A method and apparatus for limb training based on a reference trajectory and impedance

Abstract

The application provides a limb training method and device based on a reference trajectory and impedance, and the method comprises the following steps: acquiring static body parameters of a user, dynamic body parameters of the user when training, a motion trajectory of the user when training, and a torque generated between the user and an exoskeleton robot; determining first impedance parameters according to the static body parameters, the dynamic body parameters, the motion trajectory and the torque; determining a first reference trajectory according to the first impedance parameters, the motion trajectory and the torque; and controlling the exoskeleton robot to assist the user in limb training according to the first impedance parameters and the first reference trajectory. The technical problem that the prior art cannot effectively utilize training parameters generated by the user when training to evaluate the training effect and adjust the training mode is solved, and the technical effect that the user is given impedance and a trajectory more suitable for the current body state for limb training is achieved.

Description

Technical Field

[0001] This application relates to the field of limb training, and in particular to a limb training method and apparatus based on reference trajectory and impedance. Background Technology

[0002] With the development of science and technology, exoskeleton robots are often used to assist users in limb training. However, existing limb training programs fail to provide targeted training based on the individual characteristics of users, and cannot effectively utilize the training parameters generated during user training to evaluate the training effect and adjust the training mode, resulting in low training efficiency and mediocre training effect. Summary of the Invention

[0003] In view of the above problems, this application is made to provide a limb training method and apparatus based on reference trajectory and impedance that overcomes or at least partially solves the above problems.

[0004] This application discloses a limb training method based on a reference trajectory and impedance. The method is used for limb training for non-therapeutic purposes and involves an exoskeleton robot connected to a user's limb. The method includes the following steps: acquiring the user's static body parameters, the user's dynamic body parameters during training, the user's motion trajectory during training, and the torque generated between the user and the exoskeleton robot; determining a first impedance parameter based on the static body parameters, the dynamic body parameters, the motion trajectory, and the torque; determining a first reference trajectory based on the first impedance parameter, the motion trajectory, and the torque; and controlling the exoskeleton robot to assist the user in limb training based on the first impedance parameter and the first reference trajectory.

[0005] Preferably, the step of determining the first impedance parameter based on the static body parameters, the dynamic body parameters, the motion trajectory, and the torque includes: generating a median trajectory based on the static body parameters; generating a first training evaluation parameter for the user based on the dynamic body parameters; and determining the first impedance parameter based on the median trajectory, the motion trajectory, the torque, and the first training evaluation parameter.

[0006] Preferably, the step of generating the intermediate trajectory based on the static body parameters includes: generating the intermediate trajectory through Gaussian process regression based on the static body parameters.

[0007] Preferably, the step of generating the user's first training evaluation parameters based on the dynamic body parameters includes: generating the user's first training evaluation parameters based on the dynamic body parameters using a convolutional neural network and a recurrent neural network.

[0008] Preferably, the step of determining the first impedance parameter based on the dielectric trajectory, the motion trajectory, the torque, and the first training evaluation parameter includes: determining the dielectric impedance parameter based on the dielectric trajectory, the motion trajectory, and the torque; and determining the first impedance parameter based on the dielectric impedance parameter and the first training evaluation parameter.

[0009] Preferably, the step of determining the first impedance parameter based on the dielectric impedance parameter and the first training evaluation parameter includes: determining the weight coefficient corresponding to the dielectric impedance parameter based on the first training evaluation parameter; and determining the first impedance parameter based on the weight coefficient and the dielectric impedance parameter.

[0010] Preferably, after the step of controlling the exoskeleton robot to assist the user in limb training based on the first impedance parameter and the first reference trajectory, the method further includes: determining the user's motion characteristics based on the motion trajectory and the torque; generating the user's second training evaluation parameters based on the static body parameters and the motion characteristics; determining a second impedance parameter based on the first impedance parameter and the second training evaluation parameters; determining a second reference trajectory based on the second impedance parameter, the motion trajectory, and the torque; and controlling the exoskeleton robot to assist the user in limb training based on the second impedance parameter and the second reference trajectory.

[0011] This application discloses a limb training device based on a reference trajectory and impedance. The device is used for limb training for non-therapeutic purposes. The device relates to an exoskeleton robot connected to a user's limbs, comprising: an input module for acquiring the user's static body parameters, the user's dynamic body parameters during training, the user's motion trajectory during training, and the torque generated between the user and the exoskeleton robot; a first impedance parameter determination module for determining a first impedance parameter based on the static body parameters, the dynamic body parameters, the motion trajectory, and the torque; a first reference trajectory determination module for determining a first reference trajectory based on the first impedance parameter, the motion trajectory, and the torque; and a limb training module for controlling the exoskeleton robot to assist the user in limb training based on the first impedance parameter and the first reference trajectory.

[0012] This application discloses an apparatus including a processor, a memory, and a computer program stored in the memory and capable of running on the processor, wherein the computer program, when executed by the processor, implements the method described in any embodiment of this application.

[0013] This application discloses a computer-readable storage medium storing a computer program that, when executed by a processor, implements the method described in any embodiment of this application.

[0014] This application has the following advantages:

[0015] In the embodiments of this application, compared to the prior art's inability to effectively utilize training parameters generated during user training to evaluate training effectiveness and adjust training modes, this application provides a limb training method based on reference trajectory and impedance. This method is used for non-therapeutic limb training and involves an exoskeleton robot connected to a user's limbs. The method includes the following steps: acquiring the user's static body parameters, the user's dynamic body parameters during training, the user's motion trajectory during training, and the torque generated between the user and the exoskeleton robot; determining a first impedance parameter based on the static body parameters, the dynamic body parameters, the motion trajectory, and the torque; determining a first reference trajectory based on the first impedance parameter, the motion trajectory, and the torque; and controlling the exoskeleton robot to assist the user in limb training based on the first impedance parameter and the first reference trajectory. By determining the first impedance parameter based on the static body parameters, the dynamic body parameters, the motion trajectory, and the torque; and determining the first reference trajectory based on the first impedance parameter, the motion trajectory, and the torque, this method solves the technical problem of the prior art's inability to effectively utilize training parameters generated during user training to evaluate training effectiveness and adjust training modes. It achieves the technical effect of providing the user with impedance and trajectory more suitable for their current physical state for limb training. Attached Figure Description

[0016] To more clearly illustrate the technical solution of this application, the drawings used in the description of this application will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. 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 illustrating the steps of a limb training method based on reference trajectory and impedance according to an embodiment of this application;

[0018] Figure 2 This is a schematic diagram of a Gaussian process regression framework for a limb training method based on reference trajectory and impedance provided in an embodiment of this application;

[0019] Figure 3 This is a schematic diagram of a convolutional neural network and a recurrent neural network framework for a limb training method based on reference trajectory and impedance provided in an embodiment of this application;

[0020] Figure 4 This is a schematic diagram of the XGBoost regression framework for a limb training method based on reference trajectory and impedance provided in an embodiment of this application;

[0021] Figure 5 This is a structural framework diagram of a limb training device based on reference trajectory and impedance provided in one embodiment of this application;

[0022] Figure 6 This is a schematic diagram of the structure of a computer device provided in an embodiment of this application.

[0023] The attached figures are labeled as follows:

[0024] 510. Input module; 520. First impedance parameter determination module; 530. First reference trajectory determination module; 540. Limb training module;

[0025] 12. Computer equipment; 14. Peripheral devices; 16. Processing unit; 18. Bus; 20. Network adapter; 22. Input / output interface; 24. Display; 28. Memory; 30. Random access memory; 32. Cache memory; 34. Storage system; 40. Program / utility; 42. Program module. Detailed Implementation

[0026] To make the objectives, features, and advantages of this application more apparent and understandable, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments. Obviously, the described embodiments are only some, not all, of the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0027] The inventors discovered through analysis of existing technologies that current limb training programs often neglect the user's dynamic body parameters. In addition, existing limb training methods can only provide a single training reference trajectory or impedance parameter, which makes it impossible to provide impedance and trajectory more suitable for the user's current physical state for limb training, thus failing to meet the user's training needs.

[0028] Reference Figure 1 This application illustrates a limb training method based on reference trajectory and impedance, the method being used for non-therapeutic limb training, the method relating to an exoskeleton robot connected to a user's limb, and comprising the following steps:

[0029] S110. Obtain the user's static body parameters, the user's dynamic body parameters during training, the user's movement trajectory during training, and the torque generated between the user and the exoskeleton robot.

[0030] S210. Determine the first impedance parameter based on the static body parameters, the dynamic body parameters, the motion trajectory, and the torque;

[0031] S310. Determine a first reference trajectory based on the first impedance parameter, the motion trajectory, and the torque;

[0032] S410. Based on the first impedance parameter and the first reference trajectory, control the exoskeleton robot to assist the user in limb training.

[0033] In the embodiments of this application, compared to the prior art's inability to effectively utilize training parameters generated during user training to evaluate training effectiveness and adjust training modes, this application provides a limb training method based on reference trajectory and impedance. This method is used for non-therapeutic limb training and involves an exoskeleton robot connected to a user's limbs. The method includes the following steps: acquiring the user's static body parameters, the user's dynamic body parameters during training, the user's motion trajectory during training, and the torque generated between the user and the exoskeleton robot; determining a first impedance parameter based on the static body parameters, the dynamic body parameters, the motion trajectory, and the torque; determining a first reference trajectory based on the first impedance parameter, the motion trajectory, and the torque; and controlling the exoskeleton robot to assist the user in limb training based on the first impedance parameter and the first reference trajectory. By determining the first impedance parameter based on the static body parameters, the dynamic body parameters, the motion trajectory, and the torque; and determining the first reference trajectory based on the first impedance parameter, the motion trajectory, and the torque, this method solves the technical problem of the prior art's inability to effectively utilize training parameters generated during user training to evaluate training effectiveness and adjust training modes. It achieves the technical effect of providing the user with impedance and trajectory more suitable for their current physical state for limb training.

[0034] The following will further describe a limb training method based on reference trajectory and impedance in this exemplary embodiment.

[0035] As described in step S110, the user's static body parameters, the user's dynamic body parameters during training, the user's movement trajectory during training, and the torque generated between the user and the exoskeleton robot are obtained.

[0036] It should be noted that the static body parameters can be measured and input into the exoskeleton robot. Sensors mounted on the exoskeleton robot can acquire the user's dynamic body parameters during training, the user's movement trajectory during training, and the torque generated between the user and the exoskeleton robot. This acquisition process can be real-time. The static body parameters may include: the user's height, shoulder width, arm length, weight, gender, age, and heart rate when not training. The dynamic body parameters may include: the user's electromyography (EMG) signals during training, the user's oxygen consumption during training, and the user's carbon dioxide consumption during training.

[0037] Limb training is a typical physical interaction task, and its working principle emphasizes "Assisted As Needed (AAN)," which means helping users improve their motor skills through voluntary interaction. Therefore, a certain amount of free movement space needs to be provided during the interaction process to encourage users to move voluntarily. Thus, when the user's comfort level is good and the human-machine fit is high, the exoskeleton robot should reduce its impedance parameters to provide the user with more voluntary movement space. Therefore, dynamic body parameters can reflect the user's training comfort to a certain extent and can be used as parameters that affect the final impedance and reference trajectory.

[0038] As described in step S210, the first impedance parameter is determined based on the static body parameters, the dynamic body parameters, the motion trajectory, and the torque.

[0039] As described in step S310, a first reference trajectory is determined based on the first impedance parameter, the motion trajectory, and the torque.

[0040] It should be noted that the process of determining the first impedance parameter and the first reference trajectory can involve an impedance model, the formula of which is expressed as:

[0041]

[0042] in, This represents the angle vector corresponding to the user's movement trajectory. This represents the angle vector corresponding to the reference trajectory. These represent the three variables of the impedance parameter, where n represents the number of degrees of freedom of the exoskeleton robot, and M represents the number of degrees of freedom. d C represents inertia. d K represents damping. d τ represents stiffness. e This represents the torque generated between the user and the exoskeleton robot. The embodiments of this application will employ the above impedance model multiple times.

[0043] The above impedance model can also be rewritten as:

[0044]

[0045] In the formula, the reference vector can be expressed as:

[0046]

[0047]

[0048]

[0049]

[0050]

[0051] Among them, Λ and Γ are two matrices derived from the impedance parameters, and τ l It is τ e The low-pass filtered signal. Based on the above equation, the objective of impedance control can be reformulated as z→0 and t→∞ to ensure that the desired impedance model is achieved in the low-frequency range.

[0052] As described in step S410, the exoskeleton robot is controlled to assist the user in limb training based on the first impedance parameter and the first reference trajectory.

[0053] It should be noted that the first impedance parameter is the equivalent impedance parameter deployed on the motion mechanism of the exoskeleton robot. That is, the dynamic model of the exoskeleton robot can be represented as follows:

[0054]

[0055]

[0056] in, Represents the inertia matrix. This represents the correlation matrix between centrifugal force and Coriolis force. This represents the torque vector caused by gravity. This represents the rotation vector of the drive motor. This represents the stiffness matrix of the elastic components in the device. This represents the inertia matrix of the motor in the device. This indicates the output torque of the motor in the device.

[0057] The first reference trajectory can reflect the deviation between the user's motion trajectory allowed by the exoskeleton robot and the first reference trajectory during actual training. Generally speaking, the smaller the first impedance parameter, the greater the allowable deviation. That is, compared with directly controlling the position of the user's limbs (result), the first reference trajectory is more inclined to impedance control (process), thereby allowing the user's motion trajectory to deviate from the first reference trajectory to a certain extent.

[0058] In one embodiment of this application, the specific process of "determining the first impedance parameter based on the static body parameters, the dynamic body parameters, the motion trajectory and the torque" in step S210 can be further explained in conjunction with the following description.

[0059] A median trajectory is generated based on the static body parameters; a first training evaluation parameter for the user is generated based on the dynamic body parameters; and a first impedance parameter is determined based on the median trajectory, the motion trajectory, the torque, and the first training evaluation parameter.

[0060] The specific process of "generating intermediate trajectory based on the static body parameters" is as follows:

[0061] The intermediate trajectory is generated by Gaussian process regression based on the static body parameters.

[0062] It should be noted that the intermediate trajectory described in this application embodiment is for generating intermediate and process values ​​of the first impedance parameter and the first reference trajectory. Based on this, the exoskeleton robot can also use the intermediate trajectory as the user's personalized motion trajectory and directly provide assistance to the user by tracking the intermediate trajectory. That is, the intermediate trajectory can be used as a direct parameter to assist the user in limb training.

[0063] Reference Figure 2 Considering that the time-varying function corresponding to the motion trajectory cannot be directly correlated with static body parameters, this application determines the trajectory features in the frequency domain through Fourier transform: For the motion trajectory of the training sample, the phase angle, amplitude, bias, etc. in the Fast Fourier Transform (FFT) result can be extracted to construct the motion feature vector. The output motion trajectory feature structure is compact and has low dimension, which helps to efficiently represent the motion trajectory with time-varying properties; This feature conforms to the periodicity of limb training movement, that is, the user's limb and the exoskeleton robot perform interactive movements at a certain frequency; For the generated result, the predicted motion feature vector can be directly converted into robot joint motion angles through inverse Fourier transform and deployed on the exoskeleton robot.

[0064] To comprehensively and systematically assess user conditions, this application selects multidimensional and heterogeneous static body parameters for each user, such as height, shoulder width, arm length, weight, gender, age, and heart rate when not training. Multi-parameter cross-analysis is then used to further determine the weights of different static body parameters influencing movement trajectory. When hardware resources are limited, parameters with higher weights can be selected to improve computational efficiency.

[0065] Gaussian process regression (GPR) can meticulously reconstruct a user's limb movement patterns based on the static body parameters. Furthermore, as a kernel-based statistical learning method, GPR has advantages in solving small-sample learning problems and is suitable for databases containing a small number of research subjects.

[0066] Within the framework of the GPR algorithm, the mapping relationship between the static body parameters of the multiple sets of samples and the motion feature vectors of the multiple sets of samples is estimated by the probability density function P(μ|b), and the specific formula is as follows:

[0067]

[0068] Where b = (b1, b2, ..., b N ) represents the static body parameters of multiple samples, μ=(μ1,μ2,...,μ N ) represents the motion feature vectors of multiple samples, M represents the total number of samples in the dataset, and α m Represents the coefficients of each group of data in the dataset, and is defined as follows:

[0069] Wherein, g(b; μ m ,σ m Let ) represent a Gaussian distribution and define it as:

[0070]

[0071] in, This represents the mean vector of the motion feature vectors of multiple samples. Let θ represent the covariance matrix, and θ represent the hyperparameter vector.

[0072] The Gaussian distribution g(b; μ) is used to... m ,σ m Taking the logarithm yields the cost function log[g(b; μ). m ,σ m The specific formula is as follows:

[0073]

[0074] Using the above formula, the motion feature vectors of multiple samples are determined by Fast Fourier Transform (FFT) based on the motion trajectories of multiple samples. The static body parameters and motion feature vectors of the multiple samples are then sequentially input into the cost function log[g(b; μ]. m ,σ m Training will be conducted.

[0075] When the cost function log[g(b;μ m ,σ m When the minimum value is reached, the value of the hyperparameter vector θ is determined. optimal ;

[0076] The obtained θ optimal Input the probability density function P(μ|b) to establish the first mapping model;

[0077] By inputting the user's static body parameters into the first mapping model, the user's motion feature vector can be output. Then, based on the user's motion feature vector, the intermediate trajectory can be determined by inverse Fourier transform.

[0078] In addition, this application can also collect and store static body parameters and movement trajectories of people through a cloud platform for training and storing the first mapping model and building user training profiles.

[0079] In one embodiment of this application, the specific process of "generating the user's first training evaluation parameters based on the dynamic body parameters" can be further explained in conjunction with the following description.

[0080] Based on the dynamic body parameters, the user's first training evaluation parameters are generated using a convolutional neural network and a recurrent neural network.

[0081] The specific process of "generating the user's first training evaluation parameters through a convolutional neural network and a recurrent neural network based on the dynamic body parameters" is as follows:

[0082] Based on the dynamic body parameters, a first sampling parameter is determined by sampling through a sliding window; based on the first sampling parameter, features of the first sampling parameter are extracted by a convolutional neural network; based on the features of the first sampling parameter, a user comfort index is output by a recurrent neural network; and the comfort index is determined as the first training evaluation parameter.

[0083] Reference Figure 3 It should be noted that when training the network, the electromyography (EMG) signals and other trunk feedback information of multiple sample individuals are first collected. The comfort level of the sample individuals is obtained through methods such as inquiry and survey and quantified as the first sample training evaluation parameters. Then, the network is trained by backpropagation. After the training is completed, the network is deployed to the exoskeleton robot to complete the real-time perception of user comfort by the exoskeleton robot.

[0084] The advantages of the aforementioned network mainly include: enabling end-to-end extraction of scene information to achieve state perception and understanding in human-computer interaction; effectively combining multimodal data, including electromyography (EMG) signals, body oxygen consumption, and carbon dioxide consumption, and using the above-mentioned multiple limb feedback information to jointly complete the perception of the user's training state; and designing a network structure that combines convolutional neural networks and recurrent neural networks to complete the modeling of spatial information of EMG images and temporal information of limb feedback, thereby improving the accuracy of training state assessment.

[0085] The specific process of "determining the first impedance parameter based on the dielectric trajectory, the motion trajectory, the torque, and the first training evaluation parameter" is as follows:

[0086] The dielectric impedance parameter is determined based on the dielectric trajectory, the motion trajectory, and the torque; the first impedance parameter is determined based on the dielectric impedance parameter and the first training evaluation parameter.

[0087] The specific process of "determining the dielectric impedance parameter based on the dielectric trajectory, the motion trajectory, and the torque" is as follows:

[0088] The process of "determining the dielectric impedance parameter based on the dielectric trajectory, the motion trajectory, and the torque" can involve an impedance model, the formula of which is expressed as:

[0089]

[0090] in, This represents the angle vector corresponding to the trajectory of motion. This represents the angle vector corresponding to the intermediate value trajectory. Let M represent the three variables of the dielectric impedance parameter, n represent the number of degrees of freedom of the exoskeleton robot, and M represent the number of degrees of freedom of the exoskeleton robot. d0 C represents inertia. d0 K represents damping. d0 τ represents stiffness. e This represents the torque generated between the user and the exoskeleton robot.

[0091] Specifically, q and q represent the angular acceleration vector, angular velocity vector, and angle vector corresponding to the trajectory, respectively. and q d0 These represent the angular acceleration vector, angular velocity vector, and angle vector corresponding to the intermediate trajectory, respectively.

[0092] In the above formula, the three variables M of the dielectric impedance parameter can be obtained by substituting the dielectric trajectory, the motion trajectory, and the torque as known parameters. d0 C d0 and K d0 .

[0093] The specific process of "determining the first impedance parameter based on the dielectric impedance parameter and the first training evaluation parameter" is as follows:

[0094] The weighting coefficients corresponding to the dielectric impedance parameter are determined based on the first training evaluation parameters; the first impedance parameter is determined based on the weighting coefficients and the dielectric impedance parameter.

[0095] In this regard, since the exoskeleton robot should reduce its impedance parameter when the user's comfort level is good and the human-machine matching degree is high, so as to provide the user with more voluntary movement space, the weight coefficient corresponding to the first training evaluation parameter and the dielectric impedance parameter should be negatively correlated. That is, when the user's comfort level is higher, the first training evaluation parameter is higher and the weight coefficient is lower, so as to ensure that the first impedance parameter is smaller.

[0096] Using the above formula The three variables M of the dielectric impedance parameter were determined. d0 C d0 and K d0 Then, it can be obtained through three variables M d0 C d0 and K d0 The corresponding weighting coefficients determine the three variables M of the first impedance parameter. d1 C d1 and K d1 .

[0097] In one embodiment of this application, the specific process of "determining the first reference trajectory based on the first impedance parameter, the motion trajectory, and the torque" in step S310 can be further explained in conjunction with the following description.

[0098] The process of "determining the first reference trajectory based on the first impedance parameter, the motion trajectory, and the torque" may involve an impedance model, the formula of which is expressed as:

[0099]

[0100] in, This represents the angle vector corresponding to the trajectory of motion. This represents the angle vector corresponding to the first reference trajectory. Let M represent the three variables of the dielectric impedance parameter, n represent the number of degrees of freedom of the exoskeleton robot, and M represent the number of degrees of freedom of the exoskeleton robot. d1 C represents inertia. d1 K represents damping. d1 τ represents stiffness. e This represents the torque generated between the user and the exoskeleton robot.

[0101] Specifically, q and q represent the angular acceleration vector, angular velocity vector, and angle vector corresponding to the trajectory, respectively. and q d1 These represent the angular acceleration vector, angular velocity vector, and angle vector corresponding to the first reference trajectory, respectively.

[0102] In the above formula, the first reference trajectory can be obtained by substituting the first impedance parameter, the motion trajectory, and the torque as known parameters.

[0103] In one embodiment of this application, the specific process following step 410, "controlling the exoskeleton robot to assist the user in limb training based on the first impedance parameter and the first reference trajectory," can be further described in conjunction with the following description.

[0104] The user's motion characteristics are determined based on the motion trajectory and the torque; second training evaluation parameters for the user are generated based on the static body parameters and the motion characteristics; a second impedance parameter is determined based on the first impedance parameter and the second training evaluation parameters; a second reference trajectory is determined based on the second impedance parameter, the motion trajectory, and the torque; and the exoskeleton robot is controlled to assist the user in limb training based on the second impedance parameter and the second reference trajectory.

[0105] It should be noted that the second training evaluation parameter can be the user's 6MWD (6-Minutes Walk Distance) and / or the user's FMA (Fugl-Meyer) score. 6MWD and FMA scores are indicators of a user's limb motor ability and can also be used to evaluate the user's limb training effect, i.e., as the second training evaluation parameter. In addition to 6MWD and FMA scores, the user's leg lift height and walking speed can also be used to evaluate the user's limb training effect. The static body parameters may include: the user's height, shoulder width, arm length, weight, gender, age, and the user's heart rate when not training. The exoskeleton robot's sensors may include: pressure sensors, angle sensors, torque sensors, and inertial measurement unit (IMU) sensors. The user's motion characteristics include: time-to-time ratios during training, average leg lift height, average stride length, joint angle differences, and average joint torque.

[0106] The specific process of "generating the user's second training evaluation parameters based on the static body parameters and the motion characteristics" is as follows:

[0107] The first selection parameters are selected by using an XGBoost regression model based on the degree of influence of the static body parameters and the motion characteristics on the second training evaluation parameters; the second training evaluation parameters for the user are generated by using an XGBoost regression model based on the first selection parameters; wherein, the second training evaluation parameters include: the user's 6MWD and FMA scores.

[0108] Reference Figure 4It should be noted that this embodiment uses an embedded method for screening: first, an XGBoost regression model is trained to obtain the weight coefficients of each feature, and features are screened from largest to smallest based on the weight coefficients. The first screening parameters after screening are used as the final feature values ​​for model training, and the 6MWD and FMA values ​​of users collected in clinical trials are used as the target values ​​for model training. The final feature values ​​and target values ​​are input into the XGBoost regression model for training to obtain a second mapping model. The second mapping model can predict the user's 6MWD and FMA values ​​based on the first screening parameters. That is, when a user wears the exoskeleton robot and walks a short distance, the user's 6MWD and FMA scores can be predicted.

[0109] The XGBoost algorithm uses first- and second-order partial derivatives. The second derivative is beneficial for faster and more accurate gradient descent. Using Taylor expansion to obtain the second derivative form of the function as the independent variable allows for leaf splitting optimization calculations based solely on the input data values, without needing to select a specific form of the loss function. Essentially, this separates the selection of the loss function from model algorithm optimization / parameter selection. This decoupling increases the applicability of XGBoost, enabling it to select the loss function as needed; that is, the XGBoost regression model can be used for both classification / selection and regression.

[0110] Specifically, the formula for the XGBoost regression model is expressed as follows:

[0111]

[0112] Among them, Obj (t) Let l represent the objective function and l represent the loss function. Let Ω(f) represent the squared loss function. t ) represents the regularization term (including L1 and L2 regularization), constant represents the constant term, and f t (x i ) represents one of the regression trees in the XGBoost regression model;

[0113] The Taylor expansion formula involved is:

[0114]

[0115] The partial derivative is defined as:

[0116]

[0117]

[0118] For Obj (t) f in t (xi Performing a three-term Taylor expansion and substituting the partial derivatives above, we obtain the formula for the second mapping model as follows:

[0119]

[0120] In most cases, evaluators assess training effectiveness through visual inspection, which lacks precision and standardization. This leads to different evaluation results for the same user due to subjective bias. However, in the process described in this embodiment, the training evaluation for the same user is not affected by the evaluator's subjective factors. Therefore, compared to subjective evaluation results, the second training evaluation parameter is more suitable for further adjusting the impedance parameter for the user.

[0121] The specific process of "determining a second impedance parameter based on the first impedance parameter and the second training evaluation parameter; determining a second reference trajectory based on the second impedance parameter, the motion trajectory, and the torque; and controlling the exoskeleton robot to assist the user in limb training based on the second impedance parameter and the second reference trajectory" is as follows:

[0122] Since the process of determining the first impedance parameter and the first reference trajectory has been clearly explained above, and the process of determining the second impedance parameter and the second reference trajectory is similar to the above process, only a brief explanation is given below.

[0123] The specific process of "determining the second impedance parameter based on the first impedance parameter and the second training evaluation parameter" is as follows:

[0124] Based on the first impedance parameter, the impedance of the user-specific joint is adjusted according to the second training evaluation parameter to confirm the second impedance parameter. The three variables of the second impedance parameter are M... d2 C d2 ,K d2 M d2 C represents inertia. d2 K represents damping. d2 The second impedance parameter represents the stiffness and is used to determine the second reference trajectory.

[0125] The specific process of "determining the second reference trajectory based on the second impedance parameter, the motion trajectory, and the torque" is as follows:

[0126] The process of "determining the second reference trajectory based on the second impedance parameter, the motion trajectory, and the torque" may involve an impedance model, the formula of which is expressed as:

[0127]

[0128] in, This represents the angle vector corresponding to the trajectory of motion. This represents the angle vector corresponding to the second reference trajectory. These represent the three variables of the second impedance parameter, where n represents the number of degrees of freedom of the exoskeleton robot, and M... d2 C represents inertia. d2 K represents damping. d2 τ represents stiffness. e This represents the torque generated between the user and the exoskeleton robot.

[0129] Specifically, q and q represent the angular acceleration vector, angular velocity vector, and angle vector corresponding to the trajectory, respectively. and q d2 These represent the angular acceleration vector, angular velocity vector, and angle vector corresponding to the second reference trajectory, respectively.

[0130] In the above formula, the second reference trajectory can be obtained by substituting the second impedance parameter, the motion trajectory, and the torque as known parameters.

[0131] As the device embodiment is basically similar to the method embodiment, the description is relatively simple, and relevant parts can be found in the description of the method embodiment.

[0132] Reference Figure 5 A limb training device based on reference trajectory and impedance is shown, the device being used for non-therapeutic limb training, the device relating to an exoskeleton robot connected to a user's limb, comprising:

[0133] The input module 510 is used to acquire the user's static body parameters, the user's dynamic body parameters during training, the user's movement trajectory during training, and the torque generated between the user and the exoskeleton robot.

[0134] The first impedance parameter determination module 520 determines the first impedance parameter based on the static body parameters, the dynamic body parameters, the motion trajectory, and the torque.

[0135] The first reference trajectory determination module 530 is used to determine the first reference trajectory based on the first impedance parameter, the motion trajectory and the torque.

[0136] The limb training module 540 is used to control the exoskeleton robot to assist the user in limb training based on the first impedance parameter and the first reference trajectory.

[0137] Reference Figure 6 The present invention discloses a computer device for a limb training method based on reference trajectory and impedance, which may specifically include the following:

[0138] The computer device 12 described above is in the form of a general-purpose computing device. The components of the computer device 12 may include, but are not limited to: one or more processors or processing units 16, system memory 28, and bus 18 connecting different system components (including system memory 28 and processing unit 16).

[0139] Bus 18 refers to one or more of several types of bus 18 architectures, including memory bus 18 or memory controller, peripheral bus 18, graphics acceleration port, processor, or local bus 18 using any of the various bus 18 architectures. For example, these architectures include, but are not limited to, Industry Standard Architecture (ISA) bus 18, Micro Channel Architecture (MAC) bus 18, Enhanced ISA bus 18, Audio / Video Electronics Standards Association (VESA) local bus 18, and Peripheral Component Interconnect (PCI) bus 18.

[0140] Computer device 12 typically includes a variety of computer system readable media. These media can be any available media that can be accessed by computer device 12, including volatile and non-volatile media, removable and non-removable media.

[0141] System memory 28 may include computer system readable media in the form of volatile memory, such as random access memory (RAM) 30 and / or cache memory 32. Computer device 12 may further include other removable / non-removable, volatile / non-volatile computer system storage media. By way of example only, storage system 34 may be used to read and write non-removable, non-volatile magnetic media (commonly referred to as a "hard disk drive"). Figure 6 Not shown, a disk drive for reading and writing to a removable non-volatile disk (such as a "floppy disk") and an optical disk drive for reading and writing to a removable non-volatile optical disk (such as a CD-ROM, DVD-ROM, or other optical media) may be provided. In these cases, each drive may be connected to bus 18 via one or more data media interfaces. The memory may include at least one program product having a set (e.g., at least one) of program modules 42 configured to perform the functions of the embodiments of the present invention.

[0142] A program / utility 40 having a set (at least one) of program modules 42 may be stored, for example, in memory. Such program modules 42 include—but are not limited to—an operating system, one or more application programs, other program modules 42, and program data. Each or some combination of these examples may include an implementation of a network environment. Program modules 42 typically perform the functions and / or methods described in the embodiments of the present invention.

[0143] Computer device 12 can also communicate with one or more external devices 14 (e.g., keyboard, pointing device, display 24, camera, etc.), and with one or more devices that enable a user to interact with the computer device 12, and / or with any device that enables the computer device 12 to communicate with one or more other computing devices (e.g., network card, modem, etc.). This communication can be performed through input / output (I / O) interface 22. Furthermore, computer device 12 can also communicate with one or more networks (e.g., local area network (LAN)), wide area network (WAN), and / or public networks (e.g., the Internet) via network adapter 20. As shown, network adapter 20 communicates with other modules of computer device 12 via bus 18. It should be understood that, although... Figure 6 Not shown, it can be combined with computer device 12 to use other hardware and / or software modules, including but not limited to: microcode, device drivers, redundant processing unit 16, external disk drive array, RAID system, tape drive and data backup storage system 34, etc.

[0144] The processing unit 16 executes various functional applications and data processing by running programs stored in the system memory 28, such as implementing a limb training method based on reference trajectory and impedance provided in the embodiments of the present invention.

[0145] That is, when the processing unit 16 executes the above program, it achieves the following: acquiring the user's static body parameters, the user's dynamic body parameters during training, the user's motion trajectory during training, and the torque generated between the user and the exoskeleton robot; determining a first impedance parameter based on the static body parameters, the dynamic body parameters, the motion trajectory, and the torque; determining a first reference trajectory based on the first impedance parameter, the motion trajectory, and the torque; and controlling the exoskeleton robot to assist the user in limb training based on the first impedance parameter and the first reference trajectory.

[0146] In this embodiment of the invention, a computer-readable storage medium is also provided, on which a computer program is stored, which, when executed by a processor, implements the methods provided in all embodiments of this application:

[0147] That is, when the program is executed by the processor, it implements the following: acquiring the user's static body parameters, the user's dynamic body parameters during training, the user's motion trajectory during training, and the torque generated between the user and the exoskeleton robot; determining a first impedance parameter based on the static body parameters, the dynamic body parameters, the motion trajectory, and the torque; determining a first reference trajectory based on the first impedance parameter, the motion trajectory, and the torque; and controlling the exoskeleton robot to assist the user in limb training based on the first impedance parameter and the first reference trajectory.

[0148] Any combination of one or more computer-readable media may be used. A computer-readable medium may be a computer-to-signal medium or a computer-readable storage medium. A computer-readable storage medium may be, for example—but not limited to—an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples (a non-exhaustive list) of computer-readable storage media include: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In this document, a computer-readable storage medium may be any tangible medium that contains or stores a program that may be used by or in connection with an instruction execution system, apparatus, or device.

[0149] Computer-readable signal media may include data signals propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals may take various forms, including—but not limited to—electromagnetic signals, optical signals, or any suitable combination thereof. Computer-readable signal media may also be any computer-readable medium other than computer-readable storage media, capable of transmitting, propagating, or transmitting programs for use by or in connection with an instruction execution system, apparatus, or device.

[0150] Computer program code for performing the operations of this invention can be written in one or more programming languages ​​or a combination thereof. These programming languages ​​include object-oriented programming languages—such as Java, Smalltalk, and C++—and conventional procedural programming languages—such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider). The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments; similar or identical parts between embodiments can be referred to interchangeably.

[0151] Although preferred embodiments of the present application have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of the embodiments of the present application.

[0152] Finally, it should be noted that in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or terminal device that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or terminal device. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or terminal device that includes said element.

[0153] The foregoing has provided a detailed description of a limb training method and apparatus based on reference trajectory and impedance provided in this application. Specific examples have been used to illustrate the principles and implementation methods of this application. The description of the above embodiments is only for the purpose of helping to understand the method and its core ideas. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this application. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A limb training method based on reference trajectory and impedance, said method for non-therapeutic limb training, said method relating to an exoskeleton robot connected to a user's limb, characterized in that, Includes the following steps: The system acquires the user's static body parameters, the user's dynamic body parameters during training, the user's movement trajectory during training, and the torque generated between the user and the exoskeleton robot. A first impedance parameter is determined based on the static body parameters, the dynamic body parameters, the motion trajectory, and the torque; wherein, a median trajectory is generated based on the static body parameters; a first training evaluation parameter for the user is generated based on the dynamic body parameters; and the first impedance parameter is determined based on the median trajectory, the motion trajectory, the torque, and the first training evaluation parameter; wherein, the first training evaluation parameter is a user comfort index and is negatively correlated with the first impedance parameter. A first reference trajectory is determined based on the first impedance parameter, the motion trajectory, and the torque; The exoskeleton robot is controlled based on the first impedance parameter and the first reference trajectory to assist the user in limb training.

2. The method according to claim 1, characterized in that, The step of generating the intermediate trajectory based on the static body parameters includes: The intermediate trajectory is generated by Gaussian process regression based on the static body parameters.

3. The method according to claim 1, characterized in that, The step of generating the user's first training evaluation parameters based on the dynamic body parameters includes: Based on the dynamic body parameters, the user's first training evaluation parameters are generated using a convolutional neural network and a recurrent neural network.

4. The method according to claim 1, characterized in that, The step of determining the first impedance parameter based on the dielectric trajectory, the motion trajectory, the torque, and the first training evaluation parameter includes: The dielectric impedance parameters are determined based on the dielectric trajectory, the motion trajectory, and the torque. The first impedance parameter is determined based on the dielectric impedance parameter and the first training evaluation parameter.

5. The method according to claim 4, characterized in that, The step of determining the first impedance parameter based on the dielectric impedance parameter and the first training evaluation parameter includes: The weighting coefficients corresponding to the dielectric impedance parameter are determined based on the first training evaluation parameters; The first impedance parameter is determined based on the weighting coefficient and the dielectric impedance parameter.

6. The method according to claim 1, characterized in that, Following the step of controlling the exoskeleton robot to assist the user in limb training based on the first impedance parameter and the first reference trajectory, the method further includes: The user's motion characteristics are determined based on the motion trajectory and the torque; The user's second training evaluation parameters are generated based on the static body parameters and the motion characteristics; The second impedance parameter is determined based on the first impedance parameter and the second training evaluation parameter; The second reference trajectory is determined based on the second impedance parameter, the motion trajectory, and the torque; The exoskeleton robot is controlled based on the second impedance parameter and the second reference trajectory to assist the user in limb training.

7. A limb training device based on reference trajectory and impedance, the device being used for non-therapeutic limb training, the device relating to an exoskeleton robot connected to a user's limb, characterized in that, include: The input module is used to acquire the user's static body parameters, the user's dynamic body parameters during training, the user's movement trajectory during training, and the torque generated between the user and the exoskeleton robot. The first impedance parameter determination module determines a first impedance parameter based on the static body parameters, the dynamic body parameters, the motion trajectory, and the torque; wherein, a median trajectory is generated based on the static body parameters; a first training evaluation parameter for the user is generated based on the dynamic body parameters; and the first impedance parameter is determined based on the median trajectory, the motion trajectory, the torque, and the first training evaluation parameter; wherein, the first training evaluation parameter is a user comfort index and is negatively correlated with the first impedance parameter. The first reference trajectory determination module is used to determine the first reference trajectory based on the first impedance parameter, the motion trajectory, and the torque. The limb training module is used to control the exoskeleton robot to assist the user in limb training based on the first impedance parameter and the first reference trajectory.

8. A device, characterized in that, It includes a processor, a memory, and a computer program stored in the memory and capable of running on the processor, wherein the computer program, when executed by the processor, implements the method as described in any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that, A computer program is stored on the computer-readable storage medium, which, when executed by a processor, implements the method as described in any one of claims 1 to 6.

Images