Fuzzy adaptive impedance control method for upper limb rehabilitation training robot

Abstract

The invention discloses a fuzzy self-adaptive impedance control method for an upper limb rehabilitation training robot, and the method comprises the steps: sensing the motion performance and active output degree of a user in real time through a force and position sensor disposed in the robot, and carrying out the intelligent decision through a fuzzy reasoning mechanism; real-time, autonomous and smooth adjustment of impedance parameters of the robot in the upper limb rehabilitation training process is achieved. The method can accurately adapt to the dynamic changing exercise ability of the user, provides proper assistance when the user is lack of strength, and actively reduces intervention when the ability of the user is enhanced, so that the intelligent level of the rehabilitation robot can be improved, and the training safety and comfort are guaranteed.

Description

Technical Field

[0001] This invention relates to the field of medical rehabilitation robot control technology, and in particular to a fuzzy adaptive impedance control method for an upper limb rehabilitation training robot, used to achieve personalized adaptive adjustment of robot assistance force. Background Technology

[0002] Currently, upper limb motor dysfunction caused by stroke, spinal cord injury, and internal and external limb injuries severely impacts patients' quality of life. Clinical studies have shown that high-intensity, highly repetitive, task-oriented rehabilitation training is an effective means of promoting neural function remodeling and motor function recovery. Upper limb rehabilitation training robots, with their unique advantages of providing standardized, continuous, and quantifiable training, have become an important tool in modern rehabilitation medicine, greatly alleviating the physical burden on therapists. Impedance control, as a mainstream robot human-machine interaction control strategy, provides users with a relatively compliant training environment by establishing a dynamic relationship between the robot's end effector position / velocity and interaction force, and has been widely used in rehabilitation robots.

[0003] However, most existing impedance-controlled rehabilitation robot systems employ fixed-parameter impedance models, which inherently limit their control performance. Throughout the training process, the robot maintains a constant level of assistance, guiding stiffness, and motion damping, failing to adapt to dynamic fluctuations in the user's muscle strength, fatigue level, and attention level, and struggling to match the personalized needs of different rehabilitation stages. This results in a lack of intelligence in the training process: insufficient assistance may prevent users from completing training tasks when fatigued, dampening their enthusiasm; while excessive assistance may deprive users of opportunities for active movement when their abilities are good, hindering neural function reconstruction. Fixed-parameter strategies require manual intervention and adjustment by therapists, which is inefficient and makes it difficult to achieve precise "on-demand assistance," while also struggling to achieve a dynamic optimal balance between ensuring training safety and pursuing training effectiveness.

[0004] To address these issues, existing research has proposed several improved methods, such as using surface electromyography (sEMG) signals to predict movement intentions and switch control modes, or adjusting segmented parameters based on tracking errors. However, these methods typically rely on additional bioelectrical signal acquisition equipment, resulting in system complexity, high costs, and poor interference resistance; or their parameter adjustment strategies remain rigid, failing to achieve smooth, precise, and continuous adaptive changes, thus their reliability and universality in practical clinical applications need improvement.

[0005] Therefore, a novel intelligent control method is needed that can fully utilize the robot's built-in force and position sensor information to achieve real-time, autonomous, and smooth adjustment of impedance parameters without relying on complex external biosignals, thereby accurately adapting to the instantaneous changes and long-term evolution of the user's motor abilities. This is of great significance for improving the intelligence level and training safety of rehabilitation robots. Summary of the Invention

[0006] The purpose of this invention is to overcome the shortcomings of existing technologies and provide a fuzzy adaptive impedance control method for upper limb rehabilitation training robots. This method can assess the user's interaction intentions and abilities online in real time, and dynamically adjust the parameters of the impedance controller using fuzzy inference, thereby achieving personalized adaptive adjustment of the robot's assistance force. This is of great significance for improving the intelligence level and training safety of rehabilitation robots.

[0007] To achieve the above objectives, the technical solution adopted by the present invention is as follows: This invention provides a fuzzy adaptive impedance control method for an upper limb rehabilitation training robot. The method sequentially executes the following steps within one control cycle, repeating cyclically: S1: Real-time acquisition of sensor data from the upper limb rehabilitation training robot, and through positive kinematics function, gravity compensation and filtering processing, to obtain effective actual pose information and human-computer interaction force information; S2: Calculate the trajectory tracking error based on the expected trajectory and the actual pose information, and calculate the interaction force index reflecting the user's active force output based on the human-computer interaction force information; S3: Input the trajectory tracking error and the interaction force index into the fuzzy inference engine for processing; the fuzzy inference engine outputs the adjustment amount of the impedance parameter according to the preset fuzzy rule library, the impedance parameter including stiffness and damping; S4: Update the impedance parameters based on the adjustment amount and the impedance parameters of the previous control cycle, and perform smoothing filtering on the updated impedance parameters to generate impedance parameters that change smoothly in the current control cycle. S5: Based on the desired trajectory, actual pose information, and smoothed impedance parameters, the end-effector command force is calculated through the impedance control model, and the end-effector command force is mapped into joint control torque through Jacobian matrix transpose, and sent to each joint actuator of the robot to execute motion control for the current cycle.

[0008] Furthermore, in S1, firstly, in each control cycle... Obtain joint angle vectors from the robot's optical encoder or resolver. And acquire raw readings from the six-dimensional force / torque sensor. ; Then, using the robot's forward kinematics model, the end effector's position in the current control cycle is calculated. Actual pose at joint angles The calculation expression is:

[0009] in, Represents the robot's positive kinematics function; The transformation matrix from the force sensor coordinate system {S} to the robot end effector coordinate system {E} is obtained through calibration. and the gravity vector of the end link. Perform gravity compensation and calculate interaction forces:

[0010] in, This represents the interaction force after compensation, expressed in the end coordinate system {E}; Finally, a first-order Butterworth filter is used to... and Filtering is performed to obtain effective actual pose information and human-computer interaction force information.

[0011] Furthermore, in S2, the calculation process for the trajectory tracking error includes: Desired trajectory points based on the current control cycle It includes the desired position and attitude, and calculates the deviation between the actual pose and the desired pose. :

[0012] To obtain a scalar measure of error, the Euclidean norm of the error vector is calculated for position errors. and attitude error Weighted composite trajectory tracking error:

[0013] in, Indicates the trajectory tracking error. and These are the weighting coefficients; The calculation process of the interaction index includes: Calculate the interaction force vector amplitude :

[0014] in, It is the force component. It is the torque component. and These are weighting coefficients; Normalize it to The interval, normalized based on a preset maximum expected user output. :

[0015] in, It is the normalized interaction force index.

[0016] Furthermore, in step S3, the preset fuzzy rule base adopts fuzzy rules based on expert experience, and the rule form is as follows:

[0017] in, and ∈ Fuzzy set {small, medium, large}, and ∈ Fuzzy set {negative large, negative small, zero, positive small, positive large}, Indicates stiffness, Indicates damping, Indicates the stiffness adjustment amount. This indicates the amount of damping adjustment.

[0018] Furthermore, in S3, the fuzzy inference engine adopts the Mamdani model and uses the centroid method for defuzzification.

[0019] Furthermore, in step S4, a low-pass filter is used for smoothing filtering.

[0020] Furthermore, in S5, the impedance control model is:

[0021] in, It is the set inertia matrix. , These are the smoothed impedance parameters. They are the first and second derivatives, respectively. Represents the desired trajectory point, Indicates the actual pose. This represents the first derivative of the desired trajectory point. This represents the second derivative of the actual pose. The second derivative of the desired trajectory point, This represents the second derivative of the actual pose.

[0022] Compared with the prior art, the fuzzy adaptive impedance control method for an upper limb rehabilitation training robot provided by the present invention has at least the following beneficial effects: This invention uses the robot's sensors to perceive the user's movement performance and active effort level in real time, and utilizes fuzzy reasoning mechanisms for intelligent decision-making. This enables real-time, autonomous, and smooth adjustment of the robot's impedance parameters during upper limb rehabilitation training. This method facilitates precise adaptation to the user's dynamically changing motor abilities, providing appropriate assistance when the user is fatigued and proactively reducing intervention as their abilities improve. This enhances the intelligence level of the rehabilitation robot and ensures training safety and comfort.

[0023] Other features and advantages of the invention will be set forth in the following description, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures particularly pointed out in the written description and the accompanying drawings.

[0024] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0025] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0026] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used together with the embodiments of the invention to explain the invention and do not constitute a limitation thereof.

[0027] Figure 1 This is a schematic diagram of a fuzzy adaptive impedance control method for an upper limb rehabilitation training robot provided in an embodiment of the present invention.

[0028] Figure 2 A schematic diagram of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0029] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are some embodiments of the present invention, but not all embodiments.

[0030] In the description of this invention, it should be noted that some processes described in this application specification and drawings include multiple operations that appear in a specific order. However, it should be clearly understood that these operations may be performed in any order or in parallel. Furthermore, various numbers are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0031] Therefore, the following detailed description of the embodiments of the invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the invention without inventive effort are within the scope of protection of the invention.

[0032] First, let's introduce the relevant technologies involved in the embodiments of this invention: Fuzzy control is an intelligent control method based on fuzzy logic theory. It addresses the uncertainties and nonlinearities of complex systems by simulating human experience and reasoning. Unlike traditional control strategies that rely on precise mathematical models, fuzzy control transforms expert knowledge into linguistic rules (e.g., "if the temperature is high, then the cooling power is moderate"), quantifies fuzzy concepts using membership functions, and achieves a mapping from fuzzy inputs to precise outputs through fuzzy reasoning. This method is suitable for systems where models are difficult to construct precisely, where interference exists, or where there is strong nonlinearity. It has wide applications in industrial process control, intelligent home appliances, and robotics, demonstrating the advantages of human-machine collaborative processing of fuzzy information.

[0033] Upper limb rehabilitation robots are advanced medical devices integrating robotics, rehabilitation medicine, and artificial intelligence. They aim to help users with upper limb motor dysfunction due to stroke, trauma, or neurological diseases regain their motor abilities by providing high-intensity, standardized, and task-oriented training. Typically, they support the user's limb through an exoskeleton or end-effector structure. Based on real-time sensing of motor intentions and physiological states, they adaptively adjust assistive forces and training strategies to achieve neural function remodeling and motor relearning. This technology not only significantly improves the efficiency and accuracy of rehabilitation training but also provides objective evidence for treatment decisions through data-driven quantitative assessment, representing an important direction for the intelligent development of modern rehabilitation medicine.

[0034] Impedance control is a robot interaction control strategy based on the dynamic relationship between force and position. Its core objective is to adjust the impedance characteristics (i.e., stiffness, damping, and mass) of the robot's end effector to enable it to exhibit desired dynamic behavior when in contact with the environment. Unlike traditional position control, which aims for precise trajectory tracking, impedance control does not directly control the output force. Instead, it establishes a differential relationship between force and position deviation by constructing a virtual spring-damped system, enabling the robot to adaptively respond to environmental uncertainties and changes in contact forces. This method is widely used in tasks requiring compliant interaction, such as assembly, grinding, rehabilitation robot-assisted training, and human-robot collaboration, significantly improving the robot's safety and adaptability in unstructured environments.

[0035] See Figure 1 As shown, this invention provides a fuzzy adaptive impedance control method for an upper limb rehabilitation training robot, the method comprising the following steps: Sensor signal acquisition and preprocessing: The system reads the raw data from the force sensor and position encoder, and performs coordinate transformation, gravity compensation and filtering to obtain clean and usable interactive force and position information.

[0036] Fuzzy Inference Engine: The Core of Intelligent Decision Making. It takes the aforementioned performance metrics as input and uses a built-in, expert-knowledge-based fuzzy rule base to reason and make intelligent decisions on how to adjust the impedance parameters.

[0037] Calculate the impedance parameter adjustment amount: the precise stiffness output by the fuzzy inference engine ( Adjustment amount and damping ( Adjustment amount.

[0038] Update and smooth impedance parameters: Apply the adjustment to the parameter value of the previous cycle and perform low-pass filtering to generate smoothly changing impedance parameters for the current cycle, ensuring the continuity and stability of the robot's assisted behavior.

[0039] Impedance controller calculates control torque: Based on the latest smoothed impedance parameters and the deviation between the desired trajectory and the actual position, the joint control torque to be applied to the robot is calculated using the impedance control law.

[0040] Output torque to the robot: Send the calculated control torque command to the robot's joint actuators to drive the motors to perform motion, providing adaptive assistance to the user.

[0041] This invention executes the following steps sequentially within one control cycle, repeating cyclically. The system inputs are the desired trajectory and sensor readings, and the output is the joint control torque sent to the robot actuator. A detailed description follows.

[0042] 1. The first step is signal acquisition and preprocessing, which aims to obtain raw data from the sensor and perform filtering and conversion to provide a clean and usable signal for subsequent calculations.

[0043] Read sensor data in each control cycle The system reads the joint angle vector from the optical encoder or rotary transformer. Where n is the number of robot joints, and raw readings are taken from the six-dimensional force / torque sensor. .

[0044] The actual pose of the end effector is calculated using the robot's forward kinematics model, taking into account the actual pose of the end effector at the current joint angle. (Usually includes 3D position and 3D orientation, such as Euler angles or quaternions):

[0045] in The positive kinematics function representing the robot.

[0046] Interactive force calculation and filtering begin with gravity compensation to calculate pure interactive forces. It is assumed that the transformation matrix from the force sensor coordinate system {S} to the robot end-effector coordinate system {E} has been obtained through calibration. and the gravity vector of the end link. :

[0047] in This is the compensated interaction force / torque vector represented in the end-effector coordinate system {E}. Subsequently, for... and Low-pass filtering is performed to suppress high-frequency noise. A first-order Butterworth filter is used:

[0048]

[0049] in , The filtered signal For the sake of brevity, the following text will be... and Both refer to the filtered signal.

[0050] 2. The purpose of calculating evaluation indicators for athletic performance and initiative is to quantify the processed sensor data into evaluation indicators that reflect the user's athletic performance and initiative.

[0051] First, calculate the trajectory tracking error, given the desired trajectory point at the current time. (Includes desired position and orientation). Calculate the deviation between the actual pose and the desired pose:

[0052] To obtain a scalar measure of error, the Euclidean norm (2-norm) of the error vector is often calculated. For position errors... and attitude error (As represented by the angle axis), the total error can be weighted and synthesized. :

[0053] in and These are weighting coefficients used to balance the importance of position and attitude errors. It is the trajectory tracking error before normalization.

[0054] Then, the normalized interaction force index is calculated, which reflects the user's level of active effort. The interaction force vector is then calculated. Amplitude:

[0055] in It is the force component. It is the torque component. and These are weighting coefficients; To make it applicable to fuzzy reasoning, it is normalized to Interval. Normalization is based on a preset maximum expected user output. :

[0056] in It is the normalized interaction force index.

[0057] 3. Performing fuzzy inference and impedance parameter adjustment aims to address the error. and Interaction Index Through fuzzy reasoning, the adjustment amount of the impedance parameter can be obtained in real time.

[0058] First, perform fuzzification and define the input variables. and The fuzzy set and its membership function are: Define three fuzzy sets: small (S), medium (M), and large (L), as... Define three fuzzy sets: small (S), medium (M), and large (L). Calculate the current precise input value. and Membership degree of each fuzzy set :

[0059] Then, a fuzzy rule base is established, and fuzzy rules based on expert experience are created. The rule format is as follows:

[0060] Define 9 rules: IF e is S AND fint is L THEN ΔK is NL (negative large), ΔB is NL IF e is S AND f_int is M THEN ΔK is NS (negative small), ΔB is NS IF e is S AND f_int is S THEN ΔK is ZE (zero), ΔB is ZE IF e is M AND f_int is L THEN ΔK is NS, ΔB is NS IF e is M AND f_int is M THEN ΔK is ZE, ΔB is ZE IF e is M AND f_int is S THEN ΔK is PS (positive small), ΔB is PS IF e is L AND f_int is L THEN ΔK is ZE, ΔB is ZE IF e is L AND f_int is M THEN ΔK is PS, ΔB is PS IF e is L AND f_int is S THEN ΔK is PL (positive large), ΔB is PL Then, fuzzy inference and defuzzification are performed, using the Mamdani fuzzy model and the center of gravity method, with stiffness adjustment amount... For example (damping) Similarly): (1) Reasoning: For each rule i, its trigger strength The minimum value of the membership degree is input (AND operation).

[0061]

[0062] Synthesis: The output fuzzy set of each rule (from...) (Membership function definition) based on trigger strength Perform cropping, then take the union to obtain the total output fuzzy set.

[0063] (3) Defuzzification: Find the centroid of the total output fuzzy set to obtain the precise adjustment output.

[0064]

[0065] in It is a rule The corresponding output fuzzy set The center value (e.g., the center of NL is -ΔK_max, and the center of PL is ΔK_max). This is the stiffness adjustment amount for this cycle.

[0066] 4. The purpose of updating and smoothing impedance parameters is to apply the adjustment amount obtained by fuzzy inference and filter the updated parameters to ensure smooth changes.

[0067] First, add the adjustment amount to the impedance parameter of the previous cycle and update accordingly:

[0068]

[0069] The temporary parameters are then low-pass filtered to generate smooth parameters for the current control cycle. This is crucial to prevent robot jitter caused by sudden parameter changes.

[0070]

[0071] in It is the smoothing filter coefficient. The closer its value is to 1, the stronger the smoothing effect, but the slower the response.

[0072] 5. The purpose of calculating and outputting the impedance control law is to calculate the final control command based on the latest smooth impedance parameters and motion state.

[0073] First, the target impedance is calculated using position-based impedance control. The target impedance model describes the dynamic relationship between the ideal robot and its environment:

[0074] in It is the set inertia matrix (usually a diagonal matrix). and It is a scalar and The resulting diagonal matrix. These are the first-order (velocity) and second-order (acceleration) derivatives, respectively.

[0075] Then, control commands are generated. To enable the robot to exhibit the aforementioned target impedance characteristics, a reference auxiliary force needs to be calculated. Ignoring the difference between the actual robot's inertia and its model, the command force can be designed as follows:

[0076] in and It is possible Obtained by differential filtering.

[0077] Finally, through the transpose of the robot's Jacobian matrix Mapping the end-effector command force to the control torque in joint space :

[0078] Will The data is sent to the robot's joint actuators to execute motion control for the current cycle.

[0079] A typical application of this invention is in the use of an upper limb rehabilitation robot to assist users in trajectory tracking training. In this scenario, the system first uses the robot's built-in force and position sensors to continuously collect interactive force and actual position information during the user's movement in real time. Based on this real-time data, the system calculates two core indicators with clear physiological significance: one is the tracking error reflecting the accuracy of the user's movement trajectory, and the other is the interactive force index characterizing the user's active effort. These two indicators are immediately input into a fuzzy inference system with pre-set expert knowledge. This fuzzy inference system, as the core of intelligent decision-making, automatically assesses the user's current ability state and generates corresponding impedance parameter adjustment instructions accordingly. After smoothing and filtering, these instructions dynamically and continuously modify the stiffness and damping parameters in the robot's impedance control model.

[0080] Ultimately, the impedance controller uses these adaptively adjusted new parameters to calculate and output the most suitable control torque in real time. This allows the robot to smoothly transition between "high-assisted" and "low-assisted" modes, accurately matching the fluctuating muscle strength and coordination abilities of the user during movement, thus achieving truly personalized, on-demand assisted rehabilitation training. The entire process requires no human intervention, forming an autonomously operating, continuously optimizing intelligent rehabilitation closed loop.

[0081] From the description of the above embodiments, those skilled in the art will understand that the present invention provides a fuzzy adaptive impedance control method for an upper limb rehabilitation training robot, and the present invention has the following advantages: Multi-source information fusion perception: Instead of using position or force signals alone, it integrates two key indicators, trajectory tracking error (reflecting motion performance) and human-computer interaction force (reflecting active intention), as a comprehensive basis for the system to perceive the user's state.

[0082] A fuzzy logic-based intelligent decision-making mechanism: This method uses fuzzy reasoning to transform complex, nonlinear rehabilitation physician adjustment strategies into a clear "IF-THEN" rule base. It does not rely on precise mathematical models, effectively handles the uncertainty and fuzziness of motion states, and achieves a decision-making process similar to that of humans.

[0083] Real-time and autonomous adjustment of impedance parameters: Breaking through the limitations of traditional fixed impedance parameters, stiffness and damping can dynamically and adaptively change according to the user's real-time performance without manual intervention, greatly improving the machine's autonomous intelligence level and helping to enhance the personalization and effectiveness of rehabilitation training.

[0084] Smoothness of parameter changes: A smoothing filtering algorithm is introduced in the parameter update process to ensure that the change of impedance parameters is continuous and gradual, which fundamentally avoids robot shaking or discontinuous movement caused by sudden parameter changes, and ensures the safety and comfort of the training process.

[0085] Additionally, refer to Figure 2 As shown, this embodiment of the invention also provides an electronic device, which may include a processor, a memory, a communication bus and a communication interface, and may also include a computer program stored in the memory and executable on the processor. The processor executes the computer program to implement a fuzzy adaptive impedance control method for an upper limb rehabilitation training robot as described in the above method embodiment.

[0086] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, software systems, electronic devices, or computer program products, etc. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product implemented on one or more storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0087] It should be noted that the word "comprising" does not exclude the presence of components or steps not listed in the claims. The words "a" or "an" preceding a component do not exclude the presence of a plurality of such components. This invention can be implemented by means of hardware comprising several different components and by means of a suitably programmed computer.

[0088] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The same or similar parts between the various embodiments can be referred to each other.

[0089] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A fuzzy adaptive impedance control method for an upper limb rehabilitation training robot, characterized in that, This method executes the following steps sequentially within one control cycle, repeating cyclically: S1: Real-time acquisition of sensor data from the upper limb rehabilitation training robot, and through positive kinematics function, gravity compensation and filtering processing, to obtain effective actual pose information and human-computer interaction force information; S2: Calculate the trajectory tracking error based on the expected trajectory and the actual pose information, and calculate the interaction force index reflecting the user's active force output based on the human-computer interaction force information; S3: Input the trajectory tracking error and the interaction force index into the fuzzy inference engine for processing; The fuzzy inference engine outputs an adjustment amount of the impedance parameter based on a preset fuzzy rule base, and the impedance parameter includes stiffness and damping. S4: Update the impedance parameters based on the adjustment amount and the impedance parameters of the previous control cycle, and perform smoothing filtering on the updated impedance parameters to generate impedance parameters that change smoothly in the current control cycle. S5: Based on the desired trajectory, actual pose information, and smoothed impedance parameters, the end-effector command force is calculated through the impedance control model, and the end-effector command force is mapped into joint control torque through Jacobian matrix transpose, and sent to each joint actuator of the robot to execute motion control for the current cycle.

2. The fuzzy adaptive impedance control method for an upper limb rehabilitation training robot according to claim 1, characterized in that, In S1, firstly, in each control cycle Obtain joint angle vectors from the robot's optical encoder or resolver. And acquire raw readings from the six-dimensional force / torque sensor. ; Then, using the robot's forward kinematics model, the end effector's position in the current control cycle is calculated. Actual pose at joint angles The calculation expression is: in, Represents the robot's positive kinematics function; The transformation matrix from the force sensor coordinate system {S} to the robot end effector coordinate system {E} is obtained through calibration. and the gravity vector of the end link. Perform gravity compensation and calculate interaction forces: in, This represents the interaction force after compensation, expressed in the end coordinate system {E}; Finally, a first-order Butterworth filter is used to... and Filtering is performed to obtain effective actual pose information and human-computer interaction force information.

3. The fuzzy adaptive impedance control method for an upper limb rehabilitation training robot according to claim 2, characterized in that, In step S2, the calculation process for trajectory tracking error includes: Desired trajectory points based on the current control cycle It includes the desired position and attitude, and calculates the deviation between the actual pose and the desired pose. : To obtain a scalar measure of error, the Euclidean norm of the error vector is calculated for position errors. and attitude error Weighted composite trajectory tracking error: in, Indicates the trajectory tracking error. and These are the weighting coefficients; The calculation process of the interaction index includes: Calculate the magnitude of the interaction force vector : in, It is the force component. It is the torque component. and These are weighting coefficients; Normalize it to The interval, normalized based on a preset maximum expected user output. : in, It is the normalized interaction force index.

4. The fuzzy adaptive impedance control method for an upper limb rehabilitation training robot according to claim 3, characterized in that, In step S3, the preset fuzzy rule base adopts fuzzy rules based on expert experience, and the rule form is as follows: in, and ∈ Fuzzy set {small, medium, large}, and ∈ Fuzzy set {negative large, negative small, zero, positive small, positive large}, Indicates the stiffness adjustment amount. This indicates the amount of damping adjustment.

5. The fuzzy adaptive impedance control method for an upper limb rehabilitation training robot according to claim 1, characterized in that, In S3, the fuzzy inference engine adopts the Mamdani model and uses the centroid method for defuzzification.

6. The fuzzy adaptive impedance control method for an upper limb rehabilitation training robot according to claim 1, characterized in that, In step S4, a low-pass filter is used for smoothing filtering.

7. The fuzzy adaptive impedance control method for an upper limb rehabilitation training robot according to claim 3, characterized in that, In S5, the impedance control model is as follows: in, It is the set inertia matrix. , These are the smoothed impedance parameters. These are the first and second derivatives, respectively. Represents the desired trajectory point, Indicates the actual pose. This represents the first derivative of the desired trajectory point. This represents the second derivative of the actual pose. The second derivative of the desired trajectory point, This represents the second derivative of the actual pose.

8. An electronic device, characterized in that, It includes a processor and a memory, the memory storing machine-executable instructions that can be executed by the processor, the processor executing the machine-executable instructions to implement a fuzzy adaptive impedance control method for an upper limb rehabilitation training robot as described in any one of claims 1-7.

Images