and injection method. Utilize the characteristic that the imbalance point FP1 is located at the key point of roll start at t = 0.3s, and set the time 0.1 second before this moment, i.e., t = 0.2s, as the injection position of the positive perturbation on the roll axis. Injecting the perturbation in advance can generate a synergistic effect with the upcoming natural imbalance direction, enhancing the speed and accuracy of roll angle establishment. For the negative imbalance characteristic of the roll axis of the imbalance point FP2, set the negative perturbation injection position at the same moment of t = 0.2s, but the perturbation direction is opposite to that of FP1, which is used when it is necessary to slow down the roll angle establishment speed. Considering the characteristic that the imbalance point has coupled imbalance of the yaw axis and pitch axis at the key point of roll peak at t = 1.2s, determine t = 1.1s as the injection position of the yaw axis perturbation. Utilize the characteristic that the coupling effect is the strongest when the roll angle is close to the maximum value, and adjust the turning radius or enhance the turning efficiency through appropriate yaw perturbation. In an actual application scenario, when the UAV performs an emergency obstacle avoidance task, the characteristic of the imbalance point located at the key point of yaw change at t = 2.1s can be utilized, and set t = 2.0s as the injection position of the rapid yaw perturbation. By injecting the perturbation before the yaw angular velocity reaches the peak value, a faster heading change can be achieved. In the precise hovering task, according to the distribution of the imbalance points in each axial fine-tuning stage, set the perturbation injection position at the starting stage of attitude adjustment to help the UAV reach the desired hovering attitude faster. For the formation flight scenario, when detecting the imbalance points caused by the airflow interference of adjacent UAVs, a pre-compensation perturbation can be injected 0.3 seconds before the interference effect arrives to actively offset the influence of external interference. Considering the temporal correlation between different imbalance points, sufficient time intervals need to be maintained between adjacent perturbation injection positions, and the constraint condition is |t_i - t_j| ≥ Δt_min, where Δt_min = 0.4s, i ≠ j, to avoid the superposition interference of multiple perturbation effects. Determine the perturbation injection positions corresponding to each imbalance point through the analysis of the time characteristics and axial attributes of the imbalance points.

[0037] In some embodiments, generating a controllable imbalance strategy library according to the perturbation injection position includes: establishing a correspondence between the perturbation amplitude and the recovery time for each of the perturbation injection positions; determining an optimal perturbation parameter combination according to the correspondence; and classifying and storing the optimal perturbation parameter combination to generate a controllable imbalance strategy library.

[0038] First, establish the correspondence between the perturbation amplitude and the recovery time for each perturbation injection position. Based on the determined perturbation injection positions, deeply analyze the influence law of different perturbation amplitudes on the control process. For the positive perturbation injection position on the roll axis at t = 0.2 s, it is found through theoretical calculation and experimental verification that small-amplitude perturbations can quickly recover to the ideal trajectory but the control effect is relatively limited, medium-amplitude perturbations reach a better balance between the control effect and the recovery time, and large-amplitude perturbations can produce a significant control effect but require a longer recovery time. For the perturbation injection position on the yaw axis at t = 1.1 s, since this position is in the peak stage of roll and there is an obvious multi-axis coupling effect, the recovery time generated by the same amplitude of perturbation is significantly longer than that of other injection positions. Therefore, more caution is needed in the selection of the perturbation amplitude. In the search and rescue mission, the fast yaw perturbation injection position at t = 2.0 s shows unique advantageous characteristics. Utilizing the dynamic characteristics of the yaw angular velocity approaching the peak, a relatively small perturbation amplitude can produce a significant direction change effect while maintaining a relatively short recovery time. Different application scenarios have different requirements for the balance between the perturbation amplitude and the recovery time. The emergency obstacle avoidance mission can accept a longer recovery time to obtain stronger maneuverability, while the precise positioning mission requires a shorter recovery time to maintain the control accuracy. Through systematic parameter analysis and multi-scenario verification, establish the correspondence between the perturbation amplitude and the recovery time covering different amplitude levels for each perturbation injection position.

[0039] Then, determine the optimal perturbation parameter combination according to the corresponding relationship. Based on the established corresponding relationship between the perturbation amplitude and the recovery time, comprehensively consider the control effect, safety requirements, and mission characteristics, and screen out the optimal parameter configuration for each injection position. For the roll-axis perturbation injection position at t = 0.2 s, according to the analysis results in the corresponding relationship that small-amplitude perturbations recover quickly but have limited control effects, medium-amplitude perturbations have better balance, and large-amplitude perturbations have significant control effects but long recovery times, select medium-amplitude perturbations as the optimal parameters, which can not only produce an obvious roll angle establishment effect but also ensure complete recovery to the ideal trajectory within a reasonable time. For the yaw-axis perturbation injection position at t = 1.1 s, considering the characteristic that the recovery time at this position is significantly longer than other positions in the corresponding relationship, select small-amplitude perturbations as the optimal parameters to obtain sufficient control gain using the coupling effect while avoiding the influence of too long recovery time on subsequent control nodes. For the fast yaw perturbation injection position at t = 2.0 s, based on the analysis in the corresponding relationship that relatively small perturbation amplitudes can produce significant effects and the recovery time is relatively short, select small-amplitude perturbations as the optimal parameters. In the formation flight scenario, this configuration can not only achieve rapid heading adjustment but also have no impact on the overall stability of the formation. During the optimization process, the timing correlation between multiple perturbation injection positions must be considered. When multiple positions need to be used continuously, reasonably adjust the parameter configuration according to the analysis results of their respective corresponding relationships to avoid a decrease in control performance caused by the cumulative effect. Determine the optimal perturbation parameter combination through multi-objective optimization techniques.

[0040] Finally, classify and store the optimal perturbation parameter combinations to form a controllable imbalance strategy library. According to the optimal perturbation parameter combinations at each perturbation injection position determined, conduct systematic classification and organization according to the perturbation axial characteristics, application scenario types, and control target requirements. Establish a roll-axis imbalance strategy classification, integrating the medium-amplitude positive and negative perturbation parameter combinations based on the position at t = 0.2s, which are specifically applicable to various flight maneuvers that require rapid establishment or suppression of the roll angle, such as rapid turning maneuvers, lateral avoidance maneuvers, and precise attitude adjustments. Create a yaw-axis imbalance strategy classification to uniformly manage the small-amplitude coupled perturbation parameter combinations at the position of t = 1.1s and the small-amplitude rapid perturbation parameter combinations at the position of t = 2.0s, respectively targeting different control requirements for precise adjustment of the turning radius and rapid change of the heading. Establish a multi-axis collaborative imbalance strategy classification, integrating multiple optimal perturbation parameter combinations according to reasonable timing relationships and axial coordination principles to form a comprehensive control scheme applicable to complex maneuver tasks. For example, in a dynamic target tracking task, the collaborative strategy can utilize the combined effects of medium-amplitude perturbations on the roll axis and small-amplitude perturbations on the yaw axis to achieve fast and precise tracking maneuvers. Each strategy classification includes complete parameter configuration information, detailed descriptions of applicable conditions, clear explanations of safety limitations, and accurate performance expectation indicators. The strategy library establishes a multi-dimensional intelligent indexing mechanism to support rapid retrieval of the most matching imbalance control strategy according to the current flight state, specific task type, environmental condition changes, and immediate control target. Through the scientific classification and systematic storage of the optimal perturbation parameter combinations, a controllable imbalance strategy library covering multiple application scenarios is formed.

[0041] Step S130, determine the imbalance control interval according to the ideal control trajectory sequence and the controllable imbalance strategy library, generate a perturbation injection sequence according to the imbalance control interval, and establish a dynamic balance benchmark according to the perturbation injection sequence.

[0042] Specifically, the imbalance control interval is determined based on the ideal control trajectory sequence and the controllable imbalance strategy library. Based on the exact attitude target values at each 0.1 s moment from t = 0 to t = 3.5 s in the generated ideal control trajectory sequence, combined with various imbalance strategies in the controllable imbalance strategy library, the time period suitable for applying imbalance control in the trajectory sequence is analyzed. By analyzing the attitude change characteristics of the ideal control trajectory sequence, the interval with a large attitude adjustment amplitude and a long duration is identified as the candidate period for imbalance control. In the interval from t = 0.2 s to t = 1.5 s, the ideal control trajectory sequence shows that the roll angle rapidly increases from the initial state to a peak of 25°. The continuous attitude change characteristics in this interval exactly match the application conditions of the roll-axis imbalance strategy classification in the controllable imbalance strategy library. Using the medium-amplitude positive perturbation parameter combination of the roll-axis imbalance strategy in the strategy library, this time period is determined as the roll-axis-dominated imbalance control interval. In the interval from t = 1.8 s to t = 3.2 s, the ideal control trajectory sequence shows a rapid conversion process of the yaw angle changing from 127.5° to 217.5°, accompanied by a gradual decrease in the roll angle. This complex multi-axis attitude conversion characteristic exactly corresponds to the design goal of the multi-axis collaborative imbalance strategy classification in the controllable imbalance strategy library. Combining the small-amplitude rapid perturbation parameter combination of the yaw-axis imbalance strategy and the negative perturbation parameter combination of the roll-axis strategy in the strategy library, this interval is determined as the multi-axis collaborative imbalance control interval. In the target tracking task, when the ideal control trajectory sequence shows the need for frequent small-amplitude attitude adjustments, the corresponding imbalance control interval can be set using the precise control strategy category in the strategy library. For the initial acceleration phase from t = 0 to t = 0.2 s and the final stable phase from t = 3.2 s to t = 3.5 s in the ideal control trajectory sequence, since the attitude changes are relatively gentle and there are no corresponding imbalance strategies in the strategy library, these time periods are excluded from the imbalance control interval and the conventional control mode is maintained. The imbalance control interval is determined through the time characteristic analysis of the ideal control trajectory sequence and the strategy matching verification of the controllable imbalance strategy library.

[0043] Generate a disturbance injection sequence according to the imbalance control interval. Based on the determined imbalance control interval, convert the imbalance strategies within each interval into specific time-sequenced disturbance instructions. For the roll-axis dominant imbalance control interval from t = 0.2 s to t = 1.5 s, use the medium-amplitude positive disturbance parameter combination in its corresponding roll-axis imbalance strategy to design continuous disturbance injection instructions within this interval. First, inject the main roll-axis positive disturbance at t = 0.2 s, and the disturbance signal d(t) = δ*·u(t - t0)·[1 - u(t - t0 - T_d)], where δ* is the optimal disturbance amplitude, u(t) is the unit step function, t0 = 0.2 s is the injection time, and T_d is the duration. The disturbance amplitude is set at a medium level, and the duration covers the key stage of the roll angle establishment. Subsequently, inject auxiliary roll disturbances at t = 0.8 s and t = 1.2 s respectively, with the amplitude gradually decreasing, to maintain and optimize the roll angle establishment process. For the multi-axis collaborative imbalance control interval from t = 1.8 s to t = 3.2 s, design multi-axis combined disturbance injection instructions according to the combination of small-amplitude rapid disturbance of the yaw axis and negative disturbance of the roll axis it contains. At t = 1.8 s, simultaneously initiate the small-amplitude disturbance of the yaw axis and the negative disturbance of the roll axis to achieve the multi-axis collaborative imbalance effect; at t = 2.3 s, inject the enhanced disturbance of the yaw axis to accelerate the change process of the yaw angle; at t = 2.8 s, provide the finishing disturbance of the roll axis to assist the smooth reduction of the roll angle. In the forest fire monitoring task, when it is necessary to quickly adjust the observation angle, additional pitch-axis disturbance injection instructions can be added within the multi-axis collaborative imbalance control interval to form a three-axis collaborative disturbance sequence with the existing yaw and roll disturbances. For the time periods from t = 0 to t = 0.2 s and from t = 3.2 s to t = 3.5 s excluded from the imbalance control interval, the disturbance injection sequence does not contain any imbalance-related instructions, and the instruction output of the conventional control mode is maintained. The disturbance injection sequence also considers the connection relationship between each imbalance control interval, and sets a smooth transition of the disturbance intensity during the transition period from t = 1.5 s to t = 1.8 s to avoid control mutations during interval switching. Each disturbance injection instruction contains clear time markings, axis specifications, amplitude settings, and durations, forming a complete set of time-sequenced control instructions. Through the strategy conversion and time-sequencing processing of the imbalance control interval, a disturbance injection sequence covering the entire control process is generated.

[0044] In some embodiments, establishing a dynamic balance benchmark according to the disturbance injection sequence includes: determining an expected recovery trajectory after imbalance according to the disturbance injection sequence; generating a tolerance range based on the expected recovery trajectory; determining a time reference point according to the expected recovery trajectory; and combining the expected recovery trajectory, the tolerance range, and the time reference point to form a dynamic balance benchmark.

[0045] First, determine the expected recovery trajectory after imbalance according to the disturbance injection sequence. For the positive disturbance injection command on the roll axis at t = 0.2 s, through dynamic modeling and trajectory prediction, it is determined that this medium-amplitude disturbance will cause the roll angle to deviate from the ideal trajectory by approximately 4 - 6° within 0.3 s, and then gradually recover to the desired state within 0.9 s following the exponential decay law θ(t) = θ_ref + (θ_max - θ_ref)·e^(-t / τ), where θ_ref is the reference trajectory value, θ_max is the maximum deviation value, and τ is the time constant. The entire process exhibits a trajectory characteristic of rapid deviation followed by smooth convergence. For the multi-axis collaborative disturbance injection command at t = 1.8 s, since it involves a small-amplitude disturbance on the yaw axis and a negative disturbance on the roll axis simultaneously, the expected recovery trajectory exhibits complex multi-axis coupling evolution characteristics. The recovery process of the yaw angle will affect the convergence speed and trajectory shape of the roll angle through the aerodynamic coupling effect, and the overall recovery time is extended to 1.2 s. In the target tracking task, when rapid adjustment of the observation angle is required, the prediction accuracy of the expected recovery trajectory directly affects the continuity of the tracking effect. By analyzing the disturbance intensity, duration, and interaction relationship at each moment in the disturbance injection sequence, considering the superposition effect and timing influence between adjacent disturbances, establish the attitude offset, recovery time constant, and trajectory convergence characteristics corresponding to each disturbance, and form an expected recovery trajectory covering the entire control process.

[0046] Then, generate a tolerance range based on the expected recovery trajectory and determine the time reference point according to the expected recovery trajectory. Using the determined expected recovery trajectory, deeply analyze the uncertainty factors and random disturbance effects during the trajectory recovery process, and set a scientifically reasonable allowable deviation range. For the expected recovery trajectory of the roll axis disturbance at t = 0.2 s, considering actual factors such as atmospheric turbulence, sensor noise, and actuator response delay, set the tolerance range for the roll angle recovery process to ±2°, and the tolerance ranges for the yaw angle and pitch angle to ±1° to ensure that the system can still maintain stable control under normal disturbance conditions. For the expected recovery trajectory of the multi-axis collaborative disturbance at t = 1.8 s, due to the complex multi-axis coupling effect and relatively long recovery time, expand the tolerance range of the yaw angle to ±3° and set the tolerance range of the roll angle to ±2.5° to provide sufficient adjustment space for the multi-axis coordinated recovery process.

[0047] Next, based on the time evolution characteristics of each expected recovery trajectory, the time reference points are defined as follows: t_peak = argmax|θ(t) - θ_ref(t)| (deviation reference point), t_half = {t | |θ(t) - θ_ref(t)| = 0.5·|θ_max - θ_ref|} (half-way reference point), t_recover = {t | |θ(t) - θ_ref(t)| < ε_final} (recovery reference point). The key time nodes in the trajectory recovery process are identified as the monitoring and evaluation reference points. For the expected recovery trajectory of roll-axis perturbation, the moment of maximum deviation at 0.3 s is set as the deviation reference point, the moment of complete recovery at 0.9 s is set as the recovery reference point, and the moment of half-way regression at 0.6 s is set as the transition reference point. For the expected recovery trajectory of multi-axis collaborative perturbation, the moment of peak multi-axis deviation at 0.4 s is set as the composite deviation reference point, the moment of dominant-axis recovery at 0.8 s is set as the mid-term reference point, and the moment of full recovery at 1.2 s is set as the final reference point. In the formation flight scenario, the time reference points need to be synchronized with the formation coordination rhythm to ensure that the single-aircraft recovery does not affect the formation stability. Through the detailed time analysis of the expected recovery trajectory, each important time reference point is determined.

[0048] Finally, the expected recovery trajectory, tolerance range, and time reference points are combined to form a dynamic balance reference. Based on the comprehensively determined expected recovery trajectories, corresponding tolerance ranges, and key time reference points, a complete dynamic balance evaluation standard system is established. The expected recovery trajectory of the perturbation at t = 0.2 s, the roll angle tolerance range of ±2°, the deviation reference point (0.3 s), and the recovery reference point (0.9 s) are organically combined into a dynamic balance reference for roll-axis imbalance, providing a full-process monitoring standard for roll-axis imbalance control. The expected recovery trajectory of the multi-axis perturbation at t = 1.8 s, the yaw angle tolerance range of ±3° and the roll angle tolerance range of ±2.5°, the composite deviation reference point (0.4 s), and the final reference point (1.2 s) are systematically combined into a dynamic balance reference for multi-axis collaborative imbalance, providing a comprehensive evaluation basis for complex multi-axis imbalance control. Different from the traditional static balance standard, the dynamic balance reference innovatively allows the attitude to deviate as expected within a preset controllable range, and provides clear recovery path guidance and time node requirements, realizing the transformation of the control idea from passive error correction to active utilization of imbalance. In the actual control process, when the UAV attitude deviation exceeds the tolerance range of the dynamic balance reference or the recovery time exceeds the reference point requirement, the control decision-making mechanism will automatically trigger the corresponding adjustment strategy or activate the safety protection measures. In complex environments such as search and rescue missions, the dynamic balance reference can adapt to the rapidly changing flight requirements and maximize the maneuverability while ensuring safety. By the organic combination of the expected recovery trajectory, tolerance range, and time reference points, a dynamic balance reference suitable for the characteristics of imbalance control is established.

[0049] Step S140: Based on the dynamic balance benchmark, perform trajectory reconstruction to obtain an imbalance - recovery control trajectory, and obtain a set of dynamic response characteristics corresponding to the imbalance - recovery control trajectory.

[0050] Specifically, convert the dynamic balance benchmark into an executable control trajectory. Through trajectory reconstruction technology, transform the ideal linear control process into a non - linear control process that includes active imbalance and precise recovery. This reconstructed trajectory can make full use of the imbalance phenomenon to enhance the maneuverability of the UAV, while maintaining the predictability and safety of the control process.

[0051] In some embodiments, the performing trajectory reconstruction based on the dynamic balance benchmark to obtain an imbalance - recovery control trajectory includes: performing backward inference of the trajectory in the imbalance stage according to the dynamic balance benchmark to generate a backward - inferred trajectory segment; determining the conversion node of imbalance - recovery based on the backward - inferred trajectory segment; and performing trajectory splicing according to the conversion node and the dynamic balance benchmark to generate an imbalance - recovery control trajectory.

[0052] Perform backward inference of the trajectory in the imbalance stage according to the dynamic balance benchmark, and reverse - derive the development process of imbalance from the expected recovery end point. The backward - inference algorithm is based on inverse - time dynamics ẋ = - f(x, u), starting from the equilibrium state x_balance and calculating backward to the initial imbalance state x_imbalance. The backward - inference process considers the controllability and reachability constraints of the system to ensure that the generated trajectory is physically realizable. The constraint conditions include attitude - angle limits |φ| ≤ 30°, |θ| ≤ 30°, angular - velocity constraint |ω| ≤ 100° / s, and actuator - saturation limits. The boundary conditions are set based on the tolerance range of the dynamic balance benchmark to ensure the consistency between the backward - inference starting point and the expected recovery trajectory. Numerical solution uses the fourth - order Runge - Kutta method, and a step size of 0.001 seconds ensures the calculation accuracy. The time span of the backward - inferred trajectory segment is adaptively adjusted according to the severity of the imbalance, 1 - 2 seconds for mild imbalance, 2 - 3 seconds for moderate imbalance, and 3 - 5 seconds for severe imbalance. Trajectory verification ensures rationality through energy conservation and Lyapunov stability checks. For example, in the cross - wind interference scenario of agricultural plant protection operations, the complete evolution process of the initial 15° roll - imbalance state is obtained by backward - inferring 2.5 seconds from the target equilibrium state. Through inverse - dynamics calculation and multiple - constraint verification, an accurate backward - inferred trajectory segment reflecting the development law of imbalance is generated.

[0053] Based on the analysis of the inverse thrust trajectory segment, determine the imbalance-recovery transition node and identify the optimal timing for switching control strategies. The transition node identification uses a method combining phase plane analysis and energy analysis. On the attitude angle-angular velocity phase plane, find the maximum point of the trajectory curvature κ(t) = |ẋ×ẍ| / |ẋ|³, which usually corresponds to the critical position where the control direction changes. The energy criterion calculates the total energy of the system E(t) = T(t) + V(t), where T is the kinetic energy and V is the potential energy. When dE / dt changes from positive to negative, it is marked as a candidate transition point. The control input analysis detects the sign change and amplitude mutation of u(t). When sign(u(t)) ≠ sign(u(t - Δt)) and |Δu| > u_threshold, the transition is confirmed. The stability verification calculates the derivative of the Lyapunov function V(x) at the candidate point to ensure dV / dt < 0 to guarantee the stability after switching. The multi-index fusion determines the optimal transition node through weighted scoring S = w1·κ_score + w2·E_score + w3·u_score, where the weights w1 = 0.4, w2 = 0.3, and w3 = 0.3 reflect the importance of each index. The precise positioning of the node time uses the golden section search, iterating within the range of the initial estimate ±0.1 seconds with an accuracy of 0.01 seconds. For example, in a typical pitch imbalance case, the transition node is determined at t = 1.35 seconds. At this time, the pitch angle reaches the maximum deviation of 12°, the angular velocity is close to zero, and the control torque is about to reverse. Through multi-dimensional analysis and precise search, the dynamically optimized imbalance-recovery transition node is determined.

[0054] Trajectory splicing is performed according to the determined conversion nodes and the dynamic balance benchmark to generate a smooth and continuous complete control trajectory. The trajectory splicing makes full use of the three core elements of the dynamic balance benchmark: taking the expected recovery trajectory as the splicing target reference to ensure that the overall trend of the generated trajectory is consistent with the expectation; taking the tolerance range (roll ±2°, yaw ±3°, pitch ±1°) as the splicing constraint boundary; taking the time reference points (0.3 seconds away from the reference point, 0.9 seconds at the recovery reference point) as the key splicing anchor points. Cubic spline interpolation is used for splicing to ensure C² continuity, and the position, velocity, and acceleration constraints are strictly matched at the conversion nodes. The spline parameters are determined by solving the linear equation system Ax = b, where A is a tridiagonal matrix, x is the undetermined coefficient, and b is the boundary condition vector extracted from the expected recovery trajectory based on the dynamic balance benchmark. The transition zone is designed using the sigmoid weight function w(t) = 1 / (1 + exp(-(t - t_switch) / τ)), where t_switch is taken from the time reference point of the dynamic balance benchmark, and the time constant τ = 0.1 second controls the transition smoothness. The splicing verification directly uses the tolerance range of the dynamic balance benchmark for continuity checking: |x(t⁺) - x(t⁻)| < 2° (roll) or 3° (yaw), |ẋ(t⁺) - ẋ(t⁻)| < 0.01 radian / second, |ẍ(t⁺) - ẍ(t⁻)| < 0.1 radian / second². The dynamic consistency is verified by substituting the spliced trajectory into the original dynamic equation, and the residual ‖ẋ - f(x, u)‖ < δ_dyn confirms the physical feasibility. The constraint satisfaction check ensures that the entire trajectory is always within the tolerance range of the dynamic balance benchmark. Performance optimization is achieved by fine-tuning the splicing parameters to minimize the deviation of the trajectory from the expected recovery trajectory at the time reference point. The complete trajectory output includes the time series t ∈ [0, T_final], the state trajectory x(t), the control trajectory u(t), and the performance evolution J(t). For example, in the building bypass task of urban distribution, a 3.8-second smooth trajectory from an initial 20° roll imbalance to full recovery is generated through precise splicing, and the trajectory accurately passes through the preset states at the 0.3-second and 0.9-second time reference points. Through precise splicing and comprehensive verification based on the dynamic balance benchmark, a smooth and feasible imbalance-recovery control trajectory is generated.

[0055] Obtain the set of dynamic response characteristics corresponding to the imbalance - recovery control trajectory. Based on the generated imbalance - recovery control trajectory, extract the dynamic response characteristics and behavior patterns of the UAV during the trajectory execution. For the roll - axis imbalance - recovery control trajectory that includes the whole process of active imbalance and controllable recovery, starting from the disturbance injection moment at t = 0.2s, calculate the response parameters of the roll angle during the imbalance stage. It is obtained that the rise time of the roll - angle response is 0.25 seconds, the maximum deviation is 6.2°, and the instantaneous error at the conversion node at 0.3s is 4.8°. Identify the convergence law of the recovery stage from the conversion node at 0.3s to the recovery reference point at 0.9s. By fitting the second - order system response θ(t) = θss+(θ0 - θss)e^(-ζωnt)[cos(ωat)+(ζ / √(1 - ζ²))sin(ωat)], where ωn is the natural frequency, ζ is the damping ratio, and ωa = ωn√(1 - ζ²), determine that the dominant time constant of the recovery process τ = 1 / (ζωn)=0.35 seconds, and the damping ratio ζ = 0.72, showing a slightly under - damped convergence characteristic. For the multi - axis collaborative imbalance - recovery control trajectory, study the coupled response phenomenon from the multi - axis disturbance injection at t = 1.8s to the complete recovery process. It is found that the response of the yaw angle lags behind the roll angle by about 0.15 seconds, and the cross - coupling coefficient between the two axes is 0.31, reaching the maximum coupling strength at the composite deviation conversion node at 0.4s. Through the frequency - domain transformation technology, extract the frequency - response characteristics of the imbalance - recovery control trajectory. The main frequency components of the roll - axis imbalance process are concentrated around 2.1Hz, while the frequency components of the recovery process shift to the low - frequency range of 1.3Hz. In the case of multi - axis collaboration, there is a differential - frequency coupling phenomenon of 0.4Hz between the frequency responses of the yaw axis and the roll axis. In the forest fire monitoring task, this frequency characteristic can maintain the stability of image acquisition while quickly adjusting the observation angle, avoiding image blurring caused by high - frequency oscillation. Quantify the amplitude - change characteristics of the imbalance - recovery control trajectory. The maximum amplitude response of the roll - axis imbalance is 6.2°, the amplitude decay rate is - 8.3° per second, and the amplitude change rate during the recovery process follows an exponential decay law, ensuring the predictability and safety of the maneuvering process. In the formation - flight scenario, the consistency of the dynamic response characteristics helps to maintain the stability of the formation flight, and the difference in response delay between each aircraft is controlled within 0.1 second. Through time - domain calculation, frequency - domain transformation, and amplitude quantization, obtain the set of dynamic response characteristics of the imbalance - recovery control trajectory.

[0056] Step S150, obtain the balance - recovery index corresponding to the set of dynamic response characteristics, construct the imbalance - recovery spatio - temporal manifold based on the balance - recovery index, and perform feature mapping based on the imbalance - recovery spatio - temporal manifold to obtain the core imbalance - control feature set.

[0057] Specifically, based on the dynamic response feature set, key performance indicators reflecting the process of the UAV recovering from an unbalanced state to a balanced state are extracted. The dynamic response feature set includes multi-dimensional data such as time response characteristics, frequency response characteristics, amplitude change characteristics, and multi-axis coupling characteristics of the UAV during the imbalance-recovery process. The extraction of balance recovery indicators adopts feature transformation and induction methods to refine key indicators for evaluating recovery performance from the existing dynamic features. The macroscopic time indicators are directly extracted from the time response characteristics: the time constant τ is used as the benchmark indicator for the speed of recovery, and the total recovery time T_recovery = nτ, where n is the engineering convergence coefficient, usually taking 3 - 5; the rise time t_r is directly used as the response speed indicator; based on the amplitude decay rate k_decay, the convergence rate λ_conv = |k_decay| / A_max is deduced, where A_max is the maximum deviation amplitude. The mesoscopic dynamic indicators are transformed from frequency and damping characteristics: the damping ratio ζ is directly used as the oscillation suppression indicator; the recovery main frequency f_recovery is used as the system oscillation frequency; based on the maximum deviation A_max and the instantaneous error e_instant, the relative overshoot M_r = A_max / |A_max - e_instant| is calculated to reflect the overshoot degree of the dynamic response. The microscopic coupling indicators are extracted from multi-axis characteristics: the cross-coupling coefficient K_couple is used as the inter-axis influence strength indicator; based on the yaw response delay t_delay and the differential frequency coupling f_diff, the coupling complexity C_complex = t_delay × f_diff is defined to quantify the difficulty of multi-axis coordination; the ratio R_freq = f_imbalance / f_recovery of the imbalance main frequency f_imbalance to the recovery main frequency f_recovery is used as the control mode conversion feature. For example, in agricultural plant protection operations under strong wind interference, the indicators extracted based on dynamic response characteristics show that a recovery time constant of 0.35 seconds corresponds to a fast response ability, a damping ratio of 0.72 indicates good oscillation suppression, and a frequency ratio of 1.62 reflects an obvious control mode switch. Through the direct adoption, mathematical transformation, and combined derivation of dynamic response characteristics, balance recovery indicators that comprehensively reflect the performance of the imbalance recovery process are obtained.

[0058] Based on the extracted balance recovery indicators, an imbalance recovery spatio-temporal manifold is constructed to map the discrete recovery process data to a continuous geometric space. The construction of the spatio-temporal manifold takes the balance recovery indicators as coordinate dimensions, and the four most representative indicators are selected to form a four-dimensional manifold space M ⊂ R 4: The first dimension is the time constant τ, which reflects the speed of recovery. The second dimension is the damping ratio ζ, which characterizes the oscillation characteristics. The third dimension is the coupling coefficient K_couple, which depicts the multi-axis correlation. The fourth dimension is the normalized time t / T_recovery, which unifies the time scale. Each point p(τ, ζ, K_couple, t / T_recovery) on the manifold represents the comprehensive performance state at a certain moment during the recovery process. The imbalance recovery trajectory forms a continuous curve γ(s) on the manifold from the initial imbalance state p_imbalance to the final balanced state p_balance, where s is the arc length parameter. The manifold metric tensor g_μν is designed according to the physical meanings and dimensional differences of each index, and the weighted Euclidean metric is adopted to ensure the comparability between different indices. The manifold curvature reflects the complexity of the recovery process by calculating the degree of bending of the trajectory in the index space, and the high-curvature region corresponds to the key stage where the performance indices change violently. Based on the convergence rate λ_conv and the oscillation frequency f_recovery, the dynamic structure of the manifold is constructed, and the vector field V = λ_conv ∂ / ∂τ + 2πf_recovery ∂ / ∂ζ is defined to describe the evolution direction of the indices. The topological structure of the manifold is determined by analyzing the distribution of different recovery modes in the index space. The fast response mode, the stable coordination mode, and the adaptive regulation mode form different characteristic regions on the manifold. The manifold boundary is determined according to the physical constraints of the indices. For example, the damping ratio 0 < ζ < 1 ensures stability, and the coupling coefficient |K_couple| < 1 guarantees controllability. By balancing the geometric representation and topological analysis of the recovery indices, a spatio-temporal manifold reflecting the essential characteristics of the imbalance recovery process is constructed.

[0059] In some embodiments, obtaining the core imbalance control feature set by performing feature mapping based on the spatio-temporal manifold of imbalance recovery includes: performing topological analysis on the spatio-temporal manifold of imbalance recovery to obtain the geometric features of the manifold; performing spatio-temporal folding transformation based on the geometric features of the manifold to generate a compressed recovery path; extracting key points from the compressed recovery path to obtain a set of folding nodes; and performing feature mapping based on the set of folding nodes to obtain the core imbalance control feature set.

[0060] Topological analysis of the spatio-temporal manifold for imbalance recovery obtains the geometric characteristics of the manifold. Topological analysis uses differential geometry methods to systematically study the intrinsic properties and external shapes of the manifold, and extract geometric invariants that reflect the essential laws of imbalance recovery. Manifold curvature analysis characterizes the bending properties of the manifold by calculating the Gaussian curvature K = det(g^{-1}R) / 2 and the mean curvature H = tr(g^{-1}h) / 2, where g is the metric tensor, R is the Ricci tensor, and h is the second fundamental form. Regions with Gaussian curvature K > 0 correspond to stable recovery segments, regions with K < 0 are unstable transition segments, and regions with K = 0 represent a uniform recovery process. The geodesic curvature κ_g = |dT / ds| quantifies the degree to which the recovery trajectory deviates from the geodesic, and the smaller κ_g is, the more optimized the control is. Homology group calculation H_k(M) identifies the k-dimensional "hole" structure in the manifold. H_0 reflects the number of connected components, H_1 corresponds to the number of loops, and H_2 represents the number of cavities. For example, the analysis of the imbalance recovery manifold generated by a drone performing express delivery between urban buildings affected by complex airflows between buildings shows that there are 3 high-curvature regions (K > 2.5) corresponding to sharp turning points, 5 saddle points (K < -1.5) identifying the positions of control mode switches, and the first Betti number b_1 = 2 indicating the existence of two independent recovery loops. Critical point analysis identifies the extreme points, saddle points, and inflection points on the manifold by calculating the gradient ∇f = 0, where f is a scalar function defined on the manifold such as recovery time or energy consumption. Morse index calculates the number of negative eigenvalues of the Hessian matrix at the critical point, classifies the types of critical points, and predicts their stability. The Riemannian volume V_M = ∫_M√det(g)d 4 x quantifies the "size" of the state space, and the smaller the volume, the more concentrated and efficient the recovery process is. The principal curvature direction is determined by solving the characteristic equation det(h - κg) = 0, indicating the main evolution direction of the recovery process. Through systematic topological and geometric analysis, the geometric characteristics of the manifold that comprehensively describe the structural characteristics of the manifold are obtained.

[0061] Perform space-time folding transformation based on manifold geometric features to generate a compressed recovery path. The space-time folding transformation utilizes manifold geometric features and compresses the original long recovery path into a shorter equivalent path through mathematical transformation. The folding transformation is defined as a diffeomorphic mapping Φ:M→M', which preserves topological properties but changes the metric structure, such that d_M'(Φ(p),Φ(q))≤d_M(p,q) holds for all point pairs. The folding point selection is based on the curvature extreme value criterion. When |K(p)|>K_threshold = 2.0, the point p becomes a folding candidate point, and the folding direction and folding angle are determined through local analysis. The folding operation is implemented using the exponential mapping exp_p:T_pM→M, which maps the vector v in the tangent space to the point exp_p(v) on the manifold. The folding path is γ_fold(t)=exp_p(tv) where t∈[0,1]. The compression in the time dimension is achieved through the Lorentz transformation, τ'=γ_L(τ - vx / c²) where γ_L = 1 / √(1 - v² / c²) is the Lorentz factor and v is the characteristic velocity, realizing the "contraction" in the time dimension. The folding in the space dimension is driven by the Ricci flow equation ∂g_ij / ∂t = -2R_ij to evolve the metric, causing the manifold to contract towards a more regular shape. For example, in the forest fire monitoring task, the length of the original recovery path is 8.5 units, and after 3 folding transformations, it is compressed to 3.2 units, with a compression rate of 62%. The first folding uses the high curvature (K = 3.8) in the pitch-roll coupling region to shorten the path by 35%. Multiple foldings are achieved by recursively applying the transformation, and the compression effect and stability are evaluated after each folding. Through systematic folding transformation and path optimization, a space-time compressed recovery path is generated.

[0062] Key points are extracted from the compression recovery path to obtain a set of folding nodes. The key point extraction uses a recognition method that combines geometric features and dynamic significance, screening out the core nodes that determine the recovery process from the compression path. Geometric key points are identified through curvature analysis, including points with maximum curvature (κ > κ_max = 2.5) to mark sharp turn positions, points with zero curvature (|κ| < 0.1) to represent straight line segments, and points with curvature sign changes to mark steering switches. Dynamic key points are determined based on phase space analysis, including positions with special dynamic significance such as fixed points, limit cycles, and bifurcation points. Folding intersection points, as the convergence positions of multiple paths, are of great significance and are identified by calculating the distance between paths d(γ_i(t), γ_j(t)) < ε_merge = 0.2. Topological key points include representative points of homotopy equivalence classes, points corresponding to generators of the fundamental group, branch points of covering spaces, etc. For example, in the offshore wind power inspection task, 12 key points are extracted after compressing the complex imbalance recovery process caused by strong sea winds: 4 geometric sharp turn points (κ > 3.0), 3 mode switching points (control law change positions), 2 folding center points (multi-path convergence), and 3 energy extreme points (local minimum of control energy). The importance score of key points is calculated using a multi-index weighted method I_p = w_1·κ_norm + w_2·E_impact + w_3·T_influence, where κ_norm is the normalized curvature, E_impact is the energy impact factor, T_influence is the time influence range, and the weights are w_1 = 0.4, w_2 = 0.3, w_3 = 0.3. The connection relationship between nodes is represented by the reachability matrix A_ij, where A_ij = 1 indicates that node j can be directly reached from node i, forming a topological network of folding nodes. The temporal attributes of nodes record their positions on the original time axis t_original and the compressed position t_compressed, and the compression ratio r_i = t_compressed / t_original reflects the degree of time folding. The state attributes of nodes include the complete pose information, control input, constraint conditions, etc. of the point, ensuring the integrity of node information. Through multi-dimensional analysis and importance evaluation, a set of folding nodes that form the core structure of the compression path is extracted.

[0063] Based on the set of folding nodes, feature mapping is performed to obtain a set of core imbalance control features. The feature mapping uses non-linear dimensionality reduction and pattern recognition methods to project the high-dimensional node attribute space onto a low-dimensional feature space and extract the most representative control features. The mapping function is designed as F:R n →R ᵐ(m << n), which is achieved by Kernel Principal Component Analysis (KPCA). The kernel function is selected as the Gaussian radial basis k(x_i, x_j) = exp(-||x_i - x_j||² / 2σ²), and the bandwidth parameter σ is optimized through cross-validation to obtain σ_opt = 1.5. In the first step of feature extraction, the kernel matrix K_ij = k(n_i, n_j) is calculated, where n_i is the attribute vector of the i-th folding node, and then the eigenvector equation Kα = λα is solved to extract the principal eigenvectors. The proportion of the explained variance of the first m principal components reaches more than 95%. Usually, m = 5 - 8 is sufficient to capture the main features. The extracted core features include: the imbalance mode feature f_mode, which identifies typical imbalance types (uniaxial / multi-axis / burst / progressive) through cluster analysis; the recovery strategy feature f_strategy, which characterizes the controller selection mode (PID / adaptive / robust / intelligent); the time-scale feature f_time, which reflects the separation of fast and slow dynamics (fast variables < 0.5s / medium-speed variables 0.5 - 2s / slow variables > 2s); the energy distribution feature f_energy, which describes the distribution ratio of control energy in each channel; the stability margin feature f_margin, which quantifies the safety margin of the system from the instability boundary. For example, in the fine operation of power line inspection, the extracted core features show that the dominant imbalance mode is wind-induced roll (accounting for 45%), the optimal recovery strategy is adaptive sliding mode control (success rate 92%), the key time scale is 0.8 seconds (rapid response requirement), the energy proportion of the roll channel is 68% (main control channel), and the stability margin is 15° (safety boundary). The physical meaning of the features is verified through inverse mapping to ensure that the extracted features can reconstruct the key characteristics of the original control process. The discriminability of the features is evaluated by the Fisher discriminant ratio J = (μ_1 - μ_2)² / (σ_1² + σ_2²). J > 2 indicates that the features have good classification ability. Through the feature extraction and verification analysis of the system, a core imbalance control feature set that highly condenses the essential laws of imbalance control is obtained.

[0064] Step S160, establish an imbalance-recovery mapping rule according to the core imbalance control feature set, generate disturbance control parameters according to the imbalance-recovery mapping rule, and establish an adaptive mapping parameter based on the disturbance control parameters.

[0065] Specifically, the core imbalance control feature set is transformed into an operable control mapping rule, and a decision-making mechanism for imbalance control is established through pattern recognition and parameter association. This mapping rule can automatically select the optimal imbalance control strategy according to the current flight state, realizing intelligent adaptive control.

[0066] In some embodiments, establishing the imbalance - recovery mapping rule according to the core imbalance control feature set includes: identifying beneficial imbalance patterns for the core imbalance control feature set; extracting disturbance trigger feature parameters corresponding to the beneficial imbalance patterns; establishing an association relationship between the disturbance trigger feature parameters and the balance recovery index; and generating the imbalance - recovery mapping rule according to the association relationship.

[0067] Identifying beneficial imbalance patterns for the core imbalance control feature set. The principal component analysis method is used to reduce the dimensionality of the feature set, and the main feature components with the highest contribution rate are extracted. Then, the K - means clustering algorithm and hierarchical clustering analysis method are used to group the feature data, and the optimal number of clusters and cluster centers are determined by calculating the Euclidean distance and correlation coefficient between features. By analyzing the synergy relationship between the speed control feature and the accuracy control feature, it is found that when the speed control feature shows a high angle recovery rate and the accuracy control feature shows a high error convergence accuracy, the two show a significant positive synergy effect. Based on this, the fast - response imbalance pattern is identified, which is applicable to control scenarios that require rapid and precise attitude adjustment within a short time. By analyzing the matching relationship between the oscillation suppression feature and the coupling coordination feature, it is found that when the oscillation suppression feature shows an ideal damping ratio and the coupling coordination feature shows a high multi - axis synchrony, the control mechanism shows excellent stability. Based on this, the stable - coordination imbalance pattern is identified, which is mainly applicable to complex scenarios that require the smoothness of the control process and multi - axis collaborative cooperation. By analyzing the fusion relationship between the frequency adjustment feature and the amplitude adjustment feature, it is found that when the frequency adjustment feature shows good frequency conversion smoothness and the amplitude adjustment feature has high amplitude control accuracy, the adaptive - adjustment imbalance pattern is identified, which is specifically applicable to dynamic scenarios with environmental condition changes and task mode switching. For example, in the forest fire monitoring task, the fast - response mode is activated when quickly tracking the fire, the stable - coordination mode is activated when stabilizing image acquisition, and the adaptive mode is activated when adjusting the observation strategy.

[0068] Extract the disturbance trigger feature parameters corresponding to the beneficial imbalance patterns. Based on the identified three types of beneficial imbalance patterns, use feature engineering and parameter extraction techniques to accurately extract the key disturbance trigger parameters from the feature combinations of each pattern. For the fast response imbalance pattern, extract the angle recovery rate threshold as the speed trigger parameter and the error convergence accuracy threshold as the accuracy trigger parameter through statistical analysis and threshold optimization methods, and establish a dual-parameter joint discrimination mechanism. Only when both parameters meet the high-level requirements at the same time can this pattern be activated. For the stable coordination imbalance pattern, extract the damping ratio range as the stability trigger parameter and the multi-axis synchrony index as the coordination trigger parameter through sensitivity analysis and interval optimization techniques. When both parameters are within the ideal numerical range, activate this pattern. For the adaptive adjustment imbalance pattern, extract the frequency conversion rate as the frequency trigger parameter and the amplitude adjustment range as the amplitude trigger parameter through frequency domain analysis and amplitude statistics methods. When environmental changes or task switching requirements are detected, activate this pattern based on these two parameters. In special tasks such as formation flight, the trigger parameters can be dynamically adjusted according to task requirements.

[0069] Establish the correlation between the disturbance trigger feature parameters and the balance recovery indicators. Use multivariate statistical analysis and regression modeling methods to establish the quantitative relationship between the parameters. Conduct a correlation analysis between the angle recovery rate threshold and the balance recovery efficiency indicator. Through linear regression, it is found that there is a positive correlation between the two, and a linear model η_r = a·θ_th + b is established, where η_r is the recovery efficiency indicator, θ_th is the angle recovery rate threshold, and the regression coefficient a reflects the degree of influence of the threshold change on the efficiency. Establish an optimization relationship model between the damping ratio range and the stability indicator S_stab = -α(ζ - ζ_opt)² + S_max, where ζ is the damping ratio, ζ_opt is the optimal damping ratio, and S_max is the maximum stability indicator value. Through curve fitting, it is found that the stability reaches the optimal when the damping ratio is within a specific ideal range. Establish a piecewise non-linear relationship model between the frequency conversion rate and the adaptability indicator. The control process maintains good adaptability within a reasonable rate range, but the performance drops significantly when it exceeds the critical value. Establish a strong correlation between the multi-axis synchrony index and the coordination recovery indicator. The higher the synchrony, the better the coordination recovery performance. There is an optimal balance point between the amplitude adjustment range and the control accuracy, which needs to be optimized and determined.

[0070] Generate imbalance - recovery mapping rules according to the described association relationship. Based on the established association relationship, use rule reasoning and decision tree algorithms to generate executable mapping rules. Quick - response mapping rule: When it is detected that the angle recovery rate requirement exceeds the threshold and the error convergence accuracy reaches a high standard, automatically calculate the target recovery efficiency and accuracy target according to the linear association model, activate the quick - response imbalance mode and set the corresponding control parameters. Stable - coordination mapping rule: When it is detected that there is a multi - axis coordinated control requirement and the damping ratio is within the ideal range, calculate the optimal coordination parameters according to the optimization model, activate the stable - coordination imbalance mode and optimize the multi - axis cooperation strategy. Adaptive mapping rule: When the environmental conditions or task patterns change, determine the optimal conversion strategy according to the non - linear relationship model, switch to the adaptive - adjustment imbalance mode and dynamically adjust the control parameters. The mapping rules integrate intelligent mode - switching logic and can automatically and smoothly switch between different modes according to the flight conditions. Under complex meteorological conditions, the mapping rules can automatically adjust the activation priorities of each mode to enhance the anti - interference ability of the control.

[0071] Generate disturbance control parameters according to the imbalance - recovery mapping rules. Using the linear association model and control target setting in the quick - response mapping rule, when the corresponding trigger conditions are detected, calculate and generate quick - response disturbance control parameters according to the mapping rules: the disturbance amplitude δ = f1(η_target) = k_δ·η_target, the disturbance duration T_d = f2(P_target) = T_base·(P_target / P_ref)^β, the recovery time constant τ_r = f3(a,b) = (a·θ_th + b)⁻¹, where η_target is the target recovery efficiency, P_target is the accuracy target, and k_δ, T_base, β are configuration parameters. Using the relevant relationship and coordination parameters calculation in the stable - coordination mapping rule, when the multi - axis coordinated control conditions are met, generate stable - coordination disturbance control parameters according to the mapping rules: including the main - axis disturbance configuration determined based on the optimal coordination parameters, the auxiliary - axis disturbance configuration set based on the multi - axis cooperation strategy, the inter - axis timing parameters optimized based on the coordination relationship, etc. Using the non - linear relationship and balance - point principle in the adaptive mapping rule, when the environment or task changes, generate adaptive disturbance control parameters according to the mapping rules: including the frequency parameters determined based on the optimal conversion strategy, the amplitude parameters set based on the dynamic adjustment requirements, the conversion time parameters optimized based on the adaptability requirements, etc. For example, in the forest fire monitoring task, when the observation requirement switches from static monitoring to dynamic tracking, the control algorithm automatically adjusts the disturbance control parameters according to the quick - response mapping rule, switching the 5° disturbance amplitude and 1.2s recovery time in the stable - coordination mode to the 8° disturbance amplitude and 0.8s recovery time in the quick - response mode. Through the parametric conversion of the imbalance - recovery mapping rules, generate disturbance control parameters that adapt to different control requirements.

[0072] An adaptive mapping parameter is established based on the disturbance control parameter. Based on the disturbance amplitude, disturbance duration, and recovery time constant in the fast-response disturbance control parameter, a fast-response adaptive mapping parameter is established: taking these parameters as the reference values, an adaptive adjustment mechanism is set. When the actual control effect deviates from the expectation, the disturbance amplitude and recovery time constant are automatically adjusted according to the degree of deviation to ensure the stability of the fast-response performance. Based on the main-axis disturbance configuration, auxiliary-axis disturbance configuration, and inter-axis timing parameter in the stable coordination disturbance control parameter, a stable coordination adaptive mapping parameter is established: a multi-axis adaptive mechanism is constructed based on these parameters. When a change in the inter-axis coordination effect is detected, the disturbance configurations of the main axis and the auxiliary axis and the timing parameter are automatically adjusted to maintain the coordinated control effect. Based on the frequency parameter, amplitude parameter, and conversion time parameter in the adaptive disturbance control parameter, an adaptive adjustment mapping parameter is established: an environment adaptation mechanism is constructed with these parameters as the core. The frequency and amplitude parameters are dynamically adjusted according to the environmental disturbance intensity and the degree of task change to ensure control adaptability in a changing environment. Through the adaptive expansion and intelligent optimization of the disturbance control parameter, an adaptive mapping parameter with environmental adaptability and learning ability is established.

[0073] In step S170, a reverse imbalance joint control is generated according to the adaptive mapping parameter, and a flight control signal is output according to the reverse imbalance joint control to achieve the adaptive imbalance control of the UAV flight attitude.

[0074] Specifically, first, a multi-channel imbalance control matrix is established based on adaptive mapping parameters. The three types of mapping parameters, namely, fast response, stable coordination, and adaptive regulation, are used as the weight configurations for the three columns of the matrix respectively. The matrix has three rows corresponding to the pitch, roll, and yaw axes to construct a 3×3 control matrix. Each element corresponds to the imbalance control weight of a specific axis in the corresponding mode. For example, when the demand for rapid adjustment of the pitch axis in forest fire monitoring is high, a high weight is set for the first element in the first row of the matrix. Next, attitude coupling analysis is performed according to the control matrix to determine the reverse control gains of each axis. The dominant control mode of each axis is identified by extracting the maximum value in each row. The maximum weight is set as the reverse control gain reference for that axis and normalized. For example, when the weight of the rapid response of the pitch axis is the largest during payload delivery, its control gain is set accordingly. Then, the reverse control gain and the adaptive mapping parameters are adaptively fused to generate a joint control strategy. The enhanced control parameters of each axis are obtained by multiplying the gain by the corresponding mapping parameter. The control priority is determined according to the gain magnitude and a coordination rule is established. For example, when the pitch gain is the largest during formation flight, it obtains the highest priority. Finally, based on the joint control strategy, the control output intensity is adjusted in real time to form a reverse imbalance joint control. The output intensity of each axis is dynamically adjusted according to the ratio of the actual deviation to the target deviation, and a total control mechanism is established to prevent overload. For example, when the yaw demand increases under complex weather conditions, the yaw output is moderately increased within the total limit and the intensities of other axes are reduced. A complete reverse imbalance joint control is formed through parameter application and proportional adjustment.

[0075] Output flight control signals according to reverse imbalance joint control to achieve adaptive imbalance control of the UAV flight attitude. Directly convert the actual output intensity value of the pitch axis into the elevator deflection angle, map the output intensity to the specific rudder surface deflection degree and deflection rate through a preset conversion coefficient, and generate an elevator control signal. Convert the actual output intensity value of the roll axis into the aileron differential deflection angle, calculate the symmetric deflection degree of the left and right ailerons through the differential coefficient, and generate an aileron differential control signal. Convert the actual output intensity value of the yaw axis into the rudder deflection angle, determine the deflection direction and deflection amplitude of the rudder through the yaw conversion coefficient, and generate a rudder control signal. Calculate the throttle opening increment through the power conversion coefficient based on the sum of the output intensities of the three axes, and generate a throttle control signal that matches the imbalance control requirements. Perform amplitude limitation and safety checks on all control signals to ensure that the control commands of each actuator are within the designed safety range. For example, in a target tracking task, when the output intensity of the pitch axis is 6.5, the output intensity of the roll axis is 4.2, and the output intensity of the yaw axis is 3.8, specific control commands of the elevator deflecting 13 degrees upward, the left aileron deflecting 8.4 degrees downward, the right aileron deflecting 8.4 degrees upward, the rudder deflecting 11.4 degrees to the right, and the throttle opening increasing by 14.5% are generated through the conversion coefficient. Each actuator coordinates its actions according to these precise numerical commands to achieve fast and accurate attitude adjustment. Achieve adaptive imbalance control of the UAV flight attitude through the numerical conversion of the reverse imbalance joint control intensity and the precise drive of the actuator.

[0076] In order to execute the UAV flight attitude adaptive imbalance control method corresponding to the above method embodiment to achieve the corresponding functions and technical effects. Refer to Figure 2 , Figure 2 FIG. shows a structural block diagram of a UAV flight attitude adaptive imbalance control device 200 provided by an embodiment of the present application. For the sake of convenience of description, only the parts related to this embodiment are shown. The UAV flight attitude adaptive imbalance control device 200 provided by the embodiment of the present application includes: A trajectory analysis module 201, configured to obtain the current attitude state and the target attitude state of the UAV, and perform reverse trajectory analysis on the current attitude state and the target attitude state to generate an ideal control trajectory sequence; An imbalance acquisition module 202, configured to obtain the imbalance points of the ideal control trajectory sequence, determine the disturbance injection positions according to the imbalance points, and generate a controllable imbalance strategy library according to the disturbance injection positions; A reference establishment module 203, configured to determine an imbalance control interval according to the ideal control trajectory sequence and the controllable imbalance strategy library, generate a disturbance injection sequence according to the imbalance control interval, and establish a dynamic balance reference according to the disturbance injection sequence; A trajectory reconstruction module 204, configured to perform trajectory reconstruction based on the dynamic balance benchmark to obtain an imbalance-recovery control trajectory, and obtain a set of dynamic response characteristics corresponding to the imbalance-recovery control trajectory; A feature extraction module 205, configured to obtain a balance recovery index corresponding to the set of dynamic response characteristics, construct an imbalance recovery spatio-temporal manifold based on the balance recovery index, and perform feature mapping on the imbalance recovery spatio-temporal manifold to obtain a core imbalance control feature set; A mapping generation module 206, configured to establish an imbalance-recovery mapping rule according to the core imbalance control feature set, generate a perturbation control parameter according to the imbalance-recovery mapping rule, and establish an adaptive mapping parameter based on the perturbation control parameter; A control execution module 207, configured to generate a reverse imbalance joint control according to the adaptive mapping parameter, output a flight control signal according to the reverse imbalance joint control, and implement an adaptive imbalance control of the UAV flight attitude.

[0077] The above UAV flight attitude adaptive imbalance control device 200 can implement the UAV flight attitude adaptive imbalance control method in the above method embodiment. The optional items in the above method embodiment are also applicable to this embodiment, which will not be elaborated here. The remaining content of this application embodiment can refer to the content of the above method embodiment, and will not be repeated in this embodiment.

[0078] As Figure 3 shown, the third embodiment of the present invention further provides a computer device, including a memory 301, a processor 302, and a computer program stored in the memory 301 and executable on the processor 302, wherein the processor 302 implements the steps of the UAV flight attitude adaptive imbalance control method described in the first embodiment of the present invention when executing the program.

[0079] The purpose of the above embodiments is to exemplarily reproduce and deduce the technical solutions of the present invention, and to completely describe the technical solutions, purposes, and effects of the present invention. The purpose is to make the public understand the disclosed content of the present invention more thoroughly and comprehensively, and does not limit the protection scope of the present invention.

[0080] The above embodiments are not exhaustive listings based on the present invention. In addition, there may be multiple other embodiments not listed. Any replacement and improvement made without violating the concept of the present invention fall within the protection scope of the present invention.

Claims

1. A method for adaptively controlling the imbalance of the flight attitude of a drone, characterized in that, Including: Obtain the current attitude state and target attitude state of the UAV, and perform reverse trajectory analysis on the current attitude state and the target attitude state to generate an ideal control trajectory sequence; Obtain the imbalance points of the ideal control trajectory sequence, determine the disturbance injection positions according to the imbalance points, and generate a controllable imbalance strategy library according to the disturbance injection positions; Determine the imbalance control interval according to the ideal control trajectory sequence and the controllable imbalance strategy library, generate a disturbance injection sequence according to the imbalance control interval, and establish a dynamic balance benchmark according to the disturbance injection sequence; Perform trajectory reconstruction based on the dynamic balance benchmark to obtain an imbalance-recovery control trajectory, and obtain a dynamic response feature set corresponding to the imbalance-recovery control trajectory; Obtain the balance recovery index corresponding to the dynamic response feature set, construct an imbalance recovery spatio-temporal manifold based on the balance recovery index, and perform feature mapping on the imbalance recovery spatio-temporal manifold to obtain a core imbalance control feature set; Establish an imbalance-recovery mapping rule according to the core imbalance control feature set, generate disturbance control parameters according to the imbalance-recovery mapping rule, and establish an adaptive mapping parameter based on the disturbance control parameters; Generate reverse imbalance joint control according to the adaptive mapping parameter, output a flight control signal according to the reverse imbalance joint control, and realize the adaptive imbalance control of the UAV flight attitude.

2. The method according to claim 1, wherein The performing reverse trajectory analysis on the current attitude state and the target attitude state to generate an ideal control trajectory sequence includes: Perform reverse path planning from the target attitude state to the current attitude state; Determine intermediate control nodes according to the reverse path planning, and perform time sequence marking on the intermediate control nodes; Obtain the attitude change gradient of the intermediate control nodes; Generate an ideal control trajectory sequence according to the time sequence marking and the attitude change gradient.

3. The method according to claim 1, characterized in that, The obtaining the imbalance points of the ideal control trajectory sequence includes: Analyze the attitude conversion key points in the ideal control trajectory sequence; Determine the available natural imbalance direction at the attitude conversion key points; Evaluate the controllability of the natural imbalance direction to generate a controllability evaluation result; Obtain the imbalance points according to the natural imbalance direction and the controllability evaluation result.

4. The method according to claim 1, characterized in that, The generating a controllable imbalance strategy library according to the disturbance injection positions includes: Establish a corresponding relationship between the disturbance amplitude and the recovery time for each disturbance injection position; Determine the optimal disturbance parameter combination according to the corresponding relationship; Classify and store the optimal disturbance parameter combination to generate a controllable imbalance strategy library.

5. The method according to claim 1, wherein The establishing a dynamic balance benchmark according to the disturbance injection sequence includes: Determine the expected recovery trajectory after imbalance according to the disturbance injection sequence; Generate a tolerance range based on the expected recovery trajectory; Determine the time reference point according to the expected recovery trajectory; Combine the expected recovery trajectory, the tolerance range and the time reference point to form a dynamic balance benchmark.

6. The method according to claim 1, wherein The performing trajectory reconstruction based on the dynamic balance benchmark to obtain an imbalance-recovery control trajectory includes: Perform reverse inference of the trajectory in the imbalance stage according to the dynamic balance benchmark to generate a reverse inference trajectory segment; Determine the imbalance-recovery conversion node based on the reverse trajectory segment; Perform trajectory splicing according to the conversion node and the dynamic balance reference to generate an imbalance-recovery control trajectory.

7. The method according to claim 1, wherein The establishment of the imbalance-recovery mapping rule according to the core imbalance control feature set includes: Identify beneficial imbalance patterns in the core imbalance control feature set; Extract the disturbance trigger feature parameters corresponding to the beneficial imbalance patterns; Establish the correlation between the disturbance trigger feature parameters and the balance recovery index; Generate the imbalance-recovery mapping rule according to the correlation.

8. The method according to claim 1, characterized in that The generation of the reverse imbalance joint control according to the adaptive mapping parameter includes: Establish a multi-channel imbalance control matrix based on the adaptive mapping parameter; Perform attitude coupling analysis according to the multi-channel imbalance control matrix to determine the reverse control gain of each axis; Adaptive fusion of the reverse control gain and the adaptive mapping parameter to generate a joint control strategy; Based on the joint control strategy, adjust the control output intensity in real time to form a reverse imbalance joint control.

9. An adaptive imbalance control device for the flight attitude of a drone, characterized in that, Include: A trajectory analysis module for obtaining the current attitude state and the target attitude state of the UAV, and performing reverse trajectory analysis on the current attitude state and the target attitude state to generate an ideal control trajectory sequence; An imbalance acquisition module for obtaining the imbalance points of the ideal control trajectory sequence, determining the disturbance injection positions according to the imbalance points, and generating a controllable imbalance strategy library according to the disturbance injection positions; A reference establishment module for determining the imbalance control interval according to the ideal control trajectory sequence and the controllable imbalance strategy library, generating a disturbance injection sequence according to the imbalance control interval, and establishing a dynamic balance reference according to the disturbance injection sequence; A trajectory reconstruction module for performing trajectory reconstruction based on the dynamic balance reference to obtain an imbalance-recovery control trajectory, and obtaining a dynamic response feature set corresponding to the imbalance-recovery control trajectory; A feature extraction module for obtaining the balance recovery index corresponding to the dynamic response feature set, constructing an imbalance recovery spatio-temporal manifold based on the balance recovery index, and performing feature mapping on the imbalance recovery spatio-temporal manifold to obtain a core imbalance control feature set; A mapping generation module for establishing an imbalance-recovery mapping rule according to the core imbalance control feature set, generating disturbance control parameters according to the imbalance-recovery mapping rule, and establishing an adaptive mapping parameter based on the disturbance control parameters; A control execution module for generating a reverse imbalance joint control according to the adaptive mapping parameter, and outputting a flight control signal according to the reverse imbalance joint control to realize the adaptive imbalance control of the UAV flight attitude.

10. A computer device, characterized in that, Include a memory and a processor; the memory is used to store a computer program; the processor is used to execute the computer program and implement the method according to any one of claims 1 to 8 when executing the computer program.

Images