Massage chair real-time control method and system based on electromyographic signals and computer medium

Abstract

The invention discloses a massage chair real-time control method and system based on electromyographic signals and a computer medium. The method comprises the steps that the electromyographic signals of a current massage user are obtained; extracting time domain features, frequency domain features and geometric features of the electromyographic signals; obtaining a real-time muscle state according to the time domain feature, the frequency domain feature and the geometric feature of the electromyographic signal; and controlling the massage strength and rhythm of the massage chair according to the real-time muscle state and the historical massage data of the current massage user. By analyzing the long-term muscle state and massage record of the user, an instant personalized massage scheme is generated for each user. Therefore, the actual requirements of the user can be better met, the system can provide more accurate and personalized services for the user, and the optimal user experience can be achieved.

Description

Technical Field

[0001] The present invention relates to the field of intelligent massage technology, and more specifically, to a real-time control method, system and computer medium of a massage chair based on electromyographic signals. Background Art

[0002] With the accelerated pace of modern life, the increase in work pressure, and people's pursuit of healthy life, massage chairs are becoming more and more popular as a convenient home health device. Traditional massage chairs usually preset fixed massage modes in terms of massage, such as vibration, rolling, tapping, etc. These modes are fixed for the massage site, strength and rhythm, and cannot be adjusted according to the individual differences and real-time needs of users. The rise of intelligent control massage chairs is the product of the combination of modern technology and people's pursuit of a healthy and comfortable life. Compared with traditional massage chairs, intelligent control massage chairs integrate electronic technology, sensor technology and artificial intelligence algorithms to provide a more precise and personalized massage experience. Intelligent control massage chairs can adjust the massage program through artificial intelligence algorithms according to user preferences and usage habits to achieve truly personalized services.

[0003] The purpose of the existing research on the evaluation of the massage chair effect of electromyographic signals is to analyze the characteristics of electromyographic signal changes through surface electromyographic tests, so as to evaluate the massage effect of massage chairs. During massage chair massage, the electromyographic signals of the massaged parts of the subjects are detected and recorded. According to the electromyographic signals, the four common indicators of electromyography are calculated: surface electromyographic absolute value integral (iEMG), root mean square value (RMS), mean power frequency (MPF) and median frequency (MF), and then the change characteristics of each electromyographic indicator throughout the massage are analyzed and compared. The change trend of the four electromyographic indicators throughout the massage reflects whether the massage chair massage can relieve muscle fatigue. This method evaluates the effect of massage based on the four common indicators of electromyographic signals. The defect is that this method still cannot provide users with more accurate and personalized services. This method can provide better guidance for the next massage experience, but there is still no way to provide a more accurate massage experience for each different user in real time. The real-time performance is poor and there are limitations. In addition, there may be limitations in the choice of processing methods for electromyographic signals. The four commonly used indicators (iEMG, RMS, MPF, MF) calculated may not fully reflect the full picture of muscle fatigue, and there is a risk of ignoring other important signal characteristics. Finally, the data analysis method may be subjective or not objective enough, and a more rigorous and scientific data analysis method is needed to support the credibility and scientificity of the research conclusions.

[0004] In another prior art, how to find sEMG signal features that can characterize muscle fatigue is discussed. By discussing the generation mechanism, characteristics and noise problems of sEMG signals, acquisition, preprocessing, experimental design and subjective fatigue perception division are proposed. Based on the frequency domain range characteristics of sEMG signals and the time-frequency analysis method, the geometric feature extraction method is improved, and the sEMG signal is analyzed from a two-dimensional perspective to improve the efficiency of muscle fatigue recognition. The chaotic characteristics of sEMG signals are analyzed by phase space reconstruction, and geometric features are extracted to compare the changes before and after the muscle fatigue characteristics. However, this method belongs to a non-real-time detection system, and there may be a certain delay, resulting in insufficient monitoring feedback on muscle status, and it is also impossible to track changes in muscle status in real time. It is usually necessary to process and analyze the data afterwards, and it is impossible to provide real-time muscle status information immediately, resulting in the lag of monitoring results. Changes in muscle status are also discovered immediately, so it may not be timely and effective when dealing with emergencies or emergencies. Summary of the invention

[0005] One of the purposes of the present invention is to provide a real-time control method for a massage chair based on electromyographic signals, so as to solve the technical problem that the existing technology cannot make real-time adjustments during the massage process; a further purpose of the present invention is to provide a real-time control system for a massage chair based on electromyographic signals; a third purpose of the present invention is to provide a computer medium.

[0006] In order to solve the above technical problems, the technical solution of the present invention is as follows:

[0007] A first aspect of the present invention provides a real-time control method for a massage chair based on electromyographic signals, comprising:

[0008] Get the electromyographic signal of the current massage user;

[0009] Extracting time domain features, frequency domain features and geometric features of the electromyographic signal;

[0010] Obtaining real-time muscle status according to the time domain characteristics, frequency domain characteristics and geometric characteristics of the electromyographic signal;

[0011] The massage intensity and rhythm of the massage chair are controlled according to the real-time muscle state and the historical massage data of the current massage user.

[0012] In the above technical means, by acquiring electromyographic signals in real time, the subtle changes of the user's muscles are captured in real time and analyzed quickly. According to the combination of the user's historical data and real-time data, the system will instantly adjust the strength, position and rhythm of the massage chair to ensure that the massage experience is fully consistent with the user's immediate needs. At the same time, the system will also intelligently and dynamically optimize the massage strategy based on the user's historical massage data. By analyzing the user's long-term muscle state and massage records, an instant personalized massage plan is generated for each user. Make it more in line with the user's actual needs. This closed-loop adaptive control process enables the system to not only make real-time adjustments during the current massage process, optimize the comfort during the massage process, and effectively improve the massage effect, but also learn and adapt to the user's long-term needs. The system can provide users with more precise and personalized services and continuously optimize massage strategies to achieve the best user experience.

[0013] Furthermore, before extracting the time domain features, frequency domain features and geometric features of the electromyographic signal, the electromyographic signal is subjected to denoising processing.

[0014] Further, the electromyographic signal is subjected to denoising, including:

[0015] Preprocessing the electromyographic signal, including removing a DC offset operation;

[0016] The preprocessed electromyographic signal is decomposed into multiple signal components using empirical mode decomposition, each signal component corresponding to a specific frequency component of the preprocessed electromyographic signal:

[0017]

[0018] In the formula, h i (t) is the difference between the preprocessed electromyographic signal and the mean of the upper and lower envelopes, and satisfies:

[0019] (1) The number of extreme values ​​and zero points must be the same or differ by at most one;

[0020] (2) At any point, the average value of the envelope represented by the local maximum and local minimum is zero;

[0021] Perform wavelet transform on each signal component to obtain the wavelet coefficient of each signal component;

[0022] According to the distribution characteristics of the wavelet coefficients of each signal component, the signal component containing noise is identified;

[0023] The other signal components except the signal component containing noise are added together to obtain the electromyographic signal after denoising.

[0024] Furthermore, the time domain features, frequency domain features and geometric features of the electromyographic signal include:

[0025] The time domain characteristics of the electromyographic signal include the root mean square value, integrated electromyographic value, variance, zero-crossing rate and muscle tension amplitude of the electromyographic signal;

[0026] The frequency domain characteristics of the electromyographic signal include power spectrum density, average power frequency, median frequency, median frequency and spectrum moment ratio of the electromyographic signal;

[0027] The geometrical features of the electromyographic signal include the area, circularity and centroid of the electromyographic signal.

[0028] Furthermore, according to the time domain characteristics, frequency domain characteristics and geometric characteristics of the electromyographic signal, the real-time muscle state is obtained, including:

[0029] The time domain features, frequency domain features and geometric features of the electromyographic signal are input into a preset EfficientNet network to obtain the real-time muscle state.

[0030] Furthermore, the muscle state includes a muscle fatigue state, a muscle tension state and a muscle recovery state.

[0031] Furthermore, according to the real-time muscle state and the historical massage data of the current massage user, the massage intensity and rhythm of the massage chair are controlled, including:

[0032] According to the real-time muscle state, the massage preferences under the real-time muscle state are searched in the historical massage data of the current massage user, wherein the massage preferences include massage intensity, massage technique, massage area and massage time, and the massage intensity and rhythm of the massage chair are set according to the massage preferences.

[0033] Furthermore, after setting the massage intensity and rhythm of the massage chair according to the massage preference, the massage preference is also adjusted according to the feedback of the current massage user.

[0034] The second aspect of the present invention provides a real-time control system for a massage chair based on electromyographic signals, comprising:

[0035] An electromyographic signal acquisition module, wherein the electromyographic signal acquisition module acquires the electromyographic signal of the current massage user;

[0036] An electromyographic signal processing module, wherein the electromyographic signal processing module extracts time domain features, frequency domain features and geometric features of the electromyographic signal;

[0037] A muscle state judgment module, wherein the muscle state judgment module obtains the real-time muscle state according to the time domain characteristics, frequency domain characteristics and geometric characteristics of the electromyographic signal;

[0038] A massage adjustment module controls the massage intensity and rhythm of the massage chair according to the real-time muscle state and the historical massage data of the current massage user.

[0039] A third aspect of the present invention provides a computer medium having a computer program stored thereon, and when the computer program is executed by a processor, the real-time control method of a massage chair based on electromyographic signals is implemented.

[0040] Compared with the prior art, the technical solution of the present invention has the following beneficial effects:

[0041] 1. Stronger real-time performance: Current research at home and abroad mainly relies on static data (i.e. data that has been collected, recorded or stored). This research model has high stability and the results are easy to evaluate and verify, but it has low time efficiency and applicability and is not conducive to the optimization of later algorithms. This lack of real-time research model cannot capture timely changes in fatigue status, thus limiting the in-depth understanding of the dynamic characteristics of muscle fatigue. The real-time control system can adjust the massage intensity and method of the massage chair in real time according to the user's electromyographic signal, so that users can get a more personalized and comfortable massage experience.

[0042] 2. Better user experience: The real-time control system can be adjusted according to the user's real-time feedback to provide a massage effect that better meets the user's needs, thereby improving the user's comfort and satisfaction

[0043] 3. The effect is more intuitive: Through the real-time control system, users can directly feel the stimulation and relaxation effect of the massage chair on the muscles, rather than just evaluating the massage effect through the analysis of surface electromyography signals.

[0044] 4. Multi-muscle state classification processing: At present, muscle fatigue is mostly classified into two categories through machine and deep learning classification algorithms, that is, distinguishing fatigue from non-fatigue. The multi-level classification of muscle fatigue has not been clearly refined, resulting in relatively low accuracy of existing multi-level fatigue classification technology.

[0045] 5. Feature superiority: Compared with the features extracted from existing electromyographic signals, this system shows significant innovation in electromyographic signal feature extraction. The existing technology mainly extracts basic activity segment feature values, such as electromyographic integral value, signal wave length and average frequency. Although these features can be used to detect muscle status, they cannot fully capture the complex physiological activity changes of muscles. In contrast, this system uses a richer feature extraction method, including time domain, frequency domain and geometric features, and extracts a total of 13 features, which can not only reflect the overall activity level of electromyographic signals (such as root mean square value, average power frequency), but also capture the shape changes of electromyographic signals through geometric features (such as signal area). These geometric features can deeply analyze the process of muscle fatigue, especially for nonlinear and non-stationary signals of muscle activity. This system combines empirical mode decomposition (EMD) and wavelet transform for denoising, which fundamentally improves the purity of the signal and the accuracy of feature extraction. The existing technology lacks comprehensive analysis of multi-dimensional features in feature extraction. Through these innovative means, this system makes the monitoring of muscle status more accurate and diversified, ensuring that more targeted judgment basis can be provided during real-time regulation.

[0046] 6. Improved real-time performance and feedback mechanism: Compared with the existing micro-current massage control method, this system is more innovative in real-time performance and intelligent feedback control. Most of the existing technologies dynamically adjust the micro-current massage parameters through physiological feedback data. Although they can provide certain personalized regulation, their regulation methods mainly rely on slow feedback of physiological state, lacking instant judgment of muscle state and intelligent regulation combined with historical data. This feedback mechanism is relatively lagging and cannot immediately respond to changes in the user's muscle state. It fails to achieve real-time monitoring using electromyographic signals, and cannot make precise adjustments based on the user's immediate muscle state. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 A schematic flow chart of a real-time control method of a massage chair based on electromyographic signals provided by an embodiment of the present invention;

[0048] Figure 2 A schematic diagram of a module of a real-time control system for a massage chair based on electromyographic signals provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0049] The drawings are for illustrative purposes only and should not be construed as limiting the present patent;

[0050] In order to better illustrate the present embodiment, some parts in the drawings may be omitted, enlarged or reduced, and do not represent the size of the actual product;

[0051] It is understandable to those skilled in the art that some well-known structures and their descriptions may be omitted in the drawings.

[0052] The technical solution of the present invention is further described below in conjunction with the accompanying drawings and embodiments.

[0053] Example 1

[0054] This embodiment provides a real-time control method for a massage chair based on electromyographic signals. Figure 1 As shown, including:

[0055] Get the electromyographic signal of the current massage user;

[0056] Extracting time domain features, frequency domain features and geometric features of the electromyographic signal;

[0057] Obtaining real-time muscle status according to the time domain characteristics, frequency domain characteristics and geometric characteristics of the electromyographic signal;

[0058] The massage intensity and rhythm of the massage chair are controlled according to the real-time muscle state and the historical massage data of the current massage user.

[0059] In a further embodiment, before extracting the time domain features, frequency domain features and geometric features of the electromyographic signal, the electromyographic signal is further subjected to denoising processing.

[0060] Since the power supply of the device will generate power frequency interference when detecting electromyographic signals, and the influence of motion artifacts caused by human muscle movement, and other human biological signals may be mixed into the electromyographic signals during the acquisition process, in order to obtain effective information of the EMG signal, it is necessary to filter it in this embodiment, and empirical mode decomposition is a very flexible and adaptable signal processing method that can effectively process non-stationary and nonlinear signals and is suitable for signal analysis and feature extraction in various fields. When processing electromyographic signals, empirical mode decomposition can better understand the characteristics and changes of muscle activity.

[0061] At the same time, wavelet transform is a powerful signal processing tool that can be used to analyze, process and compress various signals. It can be used to analyze and process various signals, especially good at processing non-stationary signals. It has a wide range of applications in signal processing, image processing, speech recognition, medical imaging and other fields.

[0062] Therefore, in a further embodiment, the electromyographic signal is subjected to denoising, including combining wavelet transform and empirical spectrum decomposition to denoise the electromyographic signal, and the specific steps are as follows:

[0063] Preprocessing the electromyographic signal, including removing a DC offset operation, etc., to improve the effect of subsequent decomposition;

[0064] The preprocessed electromyographic signal is decomposed into multiple signal components using empirical mode decomposition, each signal component corresponding to a specific frequency component of the preprocessed electromyographic signal:

[0065]

[0066] In the formula, h i (t) is the difference between the preprocessed electromyographic signal and the mean of the upper and lower envelopes, and satisfies:

[0067] (1) The number of extreme values ​​and zero points must be the same or differ by at most one;

[0068] (2) At any point, the average value of the envelope represented by the local maximum and local minimum is zero;

[0069] Perform wavelet transform on each signal component to obtain the wavelet coefficient of each signal component; in this embodiment, select a suitable wavelet function and the number of decomposition layers, and adjust them according to the characteristics of the signal and the type of noise;

[0070] According to the distribution characteristics of the wavelet coefficients of each signal component, the signal component containing noise is identified; in this embodiment, the signal component containing noise can be identified by observing the energy distribution of the wavelet coefficients, and the wavelet coefficient energy of the noise IMF is usually concentrated in the high-frequency part;

[0071] The other signal components except the signal component containing noise are added together to obtain the electromyographic signal after denoising.

[0072] In this embodiment, if the traditional denoising method is directly used, EMG signals often have nonlinear characteristics, and some traditional linear filtering algorithms may not be able to handle this nonlinear relationship well, and may also damage important information of the signal while removing noise, which will affect the subsequent real-time judgment of muscle status. At the same time, some complex denoising algorithms may require high computing costs, which may not be practical for EMG signal processing applications with high real-time requirements.

[0073] In this embodiment, muscle fatigue will lead to an increase in the degree of synchronization of the motor units, thereby significantly increasing the amplitude of the surface electromyographic signal; at the same time, the increase in fatigue will also lead to an increase in the activity of the low-frequency transmitting units and a decrease in the activity of the high-frequency units, causing the electromyographic signal spectrum to shift to the left. By analyzing these changes in electromyographic signals, the muscle fatigue process can be effectively studied. Therefore, in a real-time massage control system based on electromyographic signals, it is crucial to select and extract effective electromyographic features that reflect changes in muscle fatigue.

[0074] In a further embodiment, the time domain features, frequency domain features and geometric features of the electromyographic signal include:

[0075] The time domain characteristics of the electromyographic signal include the root mean square value, integrated electromyographic value, variance, zero-crossing rate and muscle tension amplitude of the electromyographic signal;

[0076] In this embodiment, in the time domain, the time domain features are obtained from the time representation of the original signal. The calculation is simple, and the changes in the signal amplitude and sign slope can be directly observed. It is a feature extraction method widely used in human movement. The extracted time domain features are: root mean square value, integrated electromyographic value, variance, zero crossing rate and muscle tension amplitude. Both the root mean square value and the integrated electromyographic value can reflect the change of amplitude in the time domain. The greater the exercise intensity, the greater the increase. This change is used to judge the muscle activation intensity and fatigue state. During the fatigue process, the zero crossing rate shows a downward trend, and the muscle tension amplitude shows an upward trend;

[0077] The frequency domain characteristics of the electromyographic signal include power spectrum density, average power frequency, median frequency, median frequency and spectrum moment ratio of the electromyographic signal;

[0078] In this embodiment, in the frequency domain, the signal is divided into multiple overlapping sub-segments by the Welch method, and each sub-segment is Fourier transformed, and finally the power spectra of all sub-segments are averaged to obtain the final estimation result. This can reduce the influence of noise and improve the accuracy of spectrum estimation. The extracted frequency domain features are: power spectrum density, average power frequency, median frequency, median frequency and spectrum moment ratio. As the degree of fatigue deepens, the average power frequency and median frequency will shift from high frequency to low frequency due to the influence of muscle fiber signal conduction speed, and the heavier the fatigue, the faster the shift. By analyzing the shape and scale changes of the spectral characteristics during the fatigue process, the time required for the muscle to reach deep fatigue can be predicted, providing sufficient and effective information for subsequent massage strategy adjustments.

[0079] The geometrical features of the electromyographic signal include the area, circularity and centroid of the electromyographic signal;

[0080] In this embodiment, in terms of geometry, geometric features can effectively describe the shape, waveform and change pattern of EMG signals. These features can depict the morphology, structure and distribution characteristics of the signal from different angles, and are not affected by the shape position, rotation and size changes, reflecting the most essential characteristics of the signal. By real-time monitoring of the geometric feature changes of EMG signals, signs of muscle fatigue can be discovered in time, so that corresponding intervention measures can be taken, such as adjusting the massage intensity, massage rhythm, etc. The extracted geometric features are: area, roundness and centroid.

[0081] By extracting time domain features (such as mean, variance), frequency domain features (such as power spectrum density), etc., we can fully understand the characteristics of the signal from different angles. Therefore, we selected the time domain features, frequency domain features and geometric features of the electromyographic signal, such as the following features:

[0082] Integrated EMG value:

[0083] Power Spectral Density:

[0084] area:

[0085] In a further embodiment, the real-time muscle state is obtained according to the time domain characteristics, frequency domain characteristics and geometric characteristics of the electromyographic signal, including:

[0086] The time domain features, frequency domain features and geometric features of the electromyographic signal are input into a preset EfficientNet network to obtain the real-time muscle state.

[0087] In this embodiment, EfficientNet is an efficient neural network architecture that aims to improve the efficiency of the model without sacrificing accuracy. Combining the scaling of network depth, network width and resolution, a set of optimized network structures are obtained through automatic search, so that the model has smaller parameters and computational complexity at the same accuracy. Due to its small number of parameters and computational complexity, the model is faster in the reasoning stage and can quickly process and classify real-time data streams. This is very important for application scenarios that require real-time performance. Its network structure is automatically searched and optimized to adapt to different data sets and task requirements. The network can be fine-tuned or transferred according to actual conditions to adapt to specific real-time data stream classification tasks. At the same time, the network performs well on large-scale data sets and has strong generalization ability and stability. Therefore, the obtained features and labels are input into the constructed EfficientNet network in advance, and a real-time muscle state classification and recognition model is trained.

[0088] In this embodiment, if other deep learning networks and machine learning algorithms are used, since the real-time control system of the massage chair based on electromyographic signals needs to process real-time data streams, if other networks are selected, it may result in a large number of model parameters and high computational complexity, which is not conducive to efficient classification processing in real-time data stream scenarios. At the same time, it will also lead to a slow classification speed for real-time data streams and cannot meet the system processing requirements. Due to the huge amount of real-time data, other algorithm networks may cause unstable performance, and may fluctuate or perform poorly in actual applications.

[0089] In a further embodiment, the muscle state includes a muscle fatigue state, a muscle tension state and a muscle recovery state.

[0090] In a further embodiment, the massage intensity and rhythm of the massage chair are controlled according to the real-time muscle state and the historical massage data of the current massage user, including:

[0091] According to the real-time muscle state, the massage preferences under the real-time muscle state are searched in the historical massage data of the current massage user, wherein the massage preferences include massage intensity, massage technique, massage area and massage time, and the massage intensity and rhythm of the massage chair are set according to the massage preferences.

[0092] In a specific embodiment, when the system detects that the user's muscles are in a fatigued state through electromyographic signals, the electromyographic signals appear as continuous low-frequency and high-amplitude activities. Combined with historical data, the system will look for the user's previous preferences when in a fatigued state. Usually, at this time, the user needs a gentler massage to relieve fatigue and soreness. Therefore, the system will intelligently adjust the strength of the massage chair, using slow and gentle massage or kneading movements, and focus on muscle groups where the user often feels fatigued, such as the shoulders, neck or waist. In addition, if historical data shows that the user prefers a longer massage time to relieve fatigue, the system will extend the massage time accordingly and maintain a lower strength throughout the process;

[0093] If the system detects that the muscles are tense or stiff, the electromyographic signals usually show high-frequency and irregular signal fluctuations. Combined with historical records, the system will find the user's preferences when the muscles are tense, such as strong massage modes or focusing on relaxing specific areas. At this time, the system will increase the strength of the massage chair, using fast and powerful tapping or pressing techniques, focusing on the back, waist and other parts that are prone to tension. At the same time, the system will appropriately increase the intensity and massage time based on historical data to ensure that the massage effectively relieves muscle tension.

[0094] Finally, based on the above control results and massage effects, the myoelectric control algorithm is further adjusted and optimized, and it is actually deployed in the embedded control system of the massage chair;

[0095] For users who have just experienced exercise or have been tired for a long time, the electromyographic signal may show a lower frequency but larger fluctuations, indicating that the muscles are in the recovery period. Combined with historical records, the system will understand that the user may need a massage to promote blood circulation and recovery in this situation. At this time, the system will adjust the massage chair to a moderate strength, using kneading, beating and other movements with a more even rhythm, especially focusing on massaging the user's frequently used muscle groups, such as the legs or waist. Based on the user's previous recovery preferences, the system may shorten the massage time and reduce the strength in the later stage, providing a gentle end massage to help the muscles recover better;

[0096] In a further embodiment, based on the accumulation of long-term historical data, the system can identify the parts of the user's muscles that are often tense or painful, and perform personalized massage strategies based on the user's preferences. For example, some users' shoulders, necks or lower backs are in a state of tension for a long time. The system will give priority to these areas and select the appropriate strength and massage mode based on past feedback records. At the same time, the system will automatically select appropriate techniques such as pushing, kneading, and beating based on the user's previous preferences for specific techniques in different states, ensuring that the user is provided with the most comfortable massage experience.

[0097] Example 2

[0098] This embodiment provides a real-time control system for a massage chair based on electromyographic signals. Figure 2 As shown, the system applies the real-time control method of the massage chair based on electromyographic signals described in Example 1, including:

[0099] An electromyographic signal acquisition module, wherein the electromyographic signal acquisition module acquires the electromyographic signal of the current massage user;

[0100] An electromyographic signal processing module, wherein the electromyographic signal processing module extracts time domain features, frequency domain features and geometric features of the electromyographic signal;

[0101] A muscle state judgment module, wherein the muscle state judgment module obtains the real-time muscle state according to the time domain characteristics, frequency domain characteristics and geometric characteristics of the electromyographic signal;

[0102] A massage adjustment module controls the massage intensity and rhythm of the massage chair according to the real-time muscle state and the historical massage data of the current massage user.

[0103] Example 3

[0104] This embodiment provides a computer medium, on which a computer program is stored. When the computer program is executed by a processor, the real-time control method of a massage chair based on electromyographic signals described in Embodiment 1 is implemented.

[0105] The same or similar reference numerals correspond to the same or similar components;

[0106] The terms used in the drawings to describe positional relationships are only used for illustrative purposes and should not be construed as limiting this patent;

[0107] Obviously, the above embodiments of the present invention are merely examples for clearly illustrating the present invention, and are not intended to limit the embodiments of the present invention. For those skilled in the art, other different forms of changes or modifications can be made based on the above description. It is not necessary and impossible to list all the embodiments here. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the claims of the present invention.

Claims

1. A real-time control method for a massage chair based on electromyographic signals, characterized in that: include: Get the electromyographic signal of the current massage user; Extracting time domain features, frequency domain features and geometric features of the electromyographic signal; Obtaining real-time muscle status according to the time domain characteristics, frequency domain characteristics and geometric characteristics of the electromyographic signal; The massage intensity and rhythm of the massage chair are controlled according to the real-time muscle state and the historical massage data of the current massage user.

2. The real-time control method of a massage chair based on electromyographic signals according to claim 1, characterized in that: Before extracting the time domain features, frequency domain features and geometric features of the electromyographic signal, the electromyographic signal is subjected to denoising processing.

3. The real-time control method of a massage chair based on electromyographic signals according to claim 1, characterized in that: The electromyographic signal is subjected to denoising processing, comprising: Preprocessing the electromyographic signal, including removing a DC offset operation; The preprocessed electromyographic signal is decomposed into multiple signal components using empirical mode decomposition, each signal component corresponding to a specific frequency component of the preprocessed electromyographic signal: In the formula, h i (t) is the difference between the preprocessed electromyographic signal and the mean of the upper and lower envelopes, and satisfies: (1) The number of extreme values ​​and zero points must be the same or differ by at most one; (2) At any point, the average value of the envelope represented by the local maximum and local minimum is zero; Perform wavelet transform on each signal component to obtain the wavelet coefficient of each signal component; According to the distribution characteristics of the wavelet coefficients of each signal component, the signal component containing noise is identified; The other signal components except the signal component containing noise are added together to obtain the electromyographic signal after denoising.

4. According to the real-time control method of a massage chair based on electromyographic signals according to claim 1, it is characterized in that: The time domain features, frequency domain features and geometric features of the electromyographic signal include: The time domain characteristics of the electromyographic signal include the root mean square value, integrated electromyographic value, variance, zero-crossing rate and muscle tension amplitude of the electromyographic signal; The frequency domain characteristics of the electromyographic signal include power spectrum density, average power frequency, median frequency, median frequency and spectrum moment ratio of the electromyographic signal; The geometrical features of the electromyographic signal include the area, circularity and centroid of the electromyographic signal.

5. The real-time control method of a massage chair based on electromyographic signals according to claim 1, characterized in that: According to the time domain characteristics, frequency domain characteristics and geometric characteristics of the electromyographic signal, the real-time muscle state is obtained, including: The time domain features, frequency domain features and geometric features of the electromyographic signal are input into a preset EfficientNet network to obtain the real-time muscle state.

6. The real-time control method of a massage chair based on electromyographic signals according to claim 1, characterized in that: The muscle state includes a muscle fatigue state, a muscle tension state and a muscle recovery state.

7. The real-time control method of a massage chair based on electromyographic signals according to claim 6, characterized in that: According to the real-time muscle state and the historical massage data of the current massage user, the massage intensity and rhythm of the massage chair are controlled, including: According to the real-time muscle state, the massage preferences under the real-time muscle state are searched in the historical massage data of the current massage user, wherein the massage preferences include massage intensity, massage technique, massage area and massage time, and the massage intensity and rhythm of the massage chair are set according to the massage preferences.

8. The real-time control method of a massage chair based on electromyographic signals according to claim 7, characterized in that: After setting the massage intensity and rhythm of the massage chair according to the massage preference, the massage preference is also adjusted according to the feedback of the current massage user.

9. A real-time control system for a massage chair based on electromyographic signals, characterized in that: include: An electromyographic signal acquisition module, wherein the electromyographic signal acquisition module acquires the electromyographic signal of the current massage user; An electromyographic signal processing module, wherein the electromyographic signal processing module extracts time domain features, frequency domain features and geometric features of the electromyographic signal; A muscle state judgment module, wherein the muscle state judgment module obtains the real-time muscle state according to the time domain characteristics, frequency domain characteristics and geometric characteristics of the electromyographic signal; A massage adjustment module controls the massage intensity and rhythm of the massage chair according to the real-time muscle state and the historical massage data of the current massage user.

10. A computer medium, characterized in that The computer medium stores a computer program, and when the computer program is executed by the processor, the real-time control method of the massage chair based on electromyographic signals according to any one of claims 1 to 8 is implemented.

Images