Multi-modal control method, system and equipment for lower limb exoskeleton and medium

Through the multimodal control method, combined with the gait trajectory library, LSTM network, CPG model and adaptive RBF reverse step controller, the problem of exoskeleton control mismatch and nonlinear dynamics processing in the existing technology is solved, and the outer part of the lower limbs is realized Accurate, comfortable and flexible control of bones.

CN118963399APending Publication Date: 2024-11-15BEIHANG UNIV
View PDF 0 Cites 3 Cited by

Patent Information

Application Number
CN202411070466.9
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-06
Publication Date
2024-11-15

AI Technical Summary

Technical Problem

The existing methods of exoskeleton control for lower limbs lack consideration of individual differences, it is difficult to match the wearer's natural gait, and it does not respond quickly to changes in walking speed and pattern, so it cannot effectively deal with the nonlinear dynamic characteristics of the exoskeleton system, affecting control Accuracy and comfort.

Method used

Multimodal control methods are adopted, including gait trajectory library selection, phase estimation of pre-trained LSTM network, CPG model generation expectation trajectory, variable parameter admission controller and adaptive RBF inverse step controller, combined with human-computer interactive torque modulation Factors to achieve precise control of the exoskeleton system.

Benefits of technology

Improve the accuracy and stability of exoskeleton control, enhance the adaptability of the system and wearer's comfort, and can flexibly switch under different load conditions and operating modes, ensuring the synchronization of exoskeleton movement and wearer's gait and Nature.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118963399A_ABST
    Figure CN118963399A_ABST
Patent Text Reader

Abstract

The invention provides a multi-modal control method for a lower limb exoskeleton, and the method comprises the steps: selecting a gait track from a gait track library as an initial track according to the body feature parameters of a wearer, and carrying out the gait phase estimation through combining with a phase estimation model; using a CPG model to obtain an exoskeleton expected trajectory; inputting the exoskeleton expected trajectory and a lower limb motion trajectory in the exoskeleton model into a variable parameter admittance controller, performing modulation factor adjustment in combination with a man-machine interaction torque, and outputting a reference trajectory; and based on the reference trajectory, performing multi-modal control of the lower limb exoskeleton through a position controller based on adaptive RBF backstepping control. According to the method, in an admittance controller, a modulation factor is introduced to realize smooth transition among different control modes, so that an exoskeleton can be flexibly switched among different operation modes, accurate control of an exoskeleton system is realized through the synergistic effect of the steps, and the adaptability of the system and the comfort of a wearer are improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of lower limb exoskeleton control, and more specifically to a multimodal control method, system, device and medium for lower limb exoskeleton. Background Art

[0002] Lower limb exoskeleton is a human augmentation device that assists the wearer to complete specific tasks. It has broad application prospects in industrial production and military applications. Existing exoskeleton control methods mainly face the following challenges:

[0003] They often adopt fixed gait patterns and lack consideration for individual differences, resulting in the mismatch between the exoskeleton's movement and the wearer's natural gait, affecting the wearer's comfort and efficiency. In addition, traditional control methods usually do not consider changes in the wearer's physical characteristics, such as the decrease in leg muscle strength when fatigue accumulates, which may lead to incoordination between the exoskeleton's movement and the human body's movement.

[0004] At the same time, the wearer's walking speed and pattern changes are not responsive enough, making it difficult to adjust the gait trajectory in real time, which affects the synchronization between the exoskeleton and the wearer. The accuracy of phase control is not high, especially in complex environments and high-dynamic motion conditions, which makes it difficult for the control method to accurately adjust the movement of the exoskeleton.

[0005] In addition, existing control methods are usually based on linear assumptions and cannot effectively deal with the inherent nonlinear dynamic characteristics of exoskeleton systems. There is a lack of effective mechanisms to adjust its stiffness, damping and other parameters to adapt to the movement requirements under different load conditions.

[0006] Therefore, how to design a multimodal control method for lower limb exoskeleton to achieve precise control of the exoskeleton system and improve the adaptability of the system and the comfort of the wearer is an urgent problem that technicians in this field need to solve. Summary of the invention

[0007] In view of this, the present invention provides a multimodal control method for a lower limb exoskeleton, which achieves efficient, safe and comfortable control of the exoskeleton system through advanced algorithms and intelligent control strategies, helping to improve the wearer's quality of life.

[0008] In order to achieve the above object, the present invention adopts the following technical solution:

[0009] In a first aspect, the present invention provides a multimodal control method for a lower limb exoskeleton, comprising the following steps:

[0010] S1. According to the wearer's body characteristic parameters, a gait trajectory is selected from the gait trajectory library as the initial trajectory, and gait phase estimation is performed in combination with the phase estimation model;

[0011] S2, using the CPG model to process the initial trajectory and gait phase to obtain the desired trajectory of the exoskeleton;

[0012] S3, inputting the desired trajectory of the exoskeleton and the lower limb motion trajectory in the exoskeleton model into a variable parameter admittance controller, adjusting the modulation factor in combination with the human-computer interaction torque, and outputting a reference trajectory;

[0013] S4. Based on the reference trajectory, multi-modal control of the lower limb exoskeleton is performed through a position controller based on adaptive RBF backstepping control.

[0014] Among them, in S1, the phase estimation model adopts a pre-trained long short-term memory network to perform real-time analysis on data collected by accelerometers, gyroscopes and other sensors, integrates a feature extraction module for identifying the wearer's specific gait pattern, and combines an online learning mechanism to iteratively optimize model parameters to adapt to individual differences and environmental changes.

[0015] Furthermore, in S2, the CPG model includes:

[0016] A neural network: comprising a plurality of oscillator units, wherein each oscillator unit comprises at least two oscillators connected to each other to generate a phase difference or a frequency difference;

[0017] An interface: used to receive signals from external systems.

[0018] Among them, the CPG model is integrated with connection weights to adjust the signal transmission method between units to simulate the coordinated movements between muscles and joints; combined with the parameter adjustment mechanism, the frequency, amplitude and connection weights of the oscillator are adjusted to adjust the output gait trajectory.

[0019] Furthermore, in S3, the variable parameter admittance controller is:

[0020]

[0021] Among them, α and β are modulation factors, H a represents the adaptive inertia coefficient, B a represents the adaptive damping coefficient, K a represents the adaptive stiffness coefficient, Δx(t) represents the trajectory deviation, Δx(t)' represents the first-order derivative of the trajectory deviation Δx(t), Δx(t)" represents the second-order derivative of the trajectory deviation Δx(t), τ ext represents the human-computer interaction torque, x d represents the desired trajectory of the exoskeleton, x h represents the lower limb motion trajectory, x r represents the reference trajectory.

[0022] Further, in S3, the modulation factor includes: a first modulation factor α and a second modulation factor β;

[0023] Among them, the first modulation factor α is used to adjust the admittance parameter, α≥0; the second modulation factor β is used to adjust the proportion of the lower limb movement trajectory in the reference trajectory, 0≤β≤1.

[0024] Furthermore, in S3, the modulation factor is adjusted, including achieving a smooth transition between different modes:

[0025]

[0026] Among them, the lower limb exoskeleton does not provide assistance in Passive mode, the lower limb exoskeleton provides assistance within a preset range in Assist-as-needed mode, and the lower limb exoskeleton provides maximum assistance in Active mode.

[0027] Further, in S4, the position controller based on adaptive RBF backstepping control includes:

[0028] An RBF network is used to approximate the nonlinear dynamic model of the exoskeleton system; the RBF network includes a plurality of radial basis function units, each of which has an adjustable center point and width parameter;

[0029] a backstepping controller, which designs a control law based on the output of the RBF network so that the exoskeleton can move along a desired trajectory;

[0030] The parameters and control law of the RBF network are adjusted in combination with the adaptive learning mechanism.

[0031] In a second aspect, the present invention provides a multimodal control system for a lower limb exoskeleton, comprising:

[0032] Gait trajectory selection and phase estimation module: used to select a gait trajectory from the gait trajectory library as the initial trajectory according to the wearer's physical characteristic parameters, and perform gait phase estimation in combination with the phase estimation model;

[0033] Expected trajectory generation module: used to process the initial trajectory and gait phase using the CPG model to obtain the expected trajectory of the exoskeleton;

[0034] Reference trajectory generation module: used to input the desired trajectory of the exoskeleton and the lower limb motion trajectory in the exoskeleton model into the variable parameter admittance controller, adjust the modulation factor in combination with the human-computer interaction torque, and output the reference trajectory;

[0035] Position controller module: used for performing multi-modal control of the lower limb exoskeleton based on the reference trajectory through a position controller based on adaptive RBF backstepping control.

[0036] In a third aspect, the present invention provides an electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the above-mentioned multimodal control method for a lower limb exoskeleton when executing the computer program.

[0037] In a fourth aspect, the present invention provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the above-mentioned multimodal control method for a lower limb exoskeleton.

[0038] The descriptions of the second to fourth aspects of the present invention can refer to the detailed description of the first aspect; and the beneficial effects of the descriptions of the second to fourth aspects can refer to the analysis of the beneficial effects of the first aspect, which will not be repeated here.

[0039] It can be seen from the above technical solution that compared with the prior art, the present invention has the following beneficial effects:

[0040] 1. By analyzing the wearer's physical characteristic parameters (such as height, weight, leg length, etc.), the most appropriate gait trajectory is selected from the pre-built gait trajectory library as the initial trajectory to ensure that the exoskeleton movement is closer to the wearer's natural gait. Combined with the phase estimation model (pre-trained LSTM network), real-time monitoring of the wearer's gait phase can improve the accuracy and stability of exoskeleton control. It helps to adapt to changes in the wearer's walking pattern in a timely manner, such as speed changes, terrain changes, etc.

[0041] 2. The CPG model can generate a rhythm and pattern similar to the natural walking of the human body, thus producing a smoother and more natural exoskeleton movement. By adjusting the parameters in the CPG model, the coordination between the various parts of the exoskeleton can be optimized to reduce unnecessary vibration and instability.

[0042] 3. By combining the human-machine interaction torque, the variable parameter admittance controller can adjust its parameters (such as inertia coefficient, damping coefficient and stiffness coefficient) to cope with different load conditions and user preferences. Adjusting the modulation factor can achieve a smooth transition between different control modes, allowing the exoskeleton to flexibly switch between different operation modes to adapt to various activity requirements.

[0043] 4. RBF (Radial Basis Function) network can approximate complex nonlinear dynamic models, so that the controller can handle the inherent nonlinear characteristics of the exoskeleton system. And adaptive RBF backstepping control can adjust the parameters of the RBF network according to the real-time state of the system, so that the control strategy can adapt to changing working conditions. The control law designed by backstepping control can ensure that the exoskeleton moves according to the reference trajectory and maintains high tracking accuracy even under external interference. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0045] Figure 1 A flow chart of a multimodal control method for a lower limb exoskeleton provided by an embodiment of the present invention;

[0046] Figure 2 A multimodal control framework diagram for a lower limb exoskeleton provided by an embodiment of the present invention;

[0047] Figure 3 A schematic diagram of the structure of a multimodal control system for a lower limb exoskeleton provided by an embodiment of the present invention;

[0048] Figure 4 A schematic diagram of the structure of an electronic device is provided for an embodiment of the present invention. DETAILED DESCRIPTION

[0049] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0050] The control method provided in the embodiment of the present application can be applied to a control server. The above-mentioned control server can be hardware or software. When the control server is hardware, it can be implemented as a distributed server cluster that provides control services, or it can be implemented as a single server. When the control server is software, it can be installed in the servers listed above. It can be implemented as multiple software or software modules, or it can be implemented as a single software or software module, which is not specifically limited here.

[0051] Embodiment 1;

[0052] like Figure 1 and Figure 2 As shown, this embodiment provides a multimodal control method for a lower limb exoskeleton, comprising the following steps:

[0053] S1. According to the wearer's body characteristic parameters, a gait trajectory is selected from the gait trajectory library as the initial trajectory, and gait phase estimation is performed in combination with the phase estimation model;

[0054] S2, using the CPG model to process the initial trajectory and gait phase to obtain the desired trajectory of the exoskeleton;

[0055] S3, inputting the desired trajectory of the exoskeleton and the lower limb motion trajectory in the exoskeleton model into a variable parameter admittance controller, adjusting the modulation factor in combination with the human-computer interaction torque, and outputting a reference trajectory;

[0056] S4. Based on the reference trajectory, multi-modal control of the lower limb exoskeleton is performed through a position controller based on adaptive RBF backstepping control.

[0057] This method introduces a modulation factor in the admittance controller to achieve a smooth transition between different control modes, allowing the exoskeleton to flexibly switch between different operating modes. Through the synergy of various steps, precise control of the exoskeleton system is achieved, improving the adaptability of the system and the comfort of the wearer.

[0058] The above steps and related technical features are further described in detail below:

[0059] In this embodiment S1, the phase estimation model uses a pre-trained long short-term memory network to perform real-time analysis on data collected by accelerometers, gyroscopes and other sensors, integrates a feature extraction module for identifying the wearer's specific gait pattern, and combines an online learning mechanism to iteratively optimize model parameters to adapt to individual differences and environmental changes.

[0060] The use of a pre-trained long short-term memory (LSTM) network for real-time phase estimation improves the accuracy of phase estimation and can more accurately synchronize the gait of the exoskeleton and the wearer. The wearer's specific gait pattern is identified through a feature extraction module, and the model parameters are iteratively optimized in combination with an online learning mechanism, which can better adapt to individual differences and environmental changes and improve the personalization level of the control system.

[0061] In this embodiment S2, the CPG model simulates the oscillator network in the central nervous system, which can autonomously generate periodic movement patterns, such as walking, running, etc. In the lower limb exoskeleton control system, the CPG model is used to generate a continuous and coordinated gait trajectory, including:

[0062] A neural network: includes multiple oscillator units, where each oscillator unit includes at least two interconnected oscillators that produce a phase difference or frequency difference to mimic the interaction between muscle groups. The frequency ω of the oscillator determines the speed of the movement pattern, such as the speed of gait. The amplitude r can affect the intensity of the movement pattern, such as the force or amplitude of the gait. The phase difference Ψ between different oscillators determines the coordination relationship between them, such as the gait synchronization between the left leg and the right leg.

[0063] Connection weights: used to adjust the way signals are transmitted between units to simulate the coordinated movements of muscles and joints; changes in these weights can simulate the coordinated movements of muscles and joints, such as the swinging and support of the legs during walking.

[0064] Parameter adjustment mechanism: used to adjust the frequency, amplitude and connection weight of the oscillator to adjust the output gait trajectory. This can be done based on the wearer's real-time feedback (such as gait phase estimation, ground reaction force, etc.), thereby ensuring that the generated gait trajectory matches the wearer's actual gait.

[0065] An interface: used to receive signals from external systems, such as the wearer's movement intentions, environmental changes, or other sensor data, which can be used to adjust the parameters of the CPG model to achieve more precise gait control.

[0066] The CPG model in this step provides a powerful tool for the lower limb exoskeleton control system, which can generate and control personalized gait according to the specific situation of the wearer, while having high flexibility and adaptability.

[0067] In this embodiment S3, the variable parameter admittance controller is:

[0068]

[0069] Among them, α and β are modulation factors, H a represents the adaptive inertia coefficient, B a represents the adaptive damping coefficient, K a represents the adaptive stiffness coefficient, Δx(t) represents the trajectory deviation, Δx(t)' represents the first-order derivative of the trajectory deviation Δx(t), Δx(t)" represents the second-order derivative of the trajectory deviation Δx(t), τ ext represents the human-computer interaction torque, x d represents the desired trajectory of the exoskeleton, x h represents the lower limb motion trajectory, x r represents the reference trajectory.

[0070] The modulation factor includes a first modulation factor α for adjusting the admittance parameter, and a second modulation factor β for adjusting the proportion of the lower limb motion trajectory in the reference trajectory, wherein α≥0, 0≤β≤1.

[0071] Modulation factor adjustments, including smooth transitions between modes:

[0072]

[0073] Among them, the lower limb exoskeleton does not provide assistance in Passive mode, the lower limb exoskeleton provides assistance within a preset range in Assist-as-needed mode, and the lower limb exoskeleton provides maximum assistance in Active mode.

[0074] The following is a further description of Passive mode, Assist-as-needed mode, and Active mode:

[0075] Passive mode is a passive mode, which can be used for wearers with severe injuries who cannot move autonomously. The exoskeleton can improve the wearer's gait performance and avoid muscle atrophy and joint stiffness by assisting walking. Under this premise, the exoskeleton only needs simple position control. In this mode, α=0 and β=0. This means that the exoskeleton does not actively provide any assistance, and the user relies entirely on his own muscle strength to move.

[0076] Assist-as-needed mode is an on-demand assistance mode, mainly for wearers with partial athletic ability. Providing auxiliary support is an effective means to increase participation in sports training. The desired training trajectory can be set according to the wearer's athletic status and needs, and appropriate support and auxiliary force can be provided by changing parameters, thereby increasing the wearer's sports stability and safety. This is especially important for wearers with partial athletic ability, as it can reduce accidental injuries and instability.

[0077] In this mode, α>0 and 0<β<1. This means that the exoskeleton will provide different levels of assistance depending on the user's needs. α may be a parameter that controls the degree of assistance provided by the exoskeleton, while β determines how the exoskeleton responds to the user's input. β values ​​between 0 and 1 mean that the exoskeleton dynamically adjusts its behavior between passive and active.

[0078] By adopting the "assist-as-needed" mode in the multimodal control technology, a larger admittance parameter, that is, a larger α, is selected on a smooth road, so that the exoskeleton is more likely to participate in the wearer's "exercise intention", thereby improving the wearer's active participation in sports training; when climbing a gentle slope or after a period of sports training, a smaller admittance parameter, that is, a smaller α, is selected to increase the leading role of the exoskeleton during sports training, on the one hand to protect the wearer's safety, on the other hand to increase the wearer's single training time and improve training efficiency.

[0079] Active mode is an active mode mainly for wearers with less lower limb limitations. These wearers have strong autonomous movement ability but cannot completely get rid of assistive tools. The exoskeleton plays a protective and supportive role. By increasing the motivation and participation of the wearer, the wearer learns self-management and self-regulation skills in sports training. In this mode, α>0 and β=1. This means that the exoskeleton will actively provide the maximum level of assistance, and the user can complete the action with almost no effort. In this case, α is still the parameter that determines the size of the assistance, and β equals 1, which means that the exoskeleton is always active.

[0080] In this embodiment S4, the position controller based on adaptive RBF backstepping control adopts an adaptive RBF (radial basis function) backstepping control method, which is intended to deal with the nonlinear dynamics of the exoskeleton system and achieve precise control of the exoskeleton movement. This method combines the powerful approximation ability of the RBF network with the good robustness and tracking performance of the backstepping control, and uses an adaptive learning mechanism to optimize the controller parameters. It includes:

[0081] A RBF network is used to approximate the nonlinear dynamic model of an exoskeleton system; the RBF network includes a plurality of radial basis function units, each of which has an adjustable center point and a width parameter; the center point of each radial basis function unit determines the activation area of ​​the unit, that is, within which input range the unit will be significantly activated, and the width parameter affects the shape and activation range of the radial basis function.

[0082] A backstepping controller is used to design a control law based on the output of the RBF network so that the exoskeleton can move along the desired trajectory. Backstepping control is a recursive control design method that gradually simplifies the system into a series of first-order subsystems and designs corresponding control laws to ensure the stability of the system. Based on the output of the RBF network, a control law is designed to enable the exoskeleton to move along the desired trajectory. The purpose is to ensure that the actual position and velocity of the exoskeleton can track the desired position and velocity.

[0083] The parameters and control law of the RBF network are adjusted in combination with an adaptive learning mechanism. The parameters of the RBF network and the control law are updated using an online learning algorithm (such as the gradient descent method or the least squares method). This improves the controller's approximation accuracy to the nonlinear characteristics of the exoskeleton and enhances the system's robustness and tracking performance.

[0084] The technical solution of this embodiment proposes a multimodal control method for a lower limb exoskeleton, which selects the initial gait trajectory by combining the wearer's physical characteristic parameters, and uses a pre-trained long short-term memory network to perform real-time gait phase estimation, and then generates the expected trajectory of the exoskeleton through a CPG model.

[0085] On this basis, a variable parameter admittance controller is used in combination with human-computer interaction torque to adjust the modulation factor, output the reference trajectory, and finally realize the multimodal control of the lower limb exoskeleton through a position controller based on adaptive RBF backstepping control. This solution improves the accuracy and stability of exoskeleton control and enhances the adaptability and flexibility of the system through key technical means such as personalized gait trajectory selection, real-time phase estimation, natural gait simulation, smooth mode switching and precise control.

[0086] Embodiment 2;

[0087] like Figure 3 As shown, this embodiment provides a multimodal control system for a lower limb exoskeleton, comprising:

[0088] Gait trajectory selection and phase estimation module: used to select a gait trajectory from the gait trajectory library as the initial trajectory according to the wearer's physical characteristic parameters, and perform gait phase estimation in combination with the phase estimation model;

[0089] Expected trajectory generation module: used to process the initial trajectory and gait phase using the CPG model to obtain the expected trajectory of the exoskeleton;

[0090] Reference trajectory generation module: used to input the desired trajectory of the exoskeleton and the lower limb motion trajectory in the exoskeleton model into the variable parameter admittance controller, adjust the modulation factor in combination with the human-computer interaction torque, and output the reference trajectory;

[0091] Position controller module: used for performing multi-modal control of the lower limb exoskeleton based on the reference trajectory through a position controller based on adaptive RBF backstepping control.

[0092] Embodiment 3;

[0093] like Figure 4 As shown, this embodiment provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the above-mentioned multimodal control method when executing the computer program.

[0094] Embodiment 4;

[0095] This embodiment provides a computer-readable storage medium, wherein the storage medium stores a computer program, and when the computer program is executed by a processor, the multi-modal control method described above is implemented.

[0096] In the above embodiments provided in the present application, it should be understood that the disclosed methods, systems, devices and media can be implemented in other ways. The above described methods, systems, devices and media embodiments are only schematic. For example, the division of modules or units is only a logical function division, and there may be other division methods in actual implementation. Each functional unit can be integrated into a processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0097] The units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0098] The computer program includes computer program code, which may be in source code form, object code form, executable file or some intermediate form, etc. Computer readable media may include: any entity or device capable of carrying computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory, random access memory, electric carrier signal, telecommunication signal and software distribution medium, etc.

[0099] The above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.

Claims

1. A multimodal control method for a lower limb exoskeleton, characterized in that: The following steps are involved: S1. According to the wearer's body characteristic parameters, a gait trajectory is selected from the gait trajectory library as the initial trajectory, and gait phase estimation is performed in combination with the phase estimation model; S2, using the CPG model to process the initial trajectory and gait phase to obtain the desired trajectory of the exoskeleton; S3, inputting the desired trajectory of the exoskeleton and the lower limb motion trajectory in the exoskeleton model into a variable parameter admittance controller, adjusting the modulation factor in combination with the human-computer interaction torque, and outputting a reference trajectory; S4. Based on the reference trajectory, multi-modal control of the lower limb exoskeleton is performed through a position controller based on adaptive RBF backstepping control.

2. A multimodal control method for a lower limb exoskeleton according to claim 1, characterized in that: In S1, the phase estimation model uses a pre-trained long short-term memory network to perform real-time analysis on data collected by accelerometers, gyroscopes and other sensors, integrates a feature extraction module for identifying the wearer's specific gait pattern, and combines an online learning mechanism to iteratively optimize model parameters to adapt to individual differences and environmental changes.

3. A multimodal control method for a lower limb exoskeleton according to claim 1, characterized in that: In S2, the CPG model includes: A neural network: comprising a plurality of oscillator units, wherein each oscillator unit comprises at least two oscillators connected to each other to generate a phase difference or a frequency difference; An interface: used to receive signals from external systems. Among them, the CPG model is integrated with connection weights to adjust the signal transmission method between units to simulate the coordinated movements between muscles and joints; combined with the parameter adjustment mechanism, the frequency, amplitude and connection weights of the oscillator are adjusted to adjust the output gait trajectory.

4. A multimodal control method for a lower limb exoskeleton according to claim 3, characterized in that: In S3, the variable parameter admittance controller is: Among them, α and β are modulation factors, H a represents the adaptive inertia coefficient, B a represents the adaptive damping coefficient, K a represents the adaptive stiffness coefficient, Δx(t) represents the trajectory deviation, Δx(t)' represents the first-order derivative of the trajectory deviation Δx(t), Δx(t)" represents the second-order derivative of the trajectory deviation Δx(t), τ ext represents the human-computer interaction torque, x d represents the desired trajectory of the exoskeleton, x h represents the lower limb motion trajectory, x r represents the reference trajectory.

5. A multimodal control method for a lower limb exoskeleton according to claim 1, characterized in that: In S3, the modulation factor includes: a first modulation factor α and a second modulation factor β; Among them, the first modulation factor α is used to adjust the admittance parameter, α≥0; the second modulation factor β is used to adjust the proportion of the lower limb movement trajectory in the reference trajectory, 0≤β≤1.

6. A multimodal control method for a lower limb exoskeleton according to claim 1, characterized in that: In S3, the modulation factor is adjusted, including achieving a smooth transition between different modes: Among them, the lower limb exoskeleton does not provide assistance in Passive mode, the lower limb exoskeleton provides assistance within a preset range in Assist-as-needed mode, and the lower limb exoskeleton provides maximum assistance in Active mode.

7. A multimodal control method for a lower limb exoskeleton according to claim 1, characterized in that: In S4, the position controller based on adaptive RBF backstepping control includes: An RBF network is used to approximate the nonlinear dynamic model of the exoskeleton system; the RBF network includes a plurality of radial basis function units, each of which has an adjustable center point and width parameter; a backstepping controller, which designs a control law based on the output of the RBF network so that the exoskeleton can move along a desired trajectory; The parameters and control law of the RBF network are adjusted in combination with the adaptive learning mechanism.

8. A multimodal control system for a lower limb exoskeleton, characterized in that: include: Gait trajectory selection and phase estimation module: used to select a gait trajectory from the gait trajectory library as the initial trajectory according to the wearer's physical characteristic parameters, and perform gait phase estimation in combination with the phase estimation model; Expected trajectory generation module: used to process the initial trajectory and gait phase using the CPG model to obtain the expected trajectory of the exoskeleton; Reference trajectory generation module: used to input the desired trajectory of the exoskeleton and the lower limb motion trajectory in the exoskeleton model into the variable parameter admittance controller, adjust the modulation factor in combination with the human-computer interaction torque, and output the reference trajectory; Position controller module: used for performing multi-modal control of the lower limb exoskeleton based on the reference trajectory through a position controller based on adaptive RBF backstepping control.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the multimodal control method for a lower limb exoskeleton according to any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the multimodal control method for a lower limb exoskeleton as claimed in any one of claims 1 to 7 is implemented.

Citation Information

Cited By

  • Intelligent wearable exoskeleton self-adaptive auxiliary regulation and control method and exoskeleton system

    CN120382467A

  • Hip joint exoskeleton control method and system based on gait mode prediction and gait phase estimation, terminal and storage medium

    CN121018607A

  • Exoskeleton robot control method and system

    CN121315918A