A robot joint motor and a joint motor control method

Abstract

This invention provides a robot joint motor and a joint motor control method. The method includes: acquiring the target and actual motion states, overall task, and current state of the robot joint motors, and determining the desired reference trajectory for each joint accordingly. The error signal between the trajectory and the actual state is input to a target controller to obtain a preliminary control signal. Anti-interference compensation is applied to the preliminary control signal to obtain an anti-interference control signal. Based on the current motion and load state of each joint, the coupling torque between joints is estimated, and decoupling compensation signals are generated and fused to obtain a total decoupling compensation signal. The anti-interference control signals for each joint and the total decoupling compensation signal are combined to obtain a joint motor control signal. The joint motors are then controlled to move according to the target joint motor control signal. This invention solves the problems of slow response, large error, and difficulty in achieving precise synchronous control of each joint in traditional methods.

Description

Technical Field

[0001] This invention relates to the field of robotics, and in particular to a robot joint motor and a joint motor control method. Background Technology

[0002] With the continuous advancement of robotics technology, robot joints are crucial components that connect and drive various parts to perform movements, while the joint motors inside are the core of enabling flexible joint motion. Their control principle generally involves receiving information from sensors, adjusting control signals in real time by a controller, and outputting control actions based on the control information, thus making the motor's operation more precise.

[0003] In existing technologies, while control algorithms such as traditional PID control can achieve stable control of joint motors to a certain extent, current control methods may suffer from slow response and large control errors when the robot encounters external impacts that change its direction of movement or hinder its continued movement. In severe cases, this can lead to choppy robot movements and even damage to the mechanical structure. Furthermore, in multi-joint cooperative robot systems, the coupling effect between joints can also affect control accuracy, making it difficult to achieve precise synchronous control of the joints. Summary of the Invention

[0004] This invention provides a robot joint motor and a joint motor control method, which solves the problems of slow response, large error, and difficulty in achieving precise synchronous control of each joint in traditional methods.

[0005] In a first aspect, the present invention provides a robot joint motor control method, comprising:

[0006] The target motion state and actual motion state of each joint motor of the robot are obtained, as well as the overall task requirements and current state of the robot. Based on the overall task requirements and current state of the robot, the expected reference trajectory of each joint is determined.

[0007] For each joint, the error signal between the desired reference trajectory and the actual motion state is input into the target controller to obtain the preliminary control signal output by the target controller, and the preliminary control signal is subjected to anti-interference compensation to obtain the anti-interference control signal;

[0008] Based on the current motion state and load state of each joint, the coupling torque estimate of each joint to other joints is determined, and based on the coupling torque estimate, a decoupling compensation signal for each target joint against a specific coupling source joint is generated; the specific coupling source joint is other joints that generate coupling interference to each target joint.

[0009] For any target joint, the decoupling compensation signals of all the specific coupling source joints received for the target joint are fused to obtain the total decoupling compensation signal of the target joint;

[0010] The anti-interference control signals of each joint and the overall decoupling compensation signal are combined to obtain the joint motor control signal, and the joint motors are controlled to move according to the target joint motor control signal.

[0011] In a second aspect, the present invention also provides a robot joint motor, applied to the robot joint motor control method as described in the first aspect; the robot joint motor includes:

[0012] The trajectory generation module is used to acquire the target motion state and actual motion state of the motors of each joint of the robot, as well as the overall task requirements and current state of the robot. Based on the overall task requirements and current state of the robot, the expected reference trajectory of each joint is determined.

[0013] An anti-interference compensation module is used to input the error signal between the desired reference trajectory and the actual motion state into the target controller for each joint, obtain the preliminary control signal output by the target controller, and perform anti-interference compensation on the preliminary control signal to obtain an anti-interference control signal.

[0014] The coupling compensation generation module is used to determine the coupling torque estimate of each joint to other joints based on the current motion state and load state of each joint, and to generate a decoupling compensation signal for each target joint for a specific coupling source joint based on the coupling torque estimate; the specific coupling source joint is other joints that generate coupling interference to each target joint.

[0015] The coupling compensation fusion module is used to fuse the decoupling compensation signals of all the specific coupling source joints received for any target joint to obtain the total decoupling compensation signal of the target joint.

[0016] The execution module is used to combine the anti-interference control signal of each joint and the total decoupling compensation signal to obtain the joint motor control signal, and control the joint motors to move according to the target joint motor control signal.

[0017] Thirdly, the present invention also provides an electronic device, comprising: a memory for storing computer software programs; and a processor for reading and executing the computer software programs, thereby implementing the robot joint motor control method as described above.

[0018] Fourthly, the present invention also provides a non-transitory computer-readable storage medium storing a computer software program, which, when executed by a processor, implements the robot joint motor control method described above.

[0019] Fifthly, the present invention also provides a computer program product, including a computer program that, when executed by a processor, implements the robot joint motor control method as described above.

[0020] The robot joint motor control method provided in this invention can effectively cope with external impacts encountered by the robot during operation by inputting error signals into the target controller and performing anti-interference compensation. When external interference causes the actual motion state to deviate from the target, the anti-interference control signal can quickly adjust the motor output, shorten the system response time, and reduce control error compared with the traditional PID control algorithm. This avoids the problem of unsmooth movement caused by response lag and reduces the risk of mechanical structure damage due to impact. In addition, by estimating the coupling torque between each joint and generating decoupling compensation signals, the mutual interference during multi-joint coordinated movement can be eliminated, so that the control signal of each joint can specifically cancel the coupling effect of other joints, achieving precise synchronization of the movement of each joint. Combined with the anti-interference control signal after anti-interference compensation, the control performance is improved from two levels: external interference suppression and internal coupling elimination. This not only solves the problems of slow response and large error in traditional methods, but also further reduces the wear and tear of the mechanical structure and extends the service life of the robot through dynamic trajectory planning and coupling compensation. Attached Figure Description

[0021] Figure 1 This is a flowchart illustrating the robot joint motor control method provided in an embodiment of the present invention;

[0022] Figure 2 This is a schematic diagram of the structure of the robot joint motor provided in an embodiment of the present invention;

[0023] Figure 3 An embodiment diagram of the electronic device provided in this invention;

[0024] Figure 4 An embodiment diagram of a computer-readable storage medium provided in accordance with the present invention. Detailed Implementation

[0025] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0026] In the description of this invention, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, features defined as "first" or "second" may explicitly or implicitly include one or more features. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0027] In the description of this invention, the term "for example" is used to mean "used as an example, illustration, or description." Any embodiment described as "for example" in this invention is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is provided to enable any person skilled in the art to make and use the invention. Details are set forth in the following description for purposes of explanation. It should be understood that those skilled in the art will recognize that the invention can be made without using these specific details. In other instances, well-known structures and processes will not be described in detail to avoid obscuring the description of the invention with unnecessary detail. Therefore, the invention is not intended to be limited to the embodiments shown, but is consistent with the broadest scope of the principles and features disclosed herein.

[0028] See Figure 1 , Figure 1 This is a flowchart illustrating the robot joint motor control method provided by the present invention. In this embodiment, the execution subject of the robot joint motor control method is the joint motor. Therefore, the robot joint motor control method includes:

[0029] Step 10: Obtain the target motion state and actual motion state of each joint motor of the robot, as well as the overall task requirements and current state of the robot. Based on the overall task requirements and current state of the robot, determine the expected reference trajectory of each joint.

[0030] Optionally, the joint motors first collect the actual motion state of each joint motor in real time through sensors installed on the robot's joint motors, such as encoders and current sensors, including parameters such as motor speed, rotation angle, and current. Simultaneously, based on the robot's pre-set task, such as picking up parts on an assembly line and precisely placing them in a designated position, the robot obtains the target motion state of each joint under the corresponding task, including target position, target acceleration, and target velocity.

[0031] Furthermore, the joint motors also acquire the overall task requirements for each joint of the robot. These requirements encompass the task type (handling, assembly, etc.), motion accuracy, speed limits, etc., as well as the current state of each joint, including its current position, velocity, and surrounding environment information (such as the presence of obstacles). It's important to note that after determining the robot's overall task requirements and current state, the desired reference trajectory for each joint within each time step can be calculated based on the robot's kinematics and dynamics model, combined with path planning algorithms such as spline interpolation. For example, using fifth-order polynomial path planning, a polynomial function is constructed based on the position, velocity, acceleration, and other boundary conditions of the starting and target points, yielding a smooth curve of the joint angle changing over time, which serves as the desired reference trajectory.

[0032] In one embodiment, taking the task of a robot moving a part from point A to point B as an example, the joint angles at the starting position of the robotic arm are collected by the encoder. The joint angle at the target position is According to the task requirements, the exercise time is set as follows: A fifth-order polynomial path planning method is used. Taking joint 1 as an example, let the fifth-order polynomial be... Based on the position, velocity, and acceleration boundary conditions of the starting and target points (the starting point velocity and acceleration are 0, and the target point velocity and acceleration are 0), the coefficients are solved. The desired reference trajectory of joint 1 is obtained. Similarly, the expected reference trajectories for joints 2 and 3 can be obtained. and .

[0033] Step 20: For each joint, the error signal between the desired reference trajectory and the actual motion state is input into the target controller to obtain the preliminary control signal output by the target controller. The preliminary control signal is then subjected to anti-interference compensation to obtain the anti-interference control signal.

[0034] Optionally, the joint motor first compares the desired reference trajectory of each joint with the actual motion state to calculate the position error. Speed ​​error Error signals are identified and input into the target controller. The target controller employs advanced control algorithms, such as adaptive control and sliding mode control. Taking adaptive control as an example, the controller parameters are adjusted in real time based on the error signal, and an initial control signal is output. After obtaining the initial control signal, the robot will inevitably be subject to external interference (such as collisions, changes in friction, etc.) during operation. These external interferences will affect the accuracy of the initial control signal. Therefore, in order to compensate for these interferences, it is necessary to perform anti-interference compensation on the initial control signal to obtain a more accurate anti-interference control signal, as described in steps 201-204.

[0035] In one embodiment, taking the above-mentioned robot handling task as an example, the desired angle of joint 1 at a certain moment is... The actual angle is Then the position error The expected speed is The actual speed is Speed ​​error .Will and The input is fed into the target controller based on the adaptive control algorithm to obtain the preliminary control signal. .

[0036] Step 30: Based on the current motion state and load state of each joint, determine the coupling torque estimate of each joint to other joints, and based on the coupling torque estimate, generate a decoupling compensation signal for each target joint for a specific coupling source joint; the specific coupling source joint is other joints that generate coupling interference to each target joint.

[0037] Optionally, since there is a coupling effect between the joints when the robot moves in multi-joint cooperative motion, the motion change of one joint will affect the motion of other joints. Therefore, after obtaining the current motion state and load state of each joint, the joint motors, combined with the robot's joint dynamics model, consider factors such as the mass, inertia, velocity, and acceleration of each joint, to calculate the estimated coupling torque of each joint on other joints, as described in steps 301-305.

[0038] Furthermore, based on the obtained coupling torque estimate and the pre-built decoupling controller, the joint motor generates a decoupling compensation signal for each target joint for a specific coupling source joint. If a feedforward decoupling control method is used in the decoupling controller, the compensation signal is calculated in advance based on the coupling torque estimate to counteract coupling interference, thereby improving the control accuracy and synchronization of each joint, enabling each joint to independently and accurately perform subsequent movements according to the desired trajectory.

[0039] Step 40: For any target joint, fuse the decoupling compensation signals of all specific coupling source joints received for the target joint to obtain the total decoupling compensation signal of the target joint.

[0040] Optionally, for each target joint, the joint motor, based on the decoupling compensation signals received from all specific coupling source joints, and considering the influence of each coupling source joint on the target joint, and after denoising the decoupling compensation signals, fuses all the decoupling compensation signals in a way that adapts to complex motion scenarios, thereby obtaining a total decoupling compensation signal with higher decoupling accuracy under complex scenarios. See steps 401-405 for details.

[0041] Step 50: Combine the anti-interference control signals of each joint with the overall decoupling compensation signal to obtain the joint motor control signal, and control the movement of each joint motor according to the target joint motor control signal.

[0042] Optionally, the joint motors, based on the obtained anti-interference control signals and overall decoupling compensation signals for each joint, sum the anti-interference control signals for each joint and the overall decoupling compensation signal to obtain the final joint motor control signal. This control signal is then input to the joint motor driver, which adjusts parameters such as the motor's voltage and current according to the control signal, thereby controlling the motor's speed and rotation angle, enabling the robot's joints to move along the desired trajectory. This approach comprehensively considers external interference and the coupling effect between joints, making the motor control signal more accurately reflect actual needs.

[0043] This invention effectively addresses external impacts encountered by the robot during operation by inputting error signals into the target controller and performing anti-interference compensation. When external interference causes the actual motion state to deviate from the target, the anti-interference control signal can quickly adjust the motor output, shortening the response time. Compared with traditional PID control algorithms, this reduces control errors, avoiding the problem of unsmooth movements caused by response lag, and reducing the risk of mechanical structure damage due to impact. Furthermore, by estimating the coupling torque between each joint and generating decoupling compensation signals, mutual interference during multi-joint coordinated movement can be eliminated, enabling the control signal of each joint to specifically counteract the coupling effects of other joints, achieving precise synchronization of joint movements. Combined with the anti-interference control signal after anti-interference compensation, control performance is improved from two levels: external interference suppression and internal coupling elimination. This not only solves the problems of slow response and large errors in traditional methods, but also further reduces the wear and tear on the mechanical structure and extends the robot's service life through dynamic trajectory planning and coupling compensation.

[0044] In one embodiment, steps 201-204 are described as follows:

[0045] Step 201: Based on the dynamic model of the joint motor and the actual control signal, determine the theoretical motion state under undisturbed conditions.

[0046] Optionally, based on the actual control signals of each joint motor (such as the torque command corresponding to the voltage / current output by the motor driver) and the joint motor's dynamic model, the theoretical motion state under interference-free conditions (i.e., without external impacts, sudden load changes, or other disturbances) is first calculated. Specifically, since the joint motor's dynamic model describes the physical laws of "control signal → motor motion," taking a commonly used second-order linear model as an example, the dynamic equation can be expressed as: in, This is expressed as rotational inertia, such as the moment of inertia of a robotic arm joint motor. ; Expressed as the damping coefficient (combined equivalent of joint friction, motor internal resistance, etc., such as...) ); Represented as constant disturbance terms such as gravitational torque; It is the actual control signal; Expressed as joint angle. Expressed as angular velocity, This is expressed as angular acceleration. Furthermore, given the actual control signal... Next, substitute into the model, using known... , and The theoretical motion state (including theoretical angle, angular velocity, and angular acceleration) can be obtained by solving the dynamic equations.

[0047] Step 202: Compare the difference between the theoretical motion state and the actual motion state to obtain the estimated value of the total disturbance.

[0048] Optionally, the joint motor calculates the difference between the theoretical motion state obtained from the solution and the actual motion state collected to obtain an estimated value of the total disturbance, that is, the deviation between theory and reality, which reflects the comprehensive impact of external disturbances (such as sudden collisions or load changes).

[0049] Furthermore, the estimated total disturbance can be derived through deformation of the dynamic model. Specifically, during actual motion, the dynamic equations become as follows due to the disturbance:

[0050] ;

[0051] in, Represented as the actual disturbance quantity, which is difficult to measure directly, using... estimate.

[0052] Combining the theoretical equations under undisturbed conditions: Subtracting the two equations, we get: This can be further simplified to a formula for estimating the total disturbance (using an approximation of the difference): ;in .

[0053] Step 203: Input the estimated total disturbance value into the compensation filter to obtain the disturbance compensation signal; the compensation filter is a filter designed by inversely using the dynamic model of the articulated motor.

[0054] Optionally, filters designed inversely to the dynamic model in the joint motor are used to counteract the effects of disturbances on the system. Essentially, this involves using the inverse dynamic model to cancel out disturbances; that is, the forward process of the model is... Model The inverse model is ,in, This is represented as a compensation signal. The derivation of its inverse model formula (based on a transformation of the original model) is as follows:

[0055] Original model: Due to the need to compensate for interference Therefore, a compensation signal needs to be designed. This allows the disturbed system, plus compensation, to return to a disturbance-free state. And ideally, This achieves complete compensation; therefore, through filter optimization, the inverse model can be expressed as:

[0056] ;

[0057] in, Represented by the Laplace operator, it represents the differential element, and the filter parameters use the original model's... Designed to achieve targeted compensation.

[0058] Step 204: Determine the signal integration mode based on the disturbance type, and integrate the preliminary control signal and the disturbance compensation signal based on the signal integration mode to obtain the anti-interference control signal.

[0059] Optionally, based on the obtained disturbance compensation signal, the joint motor needs to design different signal integration modes for different disturbance types (such as time-varying disturbances and linearly superimposed disturbances) due to the different disturbance characteristics. These include a first integration mode and a second integration mode. The first integration mode is suitable for time-varying disturbances, including abrupt loads, nonlinear friction, and aperiodic external impacts. The second integration mode is suitable for linearly superimposed disturbances. Under different integration modes, the preliminary control signal and the disturbance compensation signal are integrated to obtain an anti-interference control signal, which is more adaptable to different disturbance scenarios and improves robustness. See steps 2041-2045 for details.

[0060] This invention starts from a dynamic model and directly quantifies disturbances by comparing "theoretical" and "actual" motion states, providing an accurate basis for subsequent compensation. Compared with blind compensation, it can significantly reduce the problem of "inaccurate compensation." In addition, by using the dynamic model to inversely design a filter, the influence of disturbances can theoretically be "precisely canceled." An anti-disturbance mechanism is built at the system model level, making the compensation signal highly targeted, especially suitable for robot joint motor control with known models. Finally, the integration mode is adjusted according to the type of disturbance, and the coordination between control and compensation is dynamically adjusted so that the anti-interference control can cope with various complex scenarios such as sudden load changes and nonlinear friction. This improves the stability and reliability of the robot in actual working conditions (such as factory handling and complex environment operations) and reduces the risk of motion disruption and structural damage caused by disturbances.

[0061] In one embodiment, steps 2041-2045 are described as follows:

[0062] Step 2041: If the signal integration mode is the first integration mode, then the estimated total disturbance is differentiated, and the result of the differentiation is processed by the disturbance change rate according to the central difference method to obtain the change rate sequence.

[0063] Optionally, in the first integrated mode, the joint motor first differentiates the estimated total disturbance to capture the rate of change of the disturbance, such as the rapid jump in disturbance during sudden load changes. Specifically, the derivative can be calculated using the central difference method with the disturbance estimates at different times, and the formula is as follows: ;in This is expressed as the sampling interval (e.g., in a robot control system). The sampling period is By continuously collecting multiple sets of data This allows the output of a time-varying perturbation rate sequence. The central difference method is superior to the simple forward pass method. Backward difference is more accurate and can reduce derivative errors, especially in time-varying disturbance scenarios (such as nonlinear friction with fluctuating magnitudes), making the rate of change calculation more closely match the actual disturbance trend and providing a reliable basis for subsequent gain adjustment.

[0064] In one embodiment, taking a handling robot as an example, the estimated total disturbance when grasping workpieces of different weights (sudden load change) is... It will change rapidly due to sudden changes in load. Assume a sampling interval. When grabbing heavy objects, , Then, substituting into the formula, we get:

[0065] The rate of change of the disturbance at t=1 second is obtained, and the data is continuously collected and calculated to finally obtain the rate of change sequence.

[0066] Step 2042: Based on the rate of change sequence, construct an iterative basis function for the perturbation change trend, and iteratively update the initial gain according to the direction and intensity of the perturbation rate of change based on the iterative basis function to obtain the time-varying gain.

[0067] Optionally, when constructing the iterative basis function for the disturbance change trend of the joint motor, in order to ensure that the iterative basis function can characterize the impact of the disturbance change trend on the gain, its formula is constructed as follows: ,in It is expressed as a rate of change sensitivity coefficient (for example, 0.1, which can be adjusted according to the actual disturbance intensity so that the response of the basis function to the rate of change is more adapted to the characteristics of the robot joint). Represented as a hyperbolic tangent function, it can map the rate of change of the disturbance to... The range allows for smoother gain adjustments; This is expressed as time. Then, the gain is initialized. (If set to 1 based on experience, representing the initial basic response strength), and update the gain using iterative basis functions: ,in It is expressed as the iteration step size (e.g., 0.05, which controls the step size of the gain update). Represented as a symbolic function, according to The sign of the gain determines the gain. Decrease direction. Achieve gain. The output time-varying gain sequence is dynamically adjusted according to the rate of change of the disturbance.

[0068] By iterating through the basis functions, the gain adjustment follows the perturbation trend, thus enabling the time-varying gain. It can dynamically adapt to changes in disturbances: when the disturbance becomes stronger and changes rapidly, the gain is amplified and controlled in a timely manner; when the disturbance is stable, the gain is maintained at a reasonable level. Compared with fixed gain, this greatly improves the flexibility of anti-interference.

[0069] In one embodiment, taking a conveying robot as an example: Assume hour First, calculate the basis functions: Update the gain again: At this point, the rate of change of the disturbance is large and positive, and the gain increases slightly from 1 to 1.05, enhancing the control and compensation for sudden load changes.

[0070] Step 2043: Determine the adjustment delay amount based on the robot joint movement speed, and perform delay matching processing on the disturbance compensation signal according to the adjustment delay amount to obtain the adaptation compensation signal.

[0071] Optionally, the joint motors adjust according to the robot's joint movement speed. At different times, there are differences in signal transmission and actuator response delays, requiring the determination of the adjustment delay amount. For example, during high-speed movement, the response delay of joint motors and actuators is relatively more noticeable. Specifically, this is done when calculating the adjustment delay amount. At that time, empirical formulas can be used. ;in, This is represented as a coefficient related to the mechanical properties of joints, such as those in robotic arms. Pick ,speed unit Alternatively, a delay model can be obtained through system identification. Then, the disturbance compensation signal is processed. Delay processing, such as using linear interpolation, is performed to match the effective time of the compensation signal with the actual delay of joint movement, thus obtaining an adaptive compensation signal. This solves the problem of mismatch between the compensation signal and the actual time of disturbance, improving the timeliness of anti-interference.

[0072] Step 2044: Based on the time-varying gain, the preliminary control signal and the adaptation compensation signal are fused to obtain the anti-interference control signal.

[0073] Optionally, the joint motor is adjusted according to the generated time-varying gain. , will the initial control signal and adaptation compensation signal The fusion is performed, and the fusion formula is as follows: Time-varying gain It can dynamically adjust the compensation level according to changes in disturbance, such as when the disturbance becomes stronger (larger rate of change). Increased, resulting in a higher proportion of compensation signal; weakened disturbance. Reduce to avoid overcompensation.

[0074] Step 2045: If the signal integration mode is the second integration mode, then the preliminary control signal and the disturbance compensation signal are added together to obtain the anti-interference control signal.

[0075] Optionally, in the second integrated mode, the joint motor is subjected to linear superimposed disturbances, such as stable load fluctuations (like when a robot is transporting goods at a constant speed, the weight of the goods is stable, and the friction changes are small and linear) and weak periodic disturbances (the disturbance frequency is low, the amplitude is small, and it does not produce complex nonlinearity after being superimposed with the control signal). The disturbances and control signals do not interfere with each other. Therefore, the anti-interference control signal can be obtained by directly adding the preliminary control signal and the disturbance compensation signal.

[0076] In the first integrated mode, the invention tracks the time-varying characteristics of disturbances (such as sudden loads and nonlinear friction) throughout the entire process, from disturbance rate estimation to gain adaptive adjustment and compensation signal delay matching. This ensures that the anti-interference control keeps up with the disturbance changes, the compensation signal takes effect in a timely manner, and the control error is significantly reduced. Furthermore, the time-varying gain and delay matching mechanism avoids over-compensation or under-compensation. In addition, in the second integrated mode, the invention can quickly and accurately cancel disturbances in scenarios such as stable loads and linear friction, allowing the joint movement to return to the desired trajectory.

[0077] In one embodiment, steps 301-304 are described as follows:

[0078] Step 301: Construct joint screws based on the current motion state of each joint, and correlate the joint screws through a pre-constructed homogeneous transformation matrix to obtain a joint screw model.

[0079] Optionally, the joint motor first constructs the joint screw based on screw theory according to the current motion state of each joint, such as linear velocity and angular velocity. That is, for each joint of the robot... During each joint movement, the corresponding angular velocity can be obtained. (such as the speed at which a joint rotates around its axis) and linear velocity (The velocity of movement at a point on the joint) can be used to construct the joint rotation as follows: This is used to describe the motion state of the joint. Then, to describe the pose relationships between joints, a pre-constructed homogeneous transformation matrix is ​​introduced. The homogeneous transformation matrix is ​​a 4x4 matrix that can simultaneously describe the transformation of "position" and "orientation" (such as the pose relationship between joint k and joint j). Therefore, the joint rotation from joint k to joint j... It can be achieved using a homogeneous transformation matrix. Joint rotation associated with joint j To form a joint rotation model This can be simply understood as using a transformation matrix to associate the motion characteristics of different joints.

[0080] Step 302: For each joint, differentiate the joint screw model based on the joint variables to obtain the Jacobian screw matrix.

[0081] Optionally, joint motors for each joint Its joint variables are set to (e.g., rotating the joint by an angle) The moving joint is the displacement. Here, we take a rotating joint as an example. Based on the joint spinor model obtained in step 301. That is, the joint screw model corresponding to each joint i is obtained as follows: Then, according to the definition of the Jacobian screw matrix, the joint screw model was... Regarding joint variables Taking the partial derivative, we get Since the spinor is a 6-dimensional vector, joint variables... It is n-dimensional (e.g., a rotational joint is a 1-dimensional angle, and a translational joint is a 1-dimensional displacement), therefore the Jacobian spinor matrix is ​​a 6×n matrix.

[0082] Step 303: Analyze the off-diagonal elements of the Jacobian spinor matrix to obtain the spinor coupling coefficients between each joint. Correlate the spinor coupling coefficients with the load state of each joint to obtain the coupling torque function.

[0083] Optionally, joint motors can be used when multiple joints move simultaneously, as this can lead to a coupling effect. That is, when one joint... The spinor motion of the (coupled source joint) affects the target joint through the Jacobian spinor matrix. The state of motion. Regarding the Jacobian spinor matrix. off-diagonal elements (i.e.) hour, Middle corresponding joint By analyzing the column elements of the variables, the spinor coupling coefficient is derived. Specifically, let's first examine the Jacobian spinor matrix. Regarding joint variables Find the partial derivative Then multiply by inverse matrix (need Full rank (which can be guaranteed through reasonable modeling and range of motion constraints), then multiplied by the joint. angular velocity The spinor coupling coefficient is obtained. Used to quantitatively reflect the coupling source joint For the target joint The degree of spinor coupling. Next, considering the load state of the joint, and the joint dynamics equations of the robot are as follows: ,in Represented as an inertia matrix (reflecting the forces generated by joints due to inertia) Torque characteristics), Represented as joint Angular acceleration; Represented as Coriolis force Torque term (generated by relative motion between joints); Expressed as a gravity term (determined by the robot's gravity distribution and joint positions), the obtained Represented as a coupling source joint The generalized force generated. The spinor coupling coefficient. With generalized force Correlation, constructing coupling torque function Where T represents the transpose of a matrix or vector, this allows for the combination of kinematic coupling between joints (through screws and Jacobian matrices) and dynamic loads (through generalized forces) to obtain the joint... For joints The relationship between the resulting coupling torques.

[0084] Step 304: For each joint, based on the coupling torque function, traverse all coupling source joints to obtain the coupling torque estimate of each joint on other joints.

[0085] Optionally, joint motors for each target joint Based on the coupling torque function obtained in step 303 For the coupling source joint, traverse all other coupling source joints (Right now and Calculate each corresponding Then sum them up to get the target joint. Estimation of total coupling torque If the target joint is joint 4, then calculate the coupling torques of joints 1 to 4, 2 to 4, 3 to 4, 5 to 4, and 6 to 4 in sequence. Add them together to get .

[0086] This invention unifies the multibody motion analysis framework through screw theory. It starts by constructing a screw model to establish the motion foundation, then uses the Jacobian matrix to characterize the velocity mapping, combines dynamic analysis to determine the coupling and torque relationship, and finally sums the total coupling torque. This approach has a wide range of applications and can more accurately reflect the complex coupling situation during the actual movement of the robot, thus improving the accuracy of torque estimation.

[0087] In one embodiment, steps 401-405 are described as follows:

[0088] Step 401: For any target joint, decompose each signal of the decoupling compensation signal of all specific coupling source joints for the target joint based on the Hilbert-Huang transform to obtain multiple intrinsic mode function components and a residual component.

[0089] Optionally, for any target joint, the joint motor can compensate for the decoupling signals from specific coupling source joints (e.g., robot arm joints receiving decoupling signals from joints such as the shoulder and elbow). For example, if a target joint receives signals from n coupling source joints... Let represent different coupling source joints, and j represent the target joint. Then, Empirical Mode Decomposition (EMD) is used to decompose each... Decomposed into m intrinsic mode function components , ( Each IMF corresponds to a different oscillation mode in the signal, such as high-frequency vibration, low-frequency steady change, leaving one residual component. This refers to the relatively stable, slowly changing part of the signal.

[0090] Step 402: Extract time-frequency features for each intrinsic mode function component to obtain energy features and frequency centers respectively.

[0091] Optionally, when calculating the energy characteristics of the joint motor, first for each In the time interval (e.g., the time it takes for the robot to perform the grasping action:) Within ) calculate energy Energy is represented as the integral of the square of the signal amplitude, reflecting the intensity of the component's influence on the target joint. Therefore, based on the physical definition of energy, in the time domain, signal energy = integral of instantaneous power, and instantaneous power ≈ square of amplitude, thus determining the formula:

[0092] ;in, This is represented as the energy characteristic of the i-th coupling source joint and the k-th IMF component; This is represented as the change of the k-th IMF component of the i-th coupling source joint with time t.

[0093] When calculating the frequency center, first... Perform a Fourier transform (T) to obtain the frequency domain representation. ,in Represented as frequency (the frequency domain independent variable after Fourier transform), ranging from... ,for example Its frequency center This is a weighted average frequency, where the weight is the square of the frequency domain amplitude (frequencies with larger amplitudes have a more significant impact on the signal). Therefore, based on the concept of a weighted average, the numerator is the frequency. The integral of the square of the amplitude, with the denominator being the integral of the square of the amplitude, yields the result. The formula is:

[0094] ;in, This represents the frequency center of the i-th coupling source joint and the k-th IMF component (reflecting the position of the dominant frequency of this component): It is represented as the frequency domain representation of the k-th IMF component after performing a Fourier transform (FT).

[0095] Step 403: Based on the preset energy threshold and frequency center threshold, the energy features and frequency center are filtered to obtain the filtered intrinsic mode function components.

[0096] Optionally, the joint motor first presets an energy threshold and a frequency center threshold. The energy threshold is determined experimentally to identify components that have a substantial impact on the target joint; for example, an energy level of at least 0.3 is required to filter out components with insufficient energy. The frequency center threshold is used to filter out frequencies that are too biased and can also be determined through experimental analysis. Specifically, the energy threshold is set. and frequency center threshold range filter Component: For each IMF component ,judge: and Those that meet the conditions are retained, denoted as... Discard components that do not meet the requirements (considered as noise or irrelevant components). This yields the filtered intrinsic mode function components. .

[0097] Step 404: Reconstruct the filtered intrinsic mode function components and the corresponding residual components to obtain the effective decoupling compensation signal for each coupling source joint.

[0098] Optionally, the joint motor will obtain the filtered intrinsic mode function components. With the corresponding residual components By directly adding them together, the effective decoupling compensation signal of each coupling source joint to the target joint can be reconstructed. By "filtering out noise and reconstructing the retained components," the decoupling compensation signal of each coupled source joint is made cleaner, which reduces interference during subsequent fusion.

[0099] Step 405: Fuse all effective decoupling compensation signals to obtain the total decoupling compensation signal for the target joint.

[0100] Optionally, the joint motor dynamically adjusts the influence weights based on the real-time errors of each joint according to the obtained effective decoupling compensation signal, and then performs weighted fusion to obtain a total decoupling compensation signal that better matches the joint motion. See steps 4051-4055 for details.

[0101] This invention uses HHT decomposition and energy / frequency center calculation to break down complex coupled signals into analyzable "sub-signals," distinguishing time-frequency differences more precisely than traditional methods. It captures the essence of decoupling signals at different joints and filters out weak interference and noise through a screening process, reducing computational load and making fusion more efficient. It also avoids "dirty data" dragging down decoupling accuracy. Ultimately, it effectively removes noise and unimportant components from the signal, retains key features, improves the accuracy of the overall decoupling compensation signal after fusion, and thus enhances the joint decoupling effect.

[0102] In one embodiment, steps 4051-4055 are described as follows:

[0103] Step 4051: Based on the initial motion parameters and physical kinematic model of each coupling source joint, obtain the initial influence factor of each joint.

[0104] Optionally, when fusing multiple effective decoupling compensation signals, the joint motor first determines the initial influence of each coupling source joint on the target joint. Specifically, the motion parameters of each coupling source joint (such as angular velocity) are determined. angular acceleration This will affect the target joint through mechanical structures (such as link length and joint transmission ratio). Therefore, it can be determined according to the formula: The initial influence factors for each joint were calculated. Represented as a coupling source joint Angular velocity (e.g., the rotational angular velocity of joint 1, in units of...) ); Represented as a coupling source joint Angular acceleration (e.g., the rate of change of angular velocity of joint 1, in units) ); Expressed as kinematic constants (determined by the mechanical structure, such as the mass distribution of connecting rods and the stiffness of joint networks, which can be determined through calibration or...). (Model calculations can also be used, or experiments can be conducted.) This ensures that the greater the angular velocity and angular acceleration, the stronger the impact on the target joint.

[0105] Step 4052: Perform preliminary weighted summation of the effective decoupling compensation signals and initial influence factors of each joint to obtain the intermediate signal sequence.

[0106] Optionally, the joint motor uses the effective decoupling compensation signal corresponding to each coupling source joint. (i.e., the i-th effective decoupling compensation signal for the target joint node), and compare it with the initial influence factor. We perform weighted summation to obtain the intermediate signal sequence. .

[0107] Step 4053: Input the real-time feedback error of each joint into the pre-built correction function to obtain the correction factor output by the correction function.

[0108] Optionally, during robot movement, the joint motors may experience real-time feedback errors due to discrepancies between the actual and ideal movements. (represented as the real-time feedback error of the i-th joint), and input into a pre-built correction function, which is: The correction factor for the corresponding node is obtained. This allows for the introduction of real-time errors, enabling dynamic sensing of motion deviations and adaptive compensation to better match the actual motion state.

[0109] Step 4054: Correct the initial influence factor based on the correction factor to obtain the target correction factor.

[0110] Optionally, the joint motor uses the correction factor obtained in step 4053 for each joint and modifies the initial influence factor obtained in step 4051 accordingly. This allows the decoupling compensation to specifically correct errors and enables the weights to adapt to the motion state in real time. The correction formula is as follows: This allows the weights to be changed from fixed initial values ​​to dynamically adaptive values.

[0111] Step 4055: The intermediate signal sequence is weighted and summed again according to the target correction factor to obtain the total decoupling compensation signal of the target joint.

[0112] Optionally, the joint motor further weights and sums the corrected target correction factor with the intermediate signal of each coupled source joint to obtain the total decoupling compensation signal. Specifically, the sum of all target correction factors is first calculated. , where n represents the total number of coupling source joints. Then, for each intermediate signal... The total decoupling compensation signal is obtained by weighting and summing the proportions of the target correction factors. This ensures that the total compensation takes into account both the initial kinematic effects and the real-time errors, accurately offsetting time-varying coupling interference.

[0113] This invention starts with kinematic parameters (angular velocity, angular acceleration), uses initial influence factors to quantify the basic coupling strength, and conforms to the nature of robot motion; then, it generates correction factors through real-time error feedback to make the influence factors "come alive" and cope with time-varying interference (such as sudden load changes, speed changes); finally, it first performs preliminary weighting (using initial factors) and then dynamic weighting (using corrected factors), and the two weightings ensure reasonable fusion of multi-source compensation signals and improve decoupling accuracy.

[0114] Furthermore, the robot joint motor provided by the present invention will be described below. The robot joint motor described below and the robot joint motor control method described above can be referred to in correspondence.

[0115] Optional, refer to Figure 2 , Figure 2 This is a schematic diagram of the structure of the robot joint motor provided by the present invention. The robot joint motor includes:

[0116] The trajectory generation module 210 is used to acquire the target motion state and actual motion state of the motors of each joint of the robot, as well as the overall task requirements and current state of the robot, and to determine the expected reference trajectory of each joint based on the overall task requirements and current state of the robot.

[0117] The anti-interference compensation module 220 is used to input the error signal between the desired reference trajectory and the actual motion state into the target controller for each joint, obtain the preliminary control signal output by the target controller, and perform anti-interference compensation on the preliminary control signal to obtain the anti-interference control signal.

[0118] The coupling compensation generation module 230 is used to determine the coupling torque estimate of each joint to other joints based on the current motion state and load state of each joint, and to generate a decoupling compensation signal for each target joint for a specific coupling source joint based on the coupling torque estimate; the specific coupling source joint is other joints that generate coupling interference to each target joint.

[0119] The coupling compensation fusion module 240 is used to fuse the decoupling compensation signals of all specific coupling source joints received for any target joint to obtain the total decoupling compensation signal of the target joint.

[0120] The execution module 250 is used to combine the anti-interference control signals of each joint and the overall decoupling compensation signal to obtain the joint motor control signal, and control each joint motor to move according to the target joint motor control signal.

[0121] This invention effectively addresses external impacts encountered by the robot during operation by inputting error signals into the target controller and performing anti-interference compensation. When external interference causes the actual motion state to deviate from the target, the anti-interference control signal can quickly adjust the motor output, shortening the system response time. Compared with traditional PID control algorithms, this reduces control errors, avoiding the problem of unsmooth movements caused by response lag, and reducing the risk of mechanical structure damage due to impact. Furthermore, by estimating the coupling torque between each joint and generating decoupling compensation signals, mutual interference during multi-joint coordinated movement can be eliminated, enabling the control signal of each joint to specifically counteract the coupling effects of other joints, achieving precise synchronization of joint movements. Combined with the anti-interference control signal after anti-interference compensation, control performance is improved from two levels: external interference suppression and internal coupling elimination. This not only solves the problems of slow response and large errors in traditional methods, but also further reduces the wear and tear on the mechanical structure and extends the robot's service life through dynamic trajectory planning and coupling compensation.

[0122] Please see Figure 3 , Figure 3 An embodiment diagram of an electronic device provided in accordance with the present invention. For example... Figure 3 As shown, this embodiment of the invention provides an electronic device 300, including a memory 310, a processor 320, and a computer program 311 stored in the memory 310 and executable on the processor 320. When the processor 320 executes the computer program 311, it performs the following steps:

[0123] The target motion state and actual motion state of each joint motor of the robot are obtained, as well as the overall task requirements and current state of the robot. Based on the overall task requirements and current state of the robot, the expected reference trajectory of each joint is determined.

[0124] For each joint, the error signal between the desired reference trajectory and the actual motion state is input into the target controller to obtain the preliminary control signal output by the target controller. The preliminary control signal is then compensated for to obtain the anti-interference control signal.

[0125] Based on the current motion state and load state of each joint, the coupling torque estimate of each joint to other joints is determined, and based on the coupling torque estimate, a decoupling compensation signal for each target joint against a specific coupling source joint is generated; the specific coupling source joint is other joints that generate coupling interference to each target joint.

[0126] For any target joint, the decoupling compensation signals of all specific coupling source joints received for the target joint are fused to obtain the total decoupling compensation signal of the target joint.

[0127] The anti-interference control signals of each joint and the overall decoupling compensation signal are combined to obtain the joint motor control signal, and the joint motors are controlled to move according to the target joint motor control signal.

[0128] Please see Figure 4 , Figure 4 An embodiment diagram of a computer-readable storage medium provided in accordance with an embodiment of the present invention is shown. Figure 4 As shown, this embodiment provides a computer-readable storage medium 400 on which a computer program 311 is stored. When the computer program 311 is executed by a processor, it performs the following steps:

[0129] The target motion state and actual motion state of each joint motor of the robot are obtained, as well as the overall task requirements and current state of the robot. Based on the overall task requirements and current state of the robot, the expected reference trajectory of each joint is determined.

[0130] For each joint, the error signal between the desired reference trajectory and the actual motion state is input into the target controller to obtain the preliminary control signal output by the target controller. The preliminary control signal is then compensated for to obtain the anti-interference control signal.

[0131] Based on the current motion state and load state of each joint, the coupling torque estimate of each joint to other joints is determined, and based on the coupling torque estimate, a decoupling compensation signal for each target joint against a specific coupling source joint is generated; the specific coupling source joint is other joints that generate coupling interference to each target joint.

[0132] For any target joint, the decoupling compensation signals of all specific coupling source joints received for the target joint are fused to obtain the total decoupling compensation signal of the target joint.

[0133] The anti-interference control signals of each joint and the overall decoupling compensation signal are combined to obtain the joint motor control signal, and the joint motors are controlled to move according to the target joint motor control signal.

[0134] On the other hand, the present invention also provides a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer is able to execute the robot joint motor control method provided by the above methods, the method including:

[0135] The target motion state and actual motion state of each joint motor of the robot are obtained, as well as the overall task requirements and current state of the robot. Based on the overall task requirements and current state of the robot, the expected reference trajectory of each joint is determined.

[0136] For each joint, the error signal between the desired reference trajectory and the actual motion state is input into the target controller to obtain the preliminary control signal output by the target controller. The preliminary control signal is then compensated for to obtain the anti-interference control signal.

[0137] Based on the current motion state and load state of each joint, the coupling torque estimate of each joint to other joints is determined, and based on the coupling torque estimate, a decoupling compensation signal for each target joint against a specific coupling source joint is generated; the specific coupling source joint is other joints that generate coupling interference to each target joint.

[0138] For any target joint, the decoupling compensation signals of all specific coupling source joints received for the target joint are fused to obtain the total decoupling compensation signal of the target joint.

[0139] The anti-interference control signals of each joint and the overall decoupling compensation signal are combined to obtain the joint motor control signal, and the joint motors are controlled to move according to the target joint motor control signal.

[0140] The system embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0141] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., including several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods of various embodiments or some parts of embodiments.

[0142] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for controlling robot joint motors, characterized in that, include: The target motion state and actual motion state of each joint motor of the robot are obtained, as well as the overall task requirements and current state of the robot. Based on the overall task requirements and current state of the robot, the expected reference trajectory of each joint is determined. For each joint, the error signal between the desired reference trajectory and the actual motion state is input into the target controller to obtain the preliminary control signal output by the target controller, and the preliminary control signal is subjected to anti-interference compensation to obtain the anti-interference control signal; Based on the current motion state and load state of each joint, the coupling torque estimate of each joint to other joints is determined, and based on the coupling torque estimate, a decoupling compensation signal for each target joint for a specific coupling source joint is generated. The specific coupling source joint is any other joint that generates coupling interference to each target joint; For any target joint, the decoupling compensation signals of all the specific coupling source joints received for the target joint are fused to obtain the total decoupling compensation signal of the target joint; The anti-interference control signals of each joint and the overall decoupling compensation signal are combined to obtain the joint motor control signal, and the joint motors are controlled to move according to the target joint motor control signal. The step of performing anti-interference compensation on the preliminary control signal to obtain an anti-interference control signal includes: Based on the dynamic model of the joint motor and the actual control signal, the theoretical motion state under undisturbed conditions is determined; The difference between the theoretical motion state and the actual motion state is compared to obtain the estimated value of the total disturbance. The estimated total disturbance is input into the compensation filter to obtain the disturbance compensation signal; the compensation filter is a filter designed by inversely using the dynamic model of the joint motor. The signal integration mode is determined based on the disturbance type, and the preliminary control signal and the disturbance compensation signal are integrated based on the signal integration mode to obtain the anti-interference control signal. The signal integration mode includes a first integration mode, which is characterized by being applicable to time-varying characteristic disturbances, including abrupt loads, nonlinear friction, and non-periodic external shocks. The process of integrating the preliminary control signal and the disturbance compensation signal based on the signal integration mode to obtain the anti-interference control signal includes: If the signal integration mode is the first integration mode, then the total disturbance estimate is differentiated, and the result of the differentiation is processed by the disturbance change rate according to the central difference method to obtain the change rate sequence. Based on the rate of change sequence, an iterative basis function for the perturbation change trend is constructed, and the initial gain is iteratively updated according to the direction and intensity of the perturbation rate of change based on the iterative basis function to obtain the time-varying gain; The adjustment delay is determined based on the robot joint motion speed, and the disturbance compensation signal is subjected to delay matching processing according to the adjustment delay to obtain the adaptation compensation signal. The initial control signal and the adaptation compensation signal are fused based on the time-varying gain to obtain the anti-interference control signal; The signal integration mode includes a second integration mode, which is characterized as applicable to linearly superimposed perturbations; The process of integrating the preliminary control signal and the disturbance compensation signal based on the signal integration mode to obtain the anti-interference control signal includes: If the signal integration mode is the second integration mode, then the preliminary control signal and the disturbance compensation signal are added together to obtain the anti-interference control signal.

2. The robot joint motor control method according to claim 1, characterized in that, The estimation of the coupling torque of each joint to other joints based on the current motion state and load state of each joint includes: The joint screws are constructed based on the current motion state of each joint, and the joint screws are correlated through a pre-constructed homogeneous transformation matrix to obtain a joint screw model. For each joint, the Jacobian screw matrix is ​​obtained by differentiating the joint screw model based on the joint variables. The off-diagonal elements of the Jacobian screw matrix are analyzed to obtain the screw coupling coefficients between each joint. The screw coupling coefficients are then correlated with the load state of each joint to obtain the coupling torque function. For each joint, based on the coupling torque function, all coupling source joints are traversed to obtain the coupling torque estimate of each joint on other joints.

3. The robot joint motor control method according to claim 1, characterized in that, For any target joint, the decoupling compensation signals received from all the specific coupling source joints for the target joint are fused to obtain the total decoupling compensation signal for the target joint, including: For any target joint, based on the Hilbert-Huang transform, each signal of the decoupling compensation signal received from all the specific coupling source joints for the target joint is decomposed to obtain multiple intrinsic mode function components and a residual component. Time-frequency features are extracted from each of the intrinsic mode function components to obtain energy features and frequency centers, respectively. Based on preset energy thresholds and frequency center thresholds, energy characteristics and frequency centers are filtered to obtain the filtered intrinsic mode function components. The filtered intrinsic mode function components and corresponding residual components are reconstructed to obtain the effective decoupling compensation signal for each coupling source joint. All the effective decoupling compensation signals are fused together to obtain the total decoupling compensation signal of the target joint.

4. The robot joint motor control method according to claim 3, characterized in that, The step of fusing all the effective decoupling compensation signals to obtain the total decoupling compensation signal for the target joint includes: Based on the initial motion parameters and physical kinematic model of each coupling source joint, the initial influence factor of each joint is obtained; The effective decoupling compensation signals and initial influence factors of each joint are preliminarily weighted and summed to obtain the intermediate signal sequence; The real-time feedback error of each joint is input into a pre-built correction function to obtain the correction factor output by the correction function; The initial influence factor is corrected based on the correction factor to obtain the target correction factor; The intermediate signal sequence is weighted and summed again according to the target correction factor to obtain the total decoupling compensation signal of the target joint.

5. A robot joint motor, characterized in that, Applied to the robot joint motor control method as described in any one of claims 1 to 4; the robot joint motor comprises: The trajectory generation module is used to acquire the target motion state and actual motion state of the motors of each joint of the robot, as well as the overall task requirements and current state of the robot. Based on the overall task requirements and current state of the robot, the expected reference trajectory of each joint is determined. An anti-interference compensation module is used to input the error signal between the desired reference trajectory and the actual motion state to the target controller for each joint, obtain the preliminary control signal output by the target controller, and perform anti-interference compensation on the preliminary control signal to obtain an anti-interference control signal; the anti-interference compensation of the preliminary control signal to obtain the anti-interference control signal includes: Based on the dynamic model of the joint motor and the actual control signal, the theoretical motion state under undisturbed conditions is determined; The difference between the theoretical motion state and the actual motion state is compared to obtain the estimated value of the total disturbance. The estimated total disturbance is input into the compensation filter to obtain the disturbance compensation signal; the compensation filter is a filter designed by inversely using the dynamic model of the joint motor. The signal integration mode is determined based on the disturbance type, and the preliminary control signal and the disturbance compensation signal are integrated based on the signal integration mode to obtain the anti-interference control signal. The signal integration mode includes a first integration mode, which is characterized by being applicable to time-varying characteristic disturbances, including abrupt loads, nonlinear friction, and non-periodic external shocks. The process of integrating the preliminary control signal and the disturbance compensation signal based on the signal integration mode to obtain the anti-interference control signal includes: If the signal integration mode is the first integration mode, then the total disturbance estimate is differentiated, and the result of the differentiation is processed by the disturbance change rate according to the central difference method to obtain the change rate sequence. Based on the rate of change sequence, an iterative basis function for the perturbation change trend is constructed, and the initial gain is iteratively updated according to the direction and intensity of the perturbation rate of change based on the iterative basis function to obtain the time-varying gain; The adjustment delay is determined based on the robot joint motion speed, and the disturbance compensation signal is subjected to delay matching processing according to the adjustment delay to obtain the adaptation compensation signal. The initial control signal and the adaptation compensation signal are fused based on the time-varying gain to obtain the anti-interference control signal; The signal integration mode includes a second integration mode, which is characterized as applicable to linearly superimposed perturbations; The process of integrating the preliminary control signal and the disturbance compensation signal based on the signal integration mode to obtain the anti-interference control signal includes: If the signal integration mode is the second integration mode, then the preliminary control signal and the disturbance compensation signal are added together to obtain the anti-interference control signal; The coupling compensation generation module is used to determine the coupling torque estimate of each joint to other joints based on the current motion state and load state of each joint, and to generate a decoupling compensation signal for each target joint for a specific coupling source joint based on the coupling torque estimate; the specific coupling source joint is other joints that generate coupling interference to each target joint. The coupling compensation fusion module is used to fuse the decoupling compensation signals of all the specific coupling source joints received for any target joint to obtain the total decoupling compensation signal of the target joint. The execution module is used to combine the anti-interference control signal of each joint and the total decoupling compensation signal to obtain the joint motor control signal, and control the joint motors to move according to the target joint motor control signal.

6. An electronic device, comprising: Memory, used to store computer software programs; A processor for reading and executing the computer software program, characterized in that, when the processor executes the computer software program, it implements the robot joint motor control method as described in any one of claims 1 to 4.

7. A non-transitory computer-readable storage medium, wherein a computer software program is stored therein, characterized in that, When the computer software program is executed by the processor, it implements the robot joint motor control method as described in any one of claims 1 to 4.

Images