Feedback processing device and method for upper limb rehabilitation training

Abstract

The invention discloses a feedback processing device and method for upper limb rehabilitation training. The method comprises a motion trail acquisition module, an evaluation analysis module and a feedback processing module. The motion trail acquisition module is used for acquiring an upper limb motion trail information set when a user performs upper limb rehabilitation training by using the upper limb rehabilitation training device; the evaluation and analysis module is used for evaluating and judging the upper limb movement track information set and the standard movement track information to obtain an evaluation and analysis result; the feedback processing module is used for carrying out feedback calculation processing on the evaluation analysis result, the standard motion trail information and the upper limb motion trail information set to obtain a feedback output value; the feedback output value is used for adjusting the acting force of the upper limb rehabilitation training device on the user when the user carries out upper limb rehabilitation training.

Description

Technical Field

[0001] The present invention relates to the field of exoskeleton robots and intelligent control, and in particular to a feedback processing device and method for upper limb rehabilitation training. Background Art

[0002] Rehabilitation training is essential for the recovery of patients with upper limb movement disorders. In traditional rehabilitation therapy, rehabilitation therapists usually provide one-on-one rehabilitation therapy to patients with bare hands. The rehabilitation therapists' personal treatment methods, experience differences, subjective consciousness, and fatigue level will directly affect the treatment effect. In addition, the treatment process is labor-intensive, the treatment cost is high, and the ratio of therapists to patients is seriously unbalanced, which can hardly meet the growing rehabilitation needs.

[0003] In recent years, various types of upper limb rehabilitation training robots have been developed, and rehabilitation robots have become a research hotspot in the field of upper limb rehabilitation therapy. Current rehabilitation training robots mainly provide users with a mechanical training environment, but lack the ability to evaluate the training effect of users and make corresponding changes to the training environment based on the evaluation results, which affects the training effect and efficiency. Summary of the invention

[0004] The present invention mainly solves the problem that the existing rehabilitation training robots lack the ability to evaluate the training effects of users and to make corresponding changes to the training environment according to the evaluation results. The present invention discloses a feedback processing device and method for upper limb rehabilitation training.

[0005] In a first aspect of the embodiment of the present application, a feedback processing device for upper limb rehabilitation training is disclosed, comprising: a motion trajectory acquisition module, an evaluation and analysis module, and a feedback processing module;

[0006] The motion trajectory acquisition module is used to acquire a set of upper limb motion trajectory information of a user when performing upper limb rehabilitation training using an upper limb rehabilitation training device;

[0007] The evaluation and analysis module is connected to the motion trajectory acquisition module and the feedback processing module respectively, and is used to evaluate and distinguish the upper limb motion trajectory information set and the standard motion trajectory information to obtain an evaluation and analysis result;

[0008] The feedback processing module is used to perform feedback calculation processing on the evaluation and analysis results, standard motion trajectory information and upper limb motion trajectory information set to obtain a feedback output value; the feedback output value is used to adjust the force exerted by the upper limb rehabilitation training device on the user when the user performs upper limb rehabilitation training.

[0009] The motion trajectory acquisition module is implemented by an image acquisition and analysis module or an accelerometer sensor arranged on the upper limb of the user; the upper limb motion trajectory information set includes a plurality of upper limb motion trajectory information sequences.

[0010] The evaluation and analysis module evaluates and distinguishes the upper limb motion trajectory information set and the standard motion trajectory information to obtain an evaluation and analysis result, including:

[0011] The evaluation and analysis module performs difference calculation processing on the upper limb motion trajectory information set and the standard motion trajectory information to obtain a difference value;

[0012] Determine whether the difference value is greater than the set difference discrimination threshold value to obtain a difference discrimination result; if the difference discrimination result is greater than the set difference discrimination threshold value, determine that the evaluation and analysis result is difficult to complete; if the difference discrimination result is not greater than the set difference discrimination threshold value, determine that the evaluation and analysis result can be completed.

[0013] The feedback processing module is used to perform feedback calculation processing on the evaluation and analysis results, the standard motion trajectory information and the upper limb motion trajectory information set to obtain a feedback output value, including:

[0014] When the evaluation and analysis result is difficult to complete, the feedback processing module uses the first feedback model to calculate and process the upper limb motion trajectory information set and the standard motion trajectory information to obtain a feedback output value;

[0015] When the evaluation and analysis result is that the process can be completed, the feedback processing module uses a second feedback model to calculate and process the upper limb motion trajectory information set and the standard motion trajectory information to obtain a feedback output value.

[0016] In a second aspect of an embodiment of the present invention, a feedback processing method for upper limb rehabilitation training is disclosed, which is implemented by using the feedback processing device for upper limb rehabilitation training, and includes:

[0017] S1, using the motion trajectory acquisition module to acquire a set of upper limb motion trajectory information of a user when performing upper limb rehabilitation training using an upper limb rehabilitation training device;

[0018] S2, using the evaluation and analysis module to evaluate and distinguish the upper limb motion trajectory information set and the standard motion trajectory information to obtain an evaluation and analysis result;

[0019] S3, using the feedback processing module, performing feedback calculation processing on the evaluation and analysis results, the standard motion trajectory information and the upper limb motion trajectory information set to obtain a feedback output value.

[0020] The evaluation and analysis module is used to evaluate and distinguish the upper limb motion trajectory information set and the standard motion trajectory information to obtain an evaluation and analysis result, including:

[0021] Performing difference calculation processing on the upper limb motion trajectory information set and the standard motion trajectory information to obtain a difference value;

[0022] Determine whether the difference value is greater than the set difference discrimination threshold value to obtain a difference discrimination result; if the difference discrimination result is greater than the set difference discrimination threshold value, determine that the evaluation and analysis result is difficult to complete; if the difference discrimination result is not greater than the set difference discrimination threshold value, determine that the evaluation and analysis result can be completed.

[0023] The step of performing difference calculation on the upper limb motion trajectory information set and the standard motion trajectory information to obtain a difference value includes:

[0024] Performing time alignment processing on the upper limb motion trajectory information set and the standard motion trajectory information to obtain a time-aligned upper limb motion trajectory information set and a time-aligned standard motion trajectory information;

[0025] Using the time-aligned upper limb motion trajectory information set, a motion trajectory matrix is ​​constructed; the row vector of the motion trajectory matrix is ​​an upper limb motion trajectory information sequence;

[0026] Taking the time-aligned standard motion trajectory information as a row vector, the row vector is copied in a column direction to obtain a standard motion matrix; the dimension of the standard motion matrix is ​​the same as the dimension of the motion trajectory matrix;

[0027] Subtracting the motion trajectory matrix from the standard motion matrix to obtain a difference matrix;

[0028] Performing feature mean calculation on the difference matrix to obtain a mean vector;

[0029] Performing weight vector calculation on the difference matrix to obtain a weight vector;

[0030] A weighted sum is performed on the weight vector and the mean vector to obtain a difference value.

[0031] The expression for calculating the characteristic mean is:

[0032]

[0033] Among them, b p is the pth element of the mean vector, A pq is the element in the p-th row and q-th column of the difference matrix, and n is the column dimension of the difference matrix;

[0034] The expression for calculating the weight vector is:

[0035]

[0036] Among them, v p is the pth element of the time-aligned standard motion trajectory information, f p is the p-th element of the weight vector.

[0037] The feedback calculation processing is performed on the evaluation and analysis results, the standard motion trajectory information and the upper limb motion trajectory information set to obtain a feedback output value, including:

[0038] When the evaluation and analysis result is difficult to complete, using the first feedback model, the upper limb motion trajectory information set and the standard motion trajectory information are calculated and processed to obtain a feedback output value;

[0039] When the evaluation and analysis result is that it can be completed, the upper limb motion trajectory information set and the standard motion trajectory information are calculated and processed using a second feedback model to obtain a feedback output value.

[0040] The calculation expression of the first feedback model is:

[0041]

[0042] in, represents the loss function, δ and η are the preset first adjustment factor and second adjustment factor, b max is the maximum value of the mean vector, b p is the pth element of the mean vector, m is the row dimension of the difference matrix, v p It is the pth element of the time-aligned standard motion trajectory information.

[0043] The beneficial effects of the present invention are:

[0044] The present invention solves the problem that the existing rehabilitation training robot lacks the ability to evaluate the training effect of the user and to make corresponding changes to the training environment according to the evaluation results. The present invention establishes an evaluation and analysis module to qualitatively evaluate the training effect of the user, and divides the user's training effect into two categories, "difficult to complete" and "achievable", according to the movement trajectory of the upper limbs of the user during the training process. The corresponding feedback control model is used to calculate the feedback amount in a targeted manner, thereby improving the accuracy of feedback control and the training effect and efficiency.

[0045] During the training evaluation process, the present invention obtains a difference value by performing difference calculation processing on the upper limb motion trajectory information set and the standard motion trajectory information, thereby achieving accurate and rapid evaluation of the training movements and providing a basis for achieving accurate feedback. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 It is a composition diagram of the device of the present invention;

[0047] Figure 2 It is a flow chart for implementing the method of the present invention. DETAILED DESCRIPTION

[0048] In order to better understand the content of the present invention, an embodiment is given here.

[0049] Figure 1 It is a composition diagram of the device of the present invention; Figure 2 It is a flow chart for implementing the method of the present invention.

[0050] Aiming at the problem that the existing rehabilitation training robots lack the ability to evaluate the training effects of users and make corresponding changes to the training environment according to the evaluation results, the present invention discloses a feedback processing device and method for upper limb rehabilitation training.

[0051] In a first aspect of the embodiment of the present application, a feedback processing device for upper limb rehabilitation training is disclosed, comprising: a motion trajectory acquisition module, an evaluation and analysis module, and a feedback processing module;

[0052] The motion trajectory acquisition module is used to acquire a set of upper limb motion trajectory information of a user when performing upper limb rehabilitation training using an upper limb rehabilitation training device;

[0053] The evaluation and analysis module is used to evaluate and distinguish the upper limb motion trajectory information set and the standard motion trajectory information to obtain an evaluation and analysis result;

[0054] The feedback processing module is used to perform feedback calculation processing on the evaluation and analysis results, the standard motion trajectory information and the upper limb motion trajectory information set to obtain a feedback output value; the feedback output value is used to adjust the force of the upper limb rehabilitation training device on the user when the user performs upper limb rehabilitation training;

[0055] The device of the present invention is used for upper limb rehabilitation training device;

[0056] The motion trajectory acquisition module is implemented by an image acquisition and analysis module or an accelerometer sensor disposed on the user's upper limb;

[0057] The upper limb motion trajectory information set includes a plurality of upper limb motion trajectory information sequences;

[0058] The accelerometer sensor is used to collect the upper limb motion trajectory of the user when performing upper limb rehabilitation training using the upper limb rehabilitation training device, and construct an upper limb motion trajectory information set using all collected upper limb motion trajectories;

[0059] The image acquisition and analysis module includes an image acquisition submodule and an image analysis submodule; the image acquisition submodule is used to acquire images of a user when performing upper limb rehabilitation training using an upper limb rehabilitation training device; the image analysis submodule is used to extract upper limb parts from the images of the user when performing upper limb rehabilitation training using the upper limb rehabilitation training device to acquire upper limb motion trajectories, and to construct an upper limb motion trajectory information set using all acquired upper limb motion trajectories;

[0060] The upper limb part extraction from the image of the user performing upper limb rehabilitation training using the upper limb rehabilitation training device can be implemented by using a SURF feature point detection or corner point detection algorithm, or by using an OpenPose algorithm in OpenCV.

[0061] The evaluation and analysis module evaluates and distinguishes the upper limb motion trajectory information set and the standard motion trajectory information to obtain an evaluation and analysis result, including:

[0062] The evaluation and analysis module performs difference calculation processing on the upper limb motion trajectory information set and the standard motion trajectory information to obtain a difference value;

[0063] Determine whether the difference value is greater than a set difference discrimination threshold value, and obtain a difference discrimination result; if the difference discrimination result is greater than the set difference discrimination threshold value, determine that the evaluation and analysis result is difficult to complete; if the difference discrimination result is not greater than the set difference discrimination threshold value, determine that the evaluation and analysis result can be completed;

[0064] The step of performing difference calculation on the upper limb motion trajectory information set and the standard motion trajectory information to obtain a difference value includes:

[0065] Performing time alignment processing on the upper limb motion trajectory information set and the standard motion trajectory information to obtain a time-aligned upper limb motion trajectory information set and a time-aligned standard motion trajectory information;

[0066] Using the time-aligned upper limb motion trajectory information set, a motion trajectory matrix is ​​constructed; the row vector of the motion trajectory matrix is ​​an upper limb motion trajectory information sequence;

[0067] Taking the time-aligned standard motion trajectory information as a row vector, the row vector is copied in a column direction to obtain a standard motion matrix; the dimension of the standard motion matrix is ​​the same as the dimension of the motion trajectory matrix;

[0068] Subtracting the motion trajectory matrix from the standard motion matrix to obtain a difference matrix;

[0069] Performing feature mean calculation on the difference matrix to obtain a mean vector;

[0070] The expression for calculating the characteristic mean is:

[0071]

[0072] Among them, b p is the pth element of the mean vector, A pq is the element in the p-th row and q-th column of the difference matrix, and n is the column dimension of the difference matrix;

[0073] Performing weight vector calculation on the difference matrix to obtain a weight vector;

[0074] The expression for calculating the weight vector is:

[0075]

[0076] Among them, v p is the pth element of the time-aligned standard motion trajectory information, f p is the pth element of the weight vector;

[0077] Performing weighted summation on the weight vector and the mean vector to obtain a difference value;

[0078] The feedback calculation processing is performed on the evaluation and analysis results, the standard motion trajectory information and the upper limb motion trajectory information set to obtain a feedback output value, including:

[0079] When the evaluation and analysis result is difficult to complete, using the first feedback model, the upper limb motion trajectory information set and the standard motion trajectory information are calculated and processed to obtain a feedback output value;

[0080] When the evaluation and analysis result is that the upper limb motion trajectory information set and the standard motion trajectory information are calculated and processed by using a second feedback model to obtain a feedback output value;

[0081] The calculation expression of the first feedback model is:

[0082]

[0083] in, represents the loss function, δ and η are the preset first adjustment factor and second adjustment factor, b max is the maximum value of the mean vector, b p is the pth element of the mean vector, m is the row dimension of the difference matrix, v p is the pth element of the time-aligned standard motion trajectory information, and β is the calculated feedback output value;

[0084] The second feedback model includes:

[0085] Decomposing the difference matrix to obtain a feature matrix;

[0086] The decomposition process is calculated as follows:

[0087] A=UYV,

[0088] Among them, U is the left decomposition matrix, A is the difference matrix, V is the right decomposition matrix, U and V are both orthogonal matrices, and Y is a diagonal matrix;

[0089] Extracting the diagonal elements of the feature matrix to obtain a feature vector;

[0090] Performing fitting processing on the characteristic vector and the mean vector to obtain an optimal feedback polynomial;

[0091] Substituting the difference value into the optimal feedback polynomial to obtain a feedback output value.

[0092] The fitting process uses the elements of the characteristic vector as known independent variables and the elements of the mean vector as known dependent variables, constructs the curve to be approximated using the known independent variables and known dependent variables, and uses the function approximation method to perform curve fitting on the curve to be approximated to obtain the optimal feedback polynomial f(Ix).

[0093] The curve fitting of the curve to be approximated by the function approximation method may be performed by using the best consistent linear approximation method. The best feedback polynomial f(Ix) is expressed as:

[0094] f(Ix)=α P1 (Ix) P1 +α P1-1 (Ix) P1-1 +…+α 2 (Ix) 2 +α 1 (Ix)+α 0 ,

[0095] Wherein, P1 is the order of the optimal feedback polynomial f(Ix), α0, α1, α2, …, α P1 are the coefficients of the optimal feedback polynomial f(Ix), where Ix is the independent variable;

[0096] The decomposition process can be implemented by using a matrix singular value decomposition algorithm.

[0097] The loss function may adopt a cross entropy loss function;

[0098] The time alignment process comprises:

[0099] When the collection time interval of the upper limb motion trajectory information set is greater than the time interval of the standard motion trajectory information, interpolation processing is performed on the data of the adjacent upper limb motion trajectory information sets to obtain a time-aligned upper limb motion trajectory information set;

[0100] When the collection time interval of the upper limb motion trajectory information set is less than the time interval of the standard motion trajectory information, sampling the upper limb motion trajectory information set to obtain data consistent with the time interval of the standard motion trajectory information, and using the upper limb motion trajectory information set after the sampling process as the time-aligned upper limb motion trajectory information set;

[0101] In a second aspect of an embodiment of the present invention, a feedback processing method for upper limb rehabilitation training is disclosed, which is implemented by using the feedback processing device for upper limb rehabilitation training, and includes:

[0102] S1, using the motion trajectory acquisition module to acquire a set of upper limb motion trajectory information of a user when performing upper limb rehabilitation training using an upper limb rehabilitation training device;

[0103] S2, using the evaluation and analysis module to evaluate and distinguish the upper limb motion trajectory information set and the standard motion trajectory information to obtain an evaluation and analysis result;

[0104] S3, using the feedback processing module, performing feedback calculation processing on the evaluation and analysis results, the standard motion trajectory information and the upper limb motion trajectory information set to obtain a feedback output value.

[0105] The evaluation and analysis module is used to evaluate and distinguish the upper limb motion trajectory information set and the standard motion trajectory information to obtain an evaluation and analysis result, including:

[0106] Performing difference calculation processing on the upper limb motion trajectory information set and the standard motion trajectory information to obtain a difference value;

[0107] Determine whether the difference value is greater than the set difference discrimination threshold value to obtain a difference discrimination result; if the difference discrimination result is greater than the set difference discrimination threshold value, determine that the evaluation and analysis result is difficult to complete; if the difference discrimination result is not greater than the set difference discrimination threshold value, determine that the evaluation and analysis result can be completed.

[0108] The step of performing difference calculation on the upper limb motion trajectory information set and the standard motion trajectory information to obtain a difference value includes:

[0109] Performing time alignment processing on the upper limb motion trajectory information set and the standard motion trajectory information to obtain a time-aligned upper limb motion trajectory information set and a time-aligned standard motion trajectory information;

[0110] Using the time-aligned upper limb motion trajectory information set, a motion trajectory matrix is ​​constructed; the row vector of the motion trajectory matrix is ​​an upper limb motion trajectory information sequence;

[0111] Taking the time-aligned standard motion trajectory information as a row vector, the row vector is copied in a column direction to obtain a standard motion matrix; the dimension of the standard motion matrix is ​​the same as the dimension of the motion trajectory matrix;

[0112] Subtracting the motion trajectory matrix from the standard motion matrix to obtain a difference matrix;

[0113] Performing feature mean calculation on the difference matrix to obtain a mean vector;

[0114] Performing weight vector calculation on the difference matrix to obtain a weight vector;

[0115] A weighted sum is performed on the weight vector and the mean vector to obtain a difference value.

[0116] The expression for calculating the characteristic mean is:

[0117]

[0118] Among them, b p is the pth element of the mean vector, A pq is the element in the p-th row and q-th column of the difference matrix, and n is the column dimension of the difference matrix;

[0119] The expression for calculating the weight vector is:

[0120]

[0121] Among them, v p is the pth element of the time-aligned standard motion trajectory information, f p is the p-th element of the weight vector.

[0122] The feedback calculation processing is performed on the evaluation and analysis results, the standard motion trajectory information and the upper limb motion trajectory information set to obtain a feedback output value, including:

[0123] When the evaluation and analysis result is difficult to complete, using the first feedback model, the upper limb motion trajectory information set and the standard motion trajectory information are calculated and processed to obtain a feedback output value;

[0124] When the evaluation and analysis result is that it can be completed, the upper limb motion trajectory information set and the standard motion trajectory information are calculated and processed using the second feedback model to obtain a feedback output value.

[0125] The calculation expression of the first feedback model is:

[0126]

[0127] in, represents the loss function, δ and η are the preset first adjustment factor and second adjustment factor, b max is the maximum value of the mean vector, b p is the pth element of the mean vector, m is the row dimension of the difference matrix, v p It is the pth element of the time-aligned standard motion trajectory information.

[0128] The second feedback model includes:

[0129] Decomposing the difference matrix to obtain a feature matrix;

[0130] The decomposition process is calculated as follows:

[0131] A=UYV,

[0132] Among them, U is the left decomposition matrix, A is the difference matrix, V is the right decomposition matrix, U and V are both orthogonal matrices, and Y is a diagonal matrix;

[0133] Extracting the diagonal elements of the feature matrix to obtain a feature vector;

[0134] Performing fitting processing on the characteristic vector and the mean vector to obtain an optimal feedback polynomial;

[0135] Substituting the difference value into the optimal feedback polynomial to obtain a feedback output value.

[0136] The fitting process uses the elements of the characteristic vector as known independent variables and the elements of the mean vector as known dependent variables, constructs the curve to be approximated using the known independent variables and known dependent variables, and uses the function approximation method to perform curve fitting on the curve to be approximated to obtain the optimal feedback polynomial f(Ix).

[0137] The curve fitting of the curve to be approximated by the function approximation method may be performed by using the best consistent linear approximation method. The best feedback polynomial f(Ix) is expressed as:

[0138] f(Ix)=α P1 (Ix) P1 +α P1-1 (Ix) P1-1 +…+α 2 (Ix) 2 +α 1 (Ix)+α 0 ,

[0139] Wherein, P1 is the order of the optimal feedback polynomial f(Ix), α0, α1, α2, …, α P1 are the coefficients of the optimal feedback polynomial f(Ix), where Ix is the independent variable;

[0140] The decomposition process can be implemented by using a matrix singular value decomposition algorithm.

[0141] The above is only an embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the scope of the claims of the present application.

Claims

1. A feedback processing device for upper limb rehabilitation training, characterized in that: include: Motion trajectory acquisition module, evaluation and analysis module, feedback processing module; The motion trajectory acquisition module is used to acquire a set of upper limb motion trajectory information of a user when performing upper limb rehabilitation training using an upper limb rehabilitation training device; The evaluation and analysis module is connected to the motion trajectory acquisition module and the feedback processing module respectively, and is used to evaluate and distinguish the upper limb motion trajectory information set and the standard motion trajectory information to obtain an evaluation and analysis result; The feedback processing module is used to perform feedback calculation processing on the evaluation and analysis results, the standard motion trajectory information and the upper limb motion trajectory information set to obtain a feedback output value; The feedback output value is used to adjust the force exerted by the upper limb rehabilitation training device on the user when the user performs upper limb rehabilitation training.

2. The feedback processing device for upper limb rehabilitation training according to claim 1, characterized in that: The motion trajectory acquisition module is implemented by an image acquisition and analysis module or an accelerometer sensor arranged on the upper limb of the user; the upper limb motion trajectory information set includes a plurality of upper limb motion trajectory information sequences.

3. The feedback processing device for upper limb rehabilitation training according to claim 2, characterized in that: The evaluation and analysis module evaluates and distinguishes the upper limb motion trajectory information set and the standard motion trajectory information to obtain an evaluation and analysis result, including: The evaluation and analysis module performs difference calculation processing on the upper limb motion trajectory information set and the standard motion trajectory information to obtain a difference value; Determine whether the difference value is greater than the set difference discrimination threshold value to obtain a difference discrimination result; if the difference discrimination result is greater than the set difference discrimination threshold value, determine that the evaluation and analysis result is difficult to complete; if the difference discrimination result is not greater than the set difference discrimination threshold value, determine that the evaluation and analysis result can be completed.

4. The feedback processing device for upper limb rehabilitation training according to claim 2, characterized in that: The feedback processing module is used to perform feedback calculation processing on the evaluation and analysis results, the standard motion trajectory information and the upper limb motion trajectory information set to obtain a feedback output value, including: When the evaluation and analysis result is difficult to complete, the feedback processing module uses the first feedback model to calculate and process the upper limb motion trajectory information set and the standard motion trajectory information to obtain a feedback output value; When the evaluation and analysis result is that the process can be completed, the feedback processing module uses a second feedback model to calculate and process the upper limb motion trajectory information set and the standard motion trajectory information to obtain a feedback output value.

5. A feedback processing method for upper limb rehabilitation training, characterized in that: The method is implemented by using the feedback processing device for upper limb rehabilitation training according to any one of claims 1 to 4, comprising: S1, using the motion trajectory acquisition module to acquire a set of upper limb motion trajectory information of a user when performing upper limb rehabilitation training using an upper limb rehabilitation training device; S2, using the evaluation and analysis module to evaluate and distinguish the upper limb motion trajectory information set and the standard motion trajectory information to obtain an evaluation and analysis result; S3, using the feedback processing module, performing feedback calculation processing on the evaluation and analysis results, the standard motion trajectory information and the upper limb motion trajectory information set to obtain a feedback output value.

6. The feedback processing method for upper limb rehabilitation training according to claim 5, characterized in that: The evaluation and analysis module is used to evaluate and distinguish the upper limb motion trajectory information set and the standard motion trajectory information to obtain an evaluation and analysis result, including: Performing difference calculation processing on the upper limb motion trajectory information set and the standard motion trajectory information to obtain a difference value; Determine whether the difference value is greater than the set difference discrimination threshold value to obtain a difference discrimination result; if the difference discrimination result is greater than the set difference discrimination threshold value, determine that the evaluation and analysis result is difficult to complete; if the difference discrimination result is not greater than the set difference discrimination threshold value, determine that the evaluation and analysis result can be completed.

7. The feedback processing method for upper limb rehabilitation training according to claim 6, characterized in that: The step of performing difference calculation on the upper limb motion trajectory information set and the standard motion trajectory information to obtain a difference value includes: Performing time alignment processing on the upper limb motion trajectory information set and the standard motion trajectory information to obtain a time-aligned upper limb motion trajectory information set and a time-aligned standard motion trajectory information; Using the time-aligned upper limb motion trajectory information set, a motion trajectory matrix is ​​constructed; the row vector of the motion trajectory matrix is ​​an upper limb motion trajectory information sequence; Taking the time-aligned standard motion trajectory information as a row vector, the row vector is copied in a column direction to obtain a standard motion matrix; the dimension of the standard motion matrix is ​​the same as the dimension of the motion trajectory matrix; Subtracting the motion trajectory matrix from the standard motion matrix to obtain a difference matrix; Performing feature mean calculation on the difference matrix to obtain a mean vector; Performing weight vector calculation on the difference matrix to obtain a weight vector; A weighted sum is performed on the weight vector and the mean vector to obtain a difference value.

8. The feedback processing method for upper limb rehabilitation training according to claim 7, characterized in that: The expression for calculating the characteristic mean is: Among them, b p is the pth element of the mean vector, A pq is the element in the p-th row and q-th column of the difference matrix, and n is the column dimension of the difference matrix; The expression for calculating the weight vector is: Among them, v p is the pth element of the time-aligned standard motion trajectory information, f p is the p-th element of the weight vector.

9. The feedback processing method for upper limb rehabilitation training according to claim 5, characterized in that: The feedback calculation processing is performed on the evaluation and analysis results, the standard motion trajectory information and the upper limb motion trajectory information set to obtain a feedback output value, including: When the evaluation and analysis result is difficult to complete, using the first feedback model, the upper limb motion trajectory information set and the standard motion trajectory information are calculated and processed to obtain a feedback output value; When the evaluation and analysis result is that it can be completed, the upper limb motion trajectory information set and the standard motion trajectory information are calculated and processed using the second feedback model to obtain a feedback output value.

10. The feedback processing method for upper limb rehabilitation training according to claim 9, characterized in that: The calculation expression of the first feedback model is: in, represents the loss function, δ and η are the preset first adjustment factor and second adjustment factor, b max is the maximum value of the mean vector, b p is the pth element of the mean vector, m is the row dimension of the difference matrix, v p It is the pth element of the time-aligned standard motion trajectory information.

Images