Dexterous hand variable stiffness control system and control method thereof

Abstract

The present application relates to a kind of dexterous hand variable stiffness control system and its control method, it is related to dexterous hand control technical field, it includes: finger dorsal tendon position driver, finger dorsal tendon tension controller, first control module, finger dorsal tendon tension sensor and finger pulp tendon position driver;Wherein, first control module is equipped with tension desired value inside, it can obtain tension error value f by difference between tension desired value and current tension value e . The finger dorsal tension control component in the present application is used to convert tension error value f e Into tendon position value x3, and accurately adjust the position of current finger dorsal tendon according to tendon position value x3. So as to ensure that the adjusted tension value is equal to tension desired value f d , and then realize the accurate regulation and control of finger tension value. In this way, by accurately regulating finger tension value, the stiffness of finger is accurately adjusted, so that finger has the ability of anti-deformation.

Description

Technical Field

[0001] This invention relates to the field of dexterity hand control technology, and in particular to a dexterity hand variable stiffness control system and control method thereof. Background Technology

[0002] Dexterity hand variable stiffness refers to the core technology of dynamically adjusting the stiffness of finger joints or the entire end of the hand through mechanical structure, material properties, or drive control methods, in order to balance high-precision operation and adaptive compliant gripping.

[0003] In existing technologies, dexterous hand fingers are typically driven by tendons, which control finger flexion and extension. When a finger reaches a preset flexed or extended position, external force can easily cause it to bend and deform. Furthermore, existing technologies only use angle sensors to determine if the finger is flexed. If the finger is obstructed by external force during flexion or extension, the angle sensor detects that the joint angle has not reached the target value and stops flexion or forces the finger movement. This results in a rigid opposition between the dexterous hand and the external force, potentially damaging the finger joints. Additionally, the tendons remain extended during the flexion-to-extension process. When the finger is not fully extended due to external force, the tendons are excessively extended, preventing the finger from maintaining a rigid state. In other words, on the one hand, the fingers of a dexterous hand cannot maintain the preset posture due to external interference. On the other hand, the dexterous hand cannot adjust the tension of the tendons in the finger joints according to the specific external force. Summary of the Invention

[0004] This invention provides a dexterous hand variable stiffness control system and control method to solve the problems of fingers being unable to maintain a preset posture due to external force interference, and the inability to adjust the tension of tendons and ligaments in finger joints.

[0005] This invention provides a dexterous hand variable stiffness control system, comprising: A tension sensor for the dorsal finger tendon chords is used to monitor the current tension value f of the dorsal finger tendon chords. c ; The first control module is electrically connected to the finger dorsal tendon chord tension sensor. A finger back tension control component, which is connected to the finger back tendon cord and electrically connected to the first control module; The first control module is configured with a desired tension value f. d It can convert the expected value of tension f d With the current tension value f c The tension error value f is obtained by subtraction. e The finger back tension control component can adjust the tension based on the tension error value f. e Calculate the tendon cord position value x3, and adjust the position of the tendon cord on the back of the finger according to the tendon cord position value x3 to control the tension on the back of the finger.

[0006] In one embodiment, the component further includes: a finger back tension control assembly comprising: Dorsal tendon cord position actuator, which connects to the dorsal tendon cord, is used to control the tension on the back of the finger. A dorsal finger tendon chord tension controller, which is electrically connected to the first control module, is used to control the action of the dorsal finger tendon chord position actuator. Among them, the dorsal finger tendon chord tension controller can control the tension error value f. e Convert the value to a ligament position value x3 and input it into the finger back ligament position driver for ligament actuation.

[0007] In one implementation, it further includes: A finger joint angle control component, which connects to the fingertip tendon cord, is used to control the bending of the finger joint; Finger joint angle sensor, used to monitor the current bending angle θ of the finger joint. c ; The second control module is electrically connected to the finger joint angle control component and the finger joint angle sensor; The second control module contains a predefined angle expectation value θ. d It will have the expected value of the angle θ d Compared with the current bending angle value θ c The joint angle error value θ is obtained by subtraction. e The finger joint angle control component is based on the joint angle error value θ e Calculate the tendon ligament position value x2, and adjust the position of the fingertip tendon ligament according to the tendon ligament position value x2 to control the bending of the finger joint.

[0008] In one embodiment, the finger joint angle control component includes: Finger pad tendon cord position actuator, which connects to the finger pad tendon cord, is used to control the flexion of the finger joint; A finger joint angle controller, which is electrically connected to a fingertip tendon ligament position driver and a second control module; Among them, the finger joint angle controller can measure the joint angle error value θ e Convert the value to a ligament position value x2 and input it into the fingertip ligament position driver for ligament actuation.

[0009] In one embodiment, the dexterous hand variable stiffness control system further includes: The third control module is electrically connected to the dorsal finger tendon chord tension controller and the dorsal finger tendon chord position driver. A tendon ligament position calculator, which is electrically connected to the third control module; The chord position solver is configured with an expected angle value θ. d It can be based on the expected value of the angle θd The theoretical position x1 of the tendon ligament is calculated. The third control module can add the tendon ligament position value x3 and the theoretical position x1 to obtain the tendon ligament position value x5. The finger back tendon ligament position driver adjusts the position of the finger back tendon ligament according to the tendon ligament position value x5 to control the finger back tension.

[0010] In one embodiment, the dexterous hand variable stiffness control system further includes: The fourth control module is electrically connected to the fingertip tendon ligament position driver, the finger joint angle controller, and the tendon ligament position solver. The fourth control module can add the theoretical position x1 of the tendon cord and the position value x2 of the tendon cord to obtain the position value x4 of the tendon cord. The fingertip tendon cord position driver adjusts the position of the fingertip tendon cord according to the position value x4 of the tendon cord to control the bending of the finger joint.

[0011] In one embodiment, a state observer is also included, which is electrically connected to at least one of the fourth control module, the dorsalis digitorum tendon tension sensor, the finger joint angle sensor, and the finger joint angle controller. The state observer is used to acquire the current tension value f. c Current bending angle value θ c x4 for ligament position value, and based on the current tension value f c Current bending angle value θ c The current tension of the fingertip tendon ligament is determined by the ligament position value x4.

[0012] In one embodiment, a back-of-hand tactile sensor is also included, electrically connected to the state observer, for acquiring the tactile pressure f on the back of the finger. t It then transmits this information to the state observer, which can determine the current tension value f. c Current bending angle value θ c tactile pressure on the back of the finger f t The current tension of the fingertip tendon ligament is determined by the ligament position value x4.

[0013] In one embodiment, it further includes: an online identification module for tendon ligament model parameters, which is configured with an angle expectation value θ. d The online identification module for tendon ligament model parameters is electrically connected to the second control module, the third control module, the tendon ligament position solver, and the fourth control module. It can identify the expected angle value θ based on the collected parameters. d Joint angle error value θ e The ligament position values ​​x4 and x5 are used to periodically correct the model parameters in the ligament position solver.

[0014] In one implementation, it further includes: A joint angle planner is expected to be electrically connected to a state observer and a chord position solver. The expected tension planner is electrically connected to the state observer and the first control module; The expected joint angle planner is configured with an expected angle value θ. d The joint angle planner is expected to be able to adjust the expected angle value θ according to the received control mode command. d The process is performed, and the expected value of the correction angle is input into the chordae position solver. The expected tension planner is set with an expected tension value f. d The expected tension planner can adjust the expected tension value f according to the received mode instructions. d The process is performed, and the corrected tension expectation value is input to the first control module. .

[0015] In one implementation, the control mode command is either a constant tension mode command or a snap-finger mode command. When the control mode command is a constant tension mode command, the desired correction angle value is specified. Equal to the expected value of the angle θ d Correcting the expected value of tension Equal to the expected value of tension f d When the control mode command is a snap-finger mode command, the joint angle planner is expected to dynamically plan and correct the expected angle value based on the observation state of the state observer. The expected tension planner dynamically plans and corrects the expected tension value based on the observed state of the state observer. .

[0016] This application also provides a control method for a dexterous hand variable stiffness control system, the control method employing the aforementioned dexterous hand variable stiffness control system, including: Monitor the current tension value f of the dorsal finger tendon chord c ; Call the expected tension value f d and the current tension value f c ; The expected value of tension f d With the current tension value f c The tension error value f is obtained by subtraction. e ; Based on the tension error value f e Calculate the chordae tendon position value x3; Adjust the position of the tendon cord on the back of the finger according to the tendon cord position value x3 to control the tension on the back of the finger.

[0017] This application also provides a control method for a dexterous hand variable stiffness control system, the control method employing the aforementioned dexterous hand variable stiffness control system, including: Monitor the current tension value f of the dorsal finger tendon chord c ; Call the expected tension value fd and the current tension value f c ; The expected value of tension f d With the current tension value f c The tension error value f is obtained by subtraction. e ; Based on the tension error value f e Calculate the chordae tendon position value x3; Correct the ligament position value x3; Adjust the position of the finger ligaments based on the corrected ligament position value x5; Monitor the current bending angle value θ c ; Call the current bending angle value θ c and the expected value of the angle θ d ; The expected value of the angle θ d Compared with the current bending angle value θ c The joint angle error value θ is obtained by subtraction. e ; The joint angle error value θ e Convert to ligament position value x2; Correct the ligament position value x2; Adjust the position of the fingertip tendon ligaments based on the corrected tendon ligament position value x4.

[0018] This application also provides a control method for a dexterous hand variable stiffness control system, the control method employing the aforementioned dexterous hand variable stiffness control system, including: Monitor the current tension value f of the dorsal finger tendon chord c ; Call the expected tension value f d and the current tension value f c ; The expected value of tension f d With the current tension value f c The tension error value f is obtained by subtraction. e ; Based on the tension error value f e Calculate the chordae tendon position value x3; Periodically adjust the model parameters in the chordae position solver; Correct the ligament position value x3; Adjust the position of the finger ligaments based on the corrected ligament position value x5; Monitor the current bending angle value θ c ; Call the current bending angle value θ c and the expected value of the angle θ d ; The expected value of the angle θd Compared with the current bending angle value θ c The joint angle error value θ is obtained by subtraction. e ; The joint angle error value θ e Convert to ligament position value x2; Periodically adjust the model parameters in the chordae position solver; Correct the ligament position value x2; Adjust the position of the fingertip tendon ligaments based on the corrected tendon ligament position value x4.

[0019] This application also provides a control method for a dexterous hand variable stiffness control system, the control method employing the aforementioned dexterous hand variable stiffness control system, including: Receive finger-flicking mode commands; The expected value of tension is corrected by dynamic programming based on the observed state of the state observer. ; Monitor the current tension value f of the dorsal finger tendon chord c ; Call the corrected tension expectation value and the current tension value f c ; Correct the expected value of tension With the current tension value f c The tension error value f is obtained by subtraction. e ; Based on the tension error value f e Calculate the chordae tendon position value x3; Periodically adjust the model parameters in the chordae position solver; Correct the ligament position value x3; Adjust the position of the finger ligaments based on the corrected ligament position value x5; Monitor the current bending angle value θ c ; Call the current bending angle value θ c and the expected value of the angle θ d ; The expected value of the angle θ d Compared with the current bending angle value θ c The joint angle error value θ is obtained by subtraction. e ; The joint angle error value θ e Convert to ligament position value x2; Periodically adjust the model parameters in the chordae position solver; Correct the ligament position value x2; Adjust the position of the fingertip tendon ligaments based on the corrected tendon ligament position value x4.

[0020] Compared with the prior art, the advantage of the present invention is that the finger dorsal tendon tension sensor can monitor the current tension value f of the finger in real time. c The first control module is set to control the current tension value f. c To obtain the tension error value f, correction is performed. e The finger back tension control component is set to control the tension error value f. e This is converted into a tendon ligament position value x3, and the current position of the finger back tendon ligament is precisely adjusted based on this value. This ensures that the adjusted tension value equals the desired tension value f. d This allows for precise control of finger tension. By precisely adjusting finger tension, the stiffness of the fingers can be adjusted, giving them resistance to deformation. This ensures that the fingers of a dexterous hand can maintain a preset bending posture even under external forces. Attached Figure Description

[0021] The invention will now be described in more detail with reference to embodiments and the accompanying drawings.

[0022] Figure 1 This is a schematic diagram of the module composition of the dexterous hand variable stiffness control system in Embodiment 1 of the present invention; Figure 2 This is a schematic diagram of the module composition of the dexterous hand variable stiffness control system in Embodiment 2 of the present invention; Figure 3 This is a schematic diagram of the module composition of the dexterous hand variable stiffness control system in Embodiment 3 of the present invention; Figure 4 This is a flowchart illustrating the control method steps of the dexterous hand variable stiffness control system in Embodiment 4 of the present invention. Figure 5 This is a flowchart illustrating the specific method of the control method for the dexterous hand variable stiffness control system in Embodiment 5 of the present invention. Figure 6 This is a flowchart illustrating the specific method of the control method for the dexterous hand variable stiffness control system in Embodiment Six of the present invention. Figure 7 This is a flowchart illustrating the specific control method of the dexterous hand variable stiffness control system in Embodiment 7 of the present invention. Figure label: 10. Dorsal finger tendon chord position driver; 20. Dorsal finger tendon chord tension controller; 30. First control module; 40. Dorsal finger tendon chord tension sensor; 50. Pad finger tendon chord position driver; 60. Finger joint angle controller; 70. Finger joint angle sensor; 80. Second control module; 90. Third control module; 100. Tendon chord position solver; 110. Fourth control module; 120. State observer; 130. Back of hand tactile sensor; 140. Tendon chord model parameter online identification module; 150. Desired joint angle planner; 160. Desired tension planner; 170. Control mode input module. Detailed Implementation

[0023] The invention will now be further described with reference to the accompanying drawings.

[0024] It should be noted that in this invention, a fingertip tendon cord is provided on the fingertip side, which is used to control the bending of the corresponding finger joint. A dorsal finger tendon cord is provided on the back side of the finger, which is used to control the overall return of the middle finger. For example, a dorsal finger tendon cord is provided on the back side of the middle finger of a dexterous hand to control the overall return of the middle finger; the stiffness of the middle finger can be changed by adjusting this dorsal finger tendon cord. Three fingertip tendon cords are provided on the fingertip side of the middle finger, respectively used to control the bending of the fingertip joint, the proximal palm joint, and the middle joint.

[0025] It should be noted that when the fingertip tendon ligament is taut, the finger stiffness value and the displacement of the dorsal finger tendon ligament have a certain linear relationship. In this application, each finger of the dexterous hand is provided with a dorsal finger tendon ligament. A single dorsal finger tendon ligament is used to control the overall repositioning of the finger. By adjusting the tension of the dorsal finger tendon ligament, the stiffness of the finger is changed, so that it exhibits different resistance to deformation under different stiffness control states (e.g., low stiffness compliant state, medium stiffness holding state, and high stiffness disturbance-resistant state). Among them, the low stiffness compliant state is suitable for compliant interaction when the finger comes into contact with the external environment, the medium stiffness holding state is suitable for maintaining a stable grasping posture, and the high stiffness disturbance-resistant state is suitable for withstanding external disturbance forces or achieving high-stability grasping.

[0026] In this embodiment, the number of tendon cords m and the number of joints n have a relationship of m = n + 1. That is, the number of tendon cords is 1 more than the number of joints, and only one tension sensor (finger back tendon cord tension sensor 40) needs to be installed at the tendon cord of the finger back to meet the control performance requirements.

[0027] Example 1 like Figure 1As shown, this invention provides a dexterous hand variable stiffness control system. The dexterous hand variable stiffness control system includes a dorsal finger tendon cord position actuator 10, a dorsal finger tendon cord tension controller 20, a first control module 30, and a dorsal finger tendon cord tension sensor 40. The dorsal finger tendon cord position actuator 10 is connected to the dorsal finger tendon cord and is used to control the tension on the back of the finger. The dorsal finger tendon cord tension controller 20 is used to control the movement of the dorsal finger tendon cord position actuator 10. The first control module 30 is electrically connected to the dorsal finger tendon cord tension controller 20. The dorsal finger tendon cord tension sensor 40 is electrically connected to the first control module 30 and is used to monitor the current tension value f of the dorsal finger tendon cord. c The first control module 30 is configured with a desired tension value f. d It can convert the expected value of tension f d With the current tension value f c The tension error value f is obtained by subtraction. e The dorsalis tendon tension controller 20 can adjust the tension based on the tension error value f. e The tendon cord position value x3 is calculated, and the dorsal tendon cord position actuator 10 can adjust the position of the dorsal tendon cord according to the tendon cord position value x3 to control the dorsal tension of the finger.

[0028] It should be noted that in this embodiment, the finger back tension control component consists of a finger back tendon cord position driver 10 and a finger back tendon cord tension controller 20.

[0029] In the above setup, the finger dorsal tendon tension sensor 40 is configured to monitor the current tension value f of the finger in real time. c The first control module 30 is configured to control the current tension value f. c To obtain the tension error value f, correction is performed. e The dorsal finger tendon tension controller 20 is set to control the tension error value f. e This is converted into a tendon ligament position value x3, allowing the dorsal finger tendon ligament position actuator 10 to precisely adjust the current position of the dorsal finger tendon ligament based on the tendon ligament position value x3. This ensures that the adjusted tension value equals the desired tension value f. d This allows for precise control of finger tension. By precisely adjusting finger tension, the stiffness of the fingers can be adjusted, giving them resistance to deformation. This ensures that the fingers of a dexterous hand can maintain a preset bending posture even under external forces.

[0030] It should be noted that the expected value of tension f d Set to greater than or equal to 0. The algorithm used by the dorsalis tendon tension controller 20 can be PID, sliding mode variable structure, neural network, fuzzy logic, or a combination thereof, etc., which are existing technologies and will not be described in detail here.

[0031] Specifically, such as Figure 1As shown, in one embodiment, the dexterous hand variable stiffness control system further includes a finger joint angle controller 60, a second control module 80, a finger joint angle sensor 70, and a fingertip tendon ligament position driver 50. The finger joint angle controller 60 is electrically connected to the fingertip tendon ligament position driver 50; the second control module 80 is electrically connected to the finger joint angle controller 60 and the finger joint angle sensor 70; the finger joint angle sensor 70 is used to monitor the current bending angle θ of the finger joint. c The second control module 80 has a predefined angle expectation value θ. d It will have the expected value of the angle θ d Compared with the current bending angle value θ c The joint angle error value θ is obtained by subtraction. e The finger joint angle controller 60 is based on the joint angle error value θ e The tendon ligament position value x2 is calculated, and the fingertip tendon ligament position actuator 50 adjusts the position of the fingertip tendon ligament according to the tendon ligament position value x2 to control the flexion of the finger joint.

[0032] It should be noted that in this embodiment, the finger joint angle control component consists of a fingertip tendon tract position driver 50 and a finger joint angle controller 60.

[0033] In the above configuration, the finger joint angle sensor 70 is configured to monitor the bending angle of the finger joint in real time. The second control module 80 is configured to control the current bending angle value θ. c Perform correction and obtain the joint angle error value θ e Setting the fingertip tendon ligament position actuator 50 can resolve the joint angle error value θ e This is converted into a ligament position value x2, allowing the fingertip ligament position actuator 50 to precisely adjust the position of the fingertip ligament based on this value. This ensures that the adjusted bending angle value equals the desired angle value θ. d This allows for precise bending control of each joint in the fingers.

[0034] In addition, by working together with the aforementioned dorsal tendon cord position driver 10, dorsal tendon cord tension controller 20, first control module 30 and dorsal tendon cord tension sensor 40, it is possible to precisely control the anti-deformation ability of the finger under different bending states.

[0035] It should be noted that the algorithm used by the finger joint angle controller 60 can be PID, sliding mode variable structure, neural network, fuzzy logic, or a combination thereof.

[0036] Specifically, such as Figure 1As shown, in one embodiment, the dexterous hand variable stiffness control system further includes a third control module 90 and a tendon ligament position solver 100. The third control module 90 is electrically connected to the dorsal finger tendon ligament tension controller 20 and the dorsal finger tendon ligament position driver 10; the tendon ligament position solver 100 is electrically connected to the third control module 90; wherein, the tendon ligament position solver 100 is configured with an angle expectation value θ. d It can be based on the expected value of the angle θ d The theoretical position x1 of the tendon cord is calculated. The third control module 90 can add the tendon cord position value x3 and the theoretical position x1 to obtain the tendon cord position value x5. The finger back tendon cord position driver 10 adjusts the position of the finger back tendon cord according to the tendon cord position value x5 to control the finger back tension.

[0037] In the above setup, a third control module 90 is included to correct the tendon cord position value x3. The finger back tendon cord position driver 10 adjusts the position of the finger back tendon cord based on the corrected tendon cord position value x5. This further improves the control accuracy of the finger back tension.

[0038] It should be noted that the solution method of the tendon ligament position solver 100 can be geometric modeling method, Jacobian integral method, FEM finite element method, data-driven fitting, neural network learning, lookup table (LUT) and its hybrid method. This is existing technology and will not be described in detail here.

[0039] Specifically, such as Figure 1 As shown, in one embodiment, the dexterous hand variable stiffness control system further includes a fourth control module 110, which is electrically connected to the fingertip tendon ligament position driver 50, the finger joint angle controller 60, and the tendon ligament position solver 100; the fourth control module 110 can add the theoretical tendon ligament position x1 and the tendon ligament position value x2 to obtain the tendon ligament position value x4, and the fingertip tendon ligament position driver 50 adjusts the position of the fingertip tendon ligament according to the tendon ligament position value x4 to control the bending of the finger joint.

[0040] In the above setup, a fourth control module 110 is included to correct the tendon ligament position value x2, and the fingertip tendon ligament position driver 50 adjusts the position of the fingertip tendon ligament according to the tendon ligament position value x4. This further improves the control accuracy of finger bending.

[0041] Specifically, such as Figure 1 As shown, in one embodiment, a state observer 120 is also included, which is electrically connected to the fourth control module 110, the dorsalis digitorum tendon tension sensor 40, the finger joint angle sensor 70, and the finger joint angle controller 60. The state observer 120 is used to acquire the current tension value f. c Current bending angle value θ cWith the chordae tendon position value x4, the finger joint angle controller 60 can receive the acquired current tension value f. c Current bending angle value θ c x4 for ligament position value, and based on the current tension value f c Current bending angle value θ c The current tension of the fingertip tendon ligament is determined by the ligament position value x4.

[0042] Specifically, such as Figure 1 As shown, in one embodiment, a back-of-hand tactile sensor 130 is also included, which is electrically connected to the state observer 120 for acquiring the tactile pressure f on the back of the finger. t It is then transmitted to the state observer 120, and the finger joint angle controller 60 can adjust the current tension value f. c Current bending angle value θ c tactile pressure on the back of the finger f t The current tension of the fingertip tendon ligament is determined by the ligament position value x4.

[0043] It should be noted that in this embodiment, since no tension sensor is installed on the fingertip tendon ligament, the tension of the fingertip tendon ligament cannot be determined. During the control process of the finger joint angle controller 60, situations may inevitably arise where the tendon ligament has no tension, or even worse, the tendon ligament may become entangled due to excessive spacing. To solve this problem, this embodiment adds a state observer 120 for the fingertip tendon ligament. The fingertip tendon ligament state observer 120 can collect the current tension value f. c , chordae tendon position value x4, current flexion angle value θ c and tactile pressure on the back of the fingers f t The current tension of the fingertip tendon is estimated to inform the finger joint angle controller's control decisions.

[0044] It should be noted that in the input data of the state observer 120, the tactile pressure f on the back of the finger is... t This is not a mandatory item and can be adjusted according to the actual situation.

[0045] Example 2 The differences between Example 2 and Example 1 are as follows: Specifically, such as Figure 2 As shown, in one embodiment, the dexterous hand variable stiffness control system further includes an online identification module 140 for tendon and ligament model parameters, which is configured with an angle expectation value θ. d The online identification module 140 for tendon ligament model parameters is electrically connected to the second control module 80, the third control module 90, the tendon ligament position solver 100, and the fourth control module 110. It can identify the expected angle value θ based on the collected parameters. d Joint angle error value θ eThe ligament position values ​​x4 and x5 are periodically used to correct the model parameters in the ligament position solver 100.

[0046] It should be noted that the online parameter identification module 140 for the tendon ligament model is based on the expected value θ of the joint angle. d A preset motion excitation signal is generated, and the joint angle change Δθ fed back by the second control module 80 and the tendon ligament position change Δx fed back by the fourth control module 110 are collected. The tendon ligament model parameter online identification module 140 establishes a mapping model between the tendon ligament displacement change and the joint angle change based on the Jacobian mapping relationship between the joint space and the tendon ligament space. By calculating the error between the theoretical tendon ligament space Jacobian matrix and the actual tendon ligament displacement change, the tendon ligament model parameter P is updated using the recursive least squares algorithm or gradient descent algorithm, and the updated parameters are sent to the tendon ligament position solver 100 to realize the periodic (regular) correction of the tendon ligament position solution model.

[0047] It should be noted that the parameters in the chordae tendon position solver 100 are initially theoretical model parameters, which differ from the actual model parameters. In order to adapt to the actual model, online parameter identification can be used to periodically correct the model parameters in the chordae tendon position solver 100.

[0048] It should be noted that, in order to meet the requirements for online identification of tendon ligament model parameters, the data to be collected includes, but is not limited to, the following data: Expected angle θ of finger joint at each time point d ; Finger joint curvature error θ at each time point e ; The ligament position value at each time point is multiplied by 4; The chordae tendon position value at each time point is x5.

[0049] The other structures of Example 2 are exactly the same as those of Example 1, and will not be described in detail here.

[0050] Example 3 The differences between Example 3 and Example 2 are as follows: Specifically, such as Figure 3 As shown, in one embodiment, the dexterous hand variable stiffness control system further includes a desired joint angle planner 150 and a desired tension planner 160. The desired joint angle planner 150 is electrically connected to the state observer 120 and the tendon ligament position solver 100; the desired joint angle planner 150 is configured with a desired angle value θ. d The joint angle planner 150 is expected to be able to adjust the expected angle value θ according to the received control mode command. dThe process is performed, and the expected value of the correction angle is input into the chordae position solver 100. The expected tension planner 160 has a set expected tension value f. d The expected tension planner 160 is able to adjust the expected tension value f according to the received mode command. d And input the corrected tension expectation value to the first control module 30. .

[0051] Specifically, in one embodiment, the control mode command is a variable constant tension mode command or a finger-trigger mode command. When the control mode command is a variable constant tension mode command, the desired correction angle value is... Equal to the expected value of the angle θ d Correcting the expected value of tension Equal to the expected value of tension f d When the control mode command is a snap-finger mode command, the joint angle planner 150 is expected to dynamically plan and correct the expected angle value based on the observation state of the state observer 120. The expected tension planner 160 dynamically programs and corrects the expected tension value based on the observed state of the state observer 120. This gives the dexterous hand variable stiffness control system dual operating modes, allowing operators to choose the appropriate mode based on the specific application scenario. This, in turn, expands the applicability of the dexterous hand variable stiffness control system.

[0052] It should be noted that when the state observer 120 determines that there is pressure on the back of the finger, the expected value of the tension of the tendon rope on the back of the finger is set to a relatively large value, and the expected value of the finger joint angle is set to the predicted value after the finger is snapped.

[0053] It should be noted that setting the expected tension value of the dorsal tendon to a relatively large value represents the power buildup before the finger snap. The expected finger joint angle is set to the predicted value after the snap, which is the target movement of the finger joint after the snap. This target needs to be preset before the snap.

[0054] Specifically, in one embodiment, the dexterous hand variable stiffness control system further includes a control mode input module 170, which is electrically connected to the desired joint angle planner 150 and the desired tension planner 160. The control mode input module 170 can input variable constant tension mode commands or finger snap mode commands to the desired joint angle planner 150 and the desired tension planner 160.

[0055] The other structures of Example 3 are the same as those of Example 2, and will not be described in detail here.

[0056] Example 4 like Figure 4As shown, this invention provides a control method for a dexterous hand variable stiffness control system. This dexterous hand variable stiffness control system includes a dorsal finger tendon cord position actuator 10, a dorsal finger tendon cord tension controller 20, a first control module 30, and a dorsal finger tendon cord tension sensor 40. The dorsal finger tendon cord position actuator 10 is connected to the dorsal finger tendon cord and is used to control the tension on the back of the finger. The dorsal finger tendon cord tension controller 20 is used to control the movement of the dorsal finger tendon cord position actuator 10. The first control module 30 is electrically connected to the dorsal finger tendon cord tension controller 20. The dorsal finger tendon cord tension sensor 40 is electrically connected to the first control module 30 and is used to monitor the current tension value f of the dorsal finger tendon cord. c The first control module 30 is configured with a desired tension value f. d It can convert the expected value of tension f d With the current tension value f c The tension error value f is obtained by subtraction. e The dorsalis tendon tension controller 20 can adjust the tension based on the tension error value f. e The tendon cord position value x3 is calculated, and the dorsal tendon cord position actuator 10 can adjust the position of the dorsal tendon cord according to the tendon cord position value x3 to control the dorsal tension of the finger.

[0057] Specifically, in one embodiment, the control method includes the following steps: Monitor the current tension value f of the dorsal finger tendon chord c ; Call the expected tension value f d and the current tension value f c ; The expected value of tension f d With the current tension value f c The tension error value f is obtained by subtraction. e ; Based on the tension error value f e Calculate the chordae tendon position value x3; Adjust the position of the finger ligaments based on the corrected ligament position value x3; Specifically, in one embodiment, the dorsalis digitorum tendon tension sensor 40 monitors the current tension value f of the finger in real time. c The first control module 30 is used to control the current tension value f. c To obtain the tension error value f, correction is performed. e The dorsalis tendon tension controller 20 is used to control the tension error value f. e Converted into a tendon ligament position value x3, the dorsal finger tendon ligament position driver 10 can precisely adjust the current position of the dorsal finger tendon ligament based on the tendon ligament position value x3.

[0058] Example 5 like Figure 5As shown, the present invention provides a control method for a dexterous hand variable stiffness control system. This method employs the dexterous hand variable stiffness control system described in Embodiment 1, which includes: Monitor the current tension value f of the dorsal finger tendon chord c ; Call the expected tension value f d and the current tension value f c ; The expected value of tension f d With the current tension value f c The tension error value f is obtained by subtraction. e ; Based on the tension error value f e Calculate the chordae tendon position value x3; Correct the ligament position value x3; It should be noted that the specific correction method is as follows: the chordae position solver 100 has a predefined angle expectation value θ. d It can be based on the expected value of the angle θ d The theoretical position x1 of the tendon ligament is calculated. The third control module 90 can add the tendon ligament position value x3 and the theoretical position x1 to obtain the tendon ligament position value x5.

[0059] Adjust the position of the finger ligaments based on the corrected ligament position value x5; It should be noted that the above steps are for controlling the tension on the back of the hand.

[0060] Monitor the current bending angle value θ c ; Call the current bending angle value θ c and the expected value of the angle θ d ; The expected value of the angle θ d Compared with the current bending angle value θ c The joint angle error value θ is obtained by subtraction. e ; The joint angle error value θ e Convert to ligament position value x2; Correct the ligament position value x2; It should be noted that the specific correction method is as follows: the chordae position solver 100 has a predefined angle expectation value θ. d It can be based on the expected value of the angle θ d The theoretical position x1 of the tendon ligament is calculated. The fourth control module 110 can add the theoretical position x1 of the tendon ligament and the position value x2 of the tendon ligament to obtain the position value x4 of the tendon ligament.

[0061] The position of the fingertip tendon ligament is precisely adjusted based on the corrected tendon ligament position value x4.

[0062] It should be noted that the above steps are for precise bending control of each joint of the fingers.

[0063] Following the steps outlined above ensures that the adjusted tension value equals the desired tension value f. d This allows for precise control of finger tension and ensures that the adjusted bending angle equals the desired angle value θ. d This allows for precise bending control of each joint in the fingers.

[0064] It should be noted that the above steps for controlling the tension on the back of the hand are not logically related to the above steps for precisely controlling the bending of each joint of the fingers, and there is no requirement for their order.

[0065] Specifically, in one embodiment, the dorsalis digitorum tendon tension sensor 40 monitors the current tension value f of the finger in real time. c The first control module 30 is used to control the current tension value f. c To obtain the tension error value f, correction is performed. e The dorsalis tendon tension controller 20 is used to control the tension error value f. e Convert to chord position value x3.

[0066] Specifically, in one embodiment, the finger joint angle sensor 70 monitors the bending angle of the finger joint in real time, and the second control module 80 can control the current bending angle value θ. c Perform correction and obtain the joint angle error value θ e The finger joint angle controller 60 can measure the joint angle error value θ e Convert to ligament position value x2.

[0067] Specifically, in one embodiment, the third control module 90 corrects the tendon ligament position value x3, and the dorsal tendon ligament position driver 10 adjusts the position of the dorsal tendon ligament according to the corrected tendon ligament position value x5.

[0068] Specifically, in one embodiment, the fourth control module 110 corrects the tendon cord position value x2, and the fingertip tendon cord position driver 50 adjusts the position of the fingertip tendon cord according to the tendon cord position value x4.

[0069] Specifically, in one embodiment, the control method further includes the following steps: The state observer collects the current tension value f. c Current bending angle value θ c tactile pressure on the back of the finger f t and ligament position value x4; The state observer is based on f c θ c f t x4 indicates the current tension of the fingertip tendon ligament; The finger joint angle controller receives information about the current tension and makes control decisions.

[0070] It should be noted that when the fingertip tendon ligament is taut, the finger joint angle controller 60 adjusts the joint angle error value θ. e Error convergence tracking control is implemented. When the fingertip tendon ligament is in a loose state, even if the joint angle target has not been reached, the finger joint angle controller 60 will not control the tendon ligament to continue loosening, but allows movement in the direction of tendon ligament tightening, provided that the direction of movement is within the joint angle error value θ. e Convergence process.

[0071] Example 6 The differences between Example 6 and Example 5 are as follows: Specifically, such as Figure 6 As shown, this control method employs the dexterous hand variable stiffness control system from Embodiment 2, which adds an online identification step for tendon ligament model parameters. The specific steps are as follows: The online parameter identification module for tendon ligament models acquires the expected angle θ. d Joint angle error value θ e , ligament position value x4 and ligament position value x5; The online parameter identification module for tendon ligament models identifies parameters based on the expected angle θ. d Joint angle error value θ e The model parameters of the tendon ligament position solver are periodically adjusted using the tendon ligament position value x4 and tendon ligament position value x5.

[0072] It should be noted that this embodiment employs online parameter identification, which can periodically correct the model parameters in the chordae tendon position solver 100. This improves the control accuracy of the control method. This step precedes the steps of correcting the chordae tendon position value x2 and the chordae tendon position value x3.

[0073] The other steps in Example 6 are the same as those in Example 5, and will not be described in detail here.

[0074] Example 7 Specifically, such as Figure 7 As shown, the control method adopts the dexterous hand variable stiffness control system in Embodiment 3, which adds variable constant tension mode and finger snap mode steps.

[0075] Control Mode 1 The specific steps are as follows: The control mode input module inputs variable constant tension mode commands to the desired joint angle planner and the desired tension planner. The tension planner is expected to receive commands for the variable constant tension mode. The expected tension planner will adjust the expected tension value. With tension expectation value f d equal; Monitor the current tension value f of the dorsal finger tendon chord c ; Call the corrected tension expectation value and the current tension value f c ; Correct the expected value of tension With the current tension value f c The tension error value f is obtained by subtraction. e ; Based on the tension error value f e Calculate the chordae tendon position value x3; Periodically adjust the model parameters in the chordae position solver; Correct the ligament position value x3; It should be noted that the specific correction method is as follows: the chordae position solver 100 has a predefined angle expectation value θ. d It can be based on the expected value of the angle θ d The theoretical position x1 of the tendon ligament is calculated. The third control module 90 can add the tendon ligament position value x3 and the theoretical position x1 to obtain the tendon ligament position value x5.

[0076] Adjust the position of the finger ligaments based on the corrected ligament position value x5; It should be noted that the above steps are for controlling the tension on the back of the hand.

[0077] Monitor the current bending angle value θ c ; Call the current bending angle value θ c and the expected value of the angle θ d ; The expected value of the angle θ d Compared with the current bending angle value θ c The joint angle error value θ is obtained by subtraction. e ; The joint angle error value θ e Convert to ligament position value x2; Periodically adjust the model parameters in the chordae position solver; Correct the ligament position value x2; The position of the fingertip tendon ligament is precisely adjusted based on the corrected tendon ligament position value x4.

[0078] It should be noted that the specific correction method for correcting the chordae tendon position value x2 is as follows: The joint angle planner is expected to receive the variable constant tension mode command. The joint angle planner is expected to correct the expected angle value. With the expected value of the angle θ d equal; The tendon ligament position solver 100 calculates the expected value of the correction angle. The theoretical position x1 of the tendon ligament is calculated; the fourth control module 110 can add the theoretical position x1 of the tendon ligament and the position value x2 of the tendon ligament to obtain the position value x4 of the tendon ligament.

[0079] It should be noted that the above steps are for precise bending control of each joint of the fingers, and they are not logically related to the above steps for controlling the tension on the back of the hand, and there is no requirement for their order.

[0080] Control Mode 2 The specific steps are as follows: The control mode input module inputs the snap-finger mode command to the desired joint angle planner and the desired tension planner; The tension planner is expected to receive the snap-finger mode command. The expected tension planner dynamically modifies the expected tension value based on the observed state from the state observer. ; Monitor the current tension value f of the dorsal finger tendon chord c ; Call the corrected tension expectation value and the current tension value f c ; Correct the expected value of tension With the current tension value f c The tension error value f is obtained by subtraction. e ; Based on the tension error value f e Calculate the chordae tendon position value x3; Periodically adjust the model parameters in the chordae position solver; Correct the ligament position value x3; It should be noted that the specific correction method is as follows: the chordae position solver 100 has a predefined angle expectation value θ. d It can be based on the expected value of the angle θ d The theoretical position x1 of the tendon ligament is calculated. The third control module 90 can add the tendon ligament position value x3 and the theoretical position x1 to obtain the tendon ligament position value x5.

[0081] Adjust the position of the finger ligaments based on the corrected ligament position value x5; It should be noted that the above steps are for controlling the tension on the back of the hand.

[0082] Monitor the current bending angle value θ c ; Call the current bending angle value θ c and the expected value of the correction angle ; Correct the expected value of the angle Compared with the current bending angle value θ c The joint angle error value θ is obtained by subtraction. e ; The joint angle error value θ e Convert to ligament position value x2; Periodically adjust the model parameters in the chordae position solver; Correct the ligament position value x2; The position of the fingertip tendon ligament is precisely adjusted based on the corrected tendon ligament position value x4.

[0083] It should be noted that the specific correction method for correcting the chordae tendon position value x2 is as follows: The joint angle planner is expected to receive the snap-finger mode command. The expected joint angle planner dynamically plans and corrects the expected angle value based on the observed state of the state observer. ; The tendon ligament position solver 100 calculates the expected value of the correction angle. The theoretical position x1 of the tendon ligament is calculated. The fourth control module 110 can add the theoretical position x1 of the tendon ligament and the position value x2 of the tendon ligament to obtain the position value x4 of the tendon ligament.

[0084] It should be noted that the above steps are for precise bending control of each joint of the fingers, and they are not logically related to the above steps for controlling the tension on the back of the hand, and there is no requirement for their order.

[0085] Although the invention has been described with reference to preferred embodiments, various modifications can be made and components can be replaced with equivalents without departing from the scope of the invention. In particular, the technical features mentioned in the various embodiments can be combined in any manner as long as there is no structural conflict. The invention is not limited to the specific embodiments disclosed herein, but includes all technical solutions falling within the scope of the claims.

Claims

1. A dexterous hand variable stiffness control system, characterized in that, It includes: A tension sensor for the dorsal finger tendon chords is used to monitor the current tension value f of the dorsal finger tendon chords. c ; The first control module is electrically connected to the dorsal finger tendon ligament tension sensor. A finger back tension control component, which is connected to the finger back tendon cord and electrically connected to the first control module; The first control module is configured with a desired tension value f. d And is able to express the expected tension value f d With the current tension value f c The tension error value f is obtained by subtraction. e The finger back tension control component can adjust the tension based on the tension error value f. e Calculate the tendon cord position value x3, and adjust the position of the finger back tendon cord according to the tendon cord position value x3 to control the finger back tension; The finger back tension control component includes: A dorsal finger tendon cord position actuator, which is connected to the dorsal finger tendon cord, is used to control the tension on the back of the finger. A dorsal finger tendon chord tension controller, which is electrically connected to the first control module, is used to control the action of the dorsal finger tendon chord position actuator; The dorsal finger tendon tension controller can measure the tension error value f. e Converted into a tendon cord position value x3, and input into the finger back tendon cord position driver for tendon cord actuation; The dexterity hand variable stiffness control system also includes: The third control module is electrically connected to the dorsal finger tendon cord tension controller and the dorsal finger tendon cord position driver. A tendon ligament position calculator, which is electrically connected to the third control module; The tendon ligament position solver is configured with an expected angle value θ. d It can be based on the expected value of the angle θ d The theoretical position x1 of the tendon cord is calculated. The third control module can add the tendon cord position value x3 and the theoretical position x1 to obtain the tendon cord position value x5. The finger back tendon cord position driver adjusts the position of the finger back tendon cord according to the tendon cord position value x5 to control the finger back tension.

2. The dexterous hand variable stiffness control system according to claim 1, characterized in that, Also includes: A finger joint angle control component, which connects to the fingertip tendon cord, is used to control the bending of the finger joint; A finger joint angle sensor is used to monitor the current bending angle value θ of the finger joint. c ; The second control module is electrically connected to the finger joint angle control component and the finger joint angle sensor; The second control module is configured with an expected angle value θ. d It will have the expected value of the angle θ d Compared with the current bending angle value θ c The joint angle error value θ is obtained by subtraction. e The finger joint angle control component is based on the joint angle error value θ e Calculate the tendon cord position value x2, and adjust the position of the fingertip tendon cord according to the tendon cord position value x2 to control the bending of the finger joint.

3. The dexterous hand variable stiffness control system according to claim 2, characterized in that, The finger joint angle control component includes: A fingertip tendon cord position actuator, which is connected to the fingertip tendon cord, is used to control the flexion of the finger joint; A finger joint angle controller, which is electrically connected to the fingertip tendon cord position driver and the second control module; The finger joint angle controller can measure the joint angle error value θ. e The value is converted into a tendon cord position value x2 and input into the fingertip tendon cord position driver for tendon cord actuation.

4. The dexterous hand variable stiffness control system according to claim 3, characterized in that, The dexterity hand variable stiffness control system also includes: The fourth control module is electrically connected to the fingertip tendon cord position driver, the finger joint angle controller, and the tendon cord position solver. The fourth control module can add the theoretical position x1 of the tendon cord and the position value x2 of the tendon cord to obtain the position value x4 of the tendon cord. The fingertip tendon cord position driver adjusts the position of the fingertip tendon cord according to the position value x4 of the tendon cord to control the bending of the finger joint.

5. The dexterous hand variable stiffness control system according to claim 4, characterized in that, It also includes a state observer, which is electrically connected to at least one of the fourth control module, the dorsalis digitorum tendon tension sensor, the finger joint angle sensor, and the finger joint angle controller. The state observer is used to acquire the current tension value f. c and / or the current bending angle value θ c and / or ligament position value x4, and based on the current tension value f c Current bending angle value θ c The current tension of the fingertip tendon is determined by at least one of the tendon position value x4.

6. The dexterous hand variable stiffness control system according to claim 5, characterized in that, It also includes a back-of-hand tactile sensor, which is electrically connected to the state observer, for collecting the tactile pressure f on the back of the finger. t And transmit it to the state observer, which can determine the current tension value f. c Current bending angle value θ c tactile pressure on the back of the finger f t The current tension of the fingertip tendon is determined by the tendon position value x4.

7. The dexterous hand variable stiffness control system according to claim 6, characterized in that, It also includes an online parameter identification module for tendon ligament models, which has a predefined angle expectation value θ. d The online identification module for tendon ligament model parameters is electrically connected to the second control module, the third control module, the tendon ligament position solver, and the fourth control module. It can identify parameters based on the acquired angle expectation value θ. d Joint angle error value θ e The ligament position values ​​x4 and x5 are periodically used to correct the model parameters in the ligament position solver.

8. The dexterous hand variable stiffness control system according to claim 7, characterized in that, Also includes: A joint angle planner is expected to be electrically connected to the state observer and the tendon chord position solver. A desired tension planner is electrically connected to the state observer and the first control module; The expected joint angle planner is configured with an expected angle value θ. d The desired joint angle planner can adjust the desired angle value θ according to the received control mode command. d The process is performed, and the expected value of the correction angle is input into the tendon position solver. The expected tension planner is set with an expected tension value f. d The expected tension planner can adjust the expected tension value f according to the received mode instructions. d The process is performed, and the corrected tension expectation value is input to the first control module. .

9. The dexterous hand variable stiffness control system according to claim 8, characterized in that, The control mode command is either a constant tension mode command or a snap-finger mode command. When the control mode command is a constant tension mode command, the desired correction angle value is specified. Equal to the expected value of the angle θ d Correcting the expected value of tension Equal to the expected value of tension f d When the control mode command is a snap-finger mode command, the desired joint angle planner dynamically plans and corrects the desired angle value based on the observation state of the state observer. The expected tension planner dynamically plans and corrects the expected tension value based on the observed state of the state observer. .

10. A control method for a dexterous hand variable stiffness control system, characterized in that, The control method employs the dexterous hand variable stiffness control system as described in any one of claims 1 to 9, comprising: Monitor the current tension value f of the dorsal finger tendon chord c ; Call the expected tension value f d and the current tension value f c ; The expected value of tension f d With the current tension value f c The tension error value f is obtained by subtraction. e ; Based on the tension error value f e Calculate the chordae tendon position value x3; The position of the finger back tendon cord is adjusted according to the tendon cord position value x3 to control the finger back tension.

11. A control method for a dexterous hand variable stiffness control system, characterized in that, The control method employs the dexterous hand variable stiffness control system as described in claim 6, comprising: Monitor the current tension value f of the dorsal finger tendon chord c ; Call the expected tension value f d and the current tension value f c ; The expected value of tension f d With the current tension value f c The tension error value f is obtained by subtraction. e ; Based on the tension error value f e Calculate the chordae tendon position value x3; Correct the ligament position value x3; Adjust the position of the finger ligaments based on the corrected ligament position value x5; Monitor the current bending angle value θ c ; Call the current bending angle value θ c and the expected value of the angle θ d ; The expected value of the angle θ d Compared with the current bending angle value θ c The joint angle error value θ is obtained by subtraction. e ; The joint angle error value θ e Convert to ligament position value x2; Correct the ligament position value x2; Adjust the position of the fingertip tendon ligaments based on the corrected tendon ligament position value x4.

12. A control method for a dexterous hand variable stiffness control system, characterized in that, The control method employs the dexterous hand variable stiffness control system described in claim 7, comprising: Monitor the current tension value f of the dorsal finger tendon chord c ; Call the expected tension value f d and the current tension value f c ; The expected value of tension f d With the current tension value f c The tension error value f is obtained by subtraction. e ; Based on the tension error value f e Calculate the chordae tendon position value x3; Periodically adjust the model parameters in the chordae position solver; Correct the ligament position value x3; Adjust the position of the finger ligaments based on the corrected ligament position value x5; Monitor the current bending angle value θ c ; Call the current bending angle value θ c and the expected value of the angle θ d ; The expected value of the angle θ d Compared with the current bending angle value θ c The joint angle error value θ is obtained by subtraction. e ; The joint angle error value θ e Convert to ligament position value x2; Periodically adjust the model parameters in the chordae position solver; Correct the ligament position value x2; Adjust the position of the fingertip tendon ligaments based on the corrected tendon ligament position value x4.

13. A control method for a dexterous hand variable stiffness control system, characterized in that, The control method employs the dexterous hand variable stiffness control system as described in claim 9, comprising: Receive finger-flicking mode commands; The expected value of tension is corrected by dynamic programming based on the observed state of the state observer. ; Monitor the current tension value f of the dorsal finger tendon chord c ; Call the corrected tension expectation value and the current tension value f c ; Correct the expected value of tension With the current tension value f c The tension error value f is obtained by subtraction. e ; Based on the tension error value f e Calculate the chordae tendon position value x3; Periodically adjust the model parameters in the chordae position solver; Correct the ligament position value x3; Adjust the position of the finger ligaments based on the corrected ligament position value x5; Monitor the current bending angle value θ c ; Call the current bending angle value θ c and the expected value of the angle θ d ; The expected value of the angle θ d Compared with the current bending angle value θ c The joint angle error value θ is obtained by subtraction. e ; The joint angle error value θ e Convert to ligament position value x2; Periodically adjust the model parameters in the chordae position solver; Correct the ligament position value x2; Adjust the position of the fingertip tendon ligaments based on the corrected tendon ligament position value x4.

Images