Wearable exoskeleton data collecting and processing method and system based on digital twinning

Abstract

The invention relates to the technical field of wearable exoskeleton data acquisition and processing, and discloses a wearable exoskeleton data acquisition and processing method and system based on digital twinning, and the method comprises the steps: building a simulation mapping model of exoskeleton joints through employing the digital twinning technology according to upper limb motion data; according to the upper limb motion data, executing a kinematics modeling strategy, calculating an actual pose of the end effector by using a forward kinematics algorithm, and substituting a target pose into inverse kinematics for solving, specifically, obtaining a joint angle, executing a Jacobian matrix modeling strategy, and establishing a model between the actual pose and the joint angle; the influence of the variable quantity of the joint angle on the actual pose of the end effector is obtained; if the pose error is smaller than or equal to an error threshold value, acquiring a joint angle at the moment, and recording the joint angle as a target joint angle; if the pose error is greater than an error threshold value, executing a joint angle tuning strategy, adjusting the joint angle, and repeating the steps S2-S5; the simulation precision is improved, and the man-machine coordination is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of wearable exoskeleton data acquisition and processing, and specifically provides a method and system for wearable exoskeleton data acquisition and processing based on digital twin. Background Art

[0002] With the development of human-computer interaction technology, robot control, artificial intelligence, and biomechanics, wearable exoskeletons, as emerging human-machine augmentation devices, are widely used in fields such as medical rehabilitation, industrial assistance, and military enhancement. Wearable exoskeletons are mainly used to assist the human body in performing specific actions, such as walking, carrying, and rehabilitation training, enhancing muscle strength through external power devices, reducing load, and improving work efficiency and motor ability. In the field of rehabilitation medicine, wearable exoskeletons can assist patients with mobility impairments caused by nerve injuries, muscle degeneration, etc. in performing exercise training, accelerating the rehabilitation process; with the development of human-computer interaction control, sensing technology, and artificial intelligence, exoskeleton devices are gradually developing towards the direction of intelligence, high precision, and high adaptability.

[0003] Existing upper limb exoskeleton robot simulation technologies have multiple defects when adjusting the end effector. First, the simulation system cannot accurately restore the complex motion characteristics of the human upper limb, resulting in a deviation between the simulation and the actual motion. Second, due to sensor noise and computational delay, the update speed of the simulation model cannot keep up with the actual motion, causing a response lag. In addition, the error correction ability is insufficient. Existing systems mostly adopt fixed compensation strategies and fail to dynamically adjust the model parameters, resulting in error accumulation and affecting the operation accuracy. Finally, the existing technology lacks an adaptive learning mechanism and cannot optimize the motion trajectory according to individual differences, resulting in poor human-machine coordination. Solving these problems requires improving the simulation accuracy, enhancing real-time control, optimizing error correction, and improving human-machine coordination.

[0004] This solution proposes a method and system for wearable exoskeleton data acquisition and processing based on digital twin, which uploads exoskeleton sensor data to the digital twin platform for simulation modeling in real time and adjusts the actual pose based on the target pose, solving the problems mentioned in the background art. Summary of the Invention

[0005] The present invention provides a method and system for wearable exoskeleton data acquisition and processing based on digital twin, which promotes the solution of the problems mentioned in the above background art.

[0006] In a first aspect, the present application provides a method for wearable exoskeleton data acquisition and processing based on digital twin, adopting the following technical solution: A method for wearable exoskeleton data acquisition and processing based on digital twin, comprising: Obtain the data measured by the exoskeleton joint sensor, denoted as upper limb motion data; Based on the upper limb movement data, a simulation mapping model of the exoskeleton joint is established using digital twin technology; Based on the upper limb movement data, a kinematic modeling strategy is executed, and the actual pose of the end effector is calculated using the forward kinematic algorithm. The actual pose is divided into the actual position and the actual orientation; The target pose defined by the simulation mapping model is obtained. The target pose is divided into the target position and the target orientation; Substitute the target pose into the inverse kinematics solution. Specifically: S1. Obtain the joint angles, execute the Jacobian matrix modeling strategy, establish a model between the actual pose and the joint angles, and obtain the influence of the change in joint angles on the actual pose of the end effector; S2. Establish the error between the actual pose and the target pose, denoted as the pose error; S3. Set the error threshold and compare the pose error with the error threshold; S4. If the pose error is less than or equal to the error threshold, obtain the joint angles at this time, denoted as the target joint angles; Control the exoskeleton joint to adjust the target joint angles; S5. If the pose error is greater than the error threshold, execute the joint angle optimization strategy, adjust the joint angles, and repeat S2 - S5; During the process of controlling the exoskeleton joint to adjust the target joint angles, execute the simulation model feedback adjustment strategy, compare the simulation mapping results with the actual movements of the person, and adjust the parameters of the simulation mapping model.

[0007] By obtaining the data measured by the exoskeleton joint sensors and combining digital twin technology to establish a simulation mapping model, the upper limb movement state can be accurately restored, and high-precision motion capture can be achieved. By executing the forward kinematic algorithm to calculate the actual pose of the end effector and combining the inverse kinematics to solve the target pose, the exoskeleton device can accurately follow the human movements, improving the accuracy and stability of the control. By establishing the Jacobian matrix, the influence of the joint angle change on the pose of the end effector can be analyzed, making the motion control more flexible and adjustable. By setting the error threshold and executing the joint angle optimization strategy based on error feedback, it can be ensured that the exoskeleton device is accurately adjusted within the error range, improving the motion reproduction accuracy. Through the simulation model feedback adjustment strategy, not only can the simulation mapping model be optimized to make it more in line with the actual human movement characteristics, but also the adaptability of the exoskeleton can be effectively improved, enabling it to cope with different individual human differences and achieving a more natural interaction experience.

[0008] Preferably, the establishment of the simulation mapping model of the exoskeleton joint using digital twin technology based on the upper limb movement data includes: The upper limb movement data includes the position, angle, rotation axis, and movement limit of each exoskeleton joint; Define the degrees of freedom of the exoskeleton joint in the simulation mapping model; Establish a simulation mapping model of the exoskeleton joint, specifically: Obtain the digital human 3D model and perform simplification processing on the digital human 3D model; Perform joint segmentation on the simplified digital human 3D model; Obtain the human body parameters corresponding to the digital human 3D model, and use the human body parameters to adjust the size, length, and initial position of each exoskeleton joint; Set the material and texture of the digital human 3D model and adjust the contrast color of the joints; Build the background environment of the digital human 3D model, create a background board and a floor, and adjust the lighting parameters, camera height, and angle.

[0009] By obtaining upper limb motion data and constructing a simulation mapping model of the exoskeleton joint, the motion accuracy and adaptability of the exoskeleton system can be effectively improved. First, the motion data includes the position, angle, rotation axis, and motion limits of the exoskeleton joint, providing accurate input information for the subsequent simulation model and enabling the model to truly reflect the dynamic characteristics of the exoskeleton. Defining the degrees of freedom of the exoskeleton joint in the simulation mapping model helps to refine the control and optimization of the model and provides a basis for accurately simulating the motion of the exoskeleton. Second, establishing the digital human 3D model and performing simplification processing makes the calculation more efficient, avoids waste of computing resources, and at the same time retains sufficient details for accurately simulating joint motion. Performing joint segmentation on the simplified model further improves the operability and flexibility of the model, and the motion range and posture can be adjusted for different joints. Obtaining the human body parameters corresponding to the digital human 3D model and adjusting the size, length, and initial position of the exoskeleton joint according to these parameters ensure that the exoskeleton design fits the natural structure of the human body and improves the wearing comfort and usage effect. Setting the material and texture of the digital human 3D model and adjusting the contrast color of the joints helps to visually optimize the display effect of the model and enhance the user's immersion. Finally, building the background environment and adjusting the lighting and camera angles make the simulation scene more realistic and provide more intuitive visual feedback for subsequent refined operations. This series of steps not only improves the simulation accuracy but also enhances the flexibility and adaptability of the model, providing more scientific and user-friendly support for the design and use of exoskeleton robots.

[0010] Preferably, according to the upper limb motion data, execute a kinematic modeling strategy, and use the forward kinematics algorithm to calculate the actual pose of the end effector, including: Obtain the angles θ1, θ2, …, θ n , where n is the number of exoskeleton joints, and θ i represents the angle of the i-th joint; A three-dimensional joint coordinate system is established for each joint, specifically as follows: Set the coordinate origin of the first joint, then align the coordinate origin of the second joint with the coordinate origin of the first joint, and so on to obtain the coordinate origin positions of each joint; For any joint, obtain the rotation axis of the joint and use the right-hand rule to determine the direction of the joint coordinate system; Use the homogeneous transformation matrix T i to represent the transformation action from the (i - 1)-th joint coordinate system to the i-th joint coordinate system. The said transformation action is divided into translation and rotation; where, θ i is the angle of the i-th joint, α i is the twist angle of the i-th joint, which is used to describe the angle between adjacent joint axes, a i is the link length between the i-th joint and the (i - 1)-th joint, and d i is the offset along the rotation axis of the i-th joint.

[0011] By obtaining the angles of each exoskeleton joint and establishing a three-dimensional joint coordinate system, the motion description of the exoskeleton is made more precise, achieving efficient motion control. By setting the coordinate origin of each joint, the joint coordinate systems of the entire exoskeleton system are unified, improving the consistency and computability of data processing. By using the right-hand rule to determine the direction of the joint coordinate system, the standardization of coordinate transformation is ensured, reducing calculation errors. By using the homogeneous transformation matrix to describe the translation and rotation between joints, the kinematic modeling of the entire system is made more rigorous, providing a reliable mathematical basis for subsequent motion control. By defining parameters such as joint angles, twist angles, link lengths, and offsets, the exoskeleton system can accurately describe the motion relationships between each joint, improving the accuracy of inverse kinematics solution.

[0012] Preferably, the method of performing kinematic modeling according to upper limb motion data and using the forward kinematics algorithm to calculate the actual pose of the end effector further includes: Obtain the transformation matrices T1, T2,..., T n ; Calculate the product of the transformation matrices of all joints to obtain T0 = T1 × T2 ×... × T n , where T0 is a 4*4 matrix; Obtain the matrix corresponding to the first three rows and the first three columns of T0 to get the actual pose of the end effector: where, the three column vectors respectively represent the x-axis, y-axis, and z-axis of the joint coordinate system corresponding to the joint after performing the actual pose in the direction of the joint coordinate system before performing the pose, and r ijIt represents the projection of the i-axis of the joint coordinate system corresponding to the joint after executing the actual pose on the j-axis of the joint coordinate system before executing the pose, where i, j = x, y, z; Obtain the matrix corresponding to the first column of the first three rows of T0 to get the actual position of the end effector: It represents the three-dimensional coordinates of the actual position of the end effector in the joint coordinate system.

[0013] By obtaining the transformation matrices of all joints and calculating their product, the pose of the end effector can be accurately deduced, thus realizing precise motion control. By obtaining the matrix of the first three rows and the first three columns, the pose information of the end effector can be obtained, making the pose control more intuitive and computable. By analyzing the relationship between the column vectors of the matrix, it can be ensured that the calculated pose conforms to the actual motion law and avoid pose drift caused by error accumulation. By obtaining the position of the end effector through the first column of the first three rows, the position control is made more precise, ensuring that the end effector can accurately reach the target position. By using a 4×4 transformation matrix to uniformly describe the motion transformation, the calculation efficiency can be improved and the motion control can be made more stable and reliable.

[0014] Preferably, the obtaining of the joint angles, the execution of the Jacobian matrix modeling strategy, and the establishment of the model between the actual pose and the joint angles include: The actual pose is represented as Then the actual pose is Execute the Jacobian matrix modeling strategy: ΔQ = J(θ)×Δθ, which means that the small change in the joint angle affects the actual pose of the end effector through the Jacobian matrix. Δθ is the small change in the joint angle, which is a vector composed of the angle changes of all joints. J(θ) is the Jacobian matrix, and ΔQ is the change in the actual pose of the end effector; Among them, the first 3 rows of the Jacobian matrix describe the influence of the change in the joint angle on the actual position of the end effector, and the last 3 rows describe the influence of the transformation of the joint angle on the actual pose of the end effector.

[0015] Through the Jacobian matrix modeling strategy, the small change in the joint angle can directly affect the pose of the end effector, thereby improving the flexibility and precision of motion control. By defining that the first 3 rows of the Jacobian matrix describe the position change and the last 3 rows describe the pose change, the calculation structure of the entire motion control system becomes clearer. By establishing the mathematical relationship between the change in the joint angle and the change in the pose of the end effector, the joint angle can be accurately adjusted to meet the requirements of the target pose. Through the numerical calculation of the Jacobian matrix, the motion state of the exoskeleton can be quickly adjusted, improving the response speed of the control system.

[0016] Preferably, the control of the exoskeleton joint to adjust the target joint angle includes: Calculate the pose error, which includes position error and attitude error; Calculate the position error e p =[P’ x -P x P’ y -P y P’ z -P z , where P'=[P’ x P’ y P’ z is the target position; Calculate the attitude error e R =R'×R -1 -I, where R' is the target attitude and I is the identity matrix; Obtain the pose error Obtain the error threshold ε P and ε R ; Judge whether ||e p ||≤ε P and ||e R ||≤ε R is true: If it is true, obtain the joint angle θ' at this time, denoted as the target joint angle; Control the exoskeleton joint to adjust the target joint angle.

[0017] By calculating the pose error, including position error and attitude error, the control system can clarify the error source and thus make targeted adjustments. By calculating the position error and attitude error, the system can optimize position control and attitude control respectively, improving the overall motion accuracy. By obtaining the error threshold and making a judgment, the system can operate stably within the error range, ensuring the motion control accuracy of the exoskeleton. By judging whether the error meets the threshold condition, the system can make adaptive adjustments to ensure the stability and accuracy of motion control.

[0018] Preferably, if the pose error is greater than the error threshold, execute the joint angle optimization strategy to adjust the joint angle, including: Judge whether ||e p ||≤ε P and ||e R ||≤ε R is true: If it is not true, then based on the Jacobian matrix, adjust the joint angle Δθ=J + ×e, where J +is the pseudo-inverse of the Jacobian matrix, used to adjust the joint angles, and e is the current pose error; The joint angle Δθ adjusted using the damping factor is Δθ = J T (JJ T + λI) -1 e, where λ is the damping factor used to avoid numerical instability; Update the joint angle θ” = θ' + Δθ, and θ min ≤ θ” ≤ θ max ,θ min ,θ max represents the minimum and maximum adjustable ranges of the joint.

[0019] Through the calculation based on the pseudo-inverse of the Jacobian matrix, the joint angles can be quickly adjusted, enabling the exoskeleton to converge to the target pose as soon as possible and improving the response speed. By introducing the damping factor, numerical instability can be avoided, ensuring the smoothness and controllability of the adjustment process. By updating the joint angles and ensuring they are within the adjustable range, the system can operate within a safe range, avoiding hardware damage caused by excessive motion. By continuously adjusting the joint angles, the exoskeleton can gradually optimize its motion state and improve the motion accuracy.

[0020] Preferably, during the process of controlling the exoskeleton joint to adjust the target joint angle, a simulation model feedback adjustment strategy is executed, comparing the simulation mapping result and the actual movement of the person, and adjusting the parameters of the simulation mapping model, including: Obtain the pose error e1, mechanical error e2, energy error e3, and coordination error e4 between the simulation mapping result and the actual movement of the person; calculate the target optimization function where β1, β2, β3, β4 represent error weighting factors.

[0021] By obtaining the errors between the simulation mapping result and the actual movement of the person, including pose error, mechanical error, energy error, and coordination error, the system can comprehensively evaluate the motion performance of the exoskeleton. By calculating the target optimization function and combining the error weighting factors, the optimization process can take multiple factors into account, ensuring that the exoskeleton system can achieve the optimal control effect. By optimizing the errors, the exoskeleton can execute the target motion more precisely, improving the adaptability and user experience.

[0022] In a second aspect, the present application provides a system for a data acquisition and processing method of a digital-twin-based wearable exoskeleton, adopting the following technical solution: A system for a data acquisition and processing method of a digital-twin-based wearable exoskeleton, including: Data acquisition exoskeleton module: Joint angle acquisition: 16 angle sensors cover joints such as the shoulder, elbow, forearm, and wrist. The degree-of-freedom distribution includes scapular rotation, 3 degrees of freedom in the shoulder, elbow rotation, forearm rotation, and wrist waving / tapping. Hand synchronization acquisition: Connect the Quantum Mocap Metagloves glove to achieve hand data synchronization. Digital twin simulation module: Real-time data mapping: Real-time map the collected joint angle data to the simulation mapping model. Kinematics modeling, mechanical analysis, energy management, and human-computer interaction optimization. Upper PC module: Data acquisition and processing: Receive and convert exoskeleton sensor data and perform filtering processing. Fault diagnosis: Perform anomaly monitoring based on digital twin technology. Simulation feedback optimization: Compare the simulation mapping results with the actual movements of the person and adjust the model parameters.

[0023] The present invention has the following beneficial effects: 1. This method for collecting and processing data of a wearable exoskeleton based on digital twin can effectively improve the motion accuracy and adaptability of the exoskeleton system by obtaining upper limb motion data and constructing a simulation mapping model of the exoskeleton joints. First, the motion data includes the position, angle, rotation axis, and motion limits of the exoskeleton joints, providing accurate input information for the subsequent simulation model. Defining the degrees of freedom of the exoskeleton joints in the simulation mapping model helps to refine the control and optimization of the model. Second, establish a digital human 3D model and perform simplification processing while retaining sufficient details for accurately simulating joint movements. Segment the joints of the simplified model to further improve the operability and flexibility of the model. Obtain the human body parameters corresponding to the digital human 3D model and adjust the size, length, and initial position of the exoskeleton joints according to these parameters to ensure that the exoskeleton design fits the natural structure of the human body and improves the wearing comfort and usage effect. Set the material and texture of the digital human 3D model and adjust the contrast color of the joints to enhance the user's immersion. Finally, build the background environment and adjust the lighting and camera angles to provide more intuitive visual feedback for subsequent refined operations.

[0024] 2. The method for collecting and processing data of the digital-twin-based wearable exoskeleton enables the system to accurately describe the motion state of the exoskeleton by obtaining the angles of each exoskeleton joint in the upper limb motion data, providing basic data for motion control. By establishing a three-dimensional joint coordinate system for each joint, the system can clarify the relative position relationship between different joints, improving the accuracy of motion calculation. By setting the coordinate origin of the first joint and aligning the coordinate origins of the subsequent joints, the consistency of the entire coordinate system is ensured, improving the standardization of motion modeling. By obtaining the rotation axis of the joint and using the right-hand rule to determine the direction of the joint coordinate system, the system can ensure the unity of the calculation direction and avoid calculation errors. By using the homogeneous transformation matrix to describe the transformation action from one joint coordinate system to the next, the system can accurately calculate the motion changes of the exoskeleton, improving the precision of motion control.

[0025] 3. The method for collecting and processing data of the digital-twin-based wearable exoskeleton enables the system to completely describe the motion state of the exoskeleton by obtaining the transformation matrices of all joints, improving the precision of motion control. By calculating the product of all joint transformation matrices, the system can deduce the pose of the end effector, improving the precise control of the system over the end motion. By extracting the first three rows and the first three columns of the transformation matrix, the system can calculate the attitude information of the end effector, improving the accuracy of attitude control. By analyzing the projection relationship of the matrix column vectors, the system can ensure that the calculated attitude conforms to the actual motion law and avoid the occurrence of attitude drift. By extracting the value of the first column of the first three rows of the transformation matrix, the system can calculate the actual position of the end effector, improving the precision of the motion trajectory. By using a 4×4 transformation matrix, the entire kinematic calculation process becomes more systematic, improving the calculation efficiency and control stability.

[0026] 4. The method for collecting and processing data of the digital-twin-based wearable exoskeleton enables the system to calculate how the change in joint angles affects the pose of the end effector by using the Jacobian matrix modeling strategy, thereby improving the flexibility and precision of motion control. By constructing the first 3 rows of the Jacobian matrix to describe the position change and the last 3 rows to describe the attitude change, the system can optimize the position control and attitude control respectively, improving the overall control effect. By establishing the mathematical relationship between the joint angle change and the end effector pose change, the system can accurately calculate the motion adjustment strategy, improving the adaptability of motion control. By the numerical calculation of the Jacobian matrix, the system can quickly adjust the motion state of the exoskeleton, improving the real-time response ability of the control system.

[0027] 5. The method for collecting and processing data of the wearable exoskeleton based on digital twin calculates the pose error, including position error and attitude error, enabling the system to quantify the deviation between the target pose and the actual pose, and improving the accuracy of the control system. By calculating the position error and attitude error, the system can optimize the position control and attitude control respectively, improving the overall motion precision. By obtaining the error threshold and making judgments, the system can adjust the joint angles within a reasonable error range, improving the stability of motion control. By judging whether the error meets the threshold condition, the system can dynamically adjust the motion state of the exoskeleton, improving the intelligent level of the system.

[0028] 6. The method for collecting and processing data of the wearable exoskeleton based on digital twin uses the pseudo-inverse calculation of the Jacobian matrix, enabling the system to quickly adjust the joint angles when the error exceeds the threshold, and improving the motion precision of the exoskeleton. By introducing a damping factor, the system can avoid numerical instability when adjusting the joint angles, improving the smoothness of the control process. By updating the joint angles and ensuring that they are within the adjustable range, the system can operate within a safe range, avoiding equipment damage caused by over-limit motion. By continuously optimizing the joint angles, the system can continuously improve the motion control precision and the smoothness of the motion process.

[0029] 7. The method for collecting and processing data of the wearable exoskeleton based on digital twin obtains the errors between the simulation mapping results and the actual actions of the person, including pose error, mechanical error, energy error, and coordination error, enabling the system to comprehensively evaluate the motion performance of the exoskeleton and improving the optimization ability of the control strategy. By calculating the target optimization function and combining the error weighting factor, the system can make trade-offs between multiple optimization goals, improving the rationality of the optimization. By optimizing the error, the exoskeleton can execute the target motion more precisely, improving the user experience and the reliability of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 It is a schematic diagram of the method flow of the present invention.

[0031] Figure 2 It is a schematic diagram of the system module of the present invention.

[0032] Figure 3 It is a schematic diagram of the structure of the wearable upper limb data collection exoskeleton of the present invention.

[0033] Figure 4 It is a schematic diagram of the Python script function of the digital twin simulation platform of the present invention.

[0034] Figure 5 It is a schematic diagram of the data collection process of the wearable exoskeleton data of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0035] Next, in combination with the accompanying drawings in the embodiments of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0036] Embodiment 1, referring to Figure 1 , a method for collecting and processing wearable exoskeleton data based on digital twin, including: obtaining the data measured by the exoskeleton joint sensor, denoted as upper limb movement data; According to the upper limb movement data, use digital twin technology to establish a simulation mapping model of the exoskeleton joint; According to the upper limb movement data, execute the kinematic modeling strategy, and use the forward kinematic algorithm to calculate the actual pose of the end effector, where the actual pose is divided into actual position and actual attitude; Obtain the target pose defined by the simulation mapping model, where the target pose is divided into target position and target attitude; Substitute the target pose into the inverse kinematics solution, specifically: S1. Obtain the joint angles, execute the Jacobian matrix modeling strategy, establish a model between the actual pose and the joint angles, and obtain the influence of the change in joint angles on the actual pose of the end effector; S2. Establish the error between the actual pose and the target pose, denoted as pose error; S3. Set the error threshold, and compare the pose error with the error threshold; S4. If the pose error is less than or equal to the error threshold, obtain the joint angles at this time, denoted as target joint angles; Control the exoskeleton joint to adjust the target joint angles; S5. If the pose error is greater than the error threshold, execute the joint angle optimization strategy, adjust the joint angles, and repeat S2 - S5; During the process of controlling the exoskeleton joint to adjust the target joint angles, execute the simulation model feedback adjustment strategy, compare the simulation mapping results with the actual actions of the person, and adjust the parameters of the simulation mapping model.

[0037] In this embodiment, referring to Figure 3, which is a schematic diagram of the structure of a wearable upper limb data acquisition exoskeleton. The wearable upper limb data acquisition exoskeleton mainly consists of a shoulder abduction joint, a shoulder abduction joint axis encoder, a shoulder rotation encoder, a shoulder rotation joint, a large arm abduction joint, a large arm abduction encoder, a large arm rotation encoder, a large arm sensing and driving mechanism, an elbow rotation joint, an elbow rotation encoder, a small arm fixing seat, a multi-degree-of-freedom small arm connecting rod, a palm ulnar deviation encoder, a palm connecting plate, a palm rotation encoder, a scapula base plate, a waist support seat, a waist fixing seat, and an ADC analog-to-digital conversion module, etc. This exoskeleton system contains a total of 16 angle sensors, which are used to collect 16 degrees of freedom data of the human upper limb, 8 degrees of freedom on each of the left and right sides, including scapula rotation, 3 degrees of freedom of the shoulder, elbow rotation, small arm rotation, and the waving and tapping degrees of freedom of the wrist.

[0038] The large arm part of the exoskeleton is equipped with a length adjustment mechanism, which can be adjusted according to the arm length of the operator to adapt to operators of different body types. The end of the exoskeleton is designed with a palm connecting plate, which can be closely combined with the Quantum Mocap Metagloves glove to jointly collect the angle data of the wrist and hand joints. The encoder outputs a voltage value, which needs to be converted into angle data through a Python script for subsequent data processing and recording. To ensure the smoothness of the digital human joint movement during the simulation process, it is necessary to filter the voltage signal of the encoder to reduce high-frequency noise and avoid jitter of the digital human during movement.

[0039] Using digital twin technology to establish a simulation mapping model of the exoskeleton joints based on the upper limb movement data, including: The upper limb movement data includes the position, angle, rotation axis, and movement limit of each exoskeleton joint; Define the degrees of freedom of the exoskeleton joints in the simulation mapping model; Establish a simulation mapping model of the exoskeleton joints, specifically: Obtain a digital human 3D model and perform simplification processing on the digital human 3D model; Perform joint segmentation on the simplified digital human 3D model; Obtain the human body parameters corresponding to the digital human 3D model and use the human body parameters to adjust the size, length, and initial position of each exoskeleton joint; Set the material and texture of the digital human 3D model and adjust the contrast color of the joints; Build the background environment of the digital human 3D model, create a background board and a floor, and adjust the lighting parameters, as well as the height and angle of the camera.

[0040] By acquiring upper limb motion data and constructing a simulation mapping model for exoskeleton joints, the motion accuracy and operation efficiency of the exoskeleton robot are significantly improved. By accurately obtaining the position, angle, rotation axis, and motion limits of each exoskeleton joint, detailed data support is provided to ensure that the simulation model can accurately reflect the motion characteristics of the exoskeleton. Defining the degrees of freedom of the joints makes the control of the model more refined, providing a reliable basis for subsequent kinematic modeling and adjustment. By acquiring a digital human 3D model and simplifying it, the computational load can be effectively reduced, enabling the simulation to balance accuracy and computational efficiency. The model after joint segmentation improves the operation flexibility and enhances the adaptability of the simulation system. Acquiring human body parameters and correspondingly adjusting the size, length, and initial position of the exoskeleton joints ensures a better fit between the exoskeleton and the human body, thereby enhancing the comfort and stability during wearing. Finally, setting up the background environment, adjusting the lighting and camera angles not only makes the simulation environment more realistic but also provides accurate visual feedback for actual operation, ensuring the efficient operation of the entire system.

[0041] According to the upper limb motion data, execute a kinematic modeling strategy, and use the forward kinematics algorithm to calculate the actual pose of the end effector, including: Obtain the angles θ1, θ2, …, θ of each joint of the exoskeleton in the upper limb motion data n , where n is the number of exoskeleton joints, and θ i represents the angle of the i-th joint; Establish a three-dimensional joint coordinate system for each joint, specifically: Set the coordinate origin of the first joint, then align the coordinate origin of the second joint with that of the first joint, and so on to obtain the coordinate origin positions of each joint; For any joint, obtain the rotation axis of the joint and use the right-hand rule to determine the direction of the joint coordinate system; Use the homogeneous transformation matrix T i to represent the transformation action from the (i - 1)-th joint coordinate system to the i-th joint coordinate system. The so-called transformation action is divided into translation and rotation; where θ i is the angle of the i-th joint, α i is the twist angle of the i-th joint, used to describe the angle between adjacent joint axes, a i is the link length between the i-th joint and the (i - 1)-th joint, and d i is the offset along the rotation axis of the i-th joint.

[0042] By obtaining the angles of each joint of the exoskeleton, the system can accurately describe the motion state of the joints, providing basic data for subsequent motion calculations. By establishing a three-dimensional joint coordinate system for each joint, the system can clarify the relative positions between different joints, improving the standardization of motion modeling and the accuracy of calculations. By setting the coordinate origin of the first joint and aligning the coordinate origins of the subsequent joints in sequence, the entire coordinate system has a unified reference benchmark, improving the coherence of the calculation process. By obtaining the rotation axes of the joints and using the right-hand rule to determine the directions of the joint coordinate systems, the system can maintain the consistency of the calculation directions, avoiding error accumulation. By using homogeneous transformation matrices to represent the transformation actions from one joint coordinate system to the next, the system can accurately describe the pose changes of the exoskeleton, improving the precision and predictability of motion control.

[0043] The method of performing kinematic modeling strategies based on upper limb motion data and calculating the actual pose of the end effector using forward kinematic algorithms further includes: Obtain the transformation matrices T1, T2, …, T of all joints n ; Calculate the product of the transformation matrices of all joints to obtain T0 = T1 × T2 × … × T n , where T0 is a 4*4 matrix; Obtain the matrix corresponding to the first three rows and the first three columns of T0 to get the actual pose of the end effector: Among them, the three column vectors respectively represent the directions of the x-axis, y-axis, and z-axis of the joint coordinate system corresponding to the joint after performing the actual pose in the joint coordinate system before performing the pose, and r ij represents the projection of the i-axis of the joint coordinate system corresponding to the joint after performing the actual pose on the j-axis of the joint coordinate system before performing the pose, where i, j = x, y, z; Obtain the matrix corresponding to the first three rows and the first column of T0 to get the actual position of the end effector: represents the three-dimensional coordinates of the actual position of the end effector in the joint coordinate system.

[0044] The method of obtaining joint angles, performing Jacobian matrix modeling strategies, and establishing a model between the actual pose and joint angles includes: Express the actual pose as Then the actual pose is Perform Jacobian matrix modeling strategies: ΔQ = J(θ) × Δθ indicates that the small change in joint angle affects the actual pose of the end - effector through the Jacobian matrix. Δθ is the small change in joint angle, which is a vector composed of the angle changes of all joints. J(θ) is the Jacobian matrix, and ΔQ is the change in the actual pose of the end - effector; Among them, the first 3 rows of the Jacobian matrix describe the influence of joint - angle changes on the actual position of the end - effector, and the last 3 rows describe the influence of joint - angle transformations on the actual orientation of the end - effector.

[0045] By obtaining the transformation matrices of all joints, the system can completely describe the motion state of the exoskeleton and improve the accuracy of motion control. By calculating the product of all joint - transformation matrices, the system can deduce the overall motion of the end - effector and improve the precise control ability of the end - motion. By extracting the first 3 rows and the first 3 columns of the transformation matrix, the system can accurately calculate the pose information of the end - effector and improve the accuracy of pose control. By analyzing the projection relationship of the matrix column vectors, the system can ensure that the calculated pose conforms to the actual motion law and avoid pose drift. By extracting the value of the first column of the first 3 rows of the transformation matrix, the system can accurately calculate the position of the end - effector and improve the accuracy of the motion trajectory. By using a 4×4 transformation matrix, the entire kinematic calculation process becomes more systematic, improving the calculation efficiency and control stability.

[0046] The control of the exoskeleton joint to adjust the target joint angle includes: Calculating the pose error, where the pose error includes position error and attitude error; Calculating the position error e p =[P’ x - P x P’ y - P y P’ z - P z , where P' = [P’ x P’ y P’ z is the target position; Calculating the attitude error e R =R'×R -1 - I, where R' is the target attitude and I is the identity matrix; Obtaining the pose error Obtaining the error thresholds ε P and ε R ; Judging ||e p ||≤ε P and ||e R ||≤ε RIs it true: If it is true, obtain the joint angle θ' at this time, denoted as the target joint angle; Control the exoskeleton joint to adjust the target joint angle.

[0047] By calculating the pose error, including position error and attitude error, the system can quantify the deviation between the target pose and the actual pose, improving the accuracy of the control system. By calculating the position error and attitude error, the system can optimize the position control and attitude control respectively, improving the overall motion precision. By obtaining the error threshold and making a judgment, the system can adjust the joint angle within a reasonable error range, improving the stability of the motion control. By judging whether the error meets the threshold condition, the system can dynamically adjust the motion state of the exoskeleton, improving the intelligence level of the system.

[0048] If the pose error is greater than the error threshold, execute the joint angle optimization strategy to adjust the joint angle, including: Judge ||e p ||≤ε P And ||e R ||≤ε R Is it true: If it is not true, then based on the Jacobian matrix, adjust the joint angle Δθ = J + ×e, where J + is the pseudo-inverse of the Jacobian matrix, used to adjust the joint angle, and e is the current pose error; Use the damping factor to stabilize the adjusted joint angle Δθ = J T (JJ T +λI) -1 e, where λ is the damping factor, used to avoid numerical instability; Update the joint angle θ” = θ' + Δθ, and θ min ≤θ”≤θ max ,θ min ,θ max represents the minimum and maximum ranges within which the joint can be adjusted.

[0049] By using the pseudo-inverse calculation of the Jacobian matrix, the system can quickly adjust the joint angle when the error exceeds the threshold, improving the motion accuracy of the exoskeleton. By introducing the damping factor, the system can avoid numerical instability when adjusting the joint angle, improving the smoothness of the control process. By updating the joint angle and ensuring it is within the adjustable range, the system can operate within a safe range, avoiding equipment damage caused by over-limit motion. By continuously optimizing the joint angle, the system can continuously improve the motion control accuracy and the smoothness of the motion process.

[0050] During the process of controlling the exoskeleton joint to adjust the target joint angle, a simulation model feedback adjustment strategy is executed. By comparing the simulation mapping results with the actual movements of the person, the parameters of the simulation mapping model are adjusted, including: Obtain the pose error e1, mechanical error e2, energy error e3, and coordination error e4 between the simulation mapping results and the actual movements of the person; calculate the target optimization function where β1, β2, β3, and β4 represent error weighting factors.

[0051] By obtaining the errors between the simulation mapping results and the actual movements of the person, including pose error, mechanical error, energy error, and coordination error, the system can comprehensively evaluate the motion performance of the exoskeleton and improve the optimization ability of the control strategy. By calculating the target optimization function and combining the error weighting factors, the system can balance between multiple optimization objectives and improve the rationality of the optimization. By optimizing the errors, the exoskeleton can execute the target motion more precisely, improving the user experience and system reliability.

[0052] Example 2, referring to Figure 2 , a system for a digital twin-based wearable exoskeleton data acquisition and processing method, including: Data acquisition exoskeleton module: Joint angle acquisition: 16 angle sensors covering joints such as the shoulder, elbow, forearm, and wrist, with the degree of freedom distribution including scapular rotation, 3 degrees of freedom of the shoulder, elbow rotation, forearm rotation, and wrist waving / tapping; Hand synchronization acquisition: Connect the Quantum Mocap Metagloves gloves to achieve hand data synchronization; Digital twin simulation module: Real-time data mapping: Map the collected joint angle data to the simulation mapping model in real time; Kinematics modeling, mechanical analysis, energy management, and human-computer interaction optimization; Upper PC module: Data acquisition and processing: Receive and convert the exoskeleton sensor data and perform filtering processing; Fault diagnosis: Perform anomaly monitoring based on digital twin technology; Simulation feedback tuning: Compare the simulation mapping results with the actual movements of the person and adjust the model parameters.

[0053] In this embodiment, the wearable exoskeleton data acquisition system based on digital twin mainly consists of a data acquisition exoskeleton entity, a digital twin simulation platform, and an upper PC. Among them, the data acquisition exoskeleton entity includes an exoskeleton device and an operator. The operator wears the exoskeleton device and completes specified actions. The angle sensors integrated on the exoskeleton collect the angle data of each joint of the operator and transmit these data to the upper PC in the form of sensor voltage values through a USB data cable.

[0054] After receiving the sensor voltage value, the upper PC transmits it to the digital twin simulation platform. Using the inverse kinematics mapping algorithm and the TCP (Transmission Control Protocol) communication protocol, the digital human model in the Unity engine can simulate the actions of the operator in real time. Finally, the digital twin simulation platform outputs the converted data in the form of joint angles and saves it in the upper PC. Through this system, the digital twin simulation platform realizes the accurate mapping of the data acquisition exoskeleton entity, supplementing and optimizing the deficiencies of the traditional data acquisition system.

[0055] Among them, in the construction of the digital twin simulation platform, the SolidWorks 3D CAD (Computer Aided Design) software is used to process the three-dimensional model, and the Unity engine under the Windows system is used as the development framework. The development environment includes C# as the development language of the simulation platform and Python as the data processing and transmission language.

[0056] In this embodiment, referring to Figure 4 , in the digital twin simulation platform, the main functions of the Python script include: 1. Read the sensor voltage value from the upper PC as the input for subsequent algorithms.

[0057] 2. Convert the voltage value collected by the sensor into angle data for inverse kinematics solution and control of the digital human joints in the bionic platform.

[0058] 3. Since the collected data contains high-frequency noise, the SG (Savitzky-Golay) filtering method is used to denoise the data. The SG filtering method is widely used in time series data and one-dimensional signal processing. By performing local polynomial fitting on the data points within the window and replacing the central point within the window with the value of the polynomial, the smoothing effect is achieved. This method can effectively remove noise interference and reduce the jitter phenomenon of the digital human joints during the simulation while retaining the data trend.

[0059] 4. Because one side of the exoskeleton arm where the sensor data is input has 8 degrees of freedom, while the data mapped to the digital human arm joints only requires 7 degrees of freedom, a mapping relationship for reducing the degrees of freedom needs to be established. Through the inverse kinematics mapping algorithm, the angle data of 7 degrees of freedom for each of the left and right arms is calculated.

[0060] 5. Save the mapped angle data to the upper PC and plot a curve of the joint angles changing with time to identify and eliminate abnormal data.

[0061] 6. Finally, transmit the mapped joint data to the Unity server through the TCP communication protocol.

[0062] In this embodiment, referring to Figure 5 , the data acquisition process of the wearable exoskeleton based on the digital twin simulation platform mainly includes: 1. The operator wears the exoskeleton device and adjusts the length adjustment mechanism of the upper arm and the tightness of the elastic fixing belt to ensure that the device is stably worn in the appropriate position, so that each angle sensor corresponds to the corresponding joint of the operator.

[0063] 2. Connect the exoskeleton device to the upper PC. The operator maintains the preset initial posture and performs initial data verification on the angle sensors of the left and right arms respectively. Observe whether the initial values of the joint angle sensors are close to the set values. If the gap is large, check whether the encoder is loose, adjust and fix it, update the set value and re-verify the initial data.

[0064] 3. Start the Unity3D digital twin simulation platform built based on the Unity engine, adjust the initial camera position and angle, and start the digital twin simulation platform.

[0065] 4. Run the Python script to convert the voltage values collected by the sensor into angle data, record and transmit them to the simulation platform to realize the real-time simulation of the digital human's actions by the operator.

[0066] 5. The operator completes the corresponding actions according to the task requirements, ensuring that the action amplitude is within the specified range. Avoid actions with too large amplitude or reaching the limit range, because such actions may cause the inverse kinematics mapping algorithm to fail to solve, generate singularities and cause the system to collect abnormal data.

[0067] 6. Observe the matching situation between the actions of the digital human in the digital twin simulation platform and the operator. If there is a large deviation, consider adjusting the parameters in the mapping relationship; if singularities occur, adjust the operator's action range; if the actions do not conform to the natural habits of the human body, adjust the joint constraint conditions.

[0068] 7. After the data collection is completed, turn off the digital twin simulation platform, and the joint data will be automatically saved in the upper PC. View the curve graph of the change of each joint angle over time drawn by the Python script, and combine it with the recorded digital human motion video to detect and eliminate abnormal data. If there is a large jump in the angle of a certain joint, it may be that the encoder has experienced a zero-crossing phenomenon. The initial value of the encoder should be corrected to avoid the zero-crossing situation. After correction, the initial data verification in step 2 needs to be executed again.

[0069] It should be noted that in this article, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device comprising a series of elements not only includes those elements, but also includes other elements not expressly listed, or also includes elements inherent to such process, method, article or device.

[0070] The above are only the preferred embodiments of the present invention. It should be pointed out that for those of ordinary skill in the art, without departing from the technical principle of the present invention, several improvements and refinements can be made, and these improvements and refinements should also be regarded as the protection scope of the present invention.

Claims

1. A wearable exoskeleton data acquisition and processing method based on digital twins, characterized in that: include: Acquire data measured by the exoskeleton joint sensors and record them as upper limb motion data; Based on the upper limb motion data, a simulation mapping model of the exoskeleton joint is established using digital twin technology; According to the upper limb motion data, a kinematic modeling strategy is executed, and a forward kinematics algorithm is used to calculate the actual posture of the end effector, wherein the actual posture is divided into an actual position and an actual posture; Acquire a target posture defined by a simulation mapping model, wherein the target posture is divided into a target position and a target posture; Substitute the target posture into the inverse kinematics solution, specifically: S1. Obtain the joint angle, execute the Jacobian matrix modeling strategy, establish a model between the actual posture and the joint angle, and obtain the influence of the change of the joint angle on the actual posture of the end effector; S2, establish the error between the actual posture and the target posture, recorded as posture error; S3, setting an error threshold, and comparing the posture error with the error threshold; S4. If the posture error is less than or equal to the error threshold, obtain the joint angle at this time and record it as the target joint angle; Control the exoskeleton joints to adjust the target joint angles; S5, if the posture error is greater than the error threshold, execute the joint angle tuning strategy, adjust the joint angle, and repeat S2-S5; In the process of controlling the exoskeleton joint to adjust the target joint angle, the simulation model feedback adjustment strategy is executed, the simulation mapping results are compared with the actual movements of the personnel, and the parameters of the simulation mapping model are adjusted.

2. The wearable exoskeleton data acquisition and processing method based on digital twin according to claim 1 is characterized in that: The method of establishing a simulation mapping model of an exoskeleton joint using digital twin technology based on upper limb motion data includes: The upper limb motion data includes the position, angle, rotation axis and motion limit of each exoskeleton joint; Define the degrees of freedom of the exoskeleton joints in the simulation mapping model; Establish a simulation mapping model of the exoskeleton joints, specifically: Acquire a 3D model of a digital human and simplify the 3D model of the digital human; Perform joint segmentation on the simplified 3D digital human model; Obtain the human body parameters corresponding to the digital human 3D model, and use the human body parameters to adjust the size, length and initial position of each exoskeleton joint; Set the material and texture of the digital human 3D model and adjust the contrast color of the joints; Build the background environment of the digital human 3D model, create the background plate and floor, and adjust the lighting parameters and camera height and angle.

3. The wearable exoskeleton data acquisition and processing method based on digital twin according to claim 1 is characterized in that: The kinematic modeling strategy is executed according to the upper limb motion data, and the actual position and posture of the end effector is calculated using the forward kinematics algorithm, including: Get the angles θ1, θ2, …, θ of each joint of the exoskeleton in the upper limb motion data n , where n is the number of exoskeleton joints, θ i represents the angle of the i-th joint; Establish a three-dimensional joint coordinate system for each joint, specifically: Set the coordinate origin of the first joint, then the coordinate origin of the second joint will be aligned with the coordinate origin of the first joint, and so on to get the coordinate origin position of each joint; For any joint, obtain the rotation axis of the joint and use the right-hand rule to determine the direction of the joint coordinate system; Use the homogeneous transformation matrix T i represents the transformation from the i-1th joint coordinate system to the i-th joint coordinate system. The transformation is divided into translation and rotation. Among them, θ i is the angle of the i-th joint, α i is the torsion angle of the i-th joint, which is used to describe the angle between adjacent joint axes, a i is the length of the connecting rod between the ith joint and the i-1th joint, d i is the offset along the rotation axis of the ith joint.

4. The wearable exoskeleton data acquisition and processing method based on digital twin according to claim 3 is characterized in that: The method further comprises: executing a kinematic modeling strategy according to the upper limb motion data and using a forward kinematics algorithm to calculate the actual position and posture of the end effector. Get the transformation matrices T1, T2, ..., T of all joints n ; Calculate the product of the transformation matrices of all joints and get T0 = T1 × T2 × … × T n , where T0 is a 4*4 matrix; Get the matrix corresponding to the first three rows and first three columns of T0 to get the actual posture of the end effector: Among them, the three column vectors Respectively represent the directions of the x-axis, y-axis, and z-axis of the joint coordinate system corresponding to the joint after the actual posture is executed in the joint coordinate system before the posture is executed, r ij Represents the projection of the i-axis of the joint coordinate system corresponding to the joint after the actual posture is executed on the j-axis of the joint coordinate system before the posture is executed, i, j = x, y, z; Get the matrix corresponding to the first three rows and the first column of T0 to get the actual position of the end effector: Represents the three-dimensional coordinates of the actual position of the end effector in the joint coordinate system.

5. The wearable exoskeleton data acquisition and processing method based on digital twin according to claim 4 is characterized in that: The method of obtaining joint angles, executing Jacobian matrix modeling strategy, and establishing a model between actual posture and joint angles includes: The actual posture Expressed as The actual posture is Implement the Jacobian matrix modeling strategy: ΔQ = J(θ) × Δθ, which means that a small change in the joint angle affects the actual position and posture of the end effector through the Jacobian matrix. Δθ is a small change in the joint angle, which is a vector composed of the angle changes of all joints. J(θ) is the Jacobian matrix, and ΔQ is the change in the actual position and posture of the end effector. Among them, the first three rows of the Jacobian matrix describe how the change of joint angle affects the actual position of the end effector, and the last three rows describe how the transformation of joint angle affects the actual posture of the end effector.

6. The wearable exoskeleton data acquisition and processing method based on digital twin according to claim 5 is characterized in that: The controlling the exoskeleton joint to adjust the target joint angle comprises: Calculating a posture error, wherein the posture error includes a position error and a posture error; Calculate the position error e p =[P x, -P x P y, -P y P z, -P z ], where P' = [P x, P y, P z, ] is the target position; Calculate the attitude error e R =R'×R -1 -I, where R' is the target posture and I is the identity matrix; Get the pose error Get the error threshold ε P and ε R ; Judgement||e p ||≤ε P And||e R ||≤ε R Is true: If true, obtain the joint angle θ' at this time and record it as the target joint angle; Control the exoskeleton joints to adjust the target joint angles.

7. The wearable exoskeleton data acquisition and processing method based on digital twin according to claim 6 is characterized in that: If the posture error is greater than the error threshold, the joint angle tuning strategy is executed to adjust the joint angle, including: Judgement||e p ||≤ε P And||e R ||≤ε R Is true: If not true, adjust the joint angle Δθ = J based on the Jacobian matrix + ×e, where J + is the pseudo-inverse of the Jacobian matrix, which is used to adjust the joint angles, and e is the current posture error; The joint angle Δθ adjusted using the damping factor is stabilized by T (JJ T +λI) -1 e, where λ is the damping factor used to avoid numerical instability; Update joint angle θ" = θ' + Δθ, and θ min ≤θ”≤θ max ,θ min ,θ max Indicates the minimum and maximum ranges within which the joint can be adjusted.

8. The wearable exoskeleton data acquisition and processing method based on digital twin according to claim 1 is characterized in that: In the process of controlling the exoskeleton joint to adjust the target joint angle, executing the simulation model feedback adjustment strategy, comparing the simulation mapping result with the actual action of the person, and adjusting the parameters of the simulation mapping model, including: Obtain the pose error e1, mechanical error e2, energy error e3 and coordination error e4 of the simulation mapping results and the actual action of the personnel; calculate the target optimization function Among them, β1, β2, β3, β4 represent error weighting factors.

9. A system for implementing the wearable exoskeleton data acquisition and processing method based on digital twins as described in claim 1, characterized in that: include: Data collection exoskeleton module: Joint angle acquisition: 16 angle sensors covering the shoulder, elbow, forearm, wrist and other joints. The degrees of freedom distribution includes scapula rotation, shoulder 3 degrees of freedom, elbow rotation, forearm rotation, wrist swing / tap; Hand synchronization acquisition: connect to Quantum Mocap Metagloves gloves to achieve hand data synchronization; Digital Twin Simulation Module: Real-time data mapping: Map the collected joint angle data to the simulation mapping model in real time; Kinematic modeling, mechanical analysis, energy management, and human-computer interaction optimization; Host PC module: Data acquisition and processing: Receive and convert exoskeleton sensor data and perform filtering; Fault diagnosis: abnormality monitoring based on digital twin technology; Simulation feedback tuning: Compare the simulation mapping results with the actual actions of the personnel and adjust the model parameters.

Images