An electromyogram action recognition and muscle fatigue monitoring bracelet and a monitoring method thereof

Abstract

The application provides a myoelectric action recognition and muscle fatigue monitoring bracelet and a monitoring method thereof, which comprises a watch dial and a watchband connected with the watch dial, the inner side of the watchband towards the skin is provided with myoelectric collection electrodes for collecting surface myoelectric signals; a main control circuit board is arranged inside the watch dial, a main control chip and a six-axis inertial measurement unit for collecting wrist movement data are integrated on the main control circuit board, the six-axis inertial measurement unit contains an accelerometer and a gyroscope; the main control chip is electrically connected with the myoelectric collection electrodes and the six-axis inertial measurement unit respectively, the main control chip receives the original surface myoelectric signals acquired by the myoelectric collection electrodes and the wrist movement data acquired by the six-axis inertial measurement unit, and after pretreatment, feature extraction and action state determination are performed to obtain muscle action recognition results and muscle fatigue state evaluation results. Precise recognition of myoelectric action and precise evaluation of muscle fatigue state are realized.

Description

Technical Field

[0001] This invention relates to the field of wristbands, and more particularly to a wristband for electromyography (EMG) action recognition and muscle fatigue monitoring, and a monitoring method thereof. Background Technology

[0002] With the continuous development of human-computer interaction and limb rehabilitation monitoring technology, motion recognition devices based on surface electromyography (EMG) signals have been extensively studied. EMG bracelets can non-invasively collect forearm muscle electrical signals and have broad application prospects in rehabilitation training and smart wearable control scenarios. Currently, most mainstream consumer-grade smart bracelets on the market do not support high-precision surface EMG signal detection, and EMG acquisition devices still have many insurmountable defects in actual use.

[0003] In existing technologies, surface electromyography (EMG) signals are highly susceptible to interference from environmental and movement conditions. During the acquisition process, power frequency electromagnetic interference and motion artifacts caused by limb movements are easily introduced, resulting in a decrease in the signal-to-noise ratio of the original signal and significant signal distortion, which affects the reliability of subsequent action recognition results. Traditional EMG acquisition generally uses a wet electrode structure, which requires the application of conductive paste. After prolonged wear, the conductive medium gradually dries out, easily causing discomfort symptoms such as dry skin, itching, and allergies. Dry electrode solutions lack a reliable fit structure, and the electrodes are prone to slipping relative to the skin during wristband wear, which cannot meet the requirements for long-term continuous monitoring. Most recognition solutions rely solely on a single EMG signal for action determination, and when muscles enter a state of sustained exertion and fatigue, the amplitude of the EMG signal changes significantly.

[0004] Therefore, in order to address the shortcomings of existing technologies, a myoelectric bracelet that is comfortable to wear, can suppress signal interference, and can accurately identify target movements is proposed, along with a method for myoelectric movement recognition and muscle fatigue monitoring. Summary of the Invention

[0005] This invention provides a wristband for electromyography (EMG) action recognition and muscle fatigue monitoring, and its monitoring method. Its main purpose is to solve the technical problems of weak anti-interference ability, poor wearing stability, low accuracy of gesture recognition, and inaccurate muscle fatigue monitoring results in the prior art.

[0006] To achieve the above objectives, the present invention provides a wristband for electromyography (EMG) action recognition and muscle fatigue monitoring, comprising a watch face and a strap connected to the watch face. The inner side of the strap facing the skin is provided with EMG electrodes for collecting surface EMG signals. A main control circuit board is disposed inside the watch face, integrating a main control chip and a six-axis inertial measurement unit (SMU) for collecting wrist movement data. The SMU includes an accelerometer and a gyroscope. The main control chip is electrically connected to both the EMG electrodes and the SMU. The main control chip receives the raw surface EMG signals acquired by the EMG electrodes and the wrist movement data acquired by the SMU, performs preprocessing, feature extraction, and motion state determination to obtain muscle action recognition results and muscle fatigue state assessment results.

[0007] Optionally, the electromyography (EMG) acquisition electrodes are multi-channel array electrodes.

[0008] Optionally, the dial is equipped with a touch screen, which is electrically connected to the main control chip, and receives and visualizes the muscle movement recognition results and the muscle fatigue state assessment results.

[0009] Optionally, the watch strap also has an adaptive elastic support on the inside of the skin.

[0010] To address the above problems, the present invention also provides a method for electromyographic movement recognition and muscle fatigue monitoring, the method comprising: Raw surface electromyography (EMG) signals are acquired using EMG acquisition electrodes, and wrist motion data are acquired using a six-axis inertial measurement unit. The wrist motion data includes the change in acceleration in the wrist swing direction and the angular velocity of the wrist. The original surface electromyography (EMG) signal is processed by high-pass filtering, notch filter filtering, full-wave rectification, and moving average filtering to obtain the target surface EMG signal. Feature extraction is performed on the target surface EMG signal to obtain time-domain features representing the degree of muscle activation, and the integral EMG value representing the degree of muscle fatigue is calculated. Muscle movement is determined based on the time-domain characteristics and wrist movement data, and muscle fatigue state is determined based on the continuous changing trend of the integrated electromyography value.

[0011] Optionally, the process of obtaining the target surface electromyography (EMG) signal from the original surface EMG signal through high-pass filtering, notch filtering, full-wave rectification, and moving average filtering, and then extracting features from the target surface EMG signal, includes: The original surface electromyography signal is subjected to high-pass filtering based on a filter with a preset cutoff frequency; Select the appropriate narrowband filter to perform notch filtering on the original electromyography signal after high-pass filtering to obtain the surface electromyography signal. The surface electromyography (EMG) signal is subjected to full-wave rectification to convert all of the surface EMG signals into positive amplitude signals to obtain rectified surface EMG signals. The rectified surface EMG signals are then subjected to moving average filtering through a preset filtering window width to obtain the target surface EMG signal. According to the signal analysis window of the preset length, sampling points on the electromyography signal of the target surface are selected and the electromyography amplitude at the sampling points is obtained. Based on the electromyography amplitude, time-domain features including root mean square value, number of zero crossings and number of slope sign changes are calculated, and time-domain integration is performed on the electromyography amplitude of the sampling points to obtain the integrated electromyography value.

[0012] Optionally, the process of determining the fist-clenching action based on the root mean square value and the change in acceleration in the wrist swing direction in the wrist movement data, and determining the muscle fatigue state based on the continuous changing trend of the integrated electromyography value, includes: The wrist spatial posture features are obtained by offset correction of the change in acceleration in the wrist swing direction based on the posture angular velocity in the wrist motion data; the temporal features and the wrist spatial posture features are spliced ​​and fused to obtain a fused feature vector, which is then input into a pre-trained muscle action classification and recognition model to output muscle action recognition results; when the integrated electromyography value continues to rise and the rise duration exceeds a preset time, the muscle is determined to be in a fatigue state, otherwise it is determined to be in a non-fatigue state.

[0013] Optionally, the pre-training process of the muscle action classification and recognition model includes: A training sample set is obtained, which includes multiple samples. Each sample contains a fusion feature vector for the muscle movement corresponding to the electromyographic signal on the target surface. The samples are labeled with the true value of the movement category. The fusion feature vector is obtained by concatenating and fusing temporal features and wrist spatial posture features. The initial muscle movement classification and recognition model is trained using the training sample set. The initial model includes a feature normalization module, a multi-branch feature fusion module, and a classification mapping output module. The multi-branch feature fusion module comprises an electromyography (EMG) feature encoding layer, a posture feature encoding layer, and a cross-attention fusion layer. The feature normalization module normalizes the fused feature vector, then the EMG feature encoding layer performs independent feature compression encoding on the EMG temporal features to obtain an EMG feature vector, and the posture feature encoding layer encodes the wrist spatial posture features to obtain a posture feature vector. The cross-attention fusion layer performs cross-modal attention weighted fusion on the EMG feature vector and the posture feature vector to obtain a fused enhanced feature vector. The model is then input into the classification mapping output module to output a predicted muscle movement category. The muscle movement classification and recognition model is obtained by constructing a cross-entropy loss function using the predicted values ​​and the true values ​​of the muscle movement categories, and then updating the parameters of the initial muscle movement classification and recognition model with the goal of minimizing the cross-entropy loss function.

[0014] To address the above problems, the present invention also provides an electronic device, the electronic device comprising: At least one processor; and, A memory communicatively connected to the at least one processor; wherein, The memory has a computer program that can be executed by the at least one processor to enable the at least one processor to perform the above-described electromyographic movement recognition and muscle fatigue monitoring method.

[0015] To address the aforementioned problems, the present invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements any of the above-described electromyographic movement recognition and muscle fatigue monitoring methods.

[0016] The beneficial effects of this invention are as follows: This invention provides a wristband for electromyography (EMG) action recognition and muscle fatigue monitoring, along with its monitoring method. It combines surface EMG signal acquisition with a six-axis inertial measurement unit (IMU) to achieve deep fusion of multi-dimensional physiological and motion data, overcoming the shortcomings of single sensors in complex motion scenarios, such as low recognition rate and susceptibility to interference. This improves the accuracy of action recognition and provides real-time, accurate feedback on muscle fatigue status. The wristband employs a multi-channel array electrode design with an adaptive elastic support, ensuring close contact between the electrodes and the skin, reducing motion artifact interference, and improving signal acquisition quality and wearing comfort. Through preprocessing, feature extraction, and action state determination steps, it achieves accurate recognition of EMG actions and precise assessment of muscle fatigue status.

[0017] Additional advantages, objects, and features of the invention will be set forth in part in the description which follows, and will also become apparent in part to those skilled in the art upon studying the description, or may be learned by practice of the invention. The objects and other advantages of the invention can be realized and obtained by means of the structures specifically pointed out in the description and drawings.

[0018] Those skilled in the art will understand that the objectives and advantages achievable with the present invention are not limited to those specifically described above, and that the above and other objectives achievable with the present invention will become clearer from the following detailed description. Attached Figure Description

[0019] Figure 1This is a schematic diagram of the structure of a wristband for electromyography recognition and muscle fatigue monitoring according to an embodiment of the present invention.

[0020] Figure 2 This is a flowchart illustrating the electromyographic movement recognition and muscle fatigue monitoring method according to an embodiment of the present invention.

[0021] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0022] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the embodiments and accompanying drawings. Here, the illustrative embodiments and descriptions of this invention are used to explain the invention, but are not intended to limit the invention.

[0023] It should also be noted that, in order to avoid obscuring the invention with unnecessary details, only the structures and / or processing steps closely related to the solution according to the invention are shown in the accompanying drawings, while other details that are not closely related to the invention are omitted.

[0024] It should be emphasized that the term "including / comprises" as used herein refers to the presence of a feature, element, step, or component, but does not exclude the presence or addition of one or more other features, elements, steps, or components.

[0025] It should also be noted that, unless otherwise specified, the term "connection" in this article can refer not only to a direct connection, but also to an indirect connection involving an intermediary.

[0026] In the following description, embodiments of the invention will be illustrated with reference to the accompanying drawings. In the drawings, the same reference numerals represent the same or similar parts, or the same or similar steps.

[0027] Figure 1 This is a schematic diagram of the structure of a wristband for electromyography (EMG) action recognition and muscle fatigue monitoring according to an embodiment of the present invention. This application provides an EMG action recognition and muscle fatigue monitoring wristband, including a watch face and a strap connected to the watch face. The inner side of the strap facing the skin has EMG acquisition electrodes for collecting surface EMG signals. A main control circuit board is disposed inside the watch face, integrating a main control chip and a six-axis inertial measurement unit (SMU) for collecting wrist movement data. The SMU includes an accelerometer and a gyroscope. The main control chip is electrically connected to both the EMG acquisition electrodes and the SMU. The main control chip receives the raw surface EMG signals acquired by the EMG acquisition electrodes and the wrist movement data acquired by the SMU. After preprocessing, feature extraction and motion state determination are performed to obtain muscle action recognition results and muscle fatigue state assessment results.

[0028] Furthermore, the six-axis inertial measurement unit integrates an accelerometer and a gyroscope, outputting wrist motion data including the change in acceleration in the wrist swing direction and the attitude angular velocity. The change in acceleration in the wrist swing direction is used for joint action verification, and the attitude angular velocity corrects for data errors in the change in acceleration in the wrist swing direction caused by the wearer's posture deviation.

[0029] In this embodiment of the invention, the electromyography (EMG) acquisition electrodes are multi-channel array electrodes. In this embodiment, an adaptive elastic support is also provided on the inner side of the watchband facing the skin. Specifically, the multi-channel array electrodes are distributed on the inner side of the watchband and cover the ulnar wrist flexor muscle region and the radial wrist extensor muscle region; the adaptive elastic support ensures that the multi-channel array electrodes fit snugly against the skin, unaffected by arm size, avoiding electrode slippage and sudden changes in contact impedance caused by loosening of the wristband, ensuring the continuity of EMG signal acquisition; the multi-channel array electrodes are arranged in sections on the inner side of the watchband, with electrodes conforming to the ulnar wrist flexor muscle surface position for acquiring EMG signals of fist clenching and wrist flexion movements, and electrodes conforming to the radial wrist extensor muscle region for acquiring EMG signals of palm extension and wrist extension movements; the adaptive elastic support includes a silicone base layer and a built-in spring steel sheet, with the multi-channel array electrodes located on the protrusions of the silicone base layer. When the watchband is tightened, the protrusions are deformed by pressure to conform to different skin types.

[0030] In this embodiment of the invention, the dial is equipped with a touch screen, which is electrically connected to the main control chip. The touch screen receives and visualizes muscle movement recognition results and muscle fatigue state assessment results. Specifically, in addition to displaying the muscle movement recognition results and muscle fatigue state assessment results in real time, the touch screen also allows modification of various preset parameters via touch control.

[0031] Figure 2 This is a flowchart illustrating an embodiment of the electromyography (EMG) action recognition and muscle fatigue monitoring method according to the present invention. Furthermore, the present invention also provides an EMG action recognition and muscle fatigue monitoring method, comprising steps S101 to S103: Step S101: Acquire raw surface electromyography signals through electromyography acquisition electrodes, and acquire wrist motion data through a six-axis inertial measurement unit; wrist motion data includes the change in acceleration in the wrist swing direction and the angular velocity of the posture.

[0032] Step S102: Obtain the target surface electromyography signal by high-pass filtering, notch filter filtering, full-wave rectification and moving average filtering of the original surface electromyography signal; extract features from the target surface electromyography signal to obtain time-domain features to represent the degree of muscle activation and calculate the integral electromyography value to represent the degree of muscle fatigue. Step S103: Determine muscle movement based on temporal characteristics and wrist movement data, and determine muscle fatigue state based on the continuous trend of integrated electromyography value.

[0033] In this embodiment of the invention, the process of obtaining the target surface electromyography (EMG) signal from the original surface EMG signal through high-pass filtering, notch filtering, full-wave rectification, and moving average filtering includes: The original surface electromyography signal is high-pass filtered according to a filter with a preset cutoff frequency.

[0034] Select the appropriate narrowband filter to perform notch filtering on the original electromyography (EMG) signal after high-pass filtering to obtain the surface EMG signal.

[0035] The surface electromyography (EMG) signal is subjected to full-wave rectification to convert all surface EMG signals into positive amplitude signals to obtain rectified surface EMG signals. The rectified surface EMG signals are then subjected to moving average filtering through a preset filtering window width to obtain the target surface EMG signal.

[0036] Based on the signal analysis window of a preset length, sampling points on the electromyography (EMG) signal of the target surface are selected and the EMG amplitude at the sampling points is obtained. Based on the EMG amplitude, time-domain features including root mean square value, number of zero crossings and number of slope sign changes are calculated, and time-domain integration is performed on the EMG amplitude of the sampling points to obtain the integrated EMG value.

[0037] Specifically, the original surface electromyography (EMG) signal is high-pass filtered using a high-pass filter with a fixed cutoff frequency to remove low-frequency motion artifacts caused by large limb movements. A notch filter is then used to perform power frequency notch filtering on the high-pass filtered original EMG signal to remove electromagnetic interference and obtain a noise-reduced EMG signal. Full-wave rectification converts the bidirectional alternating EMG signal into a positive amplitude signal to obtain a rectified EMG signal. A fixed-window moving average filter is then used to obtain the target EMG signal. Furthermore, the root mean square (RMS) value is calculated using the following expression: The expression for calculating the integral electromyographic value is: ;in, Indicates the total number of sampling points. Indicates the first signal in the signal analysis window The electromyography amplitude at each sampling point and Not less than 0; In this embodiment of the invention, the process of determining fist-clenching action based on the root mean square value and the change in acceleration in the wrist swing direction in wrist motion data, and determining muscle fatigue state based on the continuous changing trend of integrated electromyography values, includes: The wrist spatial posture features are obtained by offset correction of the change in acceleration in the wrist swing direction based on the attitude angular velocity in the wrist motion data. The temporal features and the wrist spatial posture features are then concatenated and fused to obtain a fused feature vector, which is input into a pre-trained muscle action classification and recognition model to output muscle action recognition results. When the integrated electromyography value continuously increases and the increase duration exceeds a preset time, the muscle is determined to be in a fatigue state; otherwise, it is determined to be in a non-fatigue state. Specifically, the attitude angular velocity in the wrist motion data is used to perform attitude offset correction on the acceleration data, and wrist spatial posture features including pitch angle, roll angle, and yaw angle are calculated.

[0038] In this embodiment of the invention, the pre-training process of the muscle action classification and recognition model includes: Obtain a training sample set, which includes multiple samples. Each sample contains a fusion feature vector for the muscle action corresponding to the electromyographic signal on the target surface. The samples are labeled with the true value of the action category. The fusion feature vector is obtained by concatenating and fusing temporal features and wrist spatial pose features. An initial muscle movement classification and recognition model is trained using a training sample set. The initial muscle movement classification and recognition model includes a feature normalization module, a multi-branch feature fusion module, and a classification mapping output module. The multi-branch feature fusion module includes an electromyography (EMG) feature encoding layer, a posture feature encoding layer, and a cross-attention fusion layer. After normalizing the fused feature vector, the EMG feature encoding layer performs independent feature compression encoding on the EMG temporal features to obtain the EMG feature vector. The posture feature encoding layer encodes the wrist spatial posture features to obtain the posture feature vector. The cross-attention fusion layer performs cross-modal attention weighted fusion on the EMG feature vector and the posture feature vector to obtain the fused enhanced feature vector. The input to the classification mapping output module outputs the predicted value of the muscle movement category. A cross-entropy loss function is constructed by combining the predicted and actual values ​​of muscle movement categories. The parameters of the initial muscle movement classification and recognition model are then updated with the goal of minimizing the cross-entropy loss function to obtain the final muscle movement classification and recognition model.

[0039] On the other hand, embodiments of the present invention also provide an electronic device, the electronic device including at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory has a computer program executable by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the above-described electromyographic movement recognition and muscle fatigue monitoring method.

[0040] On the other hand, embodiments of the present invention also provide a computer-readable storage medium storing at least one computer program, which is executed by a processor in an electronic device to implement the above-described electromyographic movement recognition and muscle fatigue monitoring methods.

[0041] The present invention will now be described with reference to a specific embodiment: This invention proposes a wristband for electromyographic movement recognition and muscle fatigue monitoring, and its monitoring method. A multi-channel array of electrodes is distributed on the inner side of the wristband, covering the ulnar wrist flexor and radial wrist extensor regions. An adaptive elastic support is used to ensure that the electrodes are always in close contact with the skin and are not affected by the size of the arm. A six-axis inertial measurement unit (IMU) including an accelerometer and a gyroscope is used to assist in the determination of movement based on electromyographic signals.

[0042] The temporal characteristics of surface electromyography (sEMG) signals are fused with the spatial attitude data of the six-axis inertial measurement unit and input into a support vector machine model network for classification. The electrodes are Ag / AgCl dry electrodes or conductive fabrics with a serrated surface to increase the contact area and reduce the impedance to less than 10kΩ. The analog front-end (AFE) uses an ADS1299 with a sampling rate of 1000Hz and a gain of 6x. The main control chip (MCU) uses an ESP32 to run the RTOS system. The power supply is a 250mAh lithium battery that supports magnetic charging, providing more than 16 hours of continuous monitoring.

[0043] The signal preprocessing process includes: removing motion artifacts by high-pass filtering with a cutoff frequency of 20Hz; removing power frequency interference by notch filters with a 50Hz / 60Hz frequency; and performing full-wave rectification and moving average filtering with a window width of 50ms.

[0044] The feature extraction process includes: calculating the root mean square (RMS) value used to assess muscle activation; and calculating the integral electromyography (iEMG) value used to assess muscle fatigue.

[0045] When judging a fist-clenching action, if the root mean square value is greater than threshold A and the change in acceleration in the direction of wrist swing is greater than threshold B, it is judged as a fist-clenching action; if the integrated electromyography value continues to rise and the duration is greater than 5 seconds, it is judged as a state of muscle fatigue.

[0046] In summary, this invention provides a wristband for electromyography (EMG) action recognition and muscle fatigue monitoring, and a monitoring method thereof. The wristband includes a dial and a strap connected to the dial. The inner side of the strap facing the skin is provided with EMG electrodes for collecting surface EMG signals. A main control circuit board is disposed inside the dial, integrating a main control chip and a six-axis inertial measurement unit (SMU) for collecting wrist movement data. The SMU includes an accelerometer and a gyroscope. The main control chip is electrically connected to both the EMG electrodes and the SMU. The main control chip receives the raw surface EMG signals acquired by the EMG electrodes and the wrist movement data acquired by the SMU. After preprocessing, feature extraction and motion state determination are performed to obtain muscle action recognition results and muscle fatigue state assessment results.

[0047] This invention also provides an electronic device, the electronic device comprising: At least one processor; and, A memory communicatively connected to the at least one processor; wherein, The memory has a computer program that can be executed by the at least one processor to enable the at least one processor to perform the above-described electromyographic movement recognition and muscle fatigue monitoring method.

[0048] This invention also provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the electromyographic movement recognition and muscle fatigue monitoring methods described above. The computer-readable storage medium can be a tangible storage medium, such as random access memory (RAM), main memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, floppy disks, hard disks, removable storage disks, CD-ROMs, or any other form of storage medium known in the art.

[0049] Those skilled in the art will understand that the exemplary components, systems, and methods described in conjunction with the embodiments disclosed herein can be implemented in hardware, software, or a combination of both. Whether implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this invention. When implemented in hardware, it can be, for example, electronic circuits, application-specific integrated circuits (ASICs), appropriate firmware, plug-ins, function cards, etc. When implemented in software, the elements of this invention are programs or code segments used to perform the desired tasks. The programs or code segments can be stored in a machine-readable medium or transmitted over a transmission medium or communication link via data signals carried in a carrier wave.

[0050] It should be clarified that the present invention is not limited to the specific configurations and processes described above and shown in the figures. For the sake of brevity, detailed descriptions of known methods are omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of the present invention is not limited to the specific steps described and shown. Those skilled in the art can make various changes, modifications, and additions, or change the order of steps, after understanding the spirit of the present invention.

[0051] In this invention, features described and / or illustrated for one embodiment may be used in the same or similar manner in one or more other embodiments, and / or combined with or in place of features of other embodiments.

[0052] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, various modifications and variations of the embodiments of the present invention are possible. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A wristband for electromyography (EMG) action recognition and muscle fatigue monitoring, comprising a watch face and a strap connected to the watch face, characterized in that, The watch band has electromyography (EMG) electrodes on its inner side facing the skin for collecting surface EMG signals. A main control circuit board is located inside the watch face, integrating a main control chip and a six-axis inertial measurement unit (SMU) for collecting wrist movement data. The SMU includes an accelerometer and a gyroscope. The main control chip is electrically connected to both the EMG electrodes and the SMU. The main control chip receives the raw surface EMG signals acquired by the EMG electrodes and the wrist movement data acquired by the SMU. After preprocessing, it performs feature extraction and motion state determination to obtain muscle movement recognition results and muscle fatigue state assessment results.

2. The electromyography (EMG) action recognition and muscle fatigue monitoring wristband according to claim 1, characterized in that, The electromyography (EMG) acquisition electrodes are multi-channel array electrodes.

3. The electromyography (EMG) action recognition and muscle fatigue monitoring wristband according to claim 1, characterized in that, The dial is equipped with a touch screen, which is electrically connected to the main control chip. It receives the muscle movement recognition results and the muscle fatigue state assessment results and displays them visually.

4. The electromyography (EMG) action recognition and muscle fatigue monitoring wristband according to claim 1, characterized in that, The watch strap also features an adaptive elastic support on the inner side facing the skin.

5. A method for electromyographic movement recognition and muscle fatigue monitoring, characterized in that, The method includes: Raw surface electromyography (EMG) signals are acquired using EMG acquisition electrodes, and wrist motion data are acquired using a six-axis inertial measurement unit. The wrist motion data includes the change in acceleration in the wrist swing direction and the angular velocity of the wrist. The original surface electromyography (EMG) signal is processed by high-pass filtering, notch filter filtering, full-wave rectification, and moving average filtering to obtain the target surface EMG signal. Feature extraction is performed on the target surface EMG signal to obtain time-domain features representing the degree of muscle activation, and the integral EMG value representing the degree of muscle fatigue is calculated. Muscle movement is determined based on the time-domain characteristics and wrist movement data, and muscle fatigue state is determined based on the continuous changing trend of the integrated electromyography value.

6. The method for electromyographic movement recognition and muscle fatigue monitoring according to claim 5, characterized in that, The process of obtaining the target surface electromyography (EMG) signal from the original surface EMG signal through high-pass filtering, notch filtering, full-wave rectification, and moving average filtering, and then extracting features from the target surface EMG signal, includes: The original surface electromyography signal is subjected to high-pass filtering based on a filter with a preset cutoff frequency; Select the appropriate narrowband filter to perform notch filtering on the original electromyography signal after high-pass filtering to obtain the surface electromyography signal. The surface electromyography (EMG) signal is subjected to full-wave rectification to convert all of the surface EMG signals into positive amplitude signals to obtain rectified surface EMG signals. The rectified surface EMG signals are then subjected to moving average filtering through a preset filtering window width to obtain the target surface EMG signal. According to the signal analysis window of the preset length, sampling points on the electromyography signal of the target surface are selected and the electromyography amplitude at the sampling points is obtained. Based on the electromyography amplitude, time-domain features including root mean square value, number of zero crossings and number of slope sign changes are calculated, and time-domain integration is performed on the electromyography amplitude of the sampling points to obtain the integrated electromyography value.

7. The method for electromyographic movement recognition and muscle fatigue monitoring according to claim 5, characterized in that, The process of determining muscle movement based on the time-domain features and wrist movement data, and determining muscle fatigue state based on the continuous trend of the integrated electromyography value, includes: The wrist spatial posture features are obtained by offset correction of the change in acceleration in the wrist swing direction based on the posture angular velocity in the wrist motion data; the temporal features and the wrist spatial posture features are spliced ​​and fused to obtain a fused feature vector, which is then input into a pre-trained muscle action classification and recognition model to output muscle action recognition results; when the integrated electromyography value continues to rise and the rise duration exceeds a preset time, the muscle is determined to be in a fatigue state, otherwise it is determined to be in a non-fatigue state.

8. The method for electromyographic movement recognition and muscle fatigue monitoring according to claim 7, characterized in that, The pre-training process of the muscle action classification and recognition model includes: A training sample set is obtained, which includes multiple samples. Each sample contains a fusion feature vector for the muscle movement corresponding to the electromyographic signal on the target surface. The samples are labeled with the true value of the movement category. The fusion feature vector is obtained by concatenating and fusing temporal features and wrist spatial posture features. The initial muscle movement classification and recognition model is trained using the training sample set. The initial model includes a feature normalization module, a multi-branch feature fusion module, and a classification mapping output module. The multi-branch feature fusion module comprises an electromyography (EMG) feature encoding layer, a posture feature encoding layer, and a cross-attention fusion layer. The feature normalization module normalizes the fused feature vector, then the EMG feature encoding layer performs independent feature compression encoding on the EMG temporal features to obtain an EMG feature vector, and the posture feature encoding layer encodes the wrist spatial posture features to obtain a posture feature vector. The cross-attention fusion layer performs cross-modal attention weighted fusion on the EMG feature vector and the posture feature vector to obtain a fused enhanced feature vector. The model is then input into the classification mapping output module to output a predicted muscle movement category. The muscle movement classification and recognition model is obtained by constructing a cross-entropy loss function using the predicted values ​​and the true values ​​of the muscle movement categories, and then updating the parameters of the initial muscle movement classification and recognition model with the goal of minimizing the cross-entropy loss function.

9. An electronic device, characterized in that, The electronic device includes: At least one processor; A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor to enable the at least one processor to perform the electromyographic movement recognition and muscle fatigue monitoring method as described in any one of claims 5 to 8.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the computer program implements the electromyographic movement recognition and muscle fatigue monitoring method as described in any one of claims 5 to 8.

Images