Flexible ankle exoskeleton for foot drop patients and control method thereof

Abstract

The present application belongs to the technical field of rehabilitation medical auxiliary, and discloses a flexible ankle exoskeleton for a patient with foot drop and a control method thereof, the control method comprising: calculating a tracking error between an actual position and a desired position of an ankle joint at a current time; reconstructing the tracking error to obtain a reconstructed error, wherein the normalized error is an error constraint boundary amplitude at the current time, the initial transition function is continuous and derivable, when t=0, the value of the initial transition function is in the interval (-1, 1), when 0 The above method has no limitation on the position at the initial time, and even if the position at the initial time is outside the error range, the control algorithm can still be executed and smoothly transitioned to the normal constraint condition.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of rehabilitation medical auxiliary, and more particularly, relates to a flexible ankle exoskeleton for patients with foot drop and a control method thereof. BACKGROUND

[0002] Foot drop is a common neuromuscular disease, and patients cannot actively perform ankle dorsiflexion when walking, resulting in difficulty in walking, abnormal gait, and serious impact on the quality of life. As a new emerging rehabilitation auxiliary device, the flexible exoskeleton can provide walking assistance for patients with foot drop. Among them, the rope-driven flexible exoskeleton has attracted more and more attention due to its light structure and comfortable wearing.

[0003] The existing flexible exoskeleton control method sets ankle position constraint conditions, and requires that the ankle trajectory error satisfies the constraint conditions. However, in actual application, the initial position of the ankle joint when starting assistance is random, and at the initial moment, the ankle position constraint condition is not necessarily satisfied, so that the control cannot continue to be executed, that is, the above method has limitations on the initial state, and the initial position of the ankle joint must satisfy the set ankle position constraint condition, therefore, the traditional flexible exoskeleton control method has application limitations. SUMMARY

[0004] In view of the above defects or improvement needs of the prior art, the present application provides a flexible ankle exoskeleton for patients with foot drop and a control method thereof, which aims to continue to execute the control when the initial position of the ankle joint does not satisfy the ankle position constraint condition.

[0005] To achieve the above object, the present application is proposed.

[0006] According to a first aspect of the present application, a flexible ankle exoskeleton control method for patients with foot drop is provided, which comprises:

[0007] calculating the tracking error between the actual position of the ankle joint at the current moment and the expected position ;

[0008] reconstructing the tracking error to obtain a reconstructed error , a normalized error , is the error constraint boundary amplitude at the current moment, is a preliminary transition function, and the function is continuous and derivable, when t=0, the value of is in the interval (-1, 1), when 0 the value of is smoothly transitioned from the value at 0 moment to the normalized error​ when t≥T,

[0009] The reconstruction error z is input into a control model to obtain the motor output current at the next moment, and the control model comprises an impedance model and a dynamics model of the flexible ankle exoskeleton system.

[0010] According to a second aspect of the present application, a flexible ankle exoskeleton for a patient with foot drop is provided, which comprises an exoskeleton wearing part, a motor driving device and a controller, the controller being configured to execute the flexible ankle exoskeleton control method for a patient with foot drop as described in any one of the above to adjust the motor output current of the motor driving device according to the actual position of the ankle joint, and the motor driving device being configured to drive the exoskeleton wearing part to move the ankle joint.

[0011] Overall, compared with the prior art, the above technical solutions conceived by the present application mainly have the following beneficial effects:

[0012] 1. The reconstruction algorithm of the tracking error is improved in the present application, and the reconstruction error is introduced. When t=0, , the value of is in the interval (-1, 1), which ensures that the logarithmic term of formula (6) has mathematical meaning; when 0 , the value of is smoothly transitioned from the value at time 0 to the normalized error ; when t≥T, . By designing the error reconstruction formula as above, , the value of is smoothly transitioned from the value at the initial moment to , and then the control can be performed according to the normal constraints. In this way, there is no restriction on the position at the initial moment, and even if the position at the initial moment is outside the error range, the control algorithm can still be executed and smoothly transitioned to the normal constraint condition.

[0013] 2. Further, considering that the amplitude of the position error may fluctuate greatly in some cases and exceed the above position constraint, resulting in control failure, in some embodiments, an adaptive term is introduced to adaptively adjust the amplitude of the error constraint boundary. The amplitude of the error constraint boundary can be adaptively adjusted according to the position error , and the real constraint boundary for dynamically controlling the system is adaptively adjusted, so that even if the tracking error exceeds the warning layer, the controller can still avoid failure.

[0014] 3. Further, considering that setting the stiffness as a fixed value is not conducive to the on-demand assistance, in some embodiments, the stiffness of the exoskeleton to pull the ankle joint is set as a time-varying function of the stiffness, which is closely related to the gait cycle, is continuously derivable on the whole, can realize seamless transition between gait phases, and provides multiple stiffness levels to realize the "on-demand assistance" function for the patient.

[0015] 4. Further, considering that the change of stiffness will introduce additional energy to the system and cause the system to be unstable, in some embodiments, an energy storage pool is constructed to facilitate the balanced exchange of system energy and ensure overall stability. BRIEF DESCRIPTION OF DRAWINGS

[0016] Figure 1 is a step flow chart of the flexible ankle exoskeleton control method in an embodiment of the present application.

[0017] Figure 2 is a control block diagram of the flexible ankle exoskeleton control method in an embodiment of the present application.

[0018] Figure 3 is a man-machine coupling model of the flexible ankle exoskeleton and the human body in an embodiment of the present application.

[0019] Figure 4 is a schematic diagram of the adaptive constraint boundary in an embodiment of the present application.

[0020] Figure 5 is a curve of the change of the exoskeleton stiffness in the whole gait cycle in an embodiment of the present application.

[0021] Figure 6 is an energy balance schematic diagram of the impedance model in an embodiment of the present application.

[0022] Figure 7 is a simulation result under the sinusoidal reference trajectory in an embodiment of the present application.

[0023] Figure 8 is the actual gait reference trajectory tracking effect in an embodiment of the present application. DETAILED DESCRIPTION

[0024] In order to make the objectives, technical solutions and advantages of the present application clearer, the present application will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and do not limit the present application. In addition, the technical features involved in the various embodiments of the present application described below can be combined with each other as long as they do not conflict with each other.

[0025] Embodiment 1

[0026] The application provides a flexible ankle exoskeleton control method for a patient with foot drop, which comprises the following steps Figure 1 Fig. 3 is a flow chart of the flexible ankle exoskeleton control method in an embodiment of the application, and Figure 2 Fig. 4 is a control block diagram of the flexible ankle exoskeleton control method in an embodiment of the application, and Figure 3 Fig. 5 is a man-machine coupling model of the flexible ankle exoskeleton and the human body in an embodiment of the application. The following will be described in detail.

[0027] S1, calculating a tracking error between a desired position and an actual position of an ankle joint at a current time . .

[0028] Specifically, a desired ankle joint motion trajectory needs to be determined in advance, and the control purpose is to make the actual ankle joint motion trajectory of the patient follow the desired value, that is, to ensure that the deviation between the two is within a preset deviation range.

[0029] wherein the tracking error .

[0030] S2, reconstructing the tracking error to obtain a reconstructed error , and a normalized error , is an error constraint boundary amplitude at the current time, is an initial transition function, the function is continuously derivable, when t=0, , when 0 , the value of the normalized error , when t≥T, .

[0031] Specifically, the system will set a tracking error constraint, and in order to ensure the stability of the control, the control algorithm needs to ensure that the tracking error satisfies the following tracking error constraint:

[0032] (1)

[0033] wherein is the tracking error at t, and are the lower limit and the upper limit of the tracking error at t, respectively.

[0034] In order to convert the constrained control problem into an unconstrained control problem, it is usually necessary to perform soft constraint conversion on the tracking error to obtain a reconstructed error, and then design the output torque of the motor based on the reconstructed error.​

[0035] The conventional method converts the tracking error as follows.

[0036] By presetting the performance function , the constraint boundary can be re-expressed as:

[0037] (2)

[0038] (3)

[0039] wherein, , , , and k is a positive number.

[0040] The transformation function is defined as follows to convert (1)~(3) into (4) and (5):

[0041] (4)

[0042] (5)

[0043] wherein z is the reconstruction error, is the normalized error.

[0044] After the conversion, the reconstruction error z is a logarithmic function, and the independent variable of the logarithm needs to be greater than 0, so as long as the logarithmic function is meaningful, the following condition can be met: always satisfies the constraint of formula (1), and the obtained reconstruction error can be any real number.

[0045] The above constraint also needs to satisfy the constraint of the initial deviation, that is, to satisfy: However, in actual application, the initial position of the ankle joint at the beginning of running is random and does not necessarily satisfy the above initial condition, so that the algorithm cannot continue to execute, that is, the above method has limitations on the initial state and has application limitations.

[0046] In the present application, in order to solve the above problems, the following optimized error reconstruction method is proposed.

[0047] The improved error reconstruction method is to optimize the reconstruction error expression of formula (5) so that it can smoothly transition from any initial position to formula (5). Specifically, the normalized error in formula (5) is replaced by , The value of is smoothly transitioned from any initial value at the initial time to , and then the control can be performed according to the constraint of formula (5) as above.

[0048] The expression of the optimized reconstruction error is as follows:

[0049] (6)

[0050] wherein, is a normalized error, is an initial transition function.

[0051] function is continuous and differentiable, and satisfies the following conditions:

[0052] when t = 0, the value of is in the interval (-1, 1), ensuring that the logarithmic term of equation (6) has mathematical meaning;

[0053] when 0 < t < T, the value of is smoothly transitioned from the value at time 0 to the normalized error ;

[0054] when t ≥ T, .

[0055] By designing the error reconstruction formula as above, the value of is smoothly transitioned from the value at the initial time to , and then the control can be executed according to the normal constraints. In this way, there is no restriction on the position at the initial time, and even if the position at the initial time is outside the error range, the control algorithm can still be executed and smoothly transitioned to the normal constraint condition.

[0056] In a specific embodiment, the function can be designed as follows:

[0057] (7)

[0058] wherein, is a piecewise function, which has the following form:

[0059] (8)

[0060] (9)

[0061] wherein T is a time constant, and a better value can be selected through experiments.

[0062] Specifically, equation (7) is divided into three stages:

[0063] (1) t = 0: equation (7) is , at this time is compressed to (-1, 1) to ensure that the value of the logarithmic term of equation (6) is always positive, so that it has mathematical meaning;

[0064] (2) 0 < t < T: In and between them is smoothly transitioned, which ensures has a compressive effect on throughout the time period T, and the degree of compression decreases as decreases;

[0065] (3) t≥T: In this stage , formula (6) returns to formula (5) at this time.

[0066] is continuous and derivable at t=T, so also remains continuous and derivable. Even when the position constraint condition is exceeded at t=0, formula (6) is still defined, eliminating the dependence on the initial condition .

[0067] In an embodiment, considering that the amplitude of the position error may fluctuate to a large extent in some cases, exceeding the above position constraint and leading to control failure. In an embodiment of the present application, the position constraint is further optimized as follows.

[0068] An adaptive term is introduced in the constraint boundary in formulas (2) and (3) , and the following adaptively adjusted performance function and the error constraint boundary amplitude are obtained:

[0069] (10)

[0070] In the formula, is a preset positive scaling gain used to adjust the degree of influence on the constraint boundary, is an adaptive term for adaptively adjusting the amplitude of the error constraint boundary, which is a dynamic adjustment term positively correlated with the amplitude of the tracking error .

[0071] In an embodiment, in order to be able to quickly adaptively adjust the constraint boundary with the change of the error , the adaptive term is constructed as follows:

[0072] (11)

[0073] In the formula, is the damping ratio of the adaptive adjustment link, is the natural frequency of the adaptive adjustment link, is a set proportional factor, an adaptive parameter introduced, a preset non-zero small value, ensuring continuous and derivable.

[0074] The adaptive term constructed by the above formula (11) is a function whose value range is (-1, 1), thereby limiting the range of adaptive adjustment. As shown in Figure 4 , the constraint boundary after the above adaptive adjustment can be divided into two layers, a warning layer set for the system, which only depends on time, within which the system can maintain high positioning tracking accuracy, however, the warning layer does not directly represent the actual position constraint of the system, an adaptive adjustment layer introduced, which is used to dynamically control the real constraint boundary of the system, thereby avoiding controller failure even if the tracking error exceeds the warning layer.

[0075] S3, input the reconstruction error z into the control model to obtain the motor output current at the next moment, the control model including an impedance model and a dynamics model of the flexible ankle exoskeleton system.

[0076] Specifically, the dynamics model of the flexible ankle exoskeleton system reflects the relationship between the motor output torque and the human ankle movement state, and by adjusting the motor output current, the motor output torque is adjusted, and then the human ankle movement trajectory is adjusted.

[0077] Generally, the construction process of the dynamics model of the flexible ankle exoskeleton system is as follows.

[0078] First, the dynamics of the motor is analyzed to obtain the following coupling model of the human body and the exoskeleton:

[0079] (12)

[0080] In the formula, and are the inertia and damping of the motor rotor, is the torque constant, is the motor current, and are the load torque and angular position of the motor shaft, is the reduction ratio, and are the torque and angular position output by the motor reducer.

[0081] In the process of assisting the patient with the exoskeleton, the Bowden wire core is always in a taut state, at which time the relationship between the motor output tension and the actual tension applied to the instep by the Bowden wire is as follows:

[0082] (13)

[0083] where, is the tension of the motor output, is the radius of the bobbin on the motor, represents the tension of the Bowden cable on the ankle joint, is the tension loss due to the friction between the bobbin and the cable sheath and the compliance of the cable sheath, usually represented by a nonlinear function.

[0084] At the ankle joint, the ankle torque is obtained by analyzing the geometric relationship between the exoskeleton and the tension , as follows:

[0085] (14)

[0086] where, the geometric configuration parameters , represent the ratio of the distance between the ankle anchor point and the shank anchor point, and are the distance between the ankle center and the instep anchor point, the distance between the ankle center and the shank anchor point, respectively, is the angular position of the ankle joint.

[0087] Since the bobbin is always kept under tension, the change in the length of the bobbin on the motor bobbin is equal to the change in the length of the bobbin at the instep, so the relationship between the motor rotation angle and the human ankle angle can be established as follows:

[0088] (15)

[0089] where, and are the initial angular positions of the motor and the human ankle joint, respectively, and are the angular positions of the motor and the human ankle joint at time t, respectively.

[0090] Based on equations (12)~(15), the relationship between the angular velocity and the angular acceleration of the motor angular position and the human ankle angular position is as follows:

[0091] (16)

[0092] where, is the motion transmission Jacobian function, representing the transmission ratio relationship between the motor angular velocity and the ankle angular velocity, is the transmission ratio parameter, representing the ratio of the ankle anchor point radius to the motor bobbin radius, and other parameters are described above.

[0093] Based on the above formula, the dynamics model of the flexible ankle exoskeleton system is constructed as follows:

[0094] (17)

[0095] wherein, is the inertia matrix of the human ankle before reconstruction, is the centrifugal and Coriolis matrix of the human ankle before reconstruction, is the damping matrix of the human ankle before reconstruction, is the motion transmission Jacobian function, representing the transmission ratio relationship between the motor angular velocity and the ankle angular velocity, is the motor output torque.

[0096] In order to adapt to the error reconstruction, the dynamics model of the flexible ankle exoskeleton system also needs to be reconstructed, and the specific process is as follows.

[0097] From equation (6), we have:

[0098] (18)

[0099] wherein, is the first order gain term of error mapping, is the nonlinear compression factor, is the error dynamic compensation term, used to compensate the influence caused by the time-varying transition function and adaptive boundary, and other parameters are described above.

[0100] Based on equation (18), the following relationship can be obtained:

[0101] (19)

[0102] wherein, the parameters are described above.

[0103] Finally, based on equations (6), (17), (18), (19), the dynamics model of the reconstructed flexible ankle exoskeleton system can be obtained as follows:

[0104] (20)

[0105] wherein, M Z , C Z , D Z are the inertia term, the centrifugal and Coriolis term, and the friction term, respectively, are the motor output torque after error reconstruction, the human-machine interaction torque, the friction torque, and the remaining torque item except the above three.

[0106] M Z , C Z , DZ , is expressed as follows:

[0107] (21)

[0108] wherein each parameter is introduced above.

[0109] Thus, a reconstructed dynamics model of the flexible ankle exoskeleton system adapted to error reconstruction can be obtained.

[0110] Specifically, the impedance model can adopt a traditional impedance model, and a formula of the impedance model is as follows:

[0111] (22)

[0112] wherein, , , are respectively expected inertia, damping and stiffness, and the three are all known parameters.

[0113] Combining the impedance model (22) and the dynamics model (20), (21) of the flexible ankle exoskeleton system, the motor output current i can be solved.

[0114] In a traditional control system, the stiffness is usually set as a fixed value. The stiffness size affects the impedance size of the exoskeleton dragging the ankle joint. The lower the stiffness, the lower the impedance, the weaker the traction ability of the exoskeleton to the ankle joint, and the greater the ankle joint trajectory deviation; the higher the stiffness, the higher the impedance, the stronger the traction ability of the exoskeleton to the ankle joint, and the smaller the ankle joint trajectory deviation. It can be analogized as that the exoskeleton drags the ankle joint through an elastic variable spring. The lower the stiffness, the softer the spring, the less controlled the ankle joint, and the higher the stiffness, the harder the spring, the more controlled the ankle joint. However, in actual application, when the exoskeleton is used to assist patients in ankle joint movement, it is not expected that the ankle joint trajectory of the patient can always follow the expected trajectory. The heel landing and the toe leaving are key events when assisting the foot drop patient to walk. When the heel landing is detected, it means that the support phase is started, and the exoskeleton should reduce the impedance to enable the patient to maintain a natural gait pattern. When the toe leaving, i.e. the swing phase, is detected, the exoskeleton should increase the impedance to provide sufficient assistance for the ankle dorsiflexion. If the stiffness is set as a fixed value, the assistance demand of the foot drop patient cannot be fully met.

[0115] Based on this, in some embodiments, the stiffness of the exoskeleton dragging the ankle joint can be further set as a time-varying function of time-varying stiffness, and the time-varying function of time-varying stiffness satisfies: when entering the support phase, the stiffness is smoothly decreased to 0 or close to 0, and when entering the swing phase, the stiffness is smoothly increased to the maximum stiffness.

[0116] In a specific embodiment, the time-varying function of stiffness is in the form of:

[0117] (23)

[0118] (24)

[0119] wherein, is the time-varying function of stiffness, is the maximum stiffness, denotes the percentage position of the gait cycle at the current time t, with the value range of [0, 1], is the stiffness adjustment function, denotes the percentage position of the gait cycle at the toe-off time.

[0120] As shown in Figure 5 , the above time-varying function of stiffness is closely related to the gait cycle, is continuously derivable globally, can realize seamless transition between gait phases, and provides multiple stiffness levels to realize the “on-demand assistance” function for patients.

[0121] In the case of considering the time-varying stiffness, the following Lyapunov candidate function is introduced:

[0122] (25)

[0123] Taking the derivative of (26) , which is a constant, we have:

[0124] (26)

[0125] wherein, The existence of the term means that the system has deviated from passivity, i.e., there is a risk of instability, because the stiffness change introduces additional energy into the system, causing the system to be unstable. If the energy dissipated by the system exceeds the energy injected by the stiffness change, the system can still maintain passivity.

[0126] Therefore, on the basis of the adjustable stiffness scheme, to ensure the stability of the system, a energy storage pool can be further introduced to store the energy dissipated by the system , thereby supporting more flexible impedance adjustment and maintaining the overall stability of the system.

[0127] In order to establish the relationship between the dissipated energy , the energy storage pool energy and the energy injected by the stiffness change, the impedance model in equation (22) is optimized to obtain the following optimized impedance model:

[0128] (27)

[0129] (28)

[0130] where, is a parameter for controlling the charging rate of the energy storage tank, is a set value, and λ is a set normal number for adjusting the system response speed, and z t is the state variable of the energy storage tank, is a control function for adjusting the energy exchange between the energy storage tank and the impedance model, and its expression is:

[0131] (29)

[0132] where, represents the energy in the energy storage tank, is the minimum energy threshold of the energy storage tank, which is a small normal number. When the energy of the energy storage tank is insufficient, the system stiffness remains unchanged to prevent from being too small to make equation (28) singular.

[0133] For the gait assistance of patients with foot drop, it can be seen from Figure 6 that during the support phase , the energy storage tank absorbs the energy generated by and , which is beneficial to the passivity of the system. While in the swing phase , the stiffness change will inject energy into the system, which is not conducive to the passivity of the system. The support and swing phase time is roughly equal, which helps to balance the exchange of system energy and ensures the overall stability.

[0134] The dynamic model of the reconstructed flexible ankle exoskeleton system and the above-optimized impedance model can be combined to express the motor output torque as follows:

[0135] (30)

[0136] (31)

[0137] where, is an intermediate parameter.

[0138] Then, combined with , the motor output current can be calculated.

[0139] The following formula verification is performed on the above equations (30) and (31).

[0140] Let , and consider another Lyapunov function:

[0141]

[0142] Its derivative is:

[0143] (32)

[0144] Thus, we have When the control parameter of charging rate is adjusted to satisfy the following relationship:

[0145] (33)

[0146] For the input-output pair Integrating over time we have

[0147] (34)

[0148] At this time, the system (20) and (27)-(31) is passive with respect to the input-output pair .

[0149] According to equation (32), the derivative of Lyapunov function can be written as:

[0150] (35)

[0151] where is the limit of human-robot interaction torque and the charging rate satisfies equation (33), Once reaches the limit , makes eventually fall into the set .

[0152] Let its derivative is:

[0153] (36)

[0154] Therefore, will also eventually fall into the set , thus ensuring that the proposed controller (27)-(31) can maintain the position constraint condition (1).

[0155] Embodiment 2

[0156] The application further provides a flexible ankle exoskeleton, comprising an exoskeleton wearing part, a motor driving device and a controller, wherein the controller is used to execute the flexible ankle exoskeleton control method for a foot drop patient in embodiment 1 to adjust the motor output current of the motor driving device according to the actual position of the ankle joint, and the motor driving device is used to drive the exoskeleton wearing part to drive the ankle joint to move.

[0157] It can be understood that the above flexible ankle exoskeleton further comprises a collection device for collecting parameters such as the actual position of the ankle joint, human-computer interaction force, gait phase information and current running time.

[0158] Embodiment 3

[0159] The above stiffness adaptive assistance scheme is verified by specific examples.

[0160] A Simulink simulation model containing a human ankle joint and a flexible exoskeleton is established, and key system parameters are set as shown in Tables 1, 2, 3 and 4.

[0161] Table 1 is the simulation parameters related to the motor and driving, and the specific contents are as follows.

[0162]

[0163] Table 2 is the exoskeleton modeling related parameters, and the specific contents are as follows.

[0164]

[0165] Table 3 is the soft constraint related parameters, and the specific contents are as follows.

[0166]

[0167] Table 4 is the impedance control related parameters, and the specific contents are as follows.

[0168]

[0169] First, the ankle joint position reference trajectory of the system is , assuming that the support phase and swing phase each account for 50% of the entire trajectory period, the trajectory tracking of the controller under this input is as shown in Figure 7 It can be seen that the actual trajectory in the swing phase closely follows the reference trajectory, while the actual trajectory in the support phase deviates from the reference trajectory to a certain extent, indicating that the stiffness of the control algorithm changes in different gait phases, thereby causing the trajectory deviation to change. In addition, in order to be more consistent with the real use scene, the data of ten gait cycles and gait phases of a normal person when walking stably are collected as the input of the controller, and the results are as shown in Figure 8It can also be seen from the figure that there is a certain deviation between the actual trajectory and the reference trajectory output by the controller when the gait enters the support phase, indicating that the controller adjusts the reference trajectory according to the actual human-machine interaction force at this time, and is more inclined to output the assistance trajectory according to the user's intention. When the gait enters the swing phase, the actual trajectory and the reference trajectory are basically coincided, at this time, the dorsiflexion assistance is provided to the patient to help the patient lift the toes.

[0170] The technical features of the above embodiments can be combined in any manner. In order to make the description simple, all possible combinations of the technical features in the above embodiments are not described, however, as long as the combinations of the technical features do not contradict, they should be considered as the scope of the present disclosure. It should be noted that the "in an embodiment of the present application", "for example", "for example" and the like are intended to illustrate the present application, and are not used to limit the present application.

[0171] The above embodiments only express several embodiments of the present application, and the description is more specific and detailed, but it should not be understood as limiting the scope of the patent application. It should be noted that for those skilled in the art, without departing from the concept of the present application, a number of modifications and improvements can be made, which are within the scope of the present application.

Claims

1. A flexible ankle exoskeleton control method for a patient with foot drop, characterized by, Comprising: calculating a tracking error between the actual position of the ankle joint at the current time instant and the desired position ;​​ tracking error reconstruction, to obtain a reconstruction error , a normalized error , is a real error constraint boundary amplitude at the current time, is a preliminary transition function, the function is continuously derivable, when t = 0, the value of is in the interval (-1, 1), when 0 < t < T, the value of is smoothly transitioned from the value at time 0 to the normalized error , when t ≥ T, ; inputting the reconstruction error z into a control model to obtain motor output current at next time, the control model comprising an impedance model and a dynamics model of the flexible ankle exoskeleton system.

2. The control method according to claim 1, characterized by, Function is of the form: wherein is a piecewise function, and T is a set time constant.

3. The control method according to claim 1, characterized by, The control method further comprises calculating an error constraint boundary amplitude performing adaptive adjustment, the error constraint boundary amplitude The calculation formula of the error constraint boundary amplitude is In the formula, is a preset positive scaling gain, is an adaptive term for introducing adaptive adjustment of the error constraint boundary amplitude, and are respectively the error constraint boundary amplitude set by the system and the performance function and scale parameter used for conversion, satisfying , is the performance function after adaptive adjustment. Among them, the adaptive term is a dynamic adjustment term positively correlated with the tracking error , when the amplitude of the tracking error increases, the value also tends to increase as a whole, thereby dynamically relaxing the constraint boundary.

4. The control method according to claim 3, characterized by, Adaptive term is the hyperbolic tangent function, whose expression is: wherein is the damping ratio of the adaptive adjustment link, is the natural frequency of the adaptive adjustment link, is a set proportionality factor, is an introduced adaptive parameter, is a preset non-zero parameter.

5. The control method according to claim 2, characterized by, The dynamics model of the flexible ankle exoskeleton system is: where M Z , C Z , D Z are the inertia, centrifugal and Coriolis terms, and the friction term, respectively, , , , are the reconstructed motor output torque, the human-machine interaction torque, the friction torque, and the remaining torque term excluding the above three, respectively, and their calculation formulas are as follows: wherein is a first order gain term of error mapping, is a non-linear compression factor, is an error dynamic compensation term, is the inertia matrix of the human ankle before reconstruction, is the matrix of centrifugal and Coriolis forces of the human ankle before reconstruction, is the damping matrix of the human ankle before reconstruction, is the motion transmission Jacobian function, is the transmission ratio parameter, is the geometric configuration parameter, and are the inertia and damping of the motor rotor, respectively, is the reduction ratio, is the radius of the bobbin winding wheel on the motor, and are the distance between the ankle center and the instep anchor point, the distance between the ankle center and the shank anchor point, respectively, f is the tension loss due to the friction between the core and the sleeve and the compliance of the sleeve, i is the motor output current, is the torque constant, represents the pulling force of the Bowden cable on the ankle.

6. The control method according to claim 5, characterized by, The impedance model is: wherein , , are the desired inertia, damping and stiffness, respectively.

7. The control method according to claim 6, characterized by, The desired stiffness is a stiffness time-varying function, and the stiffness time-varying function satisfies: when entering the support phase, the stiffness smoothly decreases to 0 or approaches 0; and when entering the swing phase, the stiffness smoothly increases to the maximum stiffness.

8. The control method according to claim 7, characterized by, The form of the stiffness time-varying function is: wherein, is a time-varying function of stiffness, is the maximum stiffness, denotes the percentage position of the gait cycle at the current time t, with values in the interval [0, 1], is a stiffness adjustment function, denotes the percentage position of the gait cycle at the time of toe-off.

9. The control method according to claim 5, characterized by, When the desired stiffness is a stiffness time-varying function, to ensure system stability, the impedance model adopts a form with an energy storage tank: wherein, , are the desired inertia, damping, is the stiffness time-varying function, λ is a positive number set to adjust the system response speed, z t is the energy storage tank state variable, is the control function that regulates the energy exchange between the energy storage tank and the impedance model, is a parameter used to control the energy storage tank charging rate, denotes the energy in the energy storage tank, is the minimum energy threshold of the energy storage tank.

10. A flexible ankle exoskeleton for a patient with foot drop, characterized in that, An exoskeleton wearing part, a motor driving device and a controller are included, the controller is used to execute the flexible ankle exoskeleton control method for foot drop patients according to any one of claims 1 to 9 to adjust the motor output current of the motor driving device according to the actual position of the ankle joint, and the motor driving device is used to drive the exoskeleton wearing part to drive the ankle joint to move.

Images