Seven-degree-of-freedom upper limb exoskeleton and on-demand assistance control method thereof

Abstract

The application discloses a seven-freedom-degree upper limb exoskeleton and a needed-assistance control method thereof, provides mechanical structure design and joint modular design of the seven-freedom-degree upper limb exoskeleton, and deduces kinematics and dynamics models of the upper limb exoskeleton system accordingly, and provides design of a control system of the upper limb exoskeleton system on the basis, wherein an impedance model is used in an outer ring to calculate assistance force, energy function and a Horf controller are combined to perform online adjustment of training trajectory frequency, radius and assistance force, the assistance force is converted into a desired trajectory of the exoskeleton by the impedance model and is transmitted to an inner ring controller, the inner ring uses a trajectory tracking controller based on BLF design, and dynamic characteristics are estimated based on time delay estimation and RBF neural network.The application improves the adaptability of the mechanical structure of the upper limb exoskeleton to the human body, effectively reduces the influence of dynamic characteristics in the trajectory tracking process, and can online adjust the training trajectory, so that a better needed-assistance effect is achieved.

Description

Technical Field

[0001] The present invention relates to the field of upper limb exoskeletons, and in particular to a seven-degree-of-freedom upper limb exoskeleton mechanical structure and an on-demand auxiliary control method. Background Art

[0002] The aging problem is becoming increasingly serious worldwide. The requirements for elderly care, medical care, rehabilitation and other fields for the elderly are getting higher and higher. The probability of motor function impairment due to injuries is gradually increasing. Physical disability is the main problem faced in the field of medical rehabilitation. Therefore, the demand for the use of exoskeleton robots for rehabilitation medicine is increasing.

[0003] In terms of mechanical structure, current upper limb exoskeletons suffer from problems such as simple degree of freedom configuration, poor compatibility with the human body structure, and poor universality, making them unable to meet the diverse needs of different patients. For example, in the literature (Wang Yucheng. Structural Design and Compliant Control Strategy of Upper Limb Exoskeleton Rehabilitation Robot [D]. Beijing University of Posts and Telecommunications, 2023. DOI: 10.26969 / d.cnki.gbydu.2023.001187.), a six-degree-of-freedom upper limb exoskeleton mechanical structure was designed, which can be infinitely adjusted according to the patient's arm length. However, the six-degree-of-freedom design does not have redundant degrees of freedom, which may result in the exoskeleton's posture being unable to change when the end position is fixed, resulting in a slightly stiff posture. Furthermore, the shoulder joint is not well-matched to the human joints, which affects the patient's comfort during wear.

[0004] In terms of control strategy, the current control strategy of the upper limb exoskeleton system is relatively simple. For example, the literature (Jiang Jieling. Design of control system for exoskeleton upper limb rehabilitation robot [D]. Xi'an University of Technology, 2024. DOI: 10.27391 / d.cnki.gxagu.2024.000624.) adopts a PD-type iterative control algorithm to perform trajectory tracking control on the exoskeleton system. On the one hand, it is particularly dependent on the precise mathematical model of the controlled object. On the other hand, it requires multiple iterations to achieve a good control effect. In addition, it does not consider the dynamic unmodeled part of the exoskeleton system, and the convergence speed is slow. Summary of the Invention

[0005] The purpose of the present invention is to provide a seven-degree-of-freedom upper limb exoskeleton and its on-demand auxiliary control method, which solves the problems that the mechanical structure of the upper limb exoskeleton is poorly matched with the human body and the control strategy fails to achieve the on-demand auxiliary control function.

[0006] The technical solutions for achieving the purpose of the present invention are:

[0007] A seven-degree-of-freedom upper limb exoskeleton, comprising:

[0008] A shoulder joint is constituted by a base, a first motor, a second motor, a third motor, a shoulder joint scissor-type structure, an upper arm support member, and a semicircular guide rail between the upper arm and the forearm; the first motor is fixed to the base, and its output shaft is connected to one end of the shoulder joint scissor-type structure, constituting a first degree of freedom of the shoulder joint abduction / adduction; the second motor is fixed to the other end of the shoulder joint scissor-type structure, and its output shaft is connected to one end of the upper arm support member, constituting a second degree of freedom of the shoulder joint flexion / extension; the upper arm support member is connected to the semicircular guide rail between the upper arm and the forearm, and the third motor is used to drive the semicircular guide rail member between the upper arm and the forearm to rotate, constituting a third degree of freedom of the shoulder joint internal rotation / external rotation;

[0009] The elbow joint is formed by a fourth motor, a fifth motor, a forearm support member, and a semicircular guide rail between the forearm and the wrist; the fourth motor is connected to the semicircular guide rail between the upper arm and the forearm, and its output shaft is connected to the forearm support member, forming a first degree of freedom of flexion / extension of the elbow joint; the forearm support member is connected to the semicircular guide rail between the forearm and the wrist, and the fifth motor drives the semicircular guide rail between the forearm and the wrist to rotate, forming a second degree of freedom of internal / external rotation of the elbow joint;

[0010] The wrist joint is composed of the sixth motor, the seventh motor, the wrist transition piece, the wrist joint support piece, and the handle; the sixth motor is fixed on the semicircular guide rail between the forearm and the wrist joint, and its output shaft is connected to the wrist transition piece, forming the first degree of freedom of the wrist joint: flexion / extension; the seventh motor is fixed on the wrist transition piece, and its output shaft is connected to the wrist support piece, forming the second degree of freedom of the wrist joint: abduction / adduction; the handle is fixed on the wrist support piece to provide support for the hand.

[0011] A seven-degree-of-freedom upper limb exoskeleton that converts the end position of the upper limb exoskeleton into the end force through the following dynamic model of the operating space:

[0012]

[0013] Where P = J + (Θ)Θ is the six-dimensional position vector of the upper limb exoskeleton system end in the Cartesian coordinate system, M c (Θ)=J +T (Θ)M(Θ)J + (Θ) is the inertia matrix of the upper limb exoskeleton system in the operating space, is the Coriolis force and centrifugal force matrix of the upper limb exoskeleton system in the operating space, G c (Θ)=J +T (Θ)G(Θ) is the gravity term of the upper limb exoskeleton system in the operating space, J + (Θ)=J T (Θ)(J(Θ)JT (Θ)) -1 is the generalized inverse form of the Jacobian matrix, f c =J +T (Θ)τ c 、f int =J +T (Θ)τ int are the control force and human-machine interaction force on the upper limb exoskeleton end respectively; Θ = [θ1…θ i …θ7] represents the joint angles of the seven joints of the upper limb exoskeleton, where θ i is the joint angle of joint i of the upper limb exoskeleton system; M(Θ) is the inertia matrix of the exoskeleton system, are the Coriolis force and centrifugal force of the upper limb exoskeleton; G(Θ) is the gravity term; τ c is the control torque matrix provided by the node driver; τ int is the human-computer interaction torque matrix; J(Θ) is the Jacobian matrix of the upper limb exoskeleton system.

[0014] A seven-degree-of-freedom upper limb exoskeleton, which adjusts the auxiliary force f provided by the end of the upper limb exoskeleton a , frequency of rehabilitation exercises d , rehabilitation movement trajectory radius r d Realize the function of on-demand assisted rehabilitation:

[0015]

[0016] Where k is the impedance stiffness of the impedance model, is the unit vector representing the direction of the auxiliary force, y1, y2, and y3 are the y state variables of the three Horf oscillators respectively, and w1, w2, and w3 are the oscillation frequencies of the three Horf oscillators respectively. is the inverse of the radius of the rehabilitation trajectory, is the reciprocal of the frequency of rehabilitation exercises. k 、μ k , ε k , γ r 、μ r , ε r , γ w 、μ w , ε w is the corresponding coefficient that needs to be adjusted in the oscillator, is the input signal that needs to be synchronized, that is, the energy function for evaluating the patient's motor ability, where β1, β2, and β3 are the coefficients of each evaluation index.

[0017] A seven-degree-of-freedom upper limb exoskeleton, which obtains the auxiliary force f by the outer loop controller aThe desired position of the upper limb exoskeleton system is finally obtained by converting the position difference between the patient and the exoskeleton through the admittance model. The exoskeleton system tracks the desired trajectory through the inner loop controller:

[0018]

[0019]

[0020] Where e1 is the position tracking error of the exoskeleton, e2 is the speed tracking error of the exoskeleton, and k e represents the error bound of the obstacle Lyapunov function, k2 is a positive real number parameter, b is the designed virtual control quantity, are the inertia matrix M in the upper limb exoskeleton dynamic model c (Θ), Coriolis force and centrifugal force matrix Gravity matrix G c The known part of (Θ), that is, the part that can be solved by the process of building the dynamic model; is the known part of the sum of centripetal force, Coriolis force, gravity and external interaction force at time t, which can be obtained by the known dynamic model of the upper limb exoskeleton system in the above formula. It is the unknown part of the sum of centripetal force, Coriolis force, gravity and external interaction force at time t; is the weight value of each hidden node of the RBF neural network after training, h(e) is the output matrix of each hidden node of the RBF neural network, is a positive real parameter; x a (t) represents the ideal position of the end of the upper limb exoskeleton system, T0 is a very short delay, They are respectively the time T0 before the current time F0(t), Respective values; x2(t) represents the velocity term, and F0(t) is the end force term that converts the motor control torque in the joint space to the operation space.

[0021] Compared with the prior art, the present invention has the following significant advantages:

[0022] (1) The design of a spherical scissor-like structure for the shoulder joint overcomes the problem of poor matching caused by micro-motion of the center of rotation of the glenohumeral joint during training. The design of adjustable-length upper arms, lower arms, and wrist supports allows the length of the exoskeleton to be adjusted according to the needs of the user, thereby improving the overall comfort of the patient during wearing.

[0023] (2) The barrier Lyapunov controller based on time delay estimation is adopted to reduce the influence of unmodeled dynamics and parameter uncertainty, with fast convergence speed and strong robustness. At the same time, it can ensure that the error converges within a certain range and achieve high-precision trajectory tracking;

[0024] (3) Through the outer loop impedance control based on the Horf oscillator, the motor ability of different patients can be evaluated, so that the auxiliary force, the period and radius of the rehabilitation training trajectory and other parameters can be quickly adjusted during the training process according to the patient's motor ability, thereby realizing the function of on-demand assistance. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 The overall flow chart of the upper limb exoskeleton system design of the present invention.

[0026] Figure 2 An axonometric diagram of the mechanical structure of the upper limb exoskeleton system.

[0027] Figure 3 This is an axonometric drawing of the exoskeleton mechanism with the left view as the main view.

[0028] Figure 4 An axonometric view of the shoulder joint of the upper limb exoskeleton.

[0029] Figure 5 An axonometric view of the exoskeleton shoulder joint with the left view as the main view.

[0030] Figure 6 Axonometric view of the elbow joint of the upper limb exoskeleton.

[0031] Figure 7 An axonometric view of the exoskeleton elbow joint with the left view as the main view.

[0032] Figure 8 Axonometric view of the wrist joint of the upper limb exoskeleton.

[0033] Figure 9 An axonometric view of the exoskeleton wrist joint with the left view as the main view.

[0034] Figure 10 This is an axonometric drawing of a semicircular guide rail.

[0035] Figure 11 Axonometric drawing of the semicircular guide rail from the side.

[0036] Figure 12 Schematic diagram of the shoulder joint scissor structure.

[0037] Figure 13 Schematic diagram of the DH coordinate system established for the upper limb exoskeleton system.

[0038] Figure 14This is the overall block diagram of the upper limb exoskeleton control system.

[0039] Figure 15 This is the structural diagram of the RBF neural network.

[0040] Figure 16 This is the simulation diagram of the trajectory tracking error in the xy directions in the inner loop controller simulation.

[0041] Figure 17 This is a simulation diagram of the reference trajectory in the xy plane in the on-demand auxiliary control simulation.

[0042] Figure 18 To assist in controlling the motion trajectories in the x and y directions in the simulation on demand.

[0043] Figure 19 This is a simulation diagram of the tracking error between the patient's motion trajectory and the reference trajectory in the on-demand assistive control simulation.

[0044] Figure 20 A simulation diagram of the assistive force provided by the exoskeleton system in the on-demand assistive control simulation.

[0045] Figure 21 This is a simulation diagram of the energy function in the on-demand auxiliary control simulation. DETAILED DESCRIPTION

[0046] In order to make the purpose, technical solutions and advantages of this application more clear, the following further describes this application in detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain this application and are not intended to limit this application.

[0047] Combine Figure 1 First, we will introduce the mechanical structure of the seven-degree-of-freedom upper-limb exoskeleton system designed by this invention. This invention primarily uses SOLIDWORKS to design the mechanical structure of the upper-limb exoskeleton. The mechanical structure design of the upper-limb exoskeleton includes the overall design of the upper-limb exoskeleton mechanical structure and the modular design of the joints.

[0048] Figure 2 、 Figure 3The figure shows an axonometric view of the overall mechanical structure of the upper limb exoskeleton system, which mainly consists of a base 1, seven motors (respectively the first motor 2-1 to the seventh motor 2-7 in the figure, among which the first motor 2-1, the second motor 2-2 and the fourth motor 2-4 are of the same model, and the third motor 2-3, the fifth motor 2-5, the sixth motor 2-6 and the seventh motor 2-7 are of the same model), a shoulder joint scissor structure 3, an upper arm support part 4, an upper arm support part slide 5, a semicircular guide part 6 between the upper arm and the forearm, a forearm support part 7, a forearm support part slide 8, a semicircular guide part 9 between the forearm and the wrist joint, a wrist transition part 10, a wrist support part 11, a wrist support part slide 12, and a rubber grip 13.

[0049] The designed upper limb exoskeleton structure has a total of seven degrees of freedom, including three degrees of freedom at the shoulder joint, two degrees of freedom at the elbow joint, and two degrees of freedom at the wrist joint. Seven motors are configured, each responsible for one degree of freedom. The specific mechanical structure connection method of each joint part is as follows:

[0050] The base 1, the first motor 2-1, the second motor 2-2, the third motor 2-3, the shoulder joint scissor structure 3, the upper arm support 4, and the semicircular guide 6 between the upper arm and the lower arm together constitute the shoulder joint, which has three degrees of freedom. The independent shoulder joint axonometric diagram is shown in the figure below. Figure 4 and Figure 5 The connection method of its mechanical structure is as follows: the base 1 is fixed to a wall or a bracket, the first motor 2-1 is fixed to the base 1, and its output shaft is connected to one end of the shoulder joint scissor structure 3, forming the first degree of freedom of the shoulder joint abduction / adduction; the second motor 2-2 is fixed to the other end of the shoulder joint scissor structure 3, and its output shaft is connected to one end of the upper arm support member 4, forming the second degree of freedom of the shoulder joint flexion / extension; the structure of the semicircular guide member 6 between the upper arm and the forearm is as shown Figure 10 and Figure 11 As shown, it is mainly divided into three parts: a fixed part 14, a gear 15 and a movable part 16. The fixed part 14 and the movable part 16 are matched with each other through an arc-shaped slide rail structure, and the movable part 16 can move along the arc-shaped slide rail between the two. The other end of the upper arm support part 4 is connected and fixed to the fixed part 14 by a bolt. The third motor 2-3 is fixed to the fixed part 14 of the semicircular guide part 6, and its output shaft is connected to the gear 15. The gear 15 and the gear teeth of the movable part 16 are engaged with each other to realize the rotation of the movable part 16 between the upper arm and the forearm, forming the third degree of freedom of internal rotation / external rotation of the shoulder joint. The scissor structure design of the shoulder joint is as shown in FIG. Figure 12As shown, the stretchable scissor-like structure effectively compensates for the impact of shoulder joint ball-and-socket center offset during movement, improving the wearer's comfort. Furthermore, the corresponding upper arm length in the upper limb exoskeleton system can be adjusted by removing the bolts, adjusting the fixings 14 of the semicircular guide member 6 along the two slots 5 on the upper arm support member to the appropriate position, and then tightening the bolts to meet the individual needs of different patients.

[0051] The fourth motor 2-4, the fifth motor 2-5, the forearm support 7, and the semicircular guide rail 9 between the forearm and the wrist together constitute the mechanical structure of the elbow joint, which has two degrees of freedom. The independent elbow joint axonometric diagram is shown in FIG. Figure 6 and Figure 7 As shown. The connection method of its mechanical structure is as follows: the fourth motor 2-4 is fixed to the movable part 16 of the semicircular guide member 6, and its output shaft is connected to the forearm support member 7, constituting the first degree of freedom of flexion / extension of the elbow joint; the structure of the semicircular guide member 9 between the forearm and the wrist joint is similar to the semicircular guide member 6 between the upper arm and the forearm mentioned above, and one end of the forearm support member 7 is connected and fixed to the fixing part of the semicircular guide member 9 by a bolt. Similar to the above, the motor 5-2 is also fixed to the fixing part of the semicircular guide member 9 between the forearm and the wrist joint, and the output shaft is connected to a gear and meshes with the gear teeth of the movable part of the semicircular guide member 9 between the forearm and the wrist, realizing the rotation movement of the movable part of the semicircular guide member 9 between the forearm and the wrist joint, constituting the second degree of freedom of internal rotation / external rotation of the elbow joint. Similar to the above, the fixing part of the semicircular guide rail part 9 between the forearm and the wrist joint can be adjusted to the appropriate position along the two slide grooves 8 on the upper arm support part by removing the two bolts and then fixing the bolts to adjust the forearm length corresponding to the upper limb exoskeleton system to achieve the function of stepless adjustment.

[0052] Finally, the sixth motor 2-6, the seventh motor 2-7, the wrist transition piece 10, the wrist joint support piece 11, and the handle 13 together form the mechanical structure of the wrist joint, which has two degrees of freedom. The independent wrist joint axonometric diagram is shown as follows: Figure 8 and Figure 9 As shown. Its specific mechanical structure connection method is as follows: the sixth motor 2-6 is fixed to the movable part of the semicircular guide member 9 between the forearm and the wrist joint, and its output shaft is connected to the wrist transition member 10, forming the first degree of freedom of the wrist joint: flexion / extension; the seventh motor 2-7 is fixed to the wrist transition member 10, and its output shaft is connected to the wrist support member 11, forming the second degree of freedom of the wrist joint: abduction / adduction; finally, the rubber grip 13 is fixed to the wrist support member 11 by bolts to achieve hand support. Similar to the above, the length of the wrist of the upper limb exoskeleton system can also be adjusted by removing the bolts, adjusting the position of the rubber grip 13 along the slide groove 12 of the wrist support member to the appropriate position, and then fixing the bolts.

[0053] The upper limb exoskeleton is donned as follows: base 1 is fixed to a stand or wall, shoulder joint scissor structure 3 cooperates with the shoulder, the inner arc of the movable part of the semicircular guide 6 between the upper and lower arms can be attached with a belt or other device to secure the upper arm. The inner arc of the movable part of the semicircular guide 9 between the lower arm and wrist can also be attached with a belt to secure the lower arm. Finally, the patient's hand grasps the rubber grip 13. This structure of using belts to secure the patient's arm is more convenient for the patient to put on and take off, and can effectively prevent secondary injuries during the process.

[0054] In a second aspect, the present invention provides kinematic and dynamic models for the seven-degree-of-freedom upper-limb exoskeleton system. The kinematic model describes the corresponding relationships between the position and velocity of each joint of the upper-limb exoskeleton and the velocity of the end position. The dynamic model uses the Lagrange equations to solve the driving torque. The specific implementation includes the following sub-steps:

[0055] Step 1: Use the DH parameter method to describe the relationship between joint angle and end position, i.e., grip position

[0056] Since the trajectory used in actual rehabilitation training is usually based on the end position, and the forward kinematic model is the process of solving the exoskeleton end position when the joint angles and lengths are known, it is mainly solved using the DH parameter method combined with the rotation transformation matrix method.

[0057] First, the seven-degree-of-freedom upper-limb exoskeleton constructed above can be thought of as a series of rigid bodies connected by joints into a kinematic chain. We refer to these rigid bodies as links. The motors are numbered: the first motor 2-1 is called Motor 1, the second motor 2-2 is called Motor 2, and so on. The seventh motor 2-7 is called Motor 7. The links are numbered starting from the fixed base of the exoskeleton arm. The fixed base can be called Link 0. The rigid body between Motors 1 and 2 is called Link 1, the rigid body between Motors 2 and 3 is called Link 2, and so on. The link at the end of the exoskeleton is called Link 7.

[0058] Then, in order to describe the relative positional relationship between each link and the adjacent links, a fixed coordinate system is defined on each link of the seven-degree-of-freedom upper limb exoskeleton system designed above, and the fixed coordinate system is named according to the number of the link where the fixed coordinate system is located. Therefore, the fixed coordinate system fixed on link i is called coordinate system {i}. The z-axis z of coordinate system {i} is i Coincident with joint axis i (axis of motor i), origin is at axis z i and z i-1At the intersection of the common perpendicular line and the joint axis i, the y-axis y of the DH coordinate system is determined according to the right-hand rule. i , thus completing the definition of coordinate system {i}. In particular, the base (reference) coordinate system {0} can be set arbitrarily, and finally the fixed coordinate system established for the upper limb exoskeleton system is obtained as follows Figure 13 Based on the coordinate system established by the above method, four kinematic parameters (DH parameters) can be used to describe the links of the upper limb exoskeleton. Two parameters are used to describe the links themselves, and two parameters are used to describe the connection relationship between the links. They are:

[0059] (1) Joint angle θ i :around z i Axis, from x i-1 Axis rotation to x i The angle of the axis;

[0060] (2) Connecting rod torsion angle α i :around x i Axis, from z i Axis rotation to z i+1 The angle of the axis;

[0061] (3) Offset distance d i : Along z i Axis, from x i-1 Axis moves to x i Axis distance;

[0062] (4) Connecting rod length a i : Along x i Axis, from z i Axis moves to z i+1 Axis distance;

[0063] based on Figure 13 Based on the above exoskeleton mechanical structure and joint angle positions, the lengths of the shoulder joint, elbow joint, and wrist joint in the above designed exoskeleton mechanical structure are defined as L1, L2, and L3 respectively. Finally, the DH parameter table shown in the following table can be obtained.

[0064] Table 1D-H parameter table

[0065]

[0066]

[0067] Use the form of homogeneous transformation matrix to express the transformation relationship between coordinate system {i-1} and coordinate system {i}, and obtain the homogeneous transformation matrix from coordinate system {i-1} to coordinate system {i} The form is as follows:

[0068]

[0069] The above homogeneous transformation matrix can be multiplied on the right continuously to obtain the secondary transformation matrix from the reference coordinate system {0} to the terminal coordinate system {7}, that is:

[0070]

[0071] in Is the rotation matrix, which is a third-order square matrix representing the posture angle of the upper limb exoskeleton end in the base coordinate system. is the translation matrix, which is a column vector representing the position of the upper limb exoskeleton end in the base coordinate system.

[0072] Step 2: Use the Jacobian matrix to build an instantaneous kinematic model

[0073] Instantaneous kinematics also describes the mapping from joint space to action space, but the object described is the instantaneous velocity variable.

[0074] First, define the position vector motion equation of the upper limb exoskeleton as:

[0075] P=F(Θ) (3)

[0076] Where P = [p1…p6] represents the six-dimensional position vector of the upper limb exoskeleton end position in the operation space (three parameters are three-dimensional position vectors and three parameters are three-dimensional pose vectors), Θ = [θ1…θ i …θ7] represents the joint angle matrix of the seven joints of the upper limb exoskeleton, where θ i represents the joint angle of the upper limb exoskeleton joint i, and F = [f1…f6] represents the conversion relationship between the two, which can be obtained through the homogeneous transformation matrix established above.

[0077] Based on the theory of multivariate functions, by taking the derivative of both sides of the above equation, we can get the following equation:

[0078]

[0079] The above formula is expressed in vector form as follows:

[0080]

[0081] In the above formula, J(Θ) is the Jacobian matrix of the upper limb exoskeleton system, J v is the Jacobian matrix between the joint velocity and the linear velocity of the exoskeleton end, J w is the Jacobian matrix between the joint velocity and the angular velocity of the exoskeleton end.

[0082] Next, we use the vector product method to solve the Jacobian matrix, and finally we can get:

[0083]

[0084] Among them, J vi is the linear velocity Jacobian matrix at the end of the upper limb exoskeleton joint i, and J wi is the angular velocity Jacobian matrix at the end of the upper limb exoskeleton joint i. Both J vi and J wi are 6×7 matrices. is the unit vector of the z-axis of the coordinate system {i} in the base coordinate system, represents the rotation matrix of the coordinate system {i} relative to the base coordinate system, represents the position vector of the origin of the end coordinate system {7} of the exoskeleton system in the coordinate system {i}.

[0085] Step 3: Use the Lagrangian equation method to describe the dynamic model of the upper limb exoskeleton system;

[0086] The kinematic model studies how the movement of joints affects the movement of the end of the exoskeleton system, while the dynamic model studies how the forces or torques applied to the joints affect the movement of the exoskeleton, which is of great significance for the subsequent design of the exoskeleton control method. Here, the Lagrangian equation is used to establish the dynamic model of the upper limb exoskeleton system. First, define the Lagrangian function as follows:

[0087] L = T - H (7)

[0088] In the above formula, H and T respectively represent the potential energy and kinetic energy of the upper limb exoskeleton system.

[0089] Then the dynamic equation of the system can be written in the following form:

[0090]

[0091] Among them represents the differentiation with respect to time, and τ c is the control torque matrix provided by the joint drivers of the exoskeleton (i.e., the first motor to the seventh motor mentioned above).

[0092] Therefore, the establishment of the system dynamic model is transformed into the solution of the kinetic energy and potential energy of the upper limb exoskeleton system.

[0093] Step 3.1: Calculation of the total kinetic energy of the system

[0094] First, the kinetic energy of the upper limb exoskeleton system with a multi-link structure can be regarded as the sum of the kinetic energies of each link, that is:

[0095]

[0096] Among them, T i (0 < i ≤ 7) is the kinetic energy of link i, and m iis the mass of connecting rod i, I i is the moment of inertia matrix of connecting rod i, v i and w i are the three-dimensional vector of linear velocity and three-dimensional vector of angular velocity at the end of connecting rod i respectively.

[0097] From the above formula (5), we can see that for the link i, the relationship between the velocity and angular velocity of its center of mass in the operation space and the joint space can be expressed as follows:

[0098]

[0099] Among them J vi 、J wi are the linear velocity Jacobian matrix and angular velocity Jacobian matrix at the center of mass of the link i mentioned above. Therefore, the kinetic energy of the link i in the operating space can be expressed as follows:

[0100]

[0101] By adding up the kinetic energy of the seven links of the upper limb exoskeleton system, we can finally get the total kinetic energy expression of the upper limb exoskeleton system:

[0102]

[0103] in is a function related to the joint angle matrix Θ, which is defined as the inertia matrix of the exoskeleton system.

[0104] Step 3.2: Calculation of the total potential energy of the system

[0105] Since the potential energy of the exoskeleton system is mainly gravitational potential energy and other potential energy can be ignored, the total potential energy of the system can be expressed as follows:

[0106]

[0107] where m i is the mass of connecting rod i, g is the gravity coefficient matrix along the x, y, and z axes, r ci is the position of the center of mass of the connecting rod i in the base coordinate system, which is a 1*3 matrix. ci It is a function of the joint angle Θ, so the final system potential energy H can be described as a scalar function related to the joint angle Θ.

[0108] Step 3.3: Description of the joint space dynamics model

[0109] Substituting the above equations (12) and (13) into the Euler-Lagrange equation, that is, equation (8), we can obtain:

[0110]

[0111] By rearranging the above formula, we can get the following formula:

[0112]

[0113] Right now:

[0114]

[0115] In the above formula, M(Θ) is the inertia matrix of the upper limb exoskeleton system, is the Coriolis force and centrifugal force matrix of the upper limb exoskeleton, is the gravity matrix.

[0116] The above is the dynamic model of the upper limb exoskeleton. Since the upper limb exoskeleton system designed by the present invention is a human-machine coupling system, the interaction force between the human and the machine needs to be considered. Therefore, the above dynamic model can be rewritten as:

[0117]

[0118] where τ c The control torque matrix provided for the joint actuators of the exoskeleton (i.e., the first to seventh motors mentioned above), τ int It is the human-computer interaction torque matrix.

[0119] Step 3.4: Establishment of the dynamic model of the operating space

[0120] The above formula (17) describes the mapping relationship between joint torque and joint state, that is, the dynamic model of the joint space. The following is the relationship between the end state and the action force, which is the dynamic model of the operation space. It is in the following form:

[0121]

[0122] Where P = J + (Θ)Θ is the six-dimensional position vector of the upper limb exoskeleton system end in the Cartesian coordinate system, M c (Θ)=J +T (Θ)M(Θ)J + (Θ) is the inertia matrix of the upper limb exoskeleton system in the operating space, is the Coriolis force and centrifugal force matrix of the upper limb exoskeleton system in the operating space, G c (Θ)=J +T (Θ)G(Θ) is the gravity term of the upper limb exoskeleton system in the operating space, J + (Θ)=J T (Θ)(J(Θ)J T (Θ)) -1is the generalized inverse form of the Jacobian matrix, f c =J +T (Θ)τ c 、f int =J +T (Θ)τ int They are the control force and human-computer interaction force acting on the end of the upper limb exoskeleton system respectively.

[0123] At this point, the establishment of the dynamic model of the upper limb exoskeleton system has been completed, and the relationship between the end position and the end force of the upper limb exoskeleton system in the operating space has been clarified, that is, the conversion between the two can be achieved, which provides a basis for the subsequent design of the control law of the inner loop obstacle Lyapunov function controller (that is, the end force f after the motor torque is converted to the operating space). c The design of the upper limb exoskeleton system provides the mathematical model basis.

[0124] In the third aspect, the present invention also provides an internal and external loop control method for an upper limb exoskeleton system. In order to achieve the goal of on-demand auxiliary control, the control block diagram of the entire control system is designed as follows: Figure 14 As shown in the figure, the entire control system is divided into an inner loop and an outer loop. The outer loop is responsible for evaluating the patient's motor ability and calculating the optimal assistive force as well as the size of parameters such as trajectory radius and frequency. The assistive force is then converted into the position difference between the exoskeleton and the patient through the admittance model, and the desired position of the exoskeleton is transmitted to the inner loop. The inner loop achieves accurate trajectory tracking through a certain trajectory tracking method.

[0125] Step 1: First, the outer loop controller is designed. An energy function is used to evaluate the patient's current motor ability. Parameters such as the assist force and the frequency and radius of rehabilitation exercises are adjusted online through a Horf oscillator to achieve on-demand assisted rehabilitation. The specific implementation includes the following sub-steps:

[0126] Step 1.1: Calculation of auxiliary forces

[0127] First, the auxiliary force required by the patient is calculated based on the impedance model, which is the human-computer interaction force f int The force along the tracking error direction is as follows:

[0128]

[0129] Where k is the impedance stiffness of the impedance model, which is related to the patient's exercise ability, e d =||P d -P h ||2 represents the task tracking error, is the control quantity related to the error, and ξ is a positive real number parameter.

[0130] is the unit vector representing the direction of the auxiliary force, which is consistent with the direction of the task tracking error, that is:

[0131]

[0132] Among them, P d represents the trajectory of rehabilitation training, P h Represents the actual trajectory of rehabilitation trainees.

[0133] Step 1.2: Implement online adjustment of parameters such as impedance stiffness based on Horf oscillator

[0134] The Horf oscillator has the following form:

[0135]

[0136] In the above formula, x and y are the two state variables of the Horf oscillator, μ and w represent the limit cycle radius and the system oscillation frequency, respectively, γ is the tracking speed, E is the input signal to be synchronized, and ε is a positive coefficient. As the input signal E increases, the state variable x also increases, and as the input signal E decreases, the state variable x also decreases.

[0137] According to the regulation principle of the Horf oscillator mentioned above, the human body's motor ability is evaluated by combining the three motion parameters of task tracking error, auxiliary force, and motion period. The energy function E of the following form is designed as the input signal of the Horf oscillator:

[0138]

[0139] where e d is the task tracking error, f a For auxiliary force, is the phase of the rehabilitation training action, T is the time to complete the previous rehabilitation training task cycle, and β1, β2, and β3 are the coefficients of the three evaluation indicators.

[0140] It can be seen that the energy function E mainly consists of two terms. The first term is the accumulation of the error term and the auxiliary force term in the previous training cycle, and the second term is the time of the previous training cycle.

[0141] When the energy function E increases, it proves that the error in the previous cycle increases, the auxiliary force increases, and the training cycle becomes longer, which proves that the human body's motor ability becomes weaker. Therefore, it is necessary to increase the impedance stiffness and improve the auxiliary force level provided by the exoskeleton. At the same time, reduce the movement frequency of the next rehabilitation movement cycle and increase the movement radius to provide the patient with appropriate auxiliary force. When the energy function E decreases, it proves that the error in the previous cycle decreases, the auxiliary force decreases, and the training cycle becomes shorter, which proves that the human body's motor ability becomes stronger. Therefore, it is necessary to reduce the impedance stiffness and reduce the auxiliary force level provided by the exoskeleton. At the same time, increase the movement frequency of the next rehabilitation movement cycle and increase the movement radius to meet the patient's need for on-demand assistance. Based on the above analysis of the energy function, the Horf oscillator form for online adjustment of the movement trajectory radius and frequency and impedance stiffness is obtained as follows:

[0142]

[0143] Where k is the impedance stiffness of the above impedance model, y1, y2, y3 are the y state variables of the three Horf oscillators respectively, w1, w2, w3 are the oscillation frequencies of the three Horf oscillators respectively, is the inverse of the radius of the rehabilitation trajectory, is the reciprocal of the frequency of rehabilitation exercises. k 、μ k , ε k , γ r 、μ r , ε r , γ w 、μ w , ε w is the corresponding coefficient that needs to be adjusted in the oscillator.

[0144] Step 2: The above design has already completed the calculation of the assist force. This assist force is converted into the displacement difference between the exoskeleton system and the patient through the admittance model. The desired position of the exoskeleton is then calculated and transmitted to the inner loop. The inner loop uses the above-constructed dynamic model between the exoskeleton and the human body to design a controller using the obstacle Lyapunov function. Combined with the dynamic characteristics of time delay estimation, the RBF neural network is used to compensate for the time delay error, realizing the BLF control algorithm for accurate trajectory tracking of the inner loop. The specific implementation includes the following sub-steps:

[0145] Step 2.1: Establish the spatial state equation

[0146] According to the exoskeleton dynamics equation (18) in the above operation space, the following state equation can be defined:

[0147]

[0148] Where x1(t)=P is the six-dimensional position vector of the exoskeleton end in the Cartesian coordinate system, x2(t) is the derivative of x1(t) representing the velocity term, F0(t) is the end force term that converts the motor control torque in the joint space to the operation space, and N(t) is the sum of the centripetal force, Coriolis force, gravity, and external interaction forces. The mathematical expressions are as follows:

[0149]

[0150] Among them, M c (Θ) represents the inertia matrix of the upper limb exoskeleton system in the operating space, is the centripetal force and Coriolis force matrix of the upper limb exoskeleton system in the operating space, G c (Θ) is the gravity matrix of the upper limb exoskeleton system in the operating space, τ c is the input torque of the exoskeleton joint actuator, J(Θ) represents the Jacobian matrix of the upper limb exoskeleton system, and f int It is the interaction force between the upper limb exoskeleton terminal system and the human.

[0151] The position tracking error of the upper limb exoskeleton system can be expressed as:

[0152] e1=x1(t)-x a (t) (26)

[0153] where x a (t) = P de Represents the ideal position of the extremities of the upper limb exoskeleton system.

[0154] The velocity tracking error of the upper limb exoskeleton system can be expressed as:

[0155] e2=x2(t)-b (27)

[0156] Where b is a virtual control variable, whose value is given in the subsequent design.

[0157] Taking the derivative of the above e2, we can get:

[0158]

[0159] Step 2.2: Design the obstacle Lyapunov function and obtain the corresponding inner loop trajectory tracking control law

[0160] Choose the following logarithmic form of the Lyapunov function:

[0161]

[0162] where k e represents the error bound of the barrier Lyapunov function, which is a positive real number.

[0163] The differential form of the above Lyapunov function with respect to time is as follows:

[0164]

[0165] Will Substituting into the above formula, we can get:

[0166]

[0167] The control law of the obstacle Lyapunov controller is designed based on the backstepping method, and the virtual control variable b is designed as follows:

[0168]

[0169] Where k1 is a positive real number parameter.

[0170] At this time, substituting equations (32) and (28) into equation (31), we can obtain:

[0171]

[0172] Therefore, the control law of the obstacle Lyapunov controller designed based on the backstepping method can be designed as follows:

[0173]

[0174] Where k2 is a positive real number parameter.

[0175] At this time, substituting F0(t) in the above formula (34) into formula (33), we can get:

[0176]

[0177] Step 2.3: Delay Estimation Dynamics

[0178] The formal design of the BLF controller has been completed above. However, since the term N(t) contains the unknown dynamic characteristics of the upper limb exoskeleton system, the time delay estimation method is used below to process and represent the unknown term N(t), thereby realizing the construction of the BLF controller.

[0179] First, in general, the parameter matrix of the upper limb exoskeleton dynamics model can be expressed as follows:

[0180]

[0181] in are the inertia matrix M in the upper limb exoskeleton dynamic model c (Θ), Coriolis force and centrifugal force matrix Gravity matrix G cThe known part of (Θ) can be solved by the process of establishing the above dynamic model, is the inertia matrix M in the upper limb exoskeleton dynamics model c (Θ), Coriolis force and centrifugal force matrix Gravity matrix G c The unknown part of (Θ). Therefore, the variable It can be expressed as:

[0182]

[0183] The variables in the above formula (37) can be expressed as follows:

[0184]

[0185] Substituting the variable N(t) in equation (37) into equation (34), we can obtain:

[0186]

[0187] In the above formula, due to It only contains the known parts of the exoskeleton dynamic model, so it is a known quantity. is an unknown variable and can be obtained by the following delay estimation method:

[0188]

[0189] T0 in the above formula represents a very short delay. F0(t-T0), They are respectively the time T0 before the current time F0(t), respective values.

[0190] Step 2.4: RBF neural network estimation of delay error

[0191] Since the above delay is estimated using the last moment The value of is taken as the current value, and the error caused by this can be expressed as follows:

[0192]

[0193] From formula (41), we can see that the error caused by the delay estimation is related to e1 and e2. Since the RBF neural network can be used to fit unknown functions, it can also be used to estimate the error caused by the delay estimation. The structure of the RBF neural network is as follows: Figure 15 As shown in FIG, e1 and e2 are selected as input variables to construct an RBF neural network to fit the estimation error caused by the time delay estimation.

[0194] The radial basis function of the RBF network selects the following Gaussian function:

[0195]

[0196] In the above formula is the input error vector, σ k represents the center of the radial basis function, λ k Represents the width of the radial basis function, k=1,2…,l represents the nodes of the hidden layer of the RBF neural network.

[0197] Depend on Figure 15 It can be seen that the output of the RBF neural network is the linear sum of the hidden layer nodes, which is:

[0198]

[0199] where h(e)=[h1(e)…h k (e)…h l (e)] is the output matrix of each hidden node, and W is the weight value of each hidden node.

[0200] Therefore, the error caused by the above delay estimation can be expressed as:

[0201]

[0202] in is the weight value of each hidden layer node obtained after training, and δ is the approximation error.

[0203] Therefore, the complete control law form of the inner loop controller design is as follows:

[0204]

[0205] in is a positive real parameter, is the weight value of each hidden layer node obtained through training above, and its update rate is designed as follows:

[0206]

[0207] where Γ is a positive definite matrix.

[0208] At this point, the design of the inner-loop trajectory tracking controller is completed, which also completes the design of the entire closed-loop control system of the upper limb exoskeleton system.

[0209] Step 3: Finally, a joint simulation is performed using SOLIDWORKS and MATLAB, using the above-mentioned inner and outer loop control methods to verify and simulate the on-demand assistive functions of the upper limb exoskeleton system.

[0210] Since the muscle force of the human upper limb can be approximated as a spring damping model, the muscle force of the patient in the human-computer interaction system can be approximated by an impedance model, namely:

[0211]

[0212] where e d =||P d -P h ||2 represents the task tracking error, P d and P h are the ideal motion trajectory and actual motion trajectory of the patient mentioned above, K p , K d is the stiffness coefficient of the impedance model.

[0213] After obtaining the patient's end muscle force based on the above impedance model, it can be seen that the patient's muscle force is also along the tracking error direction during the simulation process. Therefore, in this simulation process, the auxiliary force f a and human-computer interaction force int are equivalent to each other. Using Jacobi transformation J a By using the Jacobian matrix corresponding to the patient's affected limb, the driving torque of each joint of the human body model can be obtained, realizing the joint simulation of the upper limb exoskeleton human-computer interaction system.

[0214] The inner loop controller adopts the above-mentioned BLF (Barrier Lyapunov) controller, and the parameter k in the inner loop controller is e Set to 0.01m, k1 = [30, 100], k2 = [450, 1000], time delay T0 = 0.001s, and the parameters of the RBF neural network are λ k =1.2, the number of hidden layer nodes is 15, σ k Uniformly distributed in the interval [-2,2], Γ is the unit matrix.

[0215] The rehabilitation motion trajectory is selected as a circular trajectory in the xy plane, with the center position selected at [0.35, 0], the initial motion radius is 0.15m, the motion frequency is 1.7rad / s, and the simulation time is selected as 50s. The human body stiffness coefficient K at the initial time p The impedance stiffness of the human body is set to 850 N / m to simulate patients with strong exercise ability. During the simulation, the impedance stiffness of the human body is gradually reduced at a rate of 16 N / m. At the end of the simulation, the impedance stiffness of the human body is 50 N / m to simulate patients with weak exercise ability. The parameters of the energy function in the above outer loop controller are β1=10, β2=0.15, β3=1.2, and the parameters of the Horf oscillator γ k =2*10 -3 、μk =6.25*10 4 , ε k =100,γ r =1, μ r =256,ε r =10,γ w =

[0216] 300, μ w =0.8,ε w =0.4.

[0217] Based on the above controller parameters, the simulation results of the upper limb exoskeleton system are as follows: Figures 16-21 shown. Figure 16 is the tracking error e of the inner loop BLF controller in the x and y directions x and e y From the simulation diagram, we can see that e x and e y Always converge to the error boundary k e Accurate trajectory tracking effect can be achieved. Figure 17 The figure shows the changes in the motion trajectory in the xy plane during the entire training process. It can be seen that when the system is running, as the patient's motor ability decreases, the patient's training motion reference trajectory gradually changes from the active training trajectory to the passive training trajectory, and the radius of the training motion reference trajectory gradually becomes smaller. Figure 18 The ideal motion trajectory and the real motion trajectory of the upper limb exoskeleton system in the x and y directions are shown. It can be seen more clearly that after the human body's motor ability weakens, the frequency w of the patient's training motion reference trajectory d Lower, movement radius r d Become smaller. Figure 19 Schematic diagram of the tracking error between the patient's actual motion trajectory and the patient's training reference trajectory. Figure 20 This is a schematic diagram of the changes in the auxiliary force provided by the exoskeleton system in the x and y directions. It can be seen that when the human body's movement ability weakens, the auxiliary force f provided by the exoskeleton system x and f y Gradually become larger to assist patients in completing rehabilitation exercises. Figure 21 The energy function E and the motion radius r are given in d , movement frequency w d , the impedance stiffness k change diagram, it can be clearly seen that after the human body's motor ability weakens, the radius r of the patient's training reference trajectory d Decrease, movement frequency w dThe patient's range of rehabilitation training activities is reduced to prevent secondary injury. At the same time, the impedance stiffness k of the upper limb exoskeleton system is increased to provide greater assistive force. It can be seen that the exoskeleton system can quickly adjust the reference trajectory and the size of the assistive force online according to the patient's motor ability, thereby achieving the function of assisting rehabilitation on demand.

[0218] The technical features of the above embodiments can be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0219] The above-described embodiments merely represent several implementation methods of the present application. While the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present application. It should be noted that those skilled in the art may make various modifications and improvements without departing from the spirit of the present application, all of which fall within the scope of protection of the present application. Therefore, the scope of protection of the present application shall be determined by the appended claims.

Claims

1. A seven-degree-of-freedom upper limb exoskeleton, characterized in that: include: A shoulder joint is constituted by a base, a first motor, a second motor, a third motor, a shoulder joint scissor-type structure, an upper arm support member, and a semicircular guide rail between the upper arm and the forearm; the first motor is fixed to the base, and its output shaft is connected to one end of the shoulder joint scissor-type structure, constituting a first degree of freedom of the shoulder joint abduction / adduction; the second motor is fixed to the other end of the shoulder joint scissor-type structure, and its output shaft is connected to one end of the upper arm support member, constituting a second degree of freedom of the shoulder joint flexion / extension; the upper arm support member is connected to the semicircular guide rail between the upper arm and the forearm, and the third motor is used to drive the semicircular guide rail member between the upper arm and the forearm to rotate, constituting a third degree of freedom of the shoulder joint internal rotation / external rotation; The elbow joint is formed by a fourth motor, a fifth motor, a forearm support member, and a semicircular guide rail between the forearm and the wrist; the fourth motor is connected to the semicircular guide rail between the upper arm and the forearm, and its output shaft is connected to the forearm support member, forming a first degree of freedom of flexion / extension of the elbow joint; the forearm support member is connected to the semicircular guide rail between the forearm and the wrist, and the fifth motor drives the semicircular guide rail between the forearm and the wrist to rotate, forming a second degree of freedom of internal / external rotation of the elbow joint; The wrist joint is composed of the sixth motor, the seventh motor, the wrist transition piece, the wrist joint support piece, and the handle; the sixth motor is fixed on the semicircular guide rail between the forearm and the wrist joint, and its output shaft is connected to the wrist transition piece, forming the first degree of freedom of the wrist joint: flexion / extension; the seventh motor is fixed on the wrist transition piece, and its output shaft is connected to the wrist support piece, forming the second degree of freedom of the wrist joint: abduction / adduction; the handle is fixed on the wrist support piece to provide support for the hand.

2. The seven-degree-of-freedom upper limb exoskeleton according to claim 1, characterized in that: The semicircular guide rail between the upper arm and the lower arm and the semicircular guide rail between the lower arm and the wrist joint have the same structure, and both include a fixed part, a gear and a movable part; the fixed part and the movable part are matched through an arc-shaped slide rail, and the movable part moves along the arc-shaped slide rail between the two; the corresponding motor is fixed on the fixed part of the semicircular guide rail, and its output shaft is connected to the gear, and the gear and the gear teeth of the movable part are engaged with each other.

3. The seven-degree-of-freedom upper limb exoskeleton according to claim 2, characterized in that: The length of the boom is adjusted by removing the bolts, adjusting the position of the fixing piece of the semicircular guide rail between the boom and the forearm along the two slide grooves on the boom support, and then tightening the bolts.

4. The seven-degree-of-freedom upper limb exoskeleton according to claim 2, characterized in that: The length of the forearm is adjusted by removing the bolts, adjusting the fixings of the semicircular guide rail between the forearm and the wrist joint along the two slide grooves on the upper arm support to a suitable position, and then fastening the bolts.

5. The seven-degree-of-freedom upper limb exoskeleton according to claim 1, characterized in that: The length of the wrist is adjusted by removing the bolts, adjusting the position of the grip along the slide groove of the wrist support, and then tightening the bolts.

6. The seven-degree-of-freedom upper limb exoskeleton according to claim 1, characterized in that: The grip is a rubber grip.

7. The seven-degree-of-freedom upper limb exoskeleton according to any one of claims 1 to 6, characterized in that: The conversion between the end position and end force of the upper limb exoskeleton is achieved through the following dynamic model of the operating space: Where P = J + (Θ)Θ is the six-dimensional position vector of the upper limb exoskeleton system end in the Cartesian coordinate system, M c (Θ)=J +T (Θ)M(Θ)J + (Θ) is the inertia matrix of the upper limb exoskeleton system in the operating space, is the Coriolis force and centrifugal force matrix of the upper limb exoskeleton system in the operating space, G c (Θ)=J +T (Θ)G(Θ) is the gravity term of the upper limb exoskeleton system in the operating space, J + (Θ)=J T (Θ)(J(Θ)J T (Θ)) -1 is the generalized inverse form of the Jacobian matrix, f c =J +T (Θ)τ c 、f int =J +T (Θ)τ int are the control force and human-machine interaction force on the upper limb exoskeleton end respectively; Θ=[θ1 … θ i ... θ7] represents the joint angles of the seven joints of the upper limb exoskeleton, where θ i is the joint angle of joint i of the upper limb exoskeleton system; M(Θ) is the inertia matrix of the exoskeleton system, are the Coriolis force and centrifugal force of the upper limb exoskeleton; G(Θ) is the gravity term; τ c is the control torque matrix provided by the node driver; τ int is the human-computer interaction torque matrix; J(Θ) is the Jacobian matrix of the upper limb exoskeleton system.

8. The seven-degree-of-freedom upper limb exoskeleton according to claim 7, characterized in that: in: m i is the mass of the equivalent link i between the joints, I i is the moment of inertia matrix of the equivalent link i; where J vi 、J wi are the linear velocity Jacobian matrix and angular velocity Jacobian matrix at the mass center of the equivalent link i, respectively; r ci is the position of the center of mass of the connecting rod i in the base coordinate system {0}, and g is the gravity coefficient matrix along the x, y, and z axes respectively.

9. The seven-degree-of-freedom upper limb exoskeleton according to any one of claims 1 to 6, characterized in that: By adjusting the auxiliary force f provided by the upper limb exoskeleton a , frequency of rehabilitation exercises d , rehabilitation movement trajectory radius r d Realize the function of on-demand assisted rehabilitation: Where k is the impedance stiffness of the impedance model, is the unit vector representing the direction of the auxiliary force, y1, y2, and y3 are the y state variables of the three Horf oscillators respectively, and w1, w2, and w3 are the oscillation frequencies of the three Horf oscillators respectively. is the inverse of the radius of the rehabilitation trajectory, is the reciprocal of the rehabilitation exercise frequency, γ k 、μ k , ε k , γ r 、μ r , ε r , γ w 、μ w , ε w is the corresponding coefficient that needs to be adjusted in the oscillator; is the input signal that needs to be synchronized, that is, the energy function for evaluating the patient's motor ability, where β1, β2, and β3 are the coefficients of each evaluation index.

10. The seven-degree-of-freedom upper limb exoskeleton according to any one of claim 9, characterized in that: The auxiliary force f obtained by the outer loop controller a The desired position of the upper limb exoskeleton system is finally obtained by converting the position difference between the patient and the exoskeleton through the admittance model. The exoskeleton system tracks the desired trajectory through the inner loop controller: Where e1 is the position tracking error of the exoskeleton, e2 is the speed tracking error of the exoskeleton, and k e represents the error bound of the obstacle Lyapunov function, k1 and k2 are positive real number parameters, b is the designed virtual control quantity, are the inertia matrix M in the upper limb exoskeleton dynamic model c (Θ), Coriolis force and centrifugal force matrix Gravity matrix G c The known part of (Θ), that is, the part that can be solved by the process of building the dynamic model; is the known part of the sum of centripetal force, Coriolis force, gravity and external interaction force at time t, which can be obtained by the known dynamic model of the upper limb exoskeleton system in the above formula. It is the unknown part of the sum of centripetal force, Coriolis force, gravity and external interaction force at time t; is the weight value of each hidden node of the RBF neural network after training, h(e) is the output matrix of each hidden node of the RBF neural network, is a positive real parameter; x a (t) represents the ideal position of the end of the upper limb exoskeleton system, T0 is a very short delay, They are respectively the time T0 before the current time Respective values; x2(t) represents the velocity term, and F0(t) is the end force term that converts the motor control torque in the joint space to the operation space.

Images