Exoskeleton robot control method and system

Abstract

The invention provides an exoskeleton robot control method and system, and the method comprises the steps: obtaining the motion data of a target and a pre-trained target model, and carrying out the track prediction based on the motion data through the target model, so as to obtain a predicted ankle joint motion track; performing trajectory processing and filtering on the predicted ankle joint motion trajectory to obtain a processed motion trajectory; performing first motion solution of the exoskeleton robot based on the processed motion track to output a first control parameter; performing second motion solution of the exoskeleton robot based on the processed motion track to output a second control parameter; and the first control parameter and the second control parameter are sent to the exoskeleton robot to complete control of the exoskeleton robot, the response performance of the motor is optimized, accurate trajectory tracking is ensured, and the auxiliary effect of the exoskeleton robot is improved.

Description

Technical Field

[0001] This invention belongs to the technical field of exoskeleton robot control, and specifically relates to an exoskeleton robot control method and system. Background Technology

[0002] Exoskeleton robots typically consist of an electrical system and sensors that measure motor torque. They assist patients with limb movement disorders in rehabilitation training and are widely used in the field of medical rehabilitation. As an emerging rehabilitation aid, exoskeleton technology has demonstrated significant therapeutic potential. Unlike traditional physical therapy, which relies on slow recovery through manual training, exoskeletons, by mimicking human gait and movement mechanisms and combining intelligent control, can provide patients with customized movement assistance, significantly improving rehabilitation outcomes. Exoskeletons not only help patients regain walking ability and strengthen muscle strength but can also adjust the level of assistance in real time according to the patient's needs, improving treatment efficiency.

[0003] Existing ankle exoskeleton robots have significant limitations, namely, they cannot recognize the movement intention based on the user's movement state, resulting in insufficient real-time response capabilities and poor human-computer interaction experience. Summary of the Invention

[0004] To address the aforementioned technical problems, this invention provides an exoskeleton robot control method and system, which solves the technical problems in the prior art.

[0005] In a first aspect, the present invention provides the following technical solution: an exoskeleton robot control method, comprising: Acquire the target's motion data and a pre-trained target model, and use the target model and the motion data to predict the trajectory of the ankle joint. The predicted ankle joint motion trajectory is processed and filtered to obtain the processed motion trajectory; Based on the processed motion trajectory, the first motion solution of the exoskeleton robot is performed to output the first control parameters; The second motion of the exoskeleton robot is solved based on the processed motion trajectory to output the second control parameters; The first control parameter and the second control parameter are sent to the exoskeleton robot to complete the control of the exoskeleton robot.

[0006] Compared with existing technologies, the beneficial effects of this invention are as follows: This invention first acquires the target's motion data and a pre-trained target model, and then performs trajectory prediction based on the target model and motion data to obtain a predicted ankle joint motion trajectory; then, it processes and filters the predicted ankle joint motion trajectory to obtain a processed motion trajectory; then, it solves the first motion of the exoskeleton robot based on the processed motion trajectory to output first control parameters; then, it solves the second motion of the exoskeleton robot based on the processed motion trajectory to output second control parameters; finally, it sends the first and second control parameters to the exoskeleton robot to complete the control of the exoskeleton robot. This invention can achieve the purpose of recognizing the target's motion intention by predicting the target's ankle joint motion trajectory, and then processes the trajectory to improve its smoothness and accuracy. Finally, it controls the target by outputting the first and second control parameters, thereby optimizing the motor's response performance, ensuring accurate trajectory tracking, and improving the assistive effect of the exoskeleton robot.

[0007] Preferably, the motion data includes the angular velocity, speed, and angle of the wearable exoskeleton robot in different motion modes.

[0008] Preferably, the step of predicting the ankle joint movement trajectory by using the target model and based on the motion data includes: The motion data is converted into vector form to obtain motion vectors. The motion vectors are then input into the encoder in the target model. The encoder maps the motion vectors to a high-dimensional space to obtain query vectors, key vectors, and value vectors. Perform a Fourier transform on the query vector, the key vector, and the value vector to obtain the transformed query vector. Transform the key vector and the transformed value vector ; Sparse self-attention is calculated based on the transformed query vector, the transformed key vector, and the transformed value vector. : ; In the formula, This is the inverse Fourier transform. For zero-filling operation, To transform the query vector Transform the key vector Dimensions To transform the query vector Sparse matrices of the same dimension; The motion vector is encoded and features are extracted by the encoder to obtain extracted features. The extracted features are then concatenated to obtain concatenated features. The concatenated features and the sparse self-attention are input into the decoder of the target model and a multi-head self-attention mechanism is used for feature decoding to obtain the decoded features; The decoded features are input into a fully connected layer in the target model and mapped to the prediction space through the fully connected layer to obtain the predicted ankle joint movement trajectory.

[0009] Preferably, the step of processing and filtering the predicted ankle joint motion trajectory to obtain the processed motion trajectory includes: Extract the predicted ankle joint motion angle from the predicted ankle joint motion trajectory, and identify the angle error between the predicted ankle joint motion angle and the actual ankle joint motion angle after wearing the exoskeleton robot. Based on the angle error, the predicted ankle joint movement trajectory is corrected using a linear translation mapping method to obtain a first adjusted trajectory; The first adjustment trajectory is decomposed into several sub-trajectories, historical motion trajectories under the same motion pattern are obtained, the similarity between the sub-trajectories and the historical motion trajectories is calculated, and the sub-trajectories with similarity greater than the similarity threshold are spliced ​​and continuously adjusted to obtain the second adjustment trajectory. The second adjustment trajectory is filtered using a preset filter to obtain the processed motion trajectory.

[0010] Preferably, the step of solving the first motion of the exoskeleton robot based on the processed motion trajectory to output the first control parameters includes: Calculate the initial lengths of the two active extension components of the legs in the exoskeleton robot: ; ; In the formula, These are the initial lengths of the first and second active telescopic leg components, respectively. , These are the coordinates of the connection points between the first and second leg active telescopic components and the lower leg connecting components, respectively. , These are the coordinates of the connection points between the first and second active retractable leg components and the foot connecting component, respectively. Calculate the coordinates of the dynamic connection points between the first and second leg active telescopic components and the lower leg connecting components: ; ; In the formula, Center of rotation To handle the ankle joint movement angle in the trajectory of motion, These are the coordinates of the dynamic connection points between the first and second leg active telescopic components and the lower leg connecting components, respectively. Based on the dynamic connection point coordinates of the first and second leg active telescopic components and the lower leg connecting component, and the initial length of the two leg active telescopic components, calculate the length change of the two leg active telescopic components: ; ; In the formula, These represent the length changes of the first and second leg active telescopic components, respectively. The number of rotations of the drive motors on both sides of the exoskeleton robot is calculated based on the change in length of the two active extension components of the legs to obtain the first control parameter: ; ; In the formula, , These represent the number of revolutions and lead of the drive motor on one side, respectively. , These are the number of rotations and lead of the drive motor on the other side, respectively.

[0011] Preferably, the step of solving the second motion of the exoskeleton robot based on the processed motion trajectory to output the second control parameters includes: Extract the predicted motion parameters from the processed motion trajectory, and construct a coefficient matrix based on the predicted motion parameters. Based on the coefficient matrix, an equilibrium equation is constructed: ; In the formula, The actual internal stress of the active extension component of the leg; Define the constraints: ; ; In the formula, For the external force matrix, For relaxation amount, The maximum internal stress that the active extension component of the leg can withstand; The equilibrium equation is modified into a force potential equation by means of the constraints, and the force potential equation is optimized and solved to output the internal stress trajectory of the active extension component of the leg, so as to obtain the second control parameter.

[0012] Secondly, the present invention provides the following technical solution: an exoskeleton robot control system, the system comprising: The prediction submodule is used to acquire the target's motion data and a pre-trained target model, and to perform trajectory prediction based on the target model and the motion data to obtain the predicted ankle joint motion trajectory. The processing submodule is used to perform trajectory processing and filtering on the predicted ankle joint movement trajectory to obtain the processed movement trajectory. The first solution submodule is used to perform the first motion solution of the exoskeleton robot based on the processed motion trajectory, so as to output the first control parameters; The second solution submodule is used to solve the second motion of the exoskeleton robot based on the processed motion trajectory, so as to output the second control parameters; The control submodule is used to send the first control parameter and the second control parameter to the exoskeleton robot to complete the control of the exoskeleton robot.

[0013] Preferably, the motion data includes the angular velocity, speed, and angle of the wearable exoskeleton robot in different motion modes.

[0014] Thirdly, the present invention provides the following technical solution: a computer, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the exoskeleton robot control method described above.

[0015] Fourthly, the present invention provides the following technical solution: a storage medium storing a computer program, wherein the computer program, when executed by a processor, implements the exoskeleton robot control method as described above. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 This is a flowchart of the exoskeleton robot control method provided in Embodiment 1 of the present invention; Figure 2 This is a structural block diagram of the exoskeleton robot control system provided in Embodiment 2 of the present invention; Figure 3 This is a schematic diagram of the hardware structure of a computer provided for another embodiment of the present invention.

[0018] The embodiments of the present invention will be further described below with reference to the accompanying drawings. Detailed Implementation

[0019] Embodiments of the present invention are described in detail below, examples of which are illustrated in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain embodiments of the present invention, and should not be construed as limiting the present invention.

[0020] Example 1 In Embodiment 1 of the present invention, as Figure 1 As shown, an exoskeleton robot control method includes: S1. Obtain the target's motion data and a pre-trained target model, and perform trajectory prediction using the target model and based on the motion data to obtain the predicted ankle joint motion trajectory. Specifically, the motion data mentioned here includes the angular velocity, speed, and angle of the wearable exoskeleton robot in different motion modes. The angular velocity, speed, and angle here specifically refer to the angular velocity, speed, and angle of the ankle joint. At the same time, the target model here is based on the Ankleformer model. The original Ankleformer model is improved by setting a decoder, encoder, and multi-head self-attention mechanism to obtain the target model.

[0021] Step S1 includes: S11. Convert the motion data into vector form to obtain motion vectors, input the motion vectors into the encoder in the target model, and map the motion vectors to a high-dimensional space through the encoder to obtain query vectors, key vectors and value vectors. Specifically, the encoder here is a structure composed of several stacked modules. Each module includes a self-attention layer, a feedforward network, residual connections, and layer normalization operations. The purpose of the encoder is to fully capture the personalized gait features exhibited by different subjects in various movement modes. Specifically, the original multidimensional time series signal is first processed through an embedding layer to map it from the original input space to a high-dimensional feature space, so that the subsequent model can more effectively learn the complex temporal structure and nonlinear relationship. Then, through a progressive feature extraction process, the dynamic change features at different time scales are gradually captured, thereby achieving comprehensive modeling of local and global gait patterns.

[0022] S12. Perform a Fourier transform on the query vector, the key vector, and the value vector to obtain the transformed query vector. Transform the key vector and the transformed value vector .

[0023] S13. Calculate sparse self-attention based on the transformed query vector, the transformed key vector, and the transformed value vector. : ; In the formula, For inverse Fourier transform, For zero-filling operation, To transform the query vector Transform the key vector Dimensions To transform the query vector Sparse matrices of the same dimension.

[0024] S14. The motion vector is encoded and features are extracted by the encoder to obtain extracted features, and the extracted features are concatenated to obtain concatenated features; Specifically, during the encoding and feature extraction process, attention weights can be allocated using the sparse self-attention calculated above. This approach can significantly improve the model's ability to model long-term dependencies and enhance its perception of temporal global structures, thereby better adapting to gait prediction tasks across subjects and movement modes. The spliced ​​features obtained here are high-level gait features.

[0025] S15. Input the spliced ​​features and the sparse self-attention into the decoder of the target model and use a multi-head self-attention mechanism to perform feature decoding to obtain the decoded features; Specifically, the decoder here has a structure similar to the encoder, which is a multi-layer stacked structure. It uses a multi-layer stacking method to gradually generate the prediction results at each time step. However, the decoder uses a multi-head self-attention mechanism, which can utilize both frequency domain information and time domain information when generating the target output. Sparse self-attention enables the decoder to capture global trends, periodic changes and local details in the input sequence, while the multi-head self-attention mechanism further enhances the ability to model different features. The stacked frequency domain and time domain attention mechanism enables the decoder to efficiently extract key features from the contextual information of the generated partial sequence and the input sequence, thereby generating a more accurate and consistent output sequence.

[0026] S16. Input the decoded features into the fully connected layer in the target model and map the decoded features to the prediction space through the fully connected layer to obtain the predicted ankle joint movement trajectory; Specifically, the decoder output is mapped to the final prediction space through a fully connected layer to generate a continuous sequence of ankle joint angle trajectories, which is the predicted ankle joint movement trajectory.

[0027] S2. Perform trajectory processing and filtering on the predicted ankle joint movement trajectory to obtain the processed movement trajectory; Step S2 includes: S21. Extract the predicted ankle joint motion angle from the predicted ankle joint motion trajectory, and identify the angle error between the predicted ankle joint motion angle and the actual ankle joint motion angle after wearing the exoskeleton robot. Specifically, in practice, due to the certain deviation between the predicted ankle joint angle and the actual ankle joint angle after wearing the exoskeleton, the predicted ankle joint angle ranges from -5° to 25°, while the range of ankle joint movement angle after wearing the exoskeleton is usually from -20° to 10°. Therefore, a linear translation mapping method is used to adjust the predicted trajectory as a whole so that its range is consistent with the actual ankle joint movement angle.

[0028] S22. Based on the angle error, the predicted ankle joint movement trajectory is corrected using a linear translation mapping method to obtain a first adjusted trajectory.

[0029] S23. Decompose the first adjustment trajectory into several sub-trajectories, obtain historical motion trajectories under the same motion mode, calculate the similarity between the sub-trajectories and the historical motion trajectories, and splice and continuously adjust the sub-trajectories with similarity greater than the similarity threshold to obtain the second adjustment trajectory. Specifically, the similarity here is cosine similarity. By calculating the similarity with historical motion trajectories, the selected second adjustment trajectory is made to better match the trajectory synchronization in real-world situations. This improves the coordination and consistency between the exoskeleton and the human gait, effectively avoiding gait incoordination or unnaturalness caused by delays. It further improves the precision and smoothness of exoskeleton control. At the same time, the continuous trajectory adjustment here is to avoid trajectory interruption and ensure trajectory continuity.

[0030] S24. The second adjustment trajectory is filtered using a preset filter to obtain the processed motion trajectory; Specifically, the preset filter here is a 6th-order Butterworth low-pass filter. This filter can effectively remove high-frequency noise components while retaining the main dynamic characteristics of the motion trajectory, ensuring the stability and accuracy of the exoskeleton control signal.

[0031] S3. Based on the processed motion trajectory, perform the first motion solution for the exoskeleton robot to output the first control parameters; Step S3 includes: S31. Calculate the initial length of the two active extension components of the legs in the exoskeleton robot: ; ; In the formula, These are the initial lengths of the first and second active telescopic leg components, respectively. , These are the coordinates of the connection points between the first and second leg active telescopic components and the lower leg connecting components, respectively. , These are the coordinates of the connection points between the first and second active retractable leg components and the foot connecting component, respectively. Specifically, for ankle exoskeleton robots, they can be abstracted as rod models. The movement assistance and rehabilitation process is achieved through the extension and retraction of the active extension and retraction components of the legs. Here, the ankle exoskeleton robot is specifically a biomimetic tensioned whole ankle exoskeleton, and the active extension and retraction components of the legs are active tensioning ropes.

[0032] S32. Calculate the coordinates of the dynamic connection points between the first and second leg active telescopic components and the lower leg connecting components: ; ; In the formula, Center of rotation To handle the ankle joint movement angle in the trajectory of motion, These are the coordinates of the dynamic connection points between the first and second leg active telescopic components and the lower leg connecting components, respectively.

[0033] S33. Calculate the length change of the two active leg extension components based on the dynamic connection point coordinates of the first and second active leg extension components and the lower leg connecting component, as well as the initial length of the two active leg extension components: ; ; In the formula, These represent the length changes of the first and second active telescopic components, respectively.

[0034] S34. Calculate the number of rotations of the drive motors on both sides of the exoskeleton robot based on the length change of the two active extension components of the legs, in order to obtain the first control parameters: ; ; In the formula, , These represent the number of revolutions and lead of the drive motor on one side, respectively. , These are the number of revolutions and the lead of the drive motor on the other side, respectively. Specifically, after obtaining the number of rotations, the number of rotations can be sent to the motor drive board of the exoskeleton robot.

[0035] S4. Based on the processed motion trajectory, solve the second motion of the exoskeleton robot to output the second control parameters; Step S4 includes: S41. Extract the predicted motion parameters from the processed motion trajectory, and construct a coefficient matrix based on the predicted motion parameters. Based on the coefficient matrix, an equilibrium equation is constructed: ; In the formula, The actual internal stress of the active extension component of the leg; Specifically, the coefficient matrix here is the rope force density matrix, which can be constructed by predicting motion parameters. In actual movement, the ankle exoskeleton is subjected to non-zero external forces from human movement and the environment. In order to accurately reflect this situation, data measured by a triaxial force sensor is used as the external force input in this step to obtain the subsequent external force matrix.

[0036] S42. Determine the constraints: ; ; In the formula, For the external force matrix, For relaxation amount, The maximum internal stress that the active extension component of the leg can withstand; In the actual motion process, there is a certain deviation between the quasi-static state and the theoretical static model. In order to prevent the solution to the optimization problem from being unsolvable or singular due to rigid equality constraints, relaxation variables are introduced into the constraint conditions, so that the constraint conditions have better adaptability. At the same time, in order to avoid the breakage of the active extension component of the leg, the maximum internal stress that the active extension component of the leg can withstand is set as a constraint.

[0037] S43. The equilibrium equation is modified into a force potential equation by means of the constraint conditions, and the force potential equation is optimized and solved to output the internal stress trajectory of the active extension component of the leg to obtain the second control parameter. Specifically, after considering the comprehensive constraints, the equilibrium equation is modified into a force potential equation. After solving the force potential equation, the target stress distribution is obtained, and the target internal force distribution is converted into the target force trajectory of the two driving ropes.

[0038] S5. Send the first control parameter and the second control parameter to the exoskeleton robot to complete the control of the exoskeleton robot; Specifically, after obtaining the first control parameter and the second control parameter, they are input into their respective PID controllers to generate corresponding signals. Then, the motors are controlled by the signals to enable the exoskeleton robot to perform corresponding motion control, thereby assisting the target's ankle joint in processing the motion trajectory.

[0039] The exoskeleton robot control method provided in Embodiment 1 of this invention first acquires the target's motion data and a pre-trained target model. Then, it predicts the ankle joint motion trajectory using the target model and the motion data. Next, it processes and filters the predicted ankle joint motion trajectory to obtain a processed motion trajectory. Finally, it solves the first motion of the exoskeleton robot based on the processed motion trajectory to output first control parameters. Then, it solves the second motion of the exoskeleton robot based on the processed motion trajectory to output second control parameters. Finally, it sends the first and second control parameters to the exoskeleton robot to complete the control of the exoskeleton robot. This invention can predict the target's ankle joint motion trajectory to achieve the purpose of recognizing the target's motion intention. After processing the trajectory, it improves the smoothness and accuracy of the trajectory. Then, it controls the robot by outputting the first and second control parameters, thereby optimizing the motor's response performance, ensuring accurate trajectory tracking, and improving the assistive effect of the exoskeleton robot.

[0040] Example 2 like Figure 2 As shown, in Embodiment 2 of the present invention, an exoskeleton robot control system is provided, the system comprising: Prediction submodule 1 is used to acquire the target's motion data and a pre-trained target model, and to perform trajectory prediction based on the target model and the motion data to obtain the predicted ankle joint motion trajectory. Processing submodule 2 is used to process and filter the predicted ankle joint movement trajectory to obtain the processed movement trajectory; The first solution submodule 3 is used to perform the first motion solution of the exoskeleton robot based on the processed motion trajectory, so as to output the first control parameters; The second solution submodule 4 is used to solve the second motion of the exoskeleton robot based on the processed motion trajectory, so as to output the second control parameters; The control submodule 5 is used to send the first control parameter and the second control parameter to the exoskeleton robot to complete the control of the exoskeleton robot.

[0041] The motion data includes the angular velocity, speed, and angle of the wearable exoskeleton robot in different motion modes; The prediction submodule 1 includes: The conversion submodule is used to convert the motion data into vector form to obtain motion vectors, input the motion vectors into the encoder in the target model, and map the motion vectors to a high-dimensional space through the encoder to obtain query vectors, key vectors and value vectors. The transformation submodule is used to perform a Fourier transform on the query vector, the key vector, and the value vector to obtain a transformed query vector. Transform the key vector and the transformed value vector ; The attention submodule is used to calculate sparse self-attention based on the transform query vector, the transform key vector, and the transform value vector. : ; In the formula, This is the inverse Fourier transform. For zero-filling operation, To transform the query vector Transform the key vector Dimensions To transform the query vector Sparse matrices of the same dimension; The splicing submodule is used to encode and extract features from the motion vector through the encoder to obtain extracted features, and to splice the extracted features to obtain spliced ​​features. The decoding submodule is used to input the spliced ​​features and the sparse self-attention into the decoder of the target model and perform feature decoding using a multi-head self-attention mechanism to obtain the decoded features; The mapping submodule is used to input the decoded features into the fully connected layer in the target model and map the decoded features to the prediction space through the fully connected layer to obtain the predicted ankle joint movement trajectory.

[0042] The processing submodule 2 includes: An extraction submodule is used to extract the predicted ankle joint motion angle from the predicted ankle joint motion trajectory and identify the angle error between the predicted ankle joint motion angle and the actual ankle joint motion angle after wearing the exoskeleton robot. The correction submodule is used to correct the predicted ankle joint movement trajectory based on the angle error and by using a linear translation mapping method to obtain a first adjusted trajectory. The adjustment submodule is used to decompose the first adjustment trajectory into several sub-trajectories, obtain historical motion trajectories under the same motion mode, calculate the similarity between the sub-trajectories and the historical motion trajectories, and splice and continuously adjust the sub-trajectories with similarity greater than the similarity threshold to obtain the second adjustment trajectory. The filtering submodule is used to filter the second adjustment trajectory using a preset filter to obtain the processed motion trajectory.

[0043] The first solution submodule 3 includes: The first calculation submodule is used to calculate the initial length of the two active extension components of the legs in the exoskeleton robot: ; ; In the formula, These are the initial lengths of the first and second active telescopic leg components, respectively. , These are the coordinates of the connection points between the first and second leg active telescopic components and the lower leg connecting components, respectively. , These are the coordinates of the connection points between the first and second active retractable leg components and the foot connecting component, respectively. The second calculation submodule is used to calculate the coordinates of the dynamic connection points between the first and second leg active telescopic components and the lower leg connecting components: ; ; In the formula, Center of rotation To handle the ankle joint movement angle in the trajectory of motion, These are the coordinates of the dynamic connection points between the first and second leg active telescopic components and the lower leg connecting components, respectively. The third calculation submodule is used to calculate the length change of the two active leg telescopic components based on the dynamic connection point coordinates of the first and second active leg telescopic components and the lower leg connecting component, as well as the initial length of the two active leg telescopic components: ; ; In the formula, These represent the length changes of the first and second leg active telescopic components, respectively. The fourth calculation submodule is used to calculate the number of rotations of the drive motors on both sides of the exoskeleton robot based on the change in length of the two active extension components of the legs, in order to obtain the first control parameter: ; ; In the formula, , These represent the number of revolutions and lead of the drive motor on one side, respectively. , These are the number of rotations and lead of the drive motor on the other side, respectively.

[0044] The second solution submodule 4 includes: The fifth calculation submodule is used to extract the predicted motion parameters from the processed motion trajectory and construct a coefficient matrix based on the predicted motion parameters. Based on the coefficient matrix, an equilibrium equation is constructed: ; In the formula, The actual internal stress of the active extension component of the leg; The sixth calculation submodule is used to determine the constraints: ; ; In the formula, For the external force matrix, For relaxation amount, The maximum internal stress that the active extension component of the leg can withstand; The seventh calculation submodule is used to modify the equilibrium equation into a force potential equation through the constraints, optimize the solution of the force potential equation, and output the internal stress trajectory of the active extension component of the leg to obtain the second control parameter.

[0045] In other embodiments of the present invention, the present invention provides the following technical solution: a computer, including a memory 102, a processor 101, and a computer program stored in the memory 102 and executable on the processor 101, wherein the processor 101 executes the computer program to implement the exoskeleton robot control method as described above.

[0046] Specifically, the processor 101 may include a central processing unit (CPU), or an application specific integrated circuit (ASIC), or one or more integrated circuits that can be configured to implement the embodiments of the present invention.

[0047] The memory 102 may include a large-capacity memory for data or instructions. For example, and not limitingly, the memory 102 may include a hard disk drive (HDD), a floppy disk drive, a solid-state drive (SSD), flash memory, an optical disk drive, a magneto-optical disk drive, magnetic tape, or a Universal Serial Bus (USB) drive, or a combination of two or more of these. Where appropriate, the memory 102 may include removable or non-removable (or fixed) media. Where appropriate, the memory 102 may be internal or external to a data processing device. In a particular embodiment, the memory 102 is non-volatile memory. In a particular embodiment, the memory 102 includes read-only memory (ROM) and random access memory (RAM). Where appropriate, the ROM may be a mask-programmed ROM, a programmable read-only memory (PROM), an erasable read-only memory (EPROM), an electrically erasable read-only memory (EEPROM), an electrically alterable read-only memory (EAROM), or flash memory, or a combination of two or more of these. Where appropriate, the RAM can be Static Random-Access Memory (SRAM) or Dynamic Random-Access Memory (DRAM). DRAM can be Fast Page Mode Dynamic Random Access Memory (FPMDRAM), Extended Data Out Dynamic Random Access Memory (EDODRAM), Synchronous Dynamic Random-Access Memory (SDRAM), etc.

[0048] The memory 102 can be used to store or cache various data files that need to be processed and / or used for communication, as well as possible computer program instructions executed by the processor 101.

[0049] The processor 101 implements the above-described exoskeleton robot control method by reading and executing computer program instructions stored in the memory 102.

[0050] In some embodiments, the computer may further include a communication interface 103 and a bus 100. For example, Figure 3 As shown, the processor 101, memory 102, and communication interface 103 are connected through bus 100 and complete communication with each other.

[0051] The communication interface 103 is used to enable communication between the various modules, devices, units, and / or equipment in the embodiments of the present invention. The communication interface 103 can also enable data communication with other components such as external devices, image / data acquisition devices, databases, external storage, and image / data processing workstations.

[0052] Bus 100 includes hardware, software, or both, that couples components of a computer device together. Bus 100 includes, but is not limited to, at least one of the following: data bus, address bus, control bus, expansion bus, and local bus. For example, and not as a limitation, bus 100 may include an Accelerated Graphics Port (AGP) or other graphics bus, an Extended Industry Standard Architecture (EISA) bus, a Front Side Bus (FSB), a Hyper Transport (HT) interconnect, an Industry Standard Architecture (ISA) bus, an InfiniBand interconnect, a Low Pin Count (LPC) bus, a memory bus, a Micro Channel Architecture (MCA) bus, a Peripheral Component Interconnect (PCI) bus, a PCI-Express (PCI-X) bus, a Serial Advanced Technology Attachment (SATA) bus, a Video Electronics Standards Association Local Bus (VLB) bus, or other suitable buses, or a combination of two or more of these. Where appropriate, bus 100 may include one or more buses. Although specific buses are described and illustrated in the embodiments of the present invention, the present invention is contemplated by any suitable bus or interconnect.

[0053] The computer can execute the exoskeleton robot control method of the present invention based on the acquired exoskeleton robot control system, thereby realizing exoskeleton robot control.

[0054] In some further embodiments of the present invention, in conjunction with the above-described exoskeleton robot control method, the present invention provides the following technical solution: a storage medium storing a computer program, wherein the computer program, when executed by a processor, implements the above-described exoskeleton robot control method.

[0055] Those skilled in the art will understand that the logic and / or steps represented in the flowcharts or otherwise described herein, for example, can be considered as a ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-included system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can mean any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device.

[0056] More specific examples of readable media (a non-exhaustive list) include: electrical connections (electronic devices) with one or more wires, portable computer disk drives (magnetic devices), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Furthermore, computer-readable media can even be paper or other suitable media on which the program can be printed, since the program can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in computer memory.

[0057] It should be understood that various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0058] The technical features of the above embodiments can be combined in any way. For the sake of brevity, 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.

[0059] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these all fall within the protection scope of the present invention. Therefore, the protection scope of this invention patent should be determined by the appended claims.

Claims

1. A control method for an exoskeleton robot, characterized in that, include: Acquire the target's motion data and a pre-trained target model, and use the target model and the motion data to predict the trajectory of the ankle joint. The predicted ankle joint motion trajectory is processed and filtered to obtain the processed motion trajectory; Based on the processed motion trajectory, the first motion solution of the exoskeleton robot is performed to output the first control parameters; The second motion of the exoskeleton robot is solved based on the processed motion trajectory to output the second control parameters; The first control parameter and the second control parameter are sent to the exoskeleton robot to complete the control of the exoskeleton robot.

2. The exoskeleton robot control method according to claim 1, characterized in that, The motion data includes the angular velocity, speed, and angle of the wearable exoskeleton robot in different motion modes.

3. The exoskeleton robot control method according to claim 1, characterized in that, The step of predicting the ankle joint trajectory by using the target model and based on the motion data includes: The motion data is converted into vector form to obtain motion vectors. The motion vectors are then input into the encoder in the target model. The encoder maps the motion vectors to a high-dimensional space to obtain query vectors, key vectors, and value vectors. Perform a Fourier transform on the query vector, the key vector, and the value vector to obtain the transformed query vector. Transform the key vector and the transformed value vector ; Sparse self-attention is calculated based on the transformed query vector, the transformed key vector, and the transformed value vector. : ； In the formula, For inverse Fourier transform, For zero-filling operation, To transform the query vector Transform the key vector Dimensions To transform the query vector Sparse matrices of the same dimension; The motion vector is encoded and features are extracted by the encoder to obtain extracted features. The extracted features are then concatenated to obtain concatenated features. The concatenated features and the sparse self-attention are input into the decoder of the target model and a multi-head self-attention mechanism is used for feature decoding to obtain the decoded features; The decoded features are input into a fully connected layer in the target model and mapped to the prediction space through the fully connected layer to obtain the predicted ankle joint movement trajectory.

4. The exoskeleton robot control method according to claim 1, characterized in that, The step of processing and filtering the predicted ankle joint movement trajectory to obtain the processed movement trajectory includes: Extract the predicted ankle joint motion angle from the predicted ankle joint motion trajectory, and identify the angle error between the predicted ankle joint motion angle and the actual ankle joint motion angle after wearing the exoskeleton robot. Based on the angle error, the predicted ankle joint movement trajectory is corrected using a linear translation mapping method to obtain a first adjusted trajectory; The first adjustment trajectory is decomposed into several sub-trajectories, historical motion trajectories under the same motion pattern are obtained, the similarity between the sub-trajectories and the historical motion trajectories is calculated, and the sub-trajectories with similarity greater than the similarity threshold are spliced ​​and continuously adjusted to obtain the second adjustment trajectory. The second adjustment trajectory is filtered using a preset filter to obtain the processed motion trajectory.

5. The exoskeleton robot control method according to claim 1, characterized in that, The step of solving the first motion of the exoskeleton robot based on the processed motion trajectory to output the first control parameters includes: Calculate the initial lengths of the two active extension components of the legs in the exoskeleton robot: ； ； In the formula, These are the initial lengths of the first and second active telescopic leg components, respectively. , These are the coordinates of the connection points between the first and second leg active telescopic components and the lower leg connecting components, respectively. , These are the coordinates of the connection points between the first and second active retractable leg components and the foot connecting component, respectively. Calculate the coordinates of the dynamic connection points between the first and second leg active telescopic components and the lower leg connecting components: ； ； In the formula, Center of rotation To handle the ankle joint movement angle in the trajectory of motion, These are the coordinates of the dynamic connection points between the first and second leg active telescopic components and the lower leg connecting components, respectively. Based on the dynamic connection point coordinates of the first and second leg active telescopic components and the lower leg connecting component, and the initial length of the two leg active telescopic components, calculate the length change of the two leg active telescopic components: ； ； In the formula, These represent the length changes of the first and second leg active telescopic components, respectively. The number of rotations of the drive motors on both sides of the exoskeleton robot is calculated based on the change in length of the two active extension components of the legs to obtain the first control parameter: ； ； In the formula, , These represent the number of revolutions and lead of the drive motor on one side, respectively. , These are the number of rotations and lead of the drive motor on the other side, respectively.

6. The exoskeleton robot control method according to claim 1, characterized in that, The step of solving the second motion of the exoskeleton robot based on the processed motion trajectory to output the second control parameters includes: Extract the predicted motion parameters from the processed motion trajectory, and construct a coefficient matrix based on the predicted motion parameters. Based on the coefficient matrix, an equilibrium equation is constructed: ； In the formula, The actual internal stress of the active extension component of the leg; Define the constraints: ； ； In the formula, For the external force matrix, For relaxation amount, The maximum internal stress that the active extension component of the leg can withstand; The equilibrium equation is modified into a force potential equation by means of the constraints, and the force potential equation is optimized and solved to output the internal stress trajectory of the active extension component of the leg, so as to obtain the second control parameter.

7. A control system for an exoskeleton robot, characterized in that, The system includes: The prediction submodule is used to acquire the target's motion data and a pre-trained target model, and to perform trajectory prediction based on the target model and the motion data to obtain the predicted ankle joint motion trajectory. The processing submodule is used to perform trajectory processing and filtering on the predicted ankle joint movement trajectory to obtain the processed movement trajectory. The first solution submodule is used to perform the first motion solution of the exoskeleton robot based on the processed motion trajectory, so as to output the first control parameters; The second solution submodule is used to solve the second motion of the exoskeleton robot based on the processed motion trajectory, so as to output the second control parameters; The control submodule is used to send the first control parameter and the second control parameter to the exoskeleton robot to complete the control of the exoskeleton robot.

8. The exoskeleton robot control system according to claim 7, characterized in that, The motion data includes the angular velocity, speed, and angle of the wearable exoskeleton robot in different motion modes.

9. A computer 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, it implements the exoskeleton robot control method as described in any one of claims 1 to 6.

10. A storage medium, characterized in that, The storage medium stores a computer program, which, when executed by a processor, implements the exoskeleton robot control method as described in any one of claims 1 to 6.

Images