Fatigue recovery detection method and device, mattress and storage medium

Abstract

The embodiment of the invention discloses a fatigue recovery detection method and device, a mattress and a storage medium, the fatigue recovery detection method and device are applied to the mattress, the mattress comprises a physiological sign sensor, and the method comprises the steps that a target heart impact BCG signal, collected by the physiological sign sensor, of a user is acquired; extracting a feature set according to the target BCG signal, wherein the feature set comprises one or more physiological sign features and / or one or more original signal features; analyzing the feature set through a fatigue recovery detection model obtained through training to obtain a fatigue recovery level; the fatigue recovery level is used for indicating the fatigue recovery degree of the user; wherein the fatigue recovery detection model is obtained by training according to multiple groups of training data, and each group of training data comprises the sample electromyographic signals and the sample BCG signals collected in the same time period and the sample fatigue recovery level corresponding to the sample electromyographic signals, so that the fatigue recovery degree of the user can be accurately evaluated; and the accuracy of fatigue recovery detection is improved.

Description

Technical Field

[0001] The present application relates to the field of smart home technology, and in particular to a fatigue recovery detection method, device, mattress and storage medium. Background Art

[0002] As the pace of modern life accelerates, people are facing more and more work pressure and lifestyle changes, causing fatigue to become a common phenomenon. Long-term fatigue not only affects personal work efficiency and quality of life, but may also cause a series of health problems, such as cardiovascular disease and decreased immune system function. Therefore, effective monitoring and evaluation of fatigue recovery is of great significance to promoting public health.

[0003] Current fatigue recovery detection mainly relies on some wearable devices. Although these wearable devices can track sleep duration and basic sleep cycles, their accuracy and reliability are still limited for in-depth assessment of fatigue recovery during sleep. Summary of the invention

[0004] The embodiments of the present application disclose a fatigue recovery detection method, device, mattress and storage medium. The mattress can accurately evaluate the fatigue recovery degree of the user and improve the accuracy of fatigue recovery detection.

[0005] In a first aspect, an embodiment of the present application discloses a fatigue recovery detection method, which is applied to a mattress, wherein the mattress includes a physiological sign sensor, and the method includes:

[0006] Acquiring the target BCG signal of the user collected by the physiological sign sensor;

[0007] Extracting a feature set according to the target BCG signal, wherein the feature set includes one or more physiological sign features and / or one or more original signal features;

[0008] The fatigue recovery detection model obtained through training is used to analyze the feature set to obtain a fatigue recovery level; the fatigue recovery level is used to indicate the user's fatigue recovery degree; wherein the fatigue recovery detection model is trained based on multiple groups of training data, each group of the training data includes sample electromyography signals and sample BCG signals collected in the same time period, and sample fatigue recovery levels corresponding to the sample electromyography signals.

[0009] As an optional implementation, in the first aspect of the embodiment of the present application, the fatigue recovery detection model includes multiple decision trees, and the fatigue recovery detection model obtained by training analyzes the feature set to obtain a fatigue recovery level, including:

[0010] Analyzing the feature set through each decision tree to obtain a predicted fatigue recovery result output by each decision tree;

[0011] The predicted fatigue recovery results output by each decision tree are integrated to determine the fatigue recovery level.

[0012] As an optional implementation, in the first aspect of the embodiment of the present application, the one or more physiological sign characteristics include at least one of a heart rate characteristic, a breathing characteristic, and a body movement characteristic; and / or,

[0013] The one or more original signal features include at least one of the zero-crossing rate corresponding to one or more frequency bands in the target BCG signal and the energy corresponding to one or more frequency bands.

[0014] As an optional implementation, in the first aspect of the embodiment of the present application, the step of acquiring the target BCG signal of the user collected by the physiological sign sensor includes:

[0015] Acquire a first target BCG signal corresponding to a first time period of the user before falling asleep, and a second target BCG signal corresponding to a second time period of the user after falling asleep, collected by the physiological sign sensor;

[0016] The feature set includes: a first feature set corresponding to the first target BCG signal, and a second feature set corresponding to the second target BCG signal.

[0017] As an optional implementation, in the first aspect of the embodiment of the present application, the mattress further includes an electromyographic sensor; the sample electromyographic signal of each group of the training data includes a first sample electromyographic signal and a second sample electromyographic signal; before acquiring the target BCG signal of the user from the physiological sign sensor, the method further includes:

[0018] Collecting, by means of an electromyographic sensor, a first sample electromyographic signal corresponding to a third time period before the user falls asleep, and a second sample electromyographic signal corresponding to a fourth time period after the user falls asleep;

[0019] A sample fatigue recovery level is determined according to the first electromyographic signal and the second sample electromyographic signal.

[0020] As an optional implementation, in the first aspect of the embodiment of the present application, determining the sample fatigue recovery level according to the first electromyographic signal and the second sample electromyographic signal includes:

[0021] Calculating a plurality of first integrated electromyographic values ​​corresponding to the third time period according to the first sample electromyographic signal;

[0022] Calculate the average of the plurality of first integrated electromyographic values ​​to obtain a first average integrated electromyographic value;

[0023] Calculating a plurality of second integrated electromyographic values ​​corresponding to the fourth time period according to the second sample electromyographic signal;

[0024] Calculate the average value of the plurality of second integrated electromyographic values ​​to obtain a second average integrated electromyographic value;

[0025] The sample fatigue recovery level is determined according to the ratio of the second average integrated electromyographic value to the first average integrated electromyographic value.

[0026] As an optional implementation, in the first aspect of the embodiment of the present application, determining the sample fatigue recovery level according to the ratio of the second average integrated electromyographic value to the first average integrated electromyographic value includes:

[0027] Determine the target recovery level interval to which the ratio belongs;

[0028] Determine the sample fatigue recovery level corresponding to the target recovery level interval;

[0029] Among them, the larger the interval threshold corresponding to the target recovery level interval is, the worse the fatigue recovery degree indicated by the sample fatigue recovery level is.

[0030] A second aspect of the present application discloses a fatigue recovery detection device, which is applied to a mattress. The mattress includes a physiological sign sensor. The device includes:

[0031] A BCG signal acquisition module, used to acquire the target BCG signal of the user collected by the physiological sign sensor;

[0032] A feature extraction module, configured to extract a feature set according to the target BCG signal, wherein the feature set includes one or more physiological sign features and / or one or more original signal features;

[0033] An analysis module is used to analyze the feature set through a fatigue recovery detection model obtained through training to obtain a fatigue recovery level; the fatigue recovery level is used to indicate the user's fatigue recovery degree; wherein the fatigue recovery detection model is trained based on multiple groups of training data, each group of the training data includes sample electromyographic signals and sample BCG signals collected in the same time period, and sample fatigue recovery levels corresponding to the sample electromyographic signals.

[0034] A third aspect of the embodiment of the present application discloses a mattress, comprising:

[0035] A memory storing executable program code;

[0036] a processor coupled to the memory;

[0037] The processor calls the executable program code stored in the memory to execute the method described in any one of the above embodiments.

[0038] A fourth aspect of an embodiment of the present application discloses a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, the processor executes the method described in any one of the above embodiments.

[0039] Compared with the related art, the embodiments of the present application include at least the following beneficial effects:

[0040] The fatigue recovery detection method, device, mattress and storage medium disclosed in the embodiment of the present application obtain the target BCG signal of the user collected by the physiological sign sensor; extract a feature set according to the target BCG signal, the feature set includes one or more physiological sign features, and / or one or more original signal features; analyze the feature set through the fatigue recovery detection model obtained by training to obtain a fatigue recovery level; the fatigue recovery level is used to indicate the fatigue recovery degree of the user; wherein the fatigue recovery detection model is obtained by training based on multiple sets of training data, each set of training data includes sample electromyographic signals and sample BCG signals collected in the same time period, and the sample fatigue recovery level corresponding to the sample electromyographic signals. In the embodiment of the present application, the mattress can collect the target BCG signal of the user through the physiological sign sensor, and extract the feature set corresponding to the target BCG signal, and obtain the fatigue recovery level by inputting the feature set into the fatigue recovery detection model for analysis, and the fatigue recovery detection model is trained by the sample electromyographic signals and sample BCG signals collected in the same time period, so as to capture the physiological changes of the user from multiple angles, reduce the interference caused by a single signal, and improve the accuracy of the fatigue recovery inspection model, so as to accurately evaluate the fatigue recovery degree of the user and improve the accuracy of fatigue recovery detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0042] Figure 1 It is a schematic diagram of an application scenario of a fatigue recovery detection method disclosed in an embodiment of the present application;

[0043] Figure 2 It is a flow chart of a fatigue recovery detection method disclosed in an embodiment of the present application;

[0044] Figure 3 This is a schematic diagram of a fatigue recovery detection method disclosed in an embodiment of the present application. Figure 2 ;

[0045] Figure 4 It is a schematic diagram of a process for marking a sample fatigue recovery level corresponding to a sample electromyographic signal disclosed in an embodiment of the present application;

[0046] Figure 5 It is a flowchart of determining the fatigue recovery level of a sample according to the first electromyographic signal and the second electromyographic signal in the steps disclosed in the embodiment of the present application;

[0047] Figure 6 It is a flow chart of determining the fatigue recovery level of a sample according to the ratio of the second average integrated electromyographic value to the first average integrated electromyographic value disclosed in an embodiment of the present application;

[0048] Figure 7 It is a structural schematic diagram of a fatigue recovery detection device disclosed in an embodiment of the present application;

[0049] Figure 8 It is a structural schematic diagram of an electronic device disclosed in an embodiment of the present application. DETAILED DESCRIPTION

[0050] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0051] It should be noted that the terms "including" and "having" and any variations thereof in the embodiments of the present application and the accompanying drawings are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device including a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to these processes, methods, products or devices.

[0052] See also Figure 1 , Figure 1 Schematic diagram of an application scenario of a fatigue recovery detection method disclosed in an embodiment of the present application. Figure 1 As shown, the schematic diagram of the application scenario includes a mattress 110 and a physiological sign sensor 120. The physiological sign sensor 120 can be set on the surface or inside of the mattress 110. Optionally, the physiological sign sensor 120 can be set close to the heart position of the user when lying down, so as to monitor the user's physiological signs such as breathing and heartbeat.

[0053] The physiological sign sensor 120 can be used to collect the user's target BCG signal. Optionally, the physiological sign sensor 120 may include a piezoelectric ceramic sensor, a piezoelectric film sensor, an optical fiber sensor, etc., which is not specifically limited in the present embodiment. The BCG signal is a signal generated by the change in the external pressure of the human body caused by the heart beat and the blood circulation of the aorta.

[0054] In the embodiment of the present application, the mattress 110 may pre-store a fatigue recovery detection model, which is obtained by training based on multiple sets of training data, each set of training data includes sample electromyographic signals and sample BCG signals collected in the same time period, and the sample fatigue recovery level corresponding to the sample electromyographic signal. The mattress 110 may obtain the target BCG signal of the user collected by the physiological sign sensor 120; extract a feature set based on the target BCG signal, the feature set including one or more physiological sign features, and / or one or more original signal features; analyze the feature set through the fatigue recovery detection model obtained by training to obtain a fatigue recovery level; the fatigue recovery level is used to indicate the degree of fatigue recovery of the user.

[0055] Please read further Figure 2 , Figure 2 1 is a flow chart of a fatigue recovery detection method disclosed in an embodiment of the present application, and the fatigue recovery detection method can be applied to the above-mentioned mattress 110. Figure 2 As shown, the fatigue recovery detection method may include the following steps:

[0056] Step 210: Acquire the target BCG signal of the user collected by the physiological sign sensor.

[0057] The physiological sign sensor may collect the user's target BCG signal at a preset interval period. For example, the physiological sign sensor may collect the user's target BCG signal once every 2 seconds.

[0058] When the mattress detects that the user is on the mattress, the physiological sign sensor can be controlled to start collecting the user's target BCG signal. Furthermore, a pressure sensor can be provided in the mattress, and the mattress can obtain the pressure signal applied to the mattress by the user collected by the pressure sensor, and determine whether the user is on the mattress by judging whether the pressure signal is greater than a preset pressure threshold. If the pressure signal is greater than or equal to the preset pressure threshold, it can be determined that the user is on the mattress; if the pressure signal is less than the preset pressure threshold, it can be determined that the user is not on the mattress.

[0059] The mattress can obtain the user's target BCG signal collected by the physiological sign sensor within a certain period of time. For example, the mattress can obtain the user's target BCG signal collected by the physiological sign sensor within 1 hour, 2 hours, etc., but not limited to this. Through the target BCG signal within a certain period of time, the user's fatigue recovery degree can be analyzed more accurately.

[0060] In some embodiments, the mattress may obtain a first target BCG signal corresponding to a first time period before the user falls asleep, and a second target BCG signal corresponding to a second time period after the user falls asleep, collected by the physiological sign sensor.

[0061] Among them, before falling asleep refers to before the user enters the sleep state; after falling asleep refers to after the user enters the sleep state; the first time period refers to the time period from the time the user goes to bed to the time the user falls asleep; the second time period refers to the time period from the time the user falls asleep to the time the user wakes up.

[0062] The mattress can distinguish the target BCG signals before and after falling asleep according to the waveform changes of the target BCG signals. It can be understood that the user's heart rate will slow down after falling asleep, and the waveform corresponding to the target BCG signal after falling asleep will be smoother than the waveform corresponding to the target BCG signal before falling asleep. Therefore, the mattress can determine the band with a smoother waveform as the second target BCG signal corresponding to the second time period after falling asleep, and correspondingly, determine the band with a larger waveform change as the first target BCG signal corresponding to the first time period before falling asleep.

[0063] In some embodiments, after acquiring the target BCG signal of the user collected by the physiological sign sensor, the mattress may use a low-pass filter to filter out the high-frequency signal in the target BCG signal and retain the low-frequency signal. Among them, the high-frequency signal may refer to a signal with a frequency greater than or equal to a preset frequency threshold, and the low-frequency signal may refer to a signal with a frequency lower than the preset frequency threshold. The low-pass filter allows the target BCG signal below the preset frequency threshold to pass, while weakening or blocking the target BCG signal above the preset frequency threshold. Optionally, the low-pass filter can be an analog low-pass filter or a digital low-pass filter. By removing the high-frequency signal in the target BCG signal through a low-pass filter, high-frequency noise can be effectively removed, the signal-to-noise ratio of the target BCG signal can be improved, and it is ensured that the target BCG signal used in subsequent analysis can more accurately reflect the user's physiological sign characteristics.

[0064] In some embodiments, after acquiring the target BCG signal of the user collected by the physiological sign sensor, the mattress can perform wavelet denoising on the target BCG signal, and use wavelet transform to reduce or eliminate the impact of noise on the target BCG signal. Specifically, the mattress can decompose the target BCG signal into wavelet coefficients of different scales, each scale corresponds to a different frequency range, and modify or delete certain wavelet coefficients according to a preset threshold to remove noise. After that, the target BCG signal is reconstructed using the wavelet coefficients after threshold processing to obtain a denoised target BCG signal. Wavelet transform can capture the changing trend of the target BCG signal at different scales. Using wavelet transform to denoise the target BCG signal can effectively reduce the impact of high-frequency noise, while better maintaining the key features of the target BCG signal, avoiding the introduction of artifacts, and ensuring the accuracy of subsequent feature extraction using the target BCG signal.

[0065] Step 220: extract a feature set according to the target BCG signal, where the feature set includes one or more physiological sign features and / or one or more original signal features.

[0066] In some embodiments, the mattress may extract one or more features based on the target BCG signal to obtain a feature set, wherein the feature set refers to a combination of features extracted from the target BCG signal that can reflect changes in the target BCG signal.

[0067] The features extracted from the target BCG signal may be physiological sign features or original signal features.

[0068] Physiological sign characteristics refer to characteristics obtained by calculating the target BCG signal and capable of reflecting the physiological condition of the user. For example, the one or more physiological characteristics may include at least one of heart rate characteristics, breathing characteristics, body movement characteristics, and the like.

[0069] Among them, the heart rate characteristics may refer to characteristics that can reflect changes in the user's heart rate.

[0070] Optionally, the heart rate characteristics may include but are not limited to at least one of heart rate variability (HRV), average heart rate value, maximum heart rate value, minimum heart rate value, etc.

[0071] Among them, HRV refers to the change of heartbeat interval (RR interval or NN interval). HRV measurement indicators may include but are not limited to SDNN (Standard Deviation of Normal-to-Normal intervals, standard deviation of NN intervals), SDANN (Standard Deviation of the Average NN intervals, standard deviation of average NN intervals), PNN50 (Percentage of NN50, percentage of adjacent NN interval differences > 50ms), RMSSD (Root Mean Square of Successive Differences, average value of the square root of adjacent NN interval differences), LF (Low Frequency, low frequency power), HF (High Frequency, high frequency power), LF / HF, etc. At least one.

[0072] SDNN refers to the standard deviation of all NN intervals, which can be used to measure the degree of change of the user's heartbeat over a certain period of time. For example, the mattress can extract all NN intervals from the target BCG signal and calculate the average value of all NN intervals. For each NN interval RR i Calculate the average value The square of the difference Add all the squares of the differences and divide by N-1 to get the variance. Take the square root of the variance to get the SDNN. The calculation formula of SDNN can refer to the following formula (1):

[0073]

[0074] Among them, RR i It can represent the i-th NN interval, It can represent the average value of all NN intervals, and N can represent the total number of NN intervals.

[0075] SDANN refers to the standard deviation of the average value of NN intervals in all adjacent preset time windows. The preset time window can be used to divide the target BCG signal into multiple time periods. For example, the mattress can divide the target BCG signal into multiple time periods according to the preset time window, and for each time period, calculate the average value of the NN interval in the time period. Furthermore, the grand average of the NN interval averages for all time periods can be calculated Then, according to the calculation formula of SDANN, the standard deviation of the average value of NN intervals in all time periods is calculated to obtain SDANN. The calculation formula of SDANN can refer to the following formula (2):

[0076]

[0077] Among them, R _ R j It can represent the average value of the NN interval in the jth time period; It can represent the grand average of the NN interval averages of all time periods; M can represent the total number of time periods.

[0078] PNN50 refers to the proportion of adjacent NN intervals with a difference greater than 50ms, which can be used to evaluate the change in the user's heartbeat between adjacent NN intervals. For example, the mattress can extract all NN intervals from the target BCG signal, and for each pair of adjacent NN intervals RR i and RR i+1 , calculate their difference |RR i+1 -RR i |, count the number of all adjacent NN interval differences exceeding 50ms, divide the logarithm of the difference exceeding 50ms by the total logarithm of NN intervals, and multiply by 100% to obtain the percentage of PNN50.

[0079] RMSSD refers to the average of the square roots of the differences between adjacent NN intervals, which can be used to evaluate the short-term changes between adjacent NN intervals. For example, the mattress can extract all NN intervals from the target BCG signal, and for each pair of adjacent NN intervals RR i+1 and RR i , calculate their difference RR i+1 -RR i , square each difference (RR i+1 -RR i ) 2 , add up all the squares of the differences and divide by N-1 to get the mean square error, and take the square root of the mean square error to get the RMSSD. The calculation formula of RMSSD can refer to the following formula (3):

[0080]

[0081] Among them, RR i and RR i+1 They represent the i-th and i+1-th NN intervals respectively; N can represent the total number of heartbeats.

[0082] The breathing feature refers to a feature that can reflect the user's breathing condition. Optionally, the breathing feature may include at least one of an average breathing rate, a maximum breathing value, a minimum breathing value, and the like.

[0083] The mattress can extract a target BCG signal related to the user's breathing from the target BCG signal. For example, the mattress can identify and remove heart-related peaks through QRS complex detection, then apply a bandpass filter to retain the target BCG signal within the respiratory frequency range, and then extract the user's breathing characteristics from the target BCG signal.

[0084] The body movement feature refers to a feature that can reflect the user's body movements (such as getting up, turning over). Optionally, the body movement feature may include the body movement number. The mattress may pre-set a BCG signal amplitude threshold, determine the sampling points in the target BCG signal whose amplitude is greater than the BCG signal amplitude threshold as body movement points, and determine the number of body movement points contained in the target BCG signal as the body movement number.

[0085] The original signal features refer to the features directly extracted from the target BCG signal that can reflect the changing rules of the target BCG signal.

[0086] In some embodiments, the one or more original signal features may include at least one of the zero-crossing rate corresponding to one or more frequency bands, the energy corresponding to one or more frequency bands, etc. The zero-crossing rate refers to the number of times the target BCG signal changes from a positive value to a negative value or from a negative value to a positive value through the zero point within a preset time window; the energy corresponding to a frequency band refers to the sum of the energies of all frequency components within the frequency range corresponding to the frequency band.

[0087] Step 230, analyze the feature set through the trained fatigue recovery detection model to obtain a fatigue recovery level; the fatigue recovery level is used to indicate the user's fatigue recovery degree; wherein, the fatigue recovery detection model is trained based on multiple groups of training data, each group of training data includes sample electromyography signals and sample BCG signals collected in the same time period, and the sample fatigue recovery level corresponding to the sample electromyography signals.

[0088] The fatigue recovery detection model can be used to analyze the feature set corresponding to the target BCG signal to obtain the fatigue recovery level corresponding to the user. Exemplarily, the fatigue recovery detection model can include but is not limited to a random forest model, a support vector machine (SVM), a gradient boosting decision tree (GBDT), a convolutional neural network (CNN), a recurrent neural network (RNN), and a graph neural network (GNN). One of the Graph Neural Networks).

[0089] The fatigue recovery level may be used to indicate the user's fatigue recovery degree, which refers to the degree to which the user gradually recovers from a fatigue state to a normal or near-normal level after a night's sleep.

[0090] In some embodiments, the fatigue recovery level can be divided into multiple levels in advance according to the degree of fatigue recovery. Exemplarily, according to the degree of fatigue recovery from high to low, it can be divided into 5 fatigue recovery levels, wherein the 5 fatigue recovery levels are respectively represented by numerical values ​​from large to small, for example, the numerical value corresponding to fatigue recovery level 1 is 5, indicating excellent recovery, the numerical value corresponding to fatigue recovery level 2 is 4, indicating good fatigue recovery, the numerical value corresponding to fatigue recovery level 3 is 3, indicating general fatigue recovery, the numerical value corresponding to fatigue recovery level 4 is 2, indicating poor fatigue recovery, and the numerical value corresponding to fatigue recovery level 5 is 1, indicating extremely poor fatigue recovery. It should be noted that the smaller the fatigue recovery level, the greater the degree of fatigue recovery, which is not limited here.

[0091] The fatigue recovery detection model may be obtained by training based on multiple sets of training data, each set of training data may include sample electromyographic signals and sample BCG signals collected in the same time period, and sample fatigue recovery levels corresponding to the sample electromyographic signals.

[0092] The sample electromyographic signal refers to an electrical signal that records the user's muscle activity and can reflect the user's muscle usage. The sample fatigue recovery level corresponding to the sample electromyographic signal can refer to the fatigue recovery level determined by labeling the sample electromyographic signal, and the sample fatigue recovery level can be understood as the real fatigue recovery level.

[0093] In some embodiments, the sample fatigue recovery level corresponding to the sample electromyographic signal can be determined by the sample electromyographic signal. The mattress can analyze the sample electromyographic signal to determine the sample fatigue recovery level corresponding to the sample electromyographic signal.

[0094] During the training process of the fatigue recovery detection model, multiple sets of training data can be input into the fatigue recovery detection model, and the fatigue recovery detection model can analyze the input sample BCG signal to obtain a predicted fatigue recovery level, and compare the predicted fatigue recovery level with the sample fatigue recovery level matched by the sample electromyographic signal corresponding to the input sample BCG signal, determine the loss, and adjust the model parameters of the fatigue recovery detection model according to the loss until the fatigue recovery detection model converges. The fatigue recovery detection model can learn the ability to analyze and determine the fatigue recovery level based on the BCG signal.

[0095] Furthermore, the mattress can extract a sample feature set according to the sample BCG signal, and input the sample feature set into the fatigue recovery detection model to be trained to obtain a predicted fatigue recovery level.

[0096] In an embodiment of the present application, the mattress can obtain a target BCG signal of the user collected by a physiological sign sensor; extract a feature set based on the target BCG signal, and the feature set includes one or more physiological sign features, and / or one or more original signal features; analyze the feature set through the fatigue recovery detection model obtained by training to obtain a fatigue recovery level; the fatigue recovery level can be used to indicate the user's fatigue recovery degree; wherein the fatigue recovery detection model is trained based on multiple groups of training data, each group of training data includes sample electromyographic signals and sample BCG signals collected in the same time period, and sample fatigue recovery levels corresponding to the sample electromyographic signals. The fatigue recovery detection model is trained by using sample electromyographic signals and sample BCG signals collected in the same time period, so as to capture the user's physiological changes from multiple angles and reduce the interference caused by a single signal, so as to accurately evaluate the user's fatigue recovery degree and improve the accuracy of fatigue recovery detection.

[0097] In some embodiments, Figure 3 As shown, a fatigue recovery detection method is provided, which may include the following steps:

[0098] Step 310, obtaining a first target BCG signal corresponding to a first time period before the user falls asleep and a second target BCG signal corresponding to a second time period after the user falls asleep, collected by a physiological sign sensor;

[0099] Step 320 , extracting a feature set according to the target BCG signal, the feature set including: a first feature set corresponding to the first target BCG signal, and a second feature set corresponding to the second target BCG signal.

[0100] Step 330: Analyze the feature set through each decision tree to obtain the predicted fatigue recovery result output by each decision tree.

[0101] Each decision tree included in the fatigue recovery detection model can analyze the feature set and output the corresponding predicted fatigue recovery result. Each decision tree can classify the feature set according to its internal node splitting rules to obtain the corresponding predicted fatigue recovery result.

[0102] Furthermore, each decision tree included in the fatigue recovery detection model can analyze the first feature set and the second feature set, and output a corresponding predicted fatigue recovery result.

[0103] Step 340: The predicted fatigue recovery results output by each decision tree are integrated to determine the fatigue recovery level.

[0104] In some embodiments, the predicted fatigue recovery result may be a fatigue recovery level label. Exemplarily, the fatigue recovery level label may include excellent fatigue recovery, good fatigue recovery, general fatigue recovery, poor fatigue recovery, and extremely poor fatigue recovery. The mattress may use a numerical value to represent the fatigue recovery level label, for example, excellent fatigue recovery may be represented as a numerical value of 5, good fatigue recovery may be represented as a numerical value of 4, general fatigue recovery may be represented as a numerical value of 3, poor fatigue recovery may be represented as a numerical value of 2, and extremely poor fatigue recovery may be represented as a numerical value of 1. The mattress may determine the fatigue recovery level label that appears most frequently in the predicted fatigue recovery test results output by multiple decision trees as the fatigue recovery level.

[0105] In some embodiments, the predicted fatigue recovery result may be a probability distribution of each fatigue recovery level. Exemplarily, the predicted fatigue recovery result output by each decision tree may be expressed as [a1, a2, a3, a4, a5], where a1 may represent the probability corresponding to excellent fatigue recovery, a2 may represent the probability corresponding to good fatigue recovery, a3 may represent the probability corresponding to average fatigue recovery, a4 may represent the probability corresponding to poor fatigue recovery, and a5 may represent the probability corresponding to extremely poor fatigue recovery. The mattress may add the probability distributions corresponding to each fatigue recovery level output by each decision tree according to the fatigue recovery level to obtain the final probability distribution, and select the fatigue recovery level with the highest probability.

[0106] In some embodiments, the predicted fatigue recovery result may be a continuous value, and the mattress may use a preset numerical range to represent the degree of fatigue recovery from extremely good to extremely poor. Exemplarily, the preset numerical range may be 0 to 2, and the predicted fatigue recovery result output by each decision tree may be a value between 0 and 2, such as 1.2, 0.8, etc. The mattress may perform a weighted average of the predicted fatigue recovery results output by each decision tree, and determine the fatigue recovery level according to the recovery level intervals corresponding to each fatigue recovery level predetermined.

[0107] In some embodiments, during the training process of the fatigue recovery detection model, the mattress can train a decision tree through multiple sets of training data. Training data can be randomly extracted with replacement from the multiple sets of training data to obtain a sub-training data set, and this process can be repeated multiple times to obtain multiple sub-training data sets, each of which can be used to train a decision tree. Multiple decision trees can be trained through multiple sub-training data sets, and these decision trees constitute the fatigue recovery detection model.

[0108] Furthermore, in the process of training the decision tree through the sub-training data sets, the mattress can divide the sub-training data sets according to each feature in the sub-training data sets until the depth of the decision tree reaches a preset maximum depth or the number of groups of training data in the leaf nodes of the decision tree is less than the preset number of groups.

[0109] In the embodiment of the present application, the mattress can analyze the feature set through each decision tree to obtain the predicted fatigue recovery results output by each decision tree; then, the predicted fatigue recovery results output by each decision tree are fused to determine the fatigue recovery level. By determining the fatigue recovery level by fusing the predicted fatigue recovery results output by multiple decision trees, the overfitting problem that may exist in a single model can be effectively reduced, thereby improving the accuracy and robustness of fatigue recovery detection.

[0110] In some embodiments, before training the fatigue recovery detection model to be trained, it is necessary to mark the sample fatigue recovery level corresponding to the sample electromyographic signal. Figure 4 As shown, before training the fatigue recovery detection model to be trained, the following steps may also be included:

[0111] Step 410: collect, through the electromyographic sensor, a first sample electromyographic signal corresponding to a third time period before the user falls asleep, and a second sample electromyographic signal corresponding to a fourth time period after the user falls asleep.

[0112] Among them, the electromyographic sensor can be used to collect sample electromyographic signals of the user.

[0113] The third time period refers to the time period between the time the user goes to bed and the time the user falls asleep; the fourth time period refers to the time period between the time the user falls asleep and the time the user wakes up.

[0114] The sample electromyographic signals of each set of training data may include a first sample electromyographic signal and a second sample electromyographic signal.

[0115] The first sample EMG signal refers to the EMG signal collected before the user falls asleep, which reflects the changes in the user's muscle activity before entering the sleep state; the second sample EMG signal refers to the EMG signal collected after the user falls asleep, which reflects the changes in the user's muscle activity after entering the sleep state.

[0116] In some embodiments, the electromyographic sensor may collect, at preset intervals, a first sample electromyographic signal corresponding to a third time period before the user falls asleep and a second sample electromyographic signal corresponding to a fourth time period after the user falls asleep.

[0117] In some embodiments, when the mattress detects that the user is on the mattress, the mattress may control the physiological sign sensor and the electromyographic sensor to respectively start collecting sample BCG signals and sample electromyographic signals of the user.

[0118] In some embodiments, after collecting the first sample electromyographic signal corresponding to the third time period before the user falls asleep and the second sample electromyographic signal corresponding to the fourth time period after the user falls asleep through the electromyographic sensor, the mattress can use a low-pass filter to filter out the high-frequency signal in the first sample electromyographic signal and the second sample electromyographic signal, and retain the low-frequency signal. Wherein, the high-frequency signal refers to a signal with a frequency greater than or equal to a preset frequency threshold, and the low-frequency signal refers to a signal with a frequency less than a preset frequency threshold. The low-pass filter can allow the first sample electromyographic signal and the second sample electromyographic signal below the preset frequency threshold to pass, while weakening or blocking the first sample electromyographic signal and the second sample electromyographic signal above the preset frequency threshold. Optionally, the low-pass filter can be an analog low-pass filter or a digital low-pass filter. By removing the high-frequency signal in the first sample electromyographic signal and the second sample electromyographic signal through the low-pass filter, high-frequency noise can be effectively removed, the signal-to-noise ratio of the first sample electromyographic signal and the second sample electromyographic signal can be improved, and the fatigue recovery level determined by the first sample electromyographic signal and the second sample electromyographic signal can be ensured to accurately reflect the fatigue recovery degree of the user.

[0119] In some embodiments, after collecting the first sample electromyographic signal corresponding to the third time period before the user falls asleep and the second sample electromyographic signal corresponding to the fourth time period after the user falls asleep through the electromyographic sensor, the mattress can perform wavelet denoising on the first sample electromyographic signal and the second sample electromyographic signal, and use wavelet transform to reduce or eliminate the influence of noise on the first sample electromyographic signal and the second sample electromyographic signal. Specifically, the mattress can decompose the first sample electromyographic signal and the second sample electromyographic signal into wavelet coefficients of different scales, each scale corresponds to a different frequency range, and modify or delete certain wavelet coefficients according to a preset threshold to remove noise, and then reconstruct the first sample electromyographic signal and the second sample electromyographic signal using the wavelet coefficients after threshold processing to obtain the denoised first sample electromyographic signal and the second sample electromyographic signal. By decomposing the first sample electromyographic signal and the second sample electromyographic signal into wavelet coefficients of different scales, the noise component in the signal can be effectively separated, and the use of a preset threshold to modify or delete specific wavelet coefficients, especially those representing noise, can greatly reduce or even eliminate random interference in the original signal, so that the reconstructed first sample electromyographic signal and the second sample electromyographic signal are more accurate.

[0120] In some embodiments, after collecting the first sample electromyographic signal corresponding to the third time period before the user falls asleep and the second sample electromyographic signal corresponding to the fourth time period after the user falls asleep through the electromyographic sensor, the mattress may normalize the first sample electromyographic signal and the second sample electromyographic signal, thereby mapping the first sample electromyographic signal and the second sample electromyographic signal to the same scale range. Exemplarily, the mattress may scale the first sample electromyographic signal and the second sample electromyographic signal to a fixed range, such as [0,1], using a minimum-maximum scaling method. The first sample electromyographic signal and the second sample electromyographic signal may differ due to individual differences, device sensitivity, and other factors. Normalization can eliminate the dimensional differences caused by these factors, so that data from different users can be compared on the same scale.

[0121] Step 420: Determine the sample fatigue recovery level according to the first sample electromyographic signal and the second sample electromyographic signal.

[0122] It can be understood that the first sample EMG signal can represent the changes in the user's muscle activity before entering the sleep state, and the second sample EMG signal can represent the changes in the user's muscle activity after entering the sleep state. In the fatigue state, the muscle contraction force is weakened and the reaction time is prolonged, so the sample fatigue recovery level can be determined by the first sample EMG signal and the second sample EMG signal.

[0123] In some embodiments, the mattress can calculate the statistical parameters corresponding to the first sample electromyographic signal and the second sample electromyographic signal, such as the integral value, mean, standard deviation, peak amplitude, frequency, etc., compare the differences before and after falling asleep to evaluate the degree of fatigue recovery and determine the sample fatigue recovery level.

[0124] In the embodiment of the present application, the mattress can collect the first sample electromyographic signal corresponding to the third time period before the user falls asleep, and the second sample electromyographic signal corresponding to the fourth time period after the user falls asleep through the electromyographic sensor; the sample fatigue recovery level is determined according to the first sample electromyographic signal and the second sample electromyographic signal. By collecting the electromyographic signals before and after falling asleep respectively, the changes in the muscle activity of the user under different physiological states can be captured, and different changes in muscle activity can reflect the fatigue recovery of the user, so as to determine the sample fatigue recovery level according to the electromyographic signals before and after falling asleep, and the fatigue recovery level of the user can be accurately evaluated.

[0125] In some embodiments, the mattress can determine the sample fatigue recovery level according to the ratio of the statistical parameters of the first sample electromyographic signal and the second sample electromyographic signal. Figure 5 As shown, the step of determining the fatigue recovery level of the sample according to the first electromyographic signal and the second electromyographic signal may include the following steps:

[0126] Step 510: Calculate a plurality of first integrated electromyographic values ​​corresponding to a third time period according to the first sample electromyographic signal.

[0127] The third time period may be equally divided into a plurality of preset time periods.

[0128] The first integrated electromyographic value refers to an integrated electromyographic value calculated from electromyographic signals in each preset time period collected by the electromyographic sensor in the third time period before falling asleep.

[0129] Exemplarily, the calculation formula of the first integrated electromyographic value may refer to the following formula (4):

[0130]

[0131] Among them, IEMG i It can represent the integrated electromyographic value of the i-th preset time period, EMG ij It can represent the electromyographic signal value of the jth sampling point in the i-th preset time period, N i It can represent the total number of sampling points in the i-th preset time period.

[0132] Step 520, calculating an average value of a plurality of first integrated electromyographic values ​​to obtain a first average integrated electromyographic value.

[0133] The first average integrated EMG value refers to the average integrated EMG value of the EMG signal in a plurality of preset time periods corresponding to the third time period before the user falls asleep, and reflects the muscle activity level of the user before falling asleep.

[0134] Exemplarily, assuming that the above-mentioned multiple first integrated electromyographic values ​​are IEMG 1 ,IEMG 2 ,...,IEMG n , then the calculation formula of the first average integrated electromyographic value can refer to the following formula (5):

[0135]

[0136] Wherein, Average-IEMG may represent the first average integrated electromyographic value, and n may represent the number of preset time periods.

[0137] Step 530: Calculate a plurality of second integrated electromyographic values ​​corresponding to a fourth time period according to the second sample electromyographic signal.

[0138] The fourth time period may be equally divided into a plurality of preset time periods.

[0139] The second integrated electromyographic value refers to an integrated electromyographic value calculated from the electromyographic signals in each preset time period collected by the electromyographic sensor in the fourth time period before falling asleep.

[0140] Step 540, calculating an average value of a plurality of second integrated electromyographic values ​​to obtain a second average integrated electromyographic value.

[0141] The second average integrated EMG value refers to the average integrated EMG value of the EMG signal in multiple preset time periods corresponding to the fourth time period after the user falls asleep, reflecting the muscle activity level of the user after entering the sleep state.

[0142] Step 550: Determine the fatigue recovery level of the sample according to the ratio of the second average integrated electromyographic value to the first average integrated electromyographic value.

[0143] It can be understood that, the larger the ratio of the second average integrated electromyographic value to the first average integrated electromyographic value is, the more muscle activity the user has during sleep, and the worse the user's fatigue recovery degree is.

[0144] In some embodiments, the mattress can determine the sample fatigue recovery level by a preset threshold. Figure 6 As shown, the step of determining the fatigue recovery level of the sample according to the ratio of the second average integrated electromyographic value to the first average integrated electromyographic value may include the following steps:

[0145] Step 610, determining the target recovery level interval to which the ratio belongs.

[0146] The mattress can be pre-set with a target recovery level interval, and the target recovery level interval to which the ratio belongs is determined according to the size of the ratio.

[0147] Step 620: Determine the sample fatigue recovery level corresponding to the target recovery level interval. The larger the interval threshold corresponding to the target recovery level interval, the worse the fatigue recovery degree indicated by the sample fatigue recovery level.

[0148] The mattress may pre-store a mapping relationship table between target recovery level intervals and sample fatigue recovery levels. Then, the mattress may query the sample fatigue recovery level corresponding to the target recovery level interval from the mapping relationship table according to the target recovery level interval to which the ratio belongs.

[0149] Exemplarily, the mattress can determine the sample fatigue recovery level corresponding to the interval where the ratio is less than 0.6 as excellent fatigue recovery; determine the sample fatigue recovery level corresponding to the interval where the ratio is between 0.6 and 0.9 as good fatigue recovery; determine the sample fatigue recovery level corresponding to the interval where the ratio is between 0.9 and 1.1 as fair fatigue recovery; determine the sample fatigue recovery level corresponding to the interval where the ratio is between 1.1 and 1.4 as poor fatigue recovery; determine the sample fatigue recovery level corresponding to the interval where the ratio is greater than 1.4 as extremely poor fatigue recovery. Then, based on the size of the ratio, such as 0.8, it can be determined that the sample fatigue recovery level corresponding to the interval where the ratio belongs to the target recovery level interval between 0.6 and 0.9 is good fatigue recovery.

[0150] In the embodiment of the present application, the mattress can determine the target recovery level interval to which the ratio belongs, and determine the sample fatigue recovery level corresponding to the target recovery level interval. The ratio reflects the changing trend of the electromyographic signal in different time periods, and can more accurately capture the process from muscle fatigue to recovery. By setting a reasonable target recovery level interval, different recovery states can be classified, thereby improving the accuracy of fatigue recovery detection.

[0151] In the embodiment of the present application, the mattress can calculate multiple first integrated electromyographic values ​​corresponding to the third time period according to the first sample electromyographic signal; calculate the average of multiple first integrated electromyographic values ​​to obtain a first average integrated electromyographic value; calculate multiple second integrated electromyographic values ​​corresponding to the fourth time period according to the second sample electromyographic signal; calculate the average of multiple second integrated electromyographic values ​​to obtain a second average integrated electromyographic value; and determine the sample fatigue recovery level according to the ratio of the second average integrated electromyographic value to the first average integrated electromyographic value. By directly calculating the ratio of the second average integrated electromyographic value to the first average integrated electromyographic value, a quantitative index reflecting the degree of fatigue recovery can be quickly and easily obtained, and by fusing the first average integrated electromyographic value and the second average integrated electromyographic value, it is possible to more comprehensively reflect the changes in the user's muscle activity, not only considering the user's muscle tension in the awake state, but also paying attention to the recovery during sleep, thereby improving the accuracy of fatigue recovery detection.

[0152] See also Figure 7 , Figure 7 is a schematic diagram of a fatigue recovery detection device disclosed in an embodiment of the present application. The fatigue recovery detection device 700 can be applied to the above-mentioned mattress 110. Figure 7 As shown, the fatigue recovery detection device 700 may include: a BCG signal acquisition module 710 , a feature extraction module 720 and an analysis module 730 .

[0153] The BCG signal acquisition module 710 is used to acquire the target BCG signal of the user collected by the physiological sign sensor;

[0154] A feature extraction module 720 is used to extract a feature set according to the target BCG signal, where the feature set includes one or more physiological sign features and / or one or more original signal features;

[0155] The analysis module 730 is used to analyze the feature set through the fatigue recovery detection model obtained through training to obtain a fatigue recovery level; the fatigue recovery level is used to indicate the user's fatigue recovery degree; wherein, the fatigue recovery detection model is trained based on multiple groups of training data, each group of training data includes sample electromyography signals and sample BCG signals collected in the same time period, and the sample fatigue recovery level corresponding to the sample electromyography signal.

[0156] In some embodiments, the fatigue recovery detection model includes multiple decision trees, and the analysis module 730 is also used to analyze the feature set through each decision tree to obtain the predicted fatigue recovery results output by each decision tree; the predicted fatigue recovery results output by each decision tree are integrated to determine the fatigue recovery level.

[0157] In some embodiments, the one or more physiological sign characteristics include at least one of a heart rate characteristic, a breathing characteristic, and a body movement characteristic; and / or,

[0158] The one or more original signal features include at least one of the zero-crossing rate corresponding to one or more frequency bands in the target BCG signal and the energy corresponding to one or more frequency bands.

[0159] In some embodiments, the BCG signal acquisition module 710 is also used to acquire a first target BCG signal corresponding to a first time period of a user before falling asleep, and a second target BCG signal corresponding to a second time period of the user after falling asleep, collected by the physiological sign sensor; the above-mentioned feature set includes: a first feature set corresponding to the above-mentioned first target BCG signal, and a second feature set corresponding to the above-mentioned second target BCG signal.

[0160] In some embodiments, the mattress further includes an electromyographic sensor; the sample electromyographic signal of each set of training data includes a first sample electromyographic signal and a second sample electromyographic signal; the fatigue recovery detection device 700 further includes a sample electromyographic signal acquisition module and a sample fatigue recovery level determination module.

[0161] A sample electromyographic signal acquisition module, used to acquire, through an electromyographic sensor, a first sample electromyographic signal corresponding to a third time period before the user falls asleep, and a second sample electromyographic signal corresponding to a fourth time period after the user falls asleep;

[0162] The sample fatigue recovery level determination module is used to determine the sample fatigue recovery level according to the first electromyographic signal and the second sample electromyographic signal.

[0163] In some embodiments, the sample fatigue recovery level determination module is also used to calculate multiple first integrated electromyographic values ​​corresponding to the above-mentioned third time period based on the above-mentioned first sample electromyographic signal; calculate the average value of the above-mentioned multiple first integrated electromyographic values ​​to obtain a first average integrated electromyographic value; calculate multiple second integrated electromyographic values ​​corresponding to the above-mentioned fourth time period based on the above-mentioned second sample electromyographic signal; calculate the average value of the above-mentioned multiple second integrated electromyographic values ​​to obtain a second average integrated electromyographic value; determine the sample fatigue recovery level based on the ratio of the above-mentioned second average integrated electromyographic value to the above-mentioned first average integrated electromyographic value.

[0164] In some embodiments, the sample fatigue recovery level determination module is also used to determine the target recovery level interval to which the above-mentioned ratio belongs; determine the sample fatigue recovery level corresponding to the above-mentioned target recovery level interval; wherein, the larger the interval threshold corresponding to the above-mentioned target recovery level interval, the worse the fatigue recovery degree indicated by the above-mentioned sample fatigue recovery level.

[0165] See also Figure 8 , Figure 8 Schematic diagram of the structure of a mattress disclosed in the embodiment of the present application. Figure 8 As shown, the mattress 800 may include:

[0166] A memory 810 storing executable program codes;

[0167] a processor 820 coupled to the memory 810;

[0168] The processor 820 calls the executable program code stored in the memory 810 to execute any fatigue recovery detection method disclosed in the embodiments of the present application.

[0169] An embodiment of the present application discloses a computer-readable storage medium storing a computer program, wherein when the computer program is executed by the above-mentioned processor, the above-mentioned processor implements any fatigue recovery detection method disclosed in the embodiment of the present application.

[0170] It should be understood that the "one embodiment" or "an embodiment" mentioned throughout the specification means that the specific features, structures or characteristics related to the embodiment are included in at least one embodiment of the present application. Therefore, "in one embodiment" or "in an embodiment" appearing throughout the specification does not necessarily refer to the same embodiment. In addition, these specific features, structures or characteristics can be combined in one or more embodiments in any suitable manner. Those skilled in the art should also be aware that the embodiments described in the specification are all optional embodiments, and the actions and modules involved are not necessarily required for the present application.

[0171] In the various embodiments of the present application, it should be understood that the size of the serial numbers of the above-mentioned processes does not necessarily mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0172] The units described above as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0173] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0174] If the above-mentioned integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-accessible memory. Based on this understanding, the technical solution of the present application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product, which is stored in a memory and includes several requests for a computer device (which can be a personal computer, a server or a network device, etc., specifically a processor in a computer device) to perform some or all of the steps of the above-mentioned methods of various embodiments of the present application.

[0175] Those skilled in the art will appreciate that all or part of the steps in the various methods of the above embodiments may be completed by instructing related hardware through a program, and the program may be stored in a computer-readable storage medium, the storage medium including a read-only memory (ROM), a random access memory (RAM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), a one-time programmable read-only memory (OTPROM), an electronically erasable programmable read-only memory (EEPROM), a compact disc read-only memory (CD-ROM) or other optical disc storage, magnetic disk storage, magnetic tape storage, or any other computer-readable medium that can be used to carry or store data.

[0176] The above is a detailed introduction to a fatigue recovery detection method, device, electronic device and storage medium disclosed in the embodiments of the present application. Specific examples are used in this article to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method and core idea of ​​the present application. At the same time, for those of ordinary skill in the art, according to the idea of ​​the present application, there will be changes in the specific implementation method and application scope. In summary, the content of this specification should not be understood as a limitation on the present application.

Claims

1. A fatigue recovery detection method, characterized in that: Applied to a mattress, the mattress includes a physiological sign sensor, and the method includes: Acquiring the target BCG signal of the user collected by the physiological sign sensor; Extracting a feature set according to the target BCG signal, wherein the feature set includes one or more physiological sign features and / or one or more original signal features; The fatigue recovery detection model obtained through training is used to analyze the feature set to obtain a fatigue recovery level; the fatigue recovery level is used to indicate the user's fatigue recovery degree; wherein the fatigue recovery detection model is trained based on multiple groups of training data, each group of the training data includes sample electromyography signals and sample BCG signals collected in the same time period, and sample fatigue recovery levels corresponding to the sample electromyography signals.

2. The method according to claim 1, characterized in that The fatigue recovery detection model includes a plurality of decision trees. The fatigue recovery detection model obtained through training analyzes the feature set to obtain a fatigue recovery level, including: Analyzing the feature set through each decision tree to obtain a predicted fatigue recovery result output by each decision tree; The predicted fatigue recovery results output by each decision tree are integrated to determine the fatigue recovery level.

3. The method according to claim 1, characterized in that The one or more physiological sign characteristics include at least one of heart rate characteristics, breathing characteristics, and body movement characteristics; and / or, The one or more original signal features include at least one of the zero-crossing rate corresponding to one or more frequency bands in the target BCG signal and the energy corresponding to one or more frequency bands.

4. The method according to claim 1, characterized in that: The step of acquiring the target BCG signal of the user collected by the physiological sign sensor comprises: Acquire a first target BCG signal corresponding to a first time period of the user before falling asleep, and a second target BCG signal corresponding to a second time period of the user after falling asleep, collected by the physiological sign sensor; The feature set includes: a first feature set corresponding to the first target BCG signal, and a second feature set corresponding to the second target BCG signal.

5. The method according to claim 1, characterized in that The mattress further includes an electromyographic sensor; each set of sample electromyographic signals of the training data includes a first sample electromyographic signal and a second sample electromyographic signal; Before acquiring the target BCG signal of the user by the physiological sign sensor, the method further includes: Collecting, by the electromyographic sensor, a first sample electromyographic signal corresponding to a third time period before the user falls asleep, and a second sample electromyographic signal corresponding to a fourth time period after the user falls asleep; A sample fatigue recovery level is determined according to the first sample electromyographic signal and the second sample electromyographic signal.

6. The method according to claim 5, characterized in that The step of determining the sample fatigue recovery level according to the first sample electromyographic signal and the second sample electromyographic signal includes: Calculating a plurality of first integrated electromyographic values ​​corresponding to the third time period according to the first sample electromyographic signal; Calculate the average of the plurality of first integrated electromyographic values ​​to obtain a first average integrated electromyographic value; Calculating a plurality of second integrated electromyographic values ​​corresponding to the fourth time period according to the second sample electromyographic signal; Calculate the average value of the plurality of second integrated electromyographic values ​​to obtain a second average integrated electromyographic value; The fatigue recovery level of the sample is determined according to the ratio of the second average integrated electromyographic value to the first average integrated electromyographic value.

7. The method according to claim 6, characterized in that Determining the fatigue recovery level of the sample according to the ratio of the second average integrated electromyographic value to the first average integrated electromyographic value includes: Determining a target recovery level interval to which the ratio belongs; Determine the sample fatigue recovery level corresponding to the target recovery level interval; Among them, the larger the interval threshold corresponding to the target recovery level interval is, the worse the fatigue recovery degree indicated by the sample fatigue recovery level is.

8. A fatigue recovery detection device, characterized in that: Applied to a mattress, the mattress includes a physiological sign sensor, and the device includes: A BCG signal acquisition module, used to acquire the target BCG signal of the user collected by the physiological sign sensor; A feature extraction module, configured to extract a feature set according to the target BCG signal, wherein the feature set includes one or more physiological sign features and / or one or more original signal features; An analysis module is used to analyze the feature set through a fatigue recovery detection model obtained through training to obtain a fatigue recovery level; the fatigue recovery level is used to indicate the user's fatigue recovery degree; wherein the fatigue recovery detection model is trained based on multiple groups of training data, each group of the training data includes sample electromyographic signals and sample BCG signals collected in the same time period, and sample fatigue recovery levels corresponding to the sample electromyographic signals.

9. A mattress, characterized in that: include: A memory storing executable program code; a processor coupled to the memory; The processor calls the executable program code stored in the memory to execute the method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, wherein when the computer program is executed by a processor, the processor is caused to perform the method according to any one of claims 1 to 7.

Images