174 results about "Joint angle" patented technology

The embodiment of the invention provides a dynamic obstacle avoidance control method and device for a redundant mechanical arm and a computer readable storage medium, and the method comprises the steps: obtaining monitoring data, collected and transmitted by a sensor on the redundant mechanical arm, of the redundant mechanical arm in each control period, the monitoring data comprises an initial joint angle of each joint and a point closest to an obstacle; constraint conditions of the redundant mechanical arm in the current control period are constructed, the constraint conditions comprise a target tracking constraint function, a dynamic obstacle avoidance constraint function, a physical constraint function and a self-collision constraint function, and the target tracking constraint function is obtained by correcting an original tracking constraint function through a pre-constructed feedback function and an avoidance vector; a cost function of the redundant mechanical arm is constructed, and the cost function and the constraint conditions are combined to obtain a quadratic programming equation set of the redundant mechanical arm; and solving the quadratic programming equation set to obtain the target joint speed of each joint of the redundant mechanical arm.

The invention discloses a seven-degree-of-freedom mechanical arm inverse kinematics solving method, system and device based on numerical iteration. The method comprises the steps that S1, a D-H kinematics model of a mechanical arm is established based on a joint configuration of the seven-degree-of-freedom mechanical arm; s2, calculating a Jacobian matrix according to the joint angle sequence; s3, introducing a weight matrix to suppress the joint speed at the edge of the joint range; s4, introducing a damping factor, and calculating to obtain an optimized Jacobian pseudo-inverse matrix; s5, the pose error between the expected tail end pose and the current tail end pose of the tail end actuator of the mechanical arm is calculated; s6, converting the pose error into a joint angle adjustment amount by using the optimized Jacobian pseudo-inverse matrix to obtain a joint angle iterative formula; and S7, based on the pose error, the joint angle of each rotating joint of the mechanical arm is determined through a numerical iteration method. According to the method, the problem of unstable solution of a traditional analytical method under the singular configuration can be avoided, and the calculation efficiency and the numerical stability are improved.

The invention discloses a space manipulator impedance control method based on joint hysteresis dynamic position compensation, and belongs to the field of space robot nonlinear dynamics control. The control method comprises the following steps: constructing a flexible joint space manipulator model based on a standard dynamic model of the space manipulator; according to the flexible joint space manipulator model, a virtual torque sensing strategy is designed based on motor side sensor information, and a joint torsion torque estimated value is obtained; building a joint hysteresis model based on a Bi-LSTM mixed hysteresis modeling method fusing multi-scale memory and an attention mechanism, and training the joint hysteresis model; inputting a joint torsion moment estimation value into the trained joint hysteresis model, and estimating a joint torsion angle; based on the estimated joint torsion angle, the joint angle measured by the motor side encoder is subjected to superposition correction, and track error compensation caused by joint hysteresis is achieved.

The invention discloses a robot joint angle calibration method and system, and relates to the technical field of robotics.The calibration method comprises the specific steps that S100, sensors are installed and data are collected, specifically, the sensors are installed at all joints of a robot, real-time compensation of thermomagnetic interference is achieved through intelligent self-adaptive model adjustment, and the calibration accuracy is improved; the method effectively reduces the influence of environmental factors on the joint angle of the robot, not only considers the influence of temperature on the joint angle, but also takes magnetic field interference into consideration, and can accurately describe the joint angle error of the robot under different working conditions by constructing a coupling model of thermomagnetic coupling and joint angle error, thereby improving the accuracy of the joint angle error. A theoretical basis is provided for subsequent thermomagnetic compensation, and meanwhile, an intelligent adaptive model adjustment mechanism enables the method to continuously learn and optimize model parameters, so that the method continuously adapts to a new environment and new conditions in the operation process of the robot, and the accuracy and reliability of calibration are further improved.

The invention relates to the technical field of mechanical arm control, and discloses a bionic self-adaptive variable-stiffness mechanical arm joint control method. The method comprises the steps that multi-source biological signals such as electromyographic signals, tendon tension signals and joint angle signals during joint operation are collected; constructing a stiffness feature vector based on the myoelectricity and tendon tension signals, and generating a motion trail sequence according to the joint angle signals; extracting a corresponding stiffness feature vector for each track point, and calculating stiffness sensitivity; constructing a bionic stiffness reference model containing a mapping relation between the stiffness feature vector and a joint angle signal, and dynamically weighting and adjusting the mapping relation by using stiffness sensitivity; predicting a joint stiffness change sequence based on the adjusted model and the real-time signal, and fusing the sequence with the current joint angle signal to generate a dynamic stiffness parameter; and calculating a driving torque correction amount according to the dynamic stiffness parameter and a preset safety threshold range, iteratively updating the model weighting parameter until the correction amount meets a convergence condition, and outputting a driving torque instruction to an execution mechanism.

The invention provides a dexterous hand finger joint angle control method, device and system and a medium, and relates to the technical field of dexterous hands. The joint angles of the fingers of the dexterous hand comprise the first joint angle and the second joint angle, the first joint angle and the second joint angle belong to active degrees of freedom, and the control method for the joint angles of the fingers of the dexterous hand comprises the steps that a control instruction for the fingers of the dexterous hand is obtained; according to the control instruction, the tracking control law of the first joint angle and the tracking control law of the second joint angle, the first joint angle and the second joint angle are controlled, the tracking control law of the first joint angle and the tracking control law of the second joint angle are determined based on a superspiral sliding mode control algorithm and a dynamic model of the finger. Through the tracking control law designed based on the super-spiral sliding mode control algorithm and the dynamic model, the stability and precision of dexterous hand finger joint angle control are improved.

The invention discloses a control method and system of a mechanical arm, electronic equipment, a medium and a program product. The control method comprises the steps that the target pose of an end effector under world coordinates and forward kinematics information of the mechanical arm are obtained; according to the first joint angle of the mechanical arm, the forward kinematics information and the first current pose of the end effector, the second current pose of the end effector under the world coordinates is determined; determining a loss function, wherein the loss function comprises a first loss function corresponding to the difference between the target pose and the second current pose; and based on the loss function, the first joint angle is updated through a gradient descent algorithm to control the mechanical arm to move. The mechanical arm is controlled to move without collecting training data in advance, and the calculation cost is reduced.

The invention relates to a mechanical arm control system and method, and belongs to the field of mechanical arm control, and the control system comprises a microkernel controller which calculates the joint angle of each joint of a mechanical arm according to the difference value between the current position and the target position of the tail end of the mechanical arm and sends the joint angle to a servo driver; the servo driver is responsible for closed-loop control of a current loop and a speed loop, accurately executes the position instruction sent by the microkernel controller and feeds back the actual state of each joint to the microkernel controller in real time; the multi-degree-of-freedom mechanical arm serves as an executing mechanism, receives the joint movement instruction from the servo driver, and drives an end effector to achieve any specified pose and complex track in a three-dimensional space through cooperative movement of a plurality of joints; and the instruction unit is used for communicating with the microkernel controller and issuing the target position information of the end effector of the multi-degree-of-freedom mechanical arm to the microkernel controller. By adopting the control method of the control system, the control precision and reliability of the mechanical arm can be improved.

The invention provides a training method and device for a VLA model, and relates to the technical field of intelligent robots with bodies, and the method comprises the steps that the VLA model outputs a joint angle sequence of a mechanical arm of a robot with a body based on visual input and language input; the differentiable forward kinematics module outputs a real-time pose of an end effector of the mechanical arm in a task space; based on joint space loss formed by the joint angle sequence and the joint angle in the teaching data, task space loss formed by the real-time pose and the tail end pose in the teaching data, and task space constraint loss, a multi-objective loss function is constructed, and total loss is calculated; calculating the gradient of the total loss to the joint angle sequence and the real-time pose through back propagation, and optimizing preset parameters of the VLA model; the above steps are repeatedly executed until the total loss meets the expectation, and a trained target VLA model is obtained; the technical problems that when a VLA model is trained in a joint space, the hardware coupling performance is high, and the generalization ability is insufficient are solved.

The invention discloses a robot dexterous arm teleoperation sharing control method and a related device, and relates to the technical field of robotics.The method comprises the steps that smooth expected pose information and expected angle state information are obtained, and pose information of a mechanical arm of a slave end at the current stage and angle state information and contact force information of a dexterous hand are obtained; based on the expected pose information and the joint angle information of the mechanical arm, the inverse kinematics of the mechanical arm is solved through a nonlinear optimization method, and a joint angle instruction of the mechanical arm is obtained; based on the expected angle state information, the angle state information of the dexterous hand and the contact force information, utilizing a fuzzy logic theory and a dynamic compliance primitive model to carry out adaptive variable impedance control on the dexterous hand to obtain a joint torque instruction of the dexterous hand; and feeding back the contact force information of the dexterous hand in the next stage to a main end operator. The success rate and the working efficiency of the teleoperation robot can be improved.

This wearable apparatus may: measure a first joint angle of a user; measure a second joint angle of the user; generate a filter input filtered by a filter on the basis of a difference between first angle data obtained by measuring the first joint angle and second angle data obtained by measuring the second joint angle; filter the generated filter input by means of the filter such that a time point at which a torque rotation direction of a driving module is changed corresponds to a peak time point of the generated filter input by delaying a phase of the generated filter input by means of the filter; generate a torque on the basis of a filter output generated by the filtering; and provide the generated torque to the user.

The invention provides a control method and device of a humanoid robot and a medium, and relates to the technical field of humanoid robots, and the control method of the humanoid robot comprises the steps that the current joint angle of a mechanical arm and the current first position of a target object are obtained; inputting the joint angle and the first position into a pre-trained prediction model for processing to obtain a change value of the joint angle; the joint angle of the mechanical arm is controlled to be adjusted according to the change value, and the current second position of the end effector is obtained; determining whether the distance between the second position and the first position is smaller than a distance threshold; and if yes, an end effector is controlled to grab the target object, and if not, the step of obtaining the current joint angle of the mechanical arm and the current first position of the target object is executed, so that the humanoid robot can be applied to the dynamic environment, and the dynamic object can be accurately grabbed.

The invention discloses an intelligent planning method for a motion track of a multi-axis mechanical arm, and relates to the technical field of automatic control, and the intelligent planning method comprises the following steps: setting a task target and initial parameters of the mechanical arm, collecting environment data to generate an environment map, and constructing a dynamic environment model, the initial parameters of the mechanical arm comprise the initial position and posture, the joint angle, the speed and the acceleration; a mechanical arm track generation model is constructed, a mechanical arm preliminary track is generated based on the mechanical arm track generation model, the dynamic constraint of the mechanical arm preliminary track is checked, and the mechanical arm preliminary track is optimized; and predicting a future track of the obstacle according to the dynamic environment model, and combining the optimized preliminary track of the mechanical arm and the future track of the obstacle to generate a motion track of the mechanical arm. According to the method, through dynamic environment modeling, accurate pre-judgment of the motion trail of the dynamic obstacle is achieved, and the real-time adaptability of the mechanical arm in a complex scene is remarkably improved.

The invention discloses a VLA task space control method and system for a multi-degree-of-freedom mechanical arm, and relates to the technical field of intelligent robot control, visual information and language instructions when the mechanical arm executes a task are obtained, and a trained VLA model is adopted to obtain a target pose of an end effector in a task space; an IK solver corresponding to the mechanical arm configuration is adopted, and according to the target pose, multiple sets of joint angle solutions of all joints of the mechanical arm are obtained; selecting an optimal joint angle of each joint from the multiple groups of joint angle solutions by adopting a preset optimization criterion; and after the optimal joint angle sequence of the mechanical arm is subjected to track smoothing processing, the mechanical arm is driven to move. Mechanical arm motion control is carried out in combination with a VLA model and an IK solver, the problems that the generalization ability between different mechanical arm configurations is poor, the data collection cost is high and the model interpretability is insufficient are solved, and the generalization ability and data efficiency of mechanical arm control are improved.

The invention relates to the technical field of mechanical arm control, and discloses a mechanical arm force control method, system and device based on a current loop, and the technical scheme is characterized in that the mechanical arm force control method comprises the following steps: S1, obtaining the current value of each mechanical arm joint, and obtaining the joint torque according to the current-torque mapping relation; s2, according to the kinetic model, calculating a gravity item and a friction item in the joint torque to obtain the acting torque of the external force on each joint; s3, according to the acting moment of the external force on each joint, the tail end contact force is calculated in real time, the pose increment is obtained through processing, matrix transformation is conducted on the pose increment and the actual tail end pose, and the final pose is obtained; s4, performing inverse kinematics real-time solution on the final pose to obtain expected joint angle data, and performing torque feedforward control according to the expected joint angle data; the technical scheme of the invention has the advantages of low cost, fast response speed and strong anti-interference capability.

The invention relates to the technical field of artificial intelligence, can be applied to business scenes such as mechanical arm grabbing, financial science and technology and medical health, and discloses a target control method, device and equipment based on multi-modal information and a medium, and the method comprises the steps that a language instruction expressing an operation intention is obtained, and visual information of the environment where a target object is located is collected; semantic representation is generated through the language processing module, and visual features are generated through the visual processing module; inputting the visual features and the semantic representations into an action generation module to generate an action sequence; and a joint angle sequence and a joint angle value are determined based on the action sequence, and the mechanical arm is controlled to execute grabbing operation. According to the method, the accurate grabbing action sequence is generated by fusing language and visual information, the movement of the mechanical arm is controlled by combining the joint angle sequence and the accurate joint angle value, and smooth track movement and accurate grabbing are achieved. The adaptive capacity and the industrial-grade grabbing precision in the complex environment are improved, and the production efficiency and the grabbing success rate are effectively improved.

The invention belongs to the technical field of mechanical arm joint control, and particularly discloses a preset performance self-adaptive joint control method for a rope-driven super-redundant snakelike mechanical arm, and the method comprises the following steps: carrying out kinematics modeling on a rope joint link of the rope-driven super-redundant snakelike mechanical arm, determining a relation between a rope speed and a joint angle, and determining a preset performance of the rope-driven super-redundant snakelike mechanical arm; calculating a joint jacobian matrix; designing a joint angle controller, defining an angle error and constraining the angle error in a preset performance range, and obtaining control input through an error transformation and return-to-zero dynamics principle so as to realize joint angle control; according to the method, a joint Jacobian matrix is estimated on line, and the joint Jacobian matrix is dynamically updated according to a rope speed error and a return-to-zero dynamic theory so as to adapt to the flexibility, time-varying characteristics and external disturbance of a rope, and the control precision is ensured. According to the method, it can be guaranteed that the joint angle of the rope-driven super-redundant mechanical arm can be accurately controlled under different working conditions, and it is guaranteed that the mechanical arm completes high-precision operation under the complex working conditions.

The invention provides a modeling method and system for solving the problem of rope flexible deformation vibration of a rope-driven mechanical arm. The method comprises the steps that the mapping relation between a rope and a joint angle is established; rope deformation elastic potential energy is established according to the rope take-up and pay-off variable quantity driven by a motor-driven winch and the current joint rope mapping length; kinetic energy and potential energy of joint rotation and arm lever movement of a motor-driven rope transmission mechanical arm are defined, and a Lagrange function is constructed; according to the second type of Lagrange equation, a rope-driven mechanical arm system kinetic equation of the mapping relation between the rope variation control input and the mechanical arm joint angle is deduced; optimizing the model, proving the stability, and verifying the effectiveness and reliability of the algorithm. The technical problems of mechanical arm reliability servo defects, low set expected track tracking precision and high-frequency vibration caused by nonlinear characteristics such as flexible vibration, friction, rope deformation and hysteresis existing in a circuit, rope flexibility and a mechanical arm system are solved.

The invention discloses a singular avoidance control method and device for a six-axis mechanical arm based on virtual seven degrees of freedom. The tail end of the mechanical arm is provided with a virtual rotating shaft which coincides with the original point of a tail end coordinate system and is perpendicular to a rotating shaft of the tail end coordinate system. The control method comprises the steps that position information and angle information of all joints of the mechanical arm and the actual pose and expected pose of the tail end in the current detection period are obtained; based on the position information and the angle information of all the joints, the weight value of the virtual rotating shaft is correspondingly adjusted by combining the state that the mechanical arm approaches the shoulder singular point and / or the wrist singular point, a quadratic programming problem used for avoiding the singular points is reconstructed, and a to-be-executed joint angle increment vector is solved; and on the basis of the to-be-executed joint angle increment vector, the angle value of each joint of the mechanical arm in the next detection period is calculated, and the mechanical arm is controlled to move on the basis of the angle value of each joint so as to avoid shoulder singular points and / or wrist singular points of the mechanical arm.

The embodiment of the invention discloses a mechanical arm control method and device, electronic equipment and a storage medium. The method comprises the following steps: acquiring task parameters when a target mechanical arm executes a control task; according to the recurrent neural network, acquiring a joint angle sequence of each joint of the target mechanical arm from an initial position to a target position when the historical task is executed; and according to the task parameters and the historical task joint angle sequence, target joint angles of all joints of the target mechanical arm at the next moment are predicted, and all the joints are controlled to execute task instructions according to the target joint angles. By the adoption of the scheme, the joint angle of the mechanical arm can be flexibly adjusted according to different task requirements by introducing the multi-task parameters, multi-task operation is achieved, and the dynamic environment adaptability is enhanced; and by means of the processing capacity of the recurrent neural network on the time sequence, real-time prediction and adjustment of the joint angle of the mechanical arm are achieved, the operation stability and the task completion rate are improved, and high adaptability and universality are achieved in a multi-task scene.

The invention provides a mechanical arm control method and device, computer equipment and a storage medium, and the method comprises the steps that the target tail end pose and the initial joint angle of a mechanical arm are obtained; based on the target tail end pose, self-motion variables are introduced to parameterize the joint space of the mechanical arm, and multiple sets of candidate joint angles are obtained; constructing a joint motion cost function, and determining the total cost of moving from the initial joint angle to each group of candidate joint angles based on the joint motion cost function; traversing the value range of the self-motion variable, and screening out a solution with the minimum total cost from the multiple groups of candidate joint angles according to the joint motion cost function to serve as an optimal inverse kinematics solution; and a control signal is generated based on the optimal inverse kinematics solution, all joints of the mechanical arm are driven to move, and therefore the tail end of the mechanical arm reaches the target tail end pose. By the adoption of the method, the accuracy, economical efficiency and reliability of mechanical arm control can be improved.

The invention discloses a mechanical arm joint angular position self-adaptive specified time control method which comprises the following steps: step 1, establishing a dynamic model of a mechanical arm system, and initializing the state and control parameters of the mechanical arm system; step 2, designing an expected error trajectory of specified time convergence, and defining an error variable; and step 3, designing a specified time controller, and realizing rapid tracking of an expected angular position trajectory within specified time. According to the method, specified time tracking control over the angular position of the joint of the mechanical arm can be achieved, and meanwhile the controller does not need to be switched to guarantee continuity and boundaries of the controller; compared with a minimum upper bound of convergence time provided by limited / fixed time control, the method can realize convergence of the angular position error at an accurate specified time.

The invention relates to the technical field of robot motion, in particular to a variable load working condition heavy-load mechanical arm kinematics parameter calibration method and system. The method comprises the steps that joint angle parameters and theoretical Standard D-H model parameters are determined; the theoretical position of the end effector in the ideal state is determined; the actual position is determined, and the position error of the end effector is determined; the deflection under the preset load condition is determined, and corrected D-H model parameters are obtained; determining a load effect based on the parameter structure characteristics; introducing the load effect as a correction item into a parameter updating formula of a preset LM algorithm, and designing an adaptive compensation coefficient to adjust the contribution degree of the load effect; and dynamically adjusting the adaptive compensation coefficient through a convergence index in an iteration process of the LM algorithm. According to the method, the calibration precision of the kinematics parameters of the mechanical arm can be improved under the variable load working condition.

The invention relates to a joint angle prediction method and device, wearable lower limb exoskeleton equipment and electronic equipment, and relates to the technical field of exoskeleton equipment control. The method comprises the following steps: acquiring an electromyographic signal when a user walks by wearing the wearable lower limb exoskeleton equipment; classifying the electromyographic signals according to an electromyographic signal classification model, and determining a first electromyographic signal associated with the knee joint and a second electromyographic signal associated with the ankle joint; inputting the first electromyographic signal and the second electromyographic signal into a joint angle prediction model, and determining a joint angle of a knee joint and a joint angle of an ankle joint; wherein the wearable lower limb exoskeleton equipment performs joint control according to the joint angle of the knee joint and the joint angle of the ankle joint. Therefore, the electromyographic signals are disassembled into the signals related to all the lower limb joints, then the joint angles of all the lower limb joints are predicted through the signals related to all the lower limb joints, and therefore the joint angle prediction accuracy is improved.

The invention discloses a mechanical arm hierarchical motion planning method and device in a dynamic environment, and belongs to the technical field of mechanical arm motion planning. A target position of the mechanical arm is set, and a collision-free initial global path from the current position of each joint of the mechanical arm to the target position is planned; a continuous and smooth planned mechanical arm movement track is generated; generating the angular velocity control quantity of each joint of the mechanical arm; on the basis of avoiding the static obstacles, when it is detected that the moving range of the mechanical arm contains the dynamic obstacles, a dynamic obstacle avoiding frame is constructed in combination with a dynamic control obstacle function (DCBF); and a signal sequential logic STL is introduced into the dynamic obstacle avoidance framework to carry out task constraint, and the mechanical arm is controlled to carry out dynamic obstacle avoidance. The method is simple in step, high in execution efficiency and capable of guiding the mechanical arm to avoid obstacles and complete tasks in real time.

The invention relates to a position rigidity decoupling rope-driven mechanical arm variable-rigidity joint module. The position rigidity decoupling rope-driven mechanical arm variable-rigidity joint module comprises a joint angle input module, a position rigidity decoupling module and a position rigidity decoupling module, one side of the rigidity adjusting module is connected with the joint angle input module; the rigidity adjusting input module is arranged on the other side of the rigidity adjusting module, and the output of the rigidity adjusting input module is connected with the rigidity adjusting module; and the output module comprises an output gear, and the output module is connected with the joint angle input module through the output gear. The joint rigidity change is irrelevant to the joint angle change, joint position rigidity decoupling is achieved, and independent control over the joint rigidity is facilitated.

The invention discloses a task-driven active and passive fusion mechanical arm motion evaluation method and system, and the method comprises the steps: guiding a mechanical arm task, and collecting a joint angle data sequence; performing time alignment and smoothing processing to obtain an ideal demonstration track; inputting into a dynamic motion primitive model for learning to obtain model parameters; generalizing a new joint angle sequence conforming to the trajectory shape; a real-time control instruction is generated by taking the real-time control instruction as a reference track and adopting a reinforcement learning controller, and the mechanical arm is driven to track the reference track; and a multi-dimensional evaluation index system is constructed, and quantitative evaluation is conducted on the motion performance, the energy consumption efficiency and the intelligent level of task execution of the mechanical arm. According to the method, the robustness and the self-adaptive capability of the mechanical arm in different task scenes are remarkably improved, so that the efficiency and the safety of automatic operation are improved while stable operation is ensured.

The present invention provides a hierarchical multi-loop control method for a manipulator for compliant assembly of electronic components. The design reference trajectory, impedance force and the environmental impedance force of the previous cycle are input into the impedance feature knowledge base of the electronic component assembly scene to obtain the design impedance trajectory of the manipulator joint, and the difference is calculated with the position information of the manipulator joint. The control variable is predicted by the MPC cycle output model, and the auxiliary control variable is obtained by summing with the IHOSMC cycle output. The control torque of the manipulator system is obtained by the inverse dynamics control cycle equation, and then the joint angle variable and the output trajectory of the end effector are obtained. The environmental impedance force is obtained in combination with the environmental information; when the output trajectory of the end effector is equal to the design impedance trajectory and the environmental impedance force is equal to the design impedance force, the manipulator system is stable, and the assembly of electronic components is performed according to the output trajectory and environmental impedance force at this time. The present invention has the characteristics of high assembly accuracy and fast response speed.

The embodiment of the invention provides a trajectory optimization method and device, electronic equipment and a storage medium, and relates to the technical field of robots. The method comprises the steps that a preset trajectory of the tail end of a mechanical arm is obtained; interpolation is carried out based on points in a predetermined track, and a plurality of interpolation track points are obtained; aiming at a first track point in the starting point, the target point and the interpolation track point, determining a posture reachable interval which can be reached by the tail end of the mechanical arm at the position of the first track point; determining a target attitude for the first track point based on the attitude reachable interval of the first track point; and the posture of each first track point is updated to be the corresponding target posture, so that an optimized mechanical arm track is generated. Therefore, the task execution efficiency of the mechanical arm can be improved under the condition that the track point at the tail end of the mechanical arm has no inverse solution of the corresponding joint angle.

This application discloses a training method, apparatus, device, and storage medium for a joint angle prediction model, relating to the field of computer technology. The method includes: acquiring training sample data, which includes skeletal parameters, sample wrist joint pose sequences, and arm angle labels for the sample wrist joint pose sequences; obtaining arm angle feature information based on the sample wrist joint pose sequences using a selective state-space network of the joint angle prediction model to be trained; obtaining elbow position prediction information based on the arm angle feature information and skeletal parameters using a differentiable kinematic layer of the joint angle prediction model to be trained; and iteratively updating and training the joint angle prediction model to be trained based on a preset objective function, the training sample data, the arm angle feature information, and the elbow position prediction information to obtain a trained joint angle prediction model.