A multi-degree-of-freedom flying manipulator aerial operation stability control method and system

Abstract

This invention discloses a method and system for stabilizing a multi-degree-of-freedom flying robotic arm during aerial operations, relating to the field of aircraft control technology. The method includes: acquiring the attitude information and current contact force of the flying robotic arm; constructing an attitude control model based on the attitude information; constructing an impedance model based on desired inertia, damping, stiffness matrix, end-effector pose, end-effector velocity, end-effector acceleration, end-effector reference trajectory, the current contact force, and the desired contact force; calculating joint control torques based on the impedance model; calculating a comprehensive performance index based on the attitude control model and the impedance model; and performing stable control of the flying robotic arm based on the comprehensive performance index and the joint control torques. This invention can combine the attitude control model and the impedance model to calculate the comprehensive performance index, thereby adjusting the control parameters of the flying robotic arm to achieve stable control, improving control accuracy and flight stability.

Description

Technical Field

[0001] This invention relates to the field of aircraft control technology, and in particular to a method and system for stabilizing aerial operations of a multi-degree-of-freedom flying robotic arm. Background Technology

[0002] Flying robotic arms consist of a multi-rotor aircraft and a multi-degree-of-freedom robotic arm mounted beneath it. Combining the maneuverability of a multi-rotor with the operational capabilities of a robotic arm, flying robotic arms play a crucial role in high-altitude operations. However, when the robotic arm performs aerial grasping, handling, and manipulation tasks, its motion and the contact forces with the environment can interfere with the flight platform, altering the system's dynamic characteristics and leading to center-of-mass shift and attitude instability. Existing model-based control methods rely on precise mathematical models incorporating coupled dynamics. However, in real-world systems, the total mass, center of mass, and inertia change in real time with the robotic arm's configuration and load, making accurate modeling difficult and resulting in performance degradation. Interference observation-based control methods depend on observer bandwidth and nominal model accuracy, but when system parameters undergo large and rapid time-varying changes, the estimation accuracy and convergence speed are insufficient, leading to low control precision and low flight stability.

[0003] In summary, the technical problems existing in the relevant technologies need to be improved. Summary of the Invention

[0004] The main objective of this invention is to propose a method and system for stable control of a multi-degree-of-freedom flying robotic arm in aerial operations. This method combines attitude control model and impedance model to calculate comprehensive performance indicators, thereby adjusting the control parameters of the flying robotic arm to achieve stable control and improve control accuracy and flight stability.

[0005] On one hand, embodiments of the present invention provide a method for stabilizing and controlling the aerial operation of a multi-degree-of-freedom flying robotic arm, comprising the following steps:

[0006] Acquire the attitude information and current contact force of the flying robotic arm;

[0007] Based on the attitude information, an attitude control model is constructed;

[0008] An impedance model is constructed based on the desired inertia, damping, stiffness matrix, end-effector pose, end-effector velocity, end-effector acceleration, end-effector reference trajectory, current contact force, and desired contact force.

[0009] Calculate the joint control torque based on the impedance model;

[0010] Calculate the comprehensive performance index based on the attitude control model and the impedance model;

[0011] Based on the comprehensive performance indicators and the joint control torque, the flight robotic arm is stabilized and controlled.

[0012] In some embodiments, constructing the attitude control model based on the attitude information includes:

[0013] Based on the attitude information, extract the attitude angles of the previous moment, the current moment, and the next moment;

[0014] Based on the attitude angle at the previous moment and the attitude angle at the current moment, calculate the estimated value of the pseudo-Jacobi matrix;

[0015] The attitude control torque at the current moment is calculated based on the attitude angle at the current moment and the estimated value of the pseudo-Jacobi matrix.

[0016] Calculate the attitude angle increment for the next moment based on the attitude angle at the current moment and the attitude angle at the next moment;

[0017] Calculate the attitude control torque increment at the current moment based on the attitude control torque at the previous moment and the attitude control torque at the current moment;

[0018] The attitude control model is constructed based on the attitude angle increment, pseudo-Jacobi matrix, and attitude control torque increment at the current moment.

[0019] The expression for the attitude control model is as follows:

[0020] ;

[0021] In the formula, This is the attitude angle increment at the next moment. It is a pseudo-Jacobi matrix. This represents the increment of the attitude control torque at the current moment. For a moment.

[0022] In some embodiments, calculating the pseudo-Jacobi matrix estimate based on the attitude angle at the previous time step and the attitude angle at the current time step includes:

[0023] Calculate the attitude angle increment at the current moment based on the attitude angle at the previous moment and the attitude angle at the current moment;

[0024] The pseudo-Jacobi matrix estimate is calculated based on the learning gain, the attitude control torque increment of the previous time step, and the attitude angle increment of the current time step.

[0025] The formula for calculating the pseudo-Jacobi matrix estimate is as follows:

[0026] ;

[0027] In the formula, The pseudo-Jacobi matrix is ​​estimated as follows. This is the matrix estimate from the previous time step. For learning gain, This is the increment of the attitude control torque from the previous moment. The attitude angle increment at the current moment. To prevent division by zero constant, For a moment.

[0028] In some embodiments, calculating the attitude control torque at the current moment based on the attitude angle at the current moment and the estimated value of the pseudo-Jacobi matrix includes:

[0029] Calculate the attitude tracking error based on the current attitude angle and the desired attitude.

[0030] The attitude control torque at the current moment is calculated based on the attitude tracking error, control gain, weighting factor, estimated value of pseudo-Jacobi matrix, and attitude control torque at the previous moment.

[0031] The formula for calculating the attitude control torque at the current moment is:

[0032] ;

[0033] In the formula, The attitude control torque at the current moment is... The attitude control torque at the previous moment. To control the gain, The pseudo-Jacobi matrix is ​​an estimate. As a weighting factor, Let be the attitude tracking error, and be the time interval.

[0034] In some embodiments, calculating the joint control torque based on the impedance model includes:

[0035] Based on the inertia matrix, centrifugal force matrix, gravity term, joint angle, joint velocity, joint acceleration, and total disturbance, construct the joint dynamics equations of the robotic arm;

[0036] A linear expansion state observation model is constructed based on joint angles and joint velocities;

[0037] The joint control torque is calculated based on the joint dynamics equation of the robotic arm, the linear expansion state observation model, and the impedance model.

[0038] The expression for the joint dynamics equation of the robotic arm is as follows:

[0039] ;

[0040] In the formula, The inertia matrix, For the centrifugal force matrix, For gravity, For joint angle, For joint velocity, For joint acceleration, For the total disturbance, For joint control torque.

[0041] In some embodiments, constructing a linear expansion state observation model based on joint angles and joint velocities includes:

[0042] Calculate the estimation error based on the estimated joint angle and the joint angle itself;

[0043] The first observation parameter is calculated based on the joint velocity estimate, the first observer gain, and the estimation error;

[0044] The second observation parameters are calculated based on the total disturbance estimate, the second observer gain, the estimation error, the inertia matrix estimate, the centrifugal force matrix estimate, the gravity term estimate, the joint angle, the joint velocity, and the joint control torque.

[0045] The third observation parameter is calculated based on the third observer gain and the estimation error.

[0046] Based on the estimation error, the first observation parameter, the second observation parameter, and the third observation parameter, the linear expansion state observation model is constructed.

[0047] The expression for the linearly extended state observation model is as follows:

[0048] ;

[0049] In the formula, To estimate the error, This is an estimated value for the joint angle. This is an estimate of the joint velocity. This is the estimated total disturbance. For the first observer gain, For the second observer gain, For the third observer gain, For joint angle, For joint velocity, The transpose of the inertia matrix estimate. For centrifugal force matrix estimation, For the estimation of the gravity term, For joint control torque, The first observation parameter, This is the second observation parameter. This is the third observation parameter.

[0050] In some embodiments, calculating the joint control torque based on the robotic arm joint dynamics equations, the linear expansion state observation model, and the impedance model includes:

[0051] Get the current joint state;

[0052] Based on the current joint state, the impedance model, and the Jacobian matrix, the inverse kinematics transformation of the robotic arm is performed to obtain the joint reference command;

[0053] The joint control torque is calculated based on the joint dynamics equations of the robotic arm, the linear expansion state observation model, the joint reference command, the positive definite gain matrix, and the total disturbance estimate.

[0054] The formula for calculating the joint control torque is as follows:

[0055] ;

[0056] In the formula, The joint control torque, For inertia matrix estimation, For centrifugal force matrix estimation, For the estimation of the gravity term, and All are positive definite gain matrices. For joint angle, For joint velocity, , and All are joint reference commands. This is the estimated total disturbance.

[0057] In some embodiments, calculating the comprehensive performance index based on the attitude control model and the impedance model includes:

[0058] The target tracking error is evaluated based on the attitude control model.

[0059] Based on the target tracking error, construct the attitude error vector;

[0060] Based on the attitude error vector, the attitude error norm is calculated using the Euclidean norm;

[0061] The contact force error is evaluated based on the impedance model.

[0062] Based on the contact force error, the contact force error norm is calculated using the Euclidean norm;

[0063] The comprehensive performance index is calculated based on the attitude error norm, the contact force error norm, and the weighting coefficient.

[0064] The formula for calculating the comprehensive performance index is as follows:

[0065] ;

[0066] In the formula, The aforementioned comprehensive performance index, These are the weighting coefficients. Let the attitude error norm be... Let the contact force error norm be... For the current contact force, For the desired contact force, For a moment.

[0067] In some embodiments, the step of performing flight robotic arm stability control based on the comprehensive performance index and the joint control torque includes:

[0068] Calculate the changing trend of the comprehensive performance index;

[0069] If the value of the comprehensive performance index is greater than the preset index threshold, or the change trend is greater than the preset trend threshold, then the gradient method is used to adjust the control parameters of the flight manipulator. The control parameters of the flight manipulator include the control gain in the attitude control model, the first observer gain, the second observer gain and the third observer gain in the linear extended state observation model, and the desired inertia, damping and stiffness matrices in the impedance model.

[0070] The flight robotic arm is controlled according to the adjusted control parameters of the flight robotic arm and the joint control torque.

[0071] On the other hand, embodiments of the present invention provide a multi-degree-of-freedom flying robotic arm aerial operation stability control system, including:

[0072] The information acquisition module is used to acquire the attitude information and current contact force of the flying robotic arm;

[0073] An attitude control model construction module is used to construct an attitude control model based on the attitude information.

[0074] The impedance model construction module is used to construct an impedance model based on the desired inertia, damping, stiffness matrix, end-effector pose, end-effector velocity, end-effector acceleration, end-effector reference trajectory, the current contact force, and the desired contact force.

[0075] The joint control torque calculation module is used to calculate the joint control torque based on the impedance model.

[0076] The comprehensive performance index calculation module is used to calculate the comprehensive performance index based on the attitude control model and the impedance model.

[0077] The stability control module is used to perform stability control of the flight robotic arm based on the comprehensive performance indicators and the joint control torque.

[0078] The embodiments of this application include at least the following beneficial effects: First, the embodiments of this application acquire the attitude information and current contact force of the flying robotic arm. Based on the attitude information, an attitude control model is constructed. Then, based on the desired inertia, damping, stiffness matrix, end-effector pose, end-effector velocity, end-effector acceleration, end-effector reference trajectory, the current contact force, and the desired contact force, an impedance model is constructed. Based on the impedance model, the joint control torque is calculated. Then, based on the attitude control model and the impedance model, a comprehensive performance index is calculated. Finally, based on the comprehensive performance index and the joint control torque, the flying robotic arm is stabilized and controlled. This allows for the calculation of comprehensive performance index by combining the attitude control model and the impedance model, thereby adjusting the control parameters of the flying robotic arm to achieve stable control, improving control accuracy and flight stability.

[0079] Other features and advantages of the invention will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures particularly pointed out in the description and the drawings. Attached Figure Description

[0080] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0081] Figure 1 This is a flowchart illustrating a method for stabilizing aerial operations of a multi-degree-of-freedom flying robotic arm, according to an embodiment of the present invention.

[0082] Figure 2 This is a schematic diagram of the structure of a multi-degree-of-freedom flying robotic arm aerial operation stability control system according to an embodiment of the present invention. Detailed Implementation

[0083] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of this application and are not intended to limit it. In the following description, when referring to the accompanying drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with those of this application; they are merely examples of systems and methods consistent with some aspects of the embodiments of this application detailed in the summary of the invention.

[0084] It is understood that the terms “first,” “second,” etc., used in this application may be used herein to describe various concepts, but unless otherwise stated, these concepts are not limited by these terms. These terms are only used to distinguish one concept from another. For example, without departing from the scope of the embodiments of this application, first information may also be referred to as second information, and similarly, second information may also be referred to as first information. Depending on the context, the words “if,” “when,” or “in response to a determination” as used herein may be interpreted as “when…” or “when…” or “in response to a determination.”

[0085] As used in this application, the terms "at least one", "multiple", "each", "any", etc., "at least one" includes one, two or more, "multiple" includes two or more, "each" refers to each of the corresponding multiples, and "any" refers to any one of the multiples.

[0086] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing embodiments of this application only and is not intended to limit this application.

[0087] Before providing a detailed description of the embodiments of this application, some of the nouns and terms involved in the embodiments of this application will be explained first. The nouns and terms involved in the embodiments of this application are subject to the following interpretations.

[0088] A multi-degree-of-freedom (DOF) flying robotic arm is a robotic system capable of flight and flexible movement in three-dimensional space. It achieves complex movements similar to a human arm through multiple independent rotational or translational joints. "Multi-degree-of-freedom" refers to the robotic arm having multiple degrees of freedom (e.g., 6 or 7), each corresponding to an independent direction of motion. This allows the robotic arm to precisely control its position and posture, such as simulating combinations of shoulder, elbow, and wrist movements. The core characteristics of a multi-DOF flying robotic arm are its flexibility and autonomy. It typically integrates end effectors (such as grippers or tool interfaces) and autonomous control algorithms, enabling it to independently complete tasks in complex environments such as on the ground or in space, such as grasping objects, performing operations, or assisting human work.

[0089] In related technologies, flying robotic arms consist of a multi-rotor aircraft and a multi-degree-of-freedom robotic arm mounted beneath it. Combining the maneuverability of a multi-rotor with the operational capabilities of a robotic arm, flying robotic arms play a crucial role in high-altitude operations. However, when the robotic arm performs aerial grasping, handling, and manipulation tasks, the arm's movement and its contact forces with the environment can interfere with the flying platform, altering the system's dynamic characteristics and leading to center-of-gravity shift and attitude instability.

[0090] Existing control methods have the following main shortcomings:

[0091] (1) Model-based control methods (such as model predictive control and backstepping) heavily rely on accurate mathematical models that include coupled dynamics. In actual systems, the total mass, center of mass, and moment of inertia change in real time with the configuration of the robotic arm and the load, making accurate modeling difficult and model mismatch leading to performance degradation.

[0092] (2) Control methods based on disturbance observation (such as disturbance observers) rely on the observer bandwidth and the accuracy of the nominal model for performance. When system parameters change rapidly and significantly, the estimation accuracy and convergence speed are insufficient.

[0093] (3) Learning-based control methods (such as neural network adaptive control) often suffer from problems such as slow convergence, need for a large amount of training data, and poor real-time performance.

[0094] Existing control methods are highly dependent on the precise mathematical model or parameters of the system, making it difficult to achieve fast and robust adaptive stability control in complex operating scenarios where model parameters are unknown, time-varying, and subject to unmeasurable contact disturbances. This results in low control accuracy and low flight stability.

[0095] In view of this, this application provides a method and system for stabilizing the aerial operation of a multi-degree-of-freedom flying robotic arm. First, this application designs an attitude control model based on compact dynamic linearization to estimate the time-varying pseudo-Jacobi matrix online to compensate for unknown dynamic changes. Second, it designs an adaptive impedance model based on an extended state observer to estimate and compensate for the contact force of the robotic arm and the flight-robotic arm coupling interference in real time. Finally, it adjusts the controller parameters online through an adaptive coordinator. This application does not require precise system parameters and can effectively ensure the autonomous and stable operation of the flying robotic arm in complex aerial operations.

[0096] This application provides a method for stabilizing and controlling a multi-degree-of-freedom (DOF) robotic arm during aerial operations, relating to the field of aircraft control technology. This method can be applied to a terminal, a server, or software running on either a terminal or a server. In some embodiments, the terminal can be a smartphone, tablet, laptop, desktop computer, smart speaker, smartwatch, or vehicle-mounted terminal, but is not limited to these. The server can be configured as an independent physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, CDN, and big data and artificial intelligence platforms. The server can also be a node server in a blockchain network. The software can be an application implementing a stabilizing and controlling method for a multi-degree-of-freedom robotic arm during aerial operations, but is not limited to the above forms.

[0097] This application can be used in a wide variety of general-purpose or special-purpose computer system environments or configurations. Examples include: personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronics, network PCs, minicomputers, mainframe computers, and distributed computing environments including any of the above systems or devices. This application can be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc., that perform specific tasks or implement specific abstract data types. This application can also be practiced in distributed computing environments where tasks are performed by remote processing devices connected via a communication network. In distributed computing environments, program modules can reside in local and remote computer storage media, including storage devices.

[0098] The embodiments of this application will be explained in detail below with reference to the accompanying drawings:

[0099] Figure 1 This is an optional flowchart of a multi-degree-of-freedom flying robotic arm aerial operation stability control method provided in the embodiments of this application. Figure 1 The method may include, but is not limited to, steps S101 to S106.

[0100] Step S101: Obtain the attitude information and current contact force of the flying robotic arm;

[0101] Step S102: Construct an attitude control model based on the attitude information;

[0102] Step S103: Construct an impedance model based on the desired inertia, damping, stiffness matrix, end-effector pose, end-effector velocity, end-effector acceleration, end-effector reference trajectory, current contact force, and desired contact force;

[0103] Step S104: Calculate the joint control torque based on the impedance model;

[0104] Step S105: Calculate the comprehensive performance index based on the attitude control model and impedance model;

[0105] Step S106: Perform stable control of the flight robotic arm based on comprehensive performance indicators and joint control torque.

[0106] Steps S101 to S106 as shown in the embodiments of this application can combine the attitude control model and the impedance model to calculate the comprehensive performance index, and adjust the control parameters of the flight manipulator to achieve stable control of the flight manipulator, thereby improving control accuracy and flight stability.

[0107] In some embodiments, steps S101-S106 can first acquire the attitude information and current contact force of the flying robotic arm. Attitude information (such as position, orientation, and angle) can be acquired through sensors. For example, an IMU (Inertial Measurement Unit), including a gyroscope (measuring angular velocity), an accelerometer (measuring linear acceleration), and a magnetometer (measuring magnetic field direction), can be directly mounted on the robotic arm joints or base. Vision or laser systems, such as cameras, depth cameras, or LiDAR, can also be used to calibrate the attitude by capturing environmental features or marker points. External positioning devices, such as GPS modules (suitable for outdoor use) or UWB (Ultra-Wideband) positioning systems, can also be used for absolute position tracking. Simultaneously, the current contact force of the flying robotic arm (such as the push-pull force or torque of the end effector) can be acquired through sensors. For example, a torque sensor (such as a six-axis sensor) can be used, mounted at the end of the robotic arm or joints, to measure the contact force and torque in real time. Strain gauges or tactile sensors can also be used to calculate the force by measuring minute deformations of the robotic arm structure.

[0108] Then, based on the attitude information, an attitude control model is constructed. The attitude angles at different times can be extracted from the attitude information, the attitude angle increments at different times can be calculated, the pseudo-Jacobi matrix can be estimated, and the attitude control torque increments and the pseudo-Jacobi matrix estimates can be combined to construct the attitude control model.

[0109] An impedance model is then constructed based on the desired inertia, damping, stiffness matrix, end-effector pose, end-effector velocity, end-effector acceleration, end-effector reference trajectory, current contact force, and desired contact force. The desired inertia, damping, and stiffness matrix are system parameters, which can be obtained through system parameter initialization or by using default values. The end-effector pose, end-effector velocity, and end-effector acceleration can be acquired through end-effector sensors. The end-effector reference trajectory is a pre-set motion trajectory used to plan the motion of the end-effector. The desired contact force is a pre-set value used to adjust the current contact force to achieve the desired contact force. The desired contact force can be dynamically adjusted based on the material properties of the robotic arm and the workpiece. The expression for the impedance model is: In the formula, For expected inertia, For damping, Here is the stiffness matrix. This is the end-effector pose. For the terminal velocity, For terminal acceleration, , and All are end-point reference trajectories. For the current contact force, The desired contact force is represented by the impedance model, which is used to calculate the end-effector acceleration command and then converted into a reference command in joint space via inverse kinematics, serving as a prerequisite for subsequent calculations of joint control torque.

[0110] Finally, the joint control torque is calculated based on the impedance model. The joint control force can be calculated by constructing the joint dynamics equations and the linear expansion state observation model of the robotic arm. Simultaneously, the comprehensive performance index is calculated based on the attitude control model and the impedance model. The attitude error norm can be calculated by constructing an attitude error vector, the contact force error norm can be calculated by evaluating the contact force error, and the comprehensive performance index can be calculated by combining weighting coefficients. Further, the flight robotic arm is stabilized and controlled based on the comprehensive performance index and the joint control torque. By judging the changing trend of the comprehensive performance index, the control parameters of the flight robotic arm can be adjusted using the gradient method, and combined with the joint control torque, the flight robotic arm can be controlled to achieve stable operation.

[0111] In some embodiments, in step S102, constructing an attitude control model based on the attitude information may include, but is not limited to, the following steps:

[0112] Step S201: Based on the attitude information, extract the attitude angles of the previous moment, the current moment, and the next moment;

[0113] Step S202: Calculate the estimated value of the pseudo-Jacobi matrix based on the attitude angle at the previous moment and the attitude angle at the current moment;

[0114] Step S203: Calculate the attitude control torque at the current moment based on the attitude angle and the estimated value of the pseudo-Jacobi matrix at the current moment;

[0115] Step S204: Calculate the attitude angle increment for the next moment based on the attitude angle at the current moment and the attitude angle at the next moment;

[0116] Step S205: Calculate the attitude control torque increment at the current moment based on the attitude control torque at the previous moment and the attitude control torque at the current moment.

[0117] Step S206: Construct an attitude control model based on the attitude angle increment at the next moment, the pseudo-Jacobi matrix, and the attitude control torque increment at the current moment.

[0118] In some embodiments, the attitude angles from the previous moment can be extracted first based on the attitude information. Attitude angle at the current moment and the attitude angle at the next moment The pseudo-Jacobi matrix estimate is calculated based on the attitude angles at the previous and current moments. Alternatively, the pseudo-Jacobi matrix estimate can be calculated by taking the attitude angle increment at the current moment and combining it with the learning gain.

[0119] Then, based on the current attitude angle and the estimated pseudo-Jacobi matrix, the attitude control torque at the current moment is calculated. The attitude control torque at the current moment can be calculated by taking the attitude tracking error into account, combined with the control gain and weighting factors.

[0120] Then, based on the attitude angle at the current moment and the attitude angle at the next moment, calculate the attitude angle increment at the next moment. The formula for calculating the attitude angle increment at the next moment is: In the formula, This is the attitude angle increment for the next moment. Let this be the attitude angle at the next moment. The attitude angle at the current moment.

[0121] Based on the attitude control torque at the previous moment and the attitude control torque at the current moment, the increment of the attitude control torque at the current moment is calculated. The formula for calculating the increment of the attitude control torque at the current moment is: In the formula, This represents the increment of the attitude control torque at the current moment. The attitude control torque at the current moment. This is the attitude control torque from the previous moment.

[0122] Finally, based on the attitude angle increment at the next moment, the pseudo-Jacobi matrix, and the attitude control torque increment at the current moment, an attitude control model is constructed. The expression for the attitude control model is: In the formula, This is the attitude angle increment for the next moment. It is a pseudo-Jacobi matrix. This represents the increment of the attitude control torque at the current moment. For the moment. It is understandable that the attitude control model, as the foundation of attitude control, directly determines the attitude error component in the overall system performance index. This index, in turn, adjusts the parameters of the attitude control model through feedback from the adaptive coordinator. The two form a dynamic correlation and closed-loop collaboration, jointly ensuring the stability and operational performance of the flying robotic arm.

[0123] In some embodiments, in step S202, calculating the pseudo-Jacobi matrix estimate based on the attitude angle at the previous moment and the attitude angle at the current moment may include, but is not limited to, the following steps:

[0124] Calculate the attitude angle increment at the current moment based on the attitude angle at the previous moment and the attitude angle at the current moment;

[0125] The pseudo-Jacobi matrix estimate is calculated based on the learning gain, the attitude control torque increment of the previous time step, and the attitude angle increment of the current time step.

[0126] In some embodiments, the attitude angle increment at the current moment can be calculated first based on the attitude angle at the previous moment and the attitude angle at the current moment. The formula for calculating the attitude angle increment at the current moment is: In the formula, This represents the attitude angle increment at the current moment. The attitude angle at the current moment. The attitude angle at the previous moment.

[0127] Then, based on the learning gain, the attitude control torque increment of the previous moment, and the attitude angle increment of the current moment, the pseudo-Jacobi matrix estimate is calculated. The formula for calculating the pseudo-Jacobi matrix estimate is as follows: In the formula, This is an estimate of the pseudo-Jacobi matrix. This is the matrix estimate from the previous time step. For the learning gain, η ∈ (0,2], This is the increment of the attitude control torque from the previous moment. This represents the attitude angle increment at the current moment. To prevent division by zero constant, For a moment.

[0128] In some embodiments, in step S203, calculating the attitude control torque at the current moment based on the current attitude angle and the estimated pseudo-Jacobi matrix may include, but is not limited to, the following steps:

[0129] Calculate the attitude tracking error based on the current attitude angle and the desired attitude.

[0130] The attitude control torque at the current moment is calculated based on the attitude tracking error, control gain, weighting factor, pseudo-Jacobi matrix estimate, and attitude control torque at the previous moment.

[0131] In some embodiments, the attitude tracking error can be calculated first based on the current attitude angle and the desired attitude, wherein the formula for calculating the attitude tracking error is: In the formula, For attitude tracking error, As desired, This represents the attitude angle at the current moment. The desired attitude can be set to 0.

[0132] Then, based on the attitude tracking error, control gain, weighting factor, pseudo-Jacobi matrix estimate, and the attitude control torque from the previous time step, the attitude control torque at the current time step is calculated. The formula for calculating the attitude control torque at the current time step is: In the formula, The attitude control torque at the current moment. The attitude control torque at the previous moment. To control the gain, , This is an estimate of the pseudo-Jacobi matrix. As a weighting factor, , For attitude tracking error, For a moment.

[0133] In some embodiments, step S104, calculating the joint control torque based on the impedance model, may include, but is not limited to, the following steps:

[0134] Step S301: Construct the joint dynamics equations of the robotic arm based on the inertia matrix, centrifugal force matrix, gravity term, joint angle, joint velocity, joint acceleration, and total disturbance;

[0135] Step S302: Construct a linear expansion state observation model based on joint angles and joint velocities;

[0136] Step S303: Calculate the joint control torque based on the joint dynamics equation of the robotic arm, the linear expansion state observation model, and the impedance model.

[0137] In some embodiments, the joint dynamics equations of the robotic arm can be constructed first based on the inertia matrix, centrifugal force matrix, gravity term, joint angle, joint velocity, joint acceleration, and total disturbance. The expression for the joint dynamics equations of the robotic arm is as follows: In the formula, The inertia matrix, For the centrifugal force matrix, For gravity, For joint angle, For joint velocity, For joint acceleration, For the total disturbance, For joint control torque. The total disturbance may include unmodeled dynamics, joint friction, and flight-manipulator coupling disturbances.

[0138] Then, based on the joint angles and joint velocities, a linear expansion state observation model is constructed. Multiple observation parameters can be calculated and integrated by calculating the estimation error to construct the linear expansion state observation model. Finally, based on the robot arm joint dynamics equations, the linear expansion state observation model, and the impedance model, the joint control torque is calculated. The joint reference command can be obtained by performing an inverse kinematic transformation of the robot arm using the current joint state, and then combined with the positive definite gain matrix to calculate the joint control torque.

[0139] Understandably, the impedance model defines the desired impedance behavior (force-motion relationship) of the robotic arm's end effector. The calculated end-effector acceleration command is converted into reference commands in joint space via inverse kinematics; these commands are precisely the tracking targets in the joint control torque calculation. The robotic arm joint dynamics equations describe the actual joint dynamics containing unknown disturbances. The linear extended state observation model estimates this disturbance in real time and feeds its estimate forward to compensate for it in the joint control torque, thereby ensuring that the robotic arm can accurately and smoothly achieve the impedance control target specified by the impedance model. Ultimately, the error between the current contact force and the desired contact force of the robotic arm directly constitutes the contact force error in the subsequent system comprehensive performance indicators. Therefore, the impedance model, the robotic arm joint dynamics equations, the linear extended state observation model, and the joint control torque calculation are interconnected through a closed-loop logic of "target definition—dynamic description—disturbance observation—feedback compensation."

[0140] In some embodiments, in step S302, constructing a linear expansion state observation model based on joint angles and joint velocities may include, but is not limited to, the following steps:

[0141] Calculate the estimation error based on the estimated joint angle and the joint angle itself;

[0142] The first observation parameter is calculated based on the joint velocity estimate, the first observer gain, and the estimation error.

[0143] The second observation parameters are calculated based on the total disturbance estimate, the second observer gain, the estimation error, the inertia matrix estimate, the centrifugal force matrix estimate, the gravity term estimate, the joint angle, the joint velocity, and the joint control torque.

[0144] Calculate the third observation parameters based on the third observer gain and estimation error;

[0145] Based on the estimation error, the first observation parameter, the second observation parameter, and the third observation parameter, a linear extended state observation model is constructed.

[0146] In some embodiments, the estimation error can be calculated first based on the estimated joint angle and the joint angle itself. The formula for calculating the estimation error is as follows: In the formula, To estimate the error, This is an estimated value for the joint angle. This refers to the joint angle.

[0147] Then, based on the joint velocity estimate, the first observer gain, and the estimation error, the first observation parameter is calculated. The formula for calculating the first observation parameter is as follows: Simultaneously, based on the total disturbance estimate, second observer gain, estimation error, inertia matrix estimate, centrifugal force matrix estimate, gravity term estimate, joint angle, joint velocity, and joint control torque, the second observation parameters are calculated. The formula for calculating the second observation parameters is as follows: Based on the gain and estimation error of the third observer, the third observation parameters are calculated, where the formula for calculating the third observation parameters is: .

[0148] Based on the estimation error, the first observation parameter, the second observation parameter, and the third observation parameter, a linear extended state observation model is constructed. The expression for the linear extended state observation model is:

[0149] ;

[0150] In the formula, To estimate the error, This is an estimated value for the joint angle. This is an estimate of the joint velocity. This is the estimated total disturbance. For the first observer gain, For the second observer gain, For the third observer gain, For joint angle, For joint velocity, The transpose of the inertia matrix estimate. For centrifugal force matrix estimation, For the estimation of the gravity term, For joint control torque, The first observation parameter, This is the second observation parameter. This is the third observation parameter.

[0151] Understandably, the linear extended state observation model can observe and quantify the unmodeled dynamics of the system, joint friction, and critical flight-manipulator coupling disturbances in real time. These estimates are directly input as feedforward compensation quantities into the calculation of joint control torque to actively cancel out disturbances, thereby ensuring that the manipulator can accurately perform the impedance tasks defined by the impedance model, and ultimately optimize the overall system performance indicators by minimizing contact force errors.

[0152] In some embodiments, step S303, calculating the joint control torque based on the robotic arm joint dynamics equation, the linear expansion state observation model, and the impedance model, may include, but is not limited to, the following steps:

[0153] Get the current joint state;

[0154] Based on the current joint state, impedance model, and Jacobian matrix, perform inverse kinematic transformation of the robotic arm to obtain joint reference commands;

[0155] The joint control torque is calculated based on the joint dynamics equations of the robotic arm, the linear expansion state observation model, the joint reference command, the positive definite gain matrix, and the total disturbance estimate.

[0156] In some embodiments, the current joint state can be obtained first. Based on the current joint state, impedance model, and Jacobian matrix, inverse kinematics transformation of the robotic arm can be performed to obtain joint reference commands. It is understood that, based on the impedance model, the end-effector pose, end-effector velocity, and end-effector acceleration of the robotic arm can be calculated. Subsequently, through inverse kinematics transformation of the robotic arm, combined with the current joint state and Jacobian matrix, relevant parameters of the robotic arm joints can be converted to obtain complete joint reference commands.

[0157] Then, based on the joint dynamics equations of the robotic arm, the linear expansion state observation model, the joint reference command, the positive definite gain matrix, and the total disturbance estimate, the joint control torque is calculated. The formula for calculating the joint control torque is as follows: In the formula, For joint control torque, For inertia matrix estimation, For centrifugal force matrix estimation, For the estimation of the gravity term, and All are positive definite gain matrices. For joint angle, For joint velocity, , and All are joint reference commands. This is the estimated total disturbance.

[0158] In some embodiments, step S105, calculating the comprehensive performance index based on the attitude control model and the impedance model, may include, but is not limited to, the following steps:

[0159] Based on the attitude control model, evaluate the target tracking error;

[0160] Construct an attitude error vector based on the target tracking error;

[0161] The attitude error norm is calculated using the Euclidean norm based on the attitude error vector.

[0162] Evaluate the contact force error based on the impedance model;

[0163] The contact force error norm is calculated using the Euclidean norm based on the contact force error.

[0164] The comprehensive performance index is calculated based on the attitude error norm, contact force error norm, and weighting coefficient.

[0165] In some embodiments, the target tracking error can be evaluated first based on the attitude control model, and an attitude error vector can be constructed based on the target tracking error, wherein the expression for the attitude error vector is: In the formula, Let be the attitude error vector. , and The target tracking error is defined as follows: The attitude error norm is then calculated using the Euclidean norm based on the attitude error vector. The formula for calculating the attitude error norm is: In the formula, Let be the attitude error norm. It can be understood that the attitude error norm directly represents the overall degree of deviation of the flight platform's attitude from the desired state, and serves as a key input quantity for the overall system performance evaluation and coordination optimization of formula (8).

[0166] Next, based on the impedance model, the contact force error is evaluated, and based on the contact force error, the Euclidean norm is used to calculate the contact force error norm. Finally, based on the attitude error norm, contact force error norm, and weighting coefficients, the comprehensive performance index is calculated. The formula for calculating the comprehensive performance index is as follows: In the formula, For comprehensive performance indicators, These are the weighting coefficients. Let the attitude error norm be... Let the contact force error norm be... For the current contact force, For the desired contact force, For a moment.

[0167] In some embodiments, step S106, based on comprehensive performance indicators and joint control torque, performs stability control of the flight robotic arm, which may include, but is not limited to, the following steps:

[0168] Calculate the changing trend of comprehensive performance indicators;

[0169] If the value of the comprehensive performance index is greater than the preset index threshold, or the trend of change is greater than the preset trend threshold, the gradient method is used to adjust the control parameters of the flight manipulator. The control parameters of the flight manipulator include the control gain in the attitude control model, the first observer gain, the second observer gain and the third observer gain in the linear extended state observation model, and the desired inertia, damping and stiffness matrices in the impedance model.

[0170] The flight robotic arm is controlled based on the adjusted control parameters and joint control torque.

[0171] In some embodiments, the changing trend of the comprehensive performance index can be calculated first, and then the value and changing trend of the comprehensive performance index can be judged. If the value of the comprehensive performance index is greater than a preset index threshold, or the changing trend is greater than a preset trend threshold, it indicates that the system stability or operational accuracy has decreased. The gradient method can be used to adjust the control parameters of the flight manipulator. The control parameters of the flight manipulator include the control gain in the attitude control model, the first observer gain, the second observer gain, and the third observer gain in the linear extended state observation model, and the desired inertia, damping, and stiffness matrices in the impedance model. The average value of the performance index can be calculated as the preset index threshold by statistical analysis of a large amount of historical flight data, and the average value of the changing trend can be calculated as the preset trend threshold. Then, based on the adjusted control parameters of the flight manipulator and the joint control torque, the flight manipulator is controlled to achieve an adaptive balance between maintaining flight stability and achieving precise and compliant operation.

[0172] In some embodiments, this embodiment applies a compact form dynamic linearization-based attitude control model to the attitude stabilization of a flight robotic arm. By estimating the pseudo-Jacobi matrix online in real time, it directly uses data-driven compensation to cover all unknown dynamic changes (such as attitude control torque) caused by the robotic arm's motion, load changes, and external contact, without requiring any prior parameters such as mass, inertia, or center of mass, thus achieving "plug-and-play" functionality.

[0173] This embodiment innovatively integrates the extended state observer with impedance control (such as the joint dynamics equations of the robotic arm and the linear extended state observation model). The extended state observer observes the base coupling interference caused by the flight platform motion in real time as part of the total disturbance, and enables the robotic arm controller to actively cancel this interference through feedforward compensation (such as the calculation formula of the joint control torque), thereby significantly reducing the reaction force on the flight platform from the root and realizing dynamic decoupling during the operation.

[0174] This embodiment constructs a two-layer adaptive architecture consisting of a "model-free adaptive attitude loop" and an "observer-based adaptive impedance loop," and introduces an upper-layer adaptive coordinator (such as the calculation formula for the comprehensive performance index) for global performance optimization. This design enables the system to quickly respond to local dynamic changes while intelligently balancing the conflict between flight stability and operational accuracy.

[0175] This embodiment is based solely on system input and output data, exhibiting strong robustness to model errors and various disturbances. Its algorithm structure is clear, and its computational burden is moderate. It is particularly suitable for real-world, complex operating environments with unknown parameters, time-varying loads, and unmeasurable contact forces, thus solving a key bottleneck in the engineering application of this technology. This embodiment provides an efficient, robust, and easily implemented solution for the stable aerial operation of flying robotic arms, demonstrating significant innovation and application value.

[0176] The beneficial effects of implementing the embodiments of the present invention include: First, the embodiments of this application obtain the attitude information and current contact force of the flying robotic arm. Based on the attitude information, an attitude control model is constructed. Then, based on the desired inertia, damping, stiffness matrix, end-effector pose, end-effector velocity, end-effector acceleration, end-effector reference trajectory, current contact force, and desired contact force, an impedance model is constructed. Based on the impedance model, the joint control torque is calculated. Then, based on the attitude control model and the impedance model, the comprehensive performance index is calculated. Finally, based on the comprehensive performance index and the joint control torque, the flying robotic arm is stabilized and controlled. Thus, the comprehensive performance index can be calculated by combining the attitude control model and the impedance model to adjust the control parameters of the flying robotic arm, thereby achieving stable control of the flying robotic arm and improving control accuracy and flight stability.

[0177] like Figure 2 As shown, this embodiment of the invention also provides a multi-degree-of-freedom flying robotic arm aerial operation stability control system, including:

[0178] Information acquisition module 401 is used to acquire the attitude information and current contact force of the flying robotic arm;

[0179] The attitude control model construction module 402 is used to construct an attitude control model based on attitude information.

[0180] Impedance model construction module 403 is used to construct an impedance model based on desired inertia, damping, stiffness matrix, end pose, end velocity, end acceleration, end reference trajectory, current contact force and desired contact force;

[0181] The joint control torque calculation module 404 is used to calculate the joint control torque based on the impedance model.

[0182] The comprehensive performance index calculation module 405 is used to calculate the comprehensive performance index based on the attitude control model and the impedance model.

[0183] The stability control module 406 is used to perform stability control of the flight robotic arm based on comprehensive performance indicators and joint control torque.

[0184] The content of the above method embodiments is applicable to this system embodiment. The specific functions implemented in this system embodiment are the same as those in the above method embodiments, and the beneficial effects achieved are also the same as those achieved in the above method embodiments.

[0185] The embodiments described in this application are for the purpose of more clearly illustrating the technical solutions of the embodiments of this application, and do not constitute a limitation on the technical solutions provided by the embodiments of this application. As those skilled in the art will know, with the evolution of technology and the emergence of new application scenarios, the technical solutions provided by the embodiments of this application are also applicable to similar technical problems.

[0186] Those skilled in the art will understand that the technical solutions shown in the figures do not constitute a limitation on the embodiments of this application, and may include more or fewer steps than shown, or combine certain steps, or different steps.

[0187] The preferred embodiments of the present application have been described above with reference to the accompanying drawings, but this is not intended to limit the scope of the embodiments of the present application. Any modifications, equivalent substitutions, and improvements made by those skilled in the art without departing from the scope and substance of the embodiments of the present application shall be within the scope of the embodiments of the present application.

Claims

1. A method for stabilizing and controlling the aerial operation of a multi-degree-of-freedom flying robotic arm, characterized in that, Includes the following steps: Acquire the attitude information and current contact force of the flying robotic arm; Based on the attitude information, an attitude control model is constructed; An impedance model is constructed based on the desired inertia, damping, stiffness matrix, end-effector pose, end-effector velocity, end-effector acceleration, end-effector reference trajectory, current contact force, and desired contact force. Calculate the joint control torque based on the impedance model; Calculate the comprehensive performance index based on the attitude control model and the impedance model; Based on the comprehensive performance indicators and the joint control torque, the flight robotic arm is stabilized and controlled. The step of constructing an attitude control model based on the attitude information includes: Based on the attitude information, extract the attitude angles of the previous moment, the current moment, and the next moment; Based on the attitude angle at the previous moment and the attitude angle at the current moment, calculate the estimated value of the pseudo-Jacobi matrix; The attitude control torque at the current moment is calculated based on the attitude angle at the current moment and the estimated value of the pseudo-Jacobi matrix. Calculate the attitude angle increment for the next moment based on the attitude angle at the current moment and the attitude angle at the next moment; Calculate the attitude control torque increment at the current moment based on the attitude control torque at the previous moment and the attitude control torque at the current moment; The attitude control model is constructed based on the attitude angle increment, pseudo-Jacobi matrix, and attitude control torque increment at the current moment. The expression for the attitude control model is as follows: ； In the formula, This is the attitude angle increment at the next moment. It is a pseudo-Jacobi matrix. This represents the increment of the attitude control torque at the current moment. For a moment.

2. The method according to claim 1, characterized in that, The step of calculating the pseudo-Jacobi matrix estimate based on the attitude angle at the previous moment and the attitude angle at the current moment includes: Calculate the attitude angle increment at the current moment based on the attitude angle at the previous moment and the attitude angle at the current moment; The pseudo-Jacobi matrix estimate is calculated based on the learning gain, the attitude control torque increment of the previous time step, and the attitude angle increment of the current time step. The formula for calculating the pseudo-Jacobi matrix estimate is as follows: ； In the formula, The pseudo-Jacobi matrix is ​​estimated as follows. This is the matrix estimate from the previous time step. For learning gain, This is the increment of the attitude control torque from the previous moment. The attitude angle increment at the current moment. To prevent division by zero constant, For a moment.

3. The method according to claim 2, characterized in that, The step of calculating the attitude control torque at the current moment based on the attitude angle at the current moment and the estimated value of the pseudo-Jacobi matrix includes: Calculate the attitude tracking error based on the current attitude angle and the desired attitude. The attitude control torque at the current moment is calculated based on the attitude tracking error, control gain, weighting factor, estimated value of pseudo-Jacobi matrix, and attitude control torque at the previous moment. The formula for calculating the attitude control torque at the current moment is: ； In the formula, The attitude control torque at the current moment is... The attitude control torque at the previous moment. To control the gain, The pseudo-Jacobi matrix is ​​estimated as follows. As a weighting factor, The attitude tracking error is... For a moment.

4. The method according to claim 1, characterized in that, The calculation of the joint control torque based on the impedance model includes: Based on the inertia matrix, centrifugal force matrix, gravity term, joint angle, joint velocity, joint acceleration, and total disturbance, construct the joint dynamics equations of the robotic arm; A linear expansion state observation model is constructed based on joint angles and joint velocities; The joint control torque is calculated based on the joint dynamics equation of the robotic arm, the linear expansion state observation model, and the impedance model. The expression for the joint dynamics equation of the robotic arm is as follows: ； In the formula, The inertia matrix, For the centrifugal force matrix, For gravity, For joint angle, For joint velocity, For joint acceleration, For the total disturbance, For joint control torque.

5. The method according to claim 4, characterized in that, The process of constructing a linear expansion state observation model based on joint angles and joint velocities includes: Calculate the estimation error based on the estimated joint angle and the joint angle itself; The first observation parameter is calculated based on the joint velocity estimate, the first observer gain, and the estimation error; The second observation parameters are calculated based on the total disturbance estimate, the second observer gain, the estimation error, the inertia matrix estimate, the centrifugal force matrix estimate, the gravity term estimate, the joint angle, the joint velocity, and the joint control torque. The third observation parameter is calculated based on the third observer gain and the estimation error. Based on the estimation error, the first observation parameter, the second observation parameter, and the third observation parameter, the linear expansion state observation model is constructed. The expression for the linearly extended state observation model is as follows: ； In the formula, To estimate the error, This is an estimated value for the joint angle. This is an estimate of the joint velocity. This is the estimated total disturbance. For the first observer gain, For the second observer gain, For the third observer gain, For joint angle, For joint velocity, The transpose of the inertia matrix estimate. For centrifugal force matrix estimation, For the estimation of the gravity term, For joint control torque, The first observation parameter, This is the second observation parameter. This is the third observation parameter.

6. The method according to claim 4, characterized in that, The calculation of the joint control torque based on the joint dynamics equations of the robotic arm, the linear expansion state observation model, and the impedance model includes: Get the current joint state; Based on the current joint state, the impedance model, and the Jacobian matrix, the inverse kinematics transformation of the robotic arm is performed to obtain the joint reference command; The joint control torque is calculated based on the joint dynamics equation of the robotic arm, the linear expansion state observation model, the joint reference command, the positive definite gain matrix, and the total disturbance estimate. The formula for calculating the joint control torque is as follows: ； In the formula, The joint control torque, For inertia matrix estimation, For centrifugal force matrix estimation, For the estimation of the gravity term, and All are positive definite gain matrices. For joint angle, For joint velocity, , and All are joint reference commands. This is the estimated total disturbance.

7. The method according to claim 1, characterized in that, The calculation of the comprehensive performance index based on the attitude control model and the impedance model includes: The target tracking error is evaluated based on the attitude control model. Based on the target tracking error, construct the attitude error vector; Based on the attitude error vector, the attitude error norm is calculated using the Euclidean norm; The contact force error is evaluated based on the impedance model. Based on the contact force error, the contact force error norm is calculated using the Euclidean norm; The comprehensive performance index is calculated based on the attitude error norm, the contact force error norm, and the weighting coefficient. The formula for calculating the comprehensive performance index is as follows: ； In the formula, The aforementioned comprehensive performance index, These are the weighting coefficients. Let the attitude error norm be... Let the contact force error norm be... For the current contact force, For the desired contact force, For a moment.

8. The method according to claim 5, characterized in that, The step of performing stable control of the flight robotic arm based on the comprehensive performance indicators and the joint control torque includes: Calculate the changing trend of the comprehensive performance index; If the value of the comprehensive performance index is greater than the preset index threshold, or the change trend is greater than the preset trend threshold, then the gradient method is used to adjust the control parameters of the flight manipulator. The control parameters of the flight manipulator include the control gain in the attitude control model, the first observer gain, the second observer gain and the third observer gain in the linear extended state observation model, and the desired inertia, damping and stiffness matrices in the impedance model. The flight robotic arm is controlled according to the adjusted control parameters of the flight robotic arm and the joint control torque.

9. A multi-degree-of-freedom flying robotic arm aerial operation stability control system, characterized in that, include: The information acquisition module is used to acquire the attitude information and current contact force of the flying robotic arm; An attitude control model construction module is used to construct an attitude control model based on the attitude information. The impedance model construction module is used to construct an impedance model based on the desired inertia, damping, stiffness matrix, end-effector pose, end-effector velocity, end-effector acceleration, end-effector reference trajectory, the current contact force, and the desired contact force. The joint control torque calculation module is used to calculate the joint control torque based on the impedance model. The comprehensive performance index calculation module is used to calculate the comprehensive performance index based on the attitude control model and the impedance model. A stability control module is used to perform stability control of the flight robotic arm based on the comprehensive performance indicators and the joint control torque. The step of constructing an attitude control model based on the attitude information includes: Based on the attitude information, extract the attitude angles of the previous moment, the current moment, and the next moment; Based on the attitude angle at the previous moment and the attitude angle at the current moment, calculate the estimated value of the pseudo-Jacobi matrix; The attitude control torque at the current moment is calculated based on the attitude angle at the current moment and the estimated value of the pseudo-Jacobi matrix. Calculate the attitude angle increment for the next moment based on the attitude angle at the current moment and the attitude angle at the next moment; Calculate the attitude control torque increment at the current moment based on the attitude control torque at the previous moment and the attitude control torque at the current moment; The attitude control model is constructed based on the attitude angle increment, pseudo-Jacobi matrix, and attitude control torque increment at the current moment. The expression for the attitude control model is as follows: ； In the formula, This is the attitude angle increment at the next moment. It is a pseudo-Jacobi matrix. This represents the increment of the attitude control torque at the current moment. For a moment.

Images