A mesh distributed massage air bag intelligent inflation and deflation method and system

Abstract

The application discloses a kind of grid distributed massage air bag intelligent inflation and deflation method and system, it is related to massage nursing technical field, including the following steps: construct space-time distribution matrix based on bed surface air pressure fluctuation field in mattress interior, the pressure change of each air bag in inflation and deflation period is converted into continuous air pressure disturbance sequence, and Fourier transform is carried out to air pressure disturbance sequence to extract main frequency component, form air bag alternate frequency spectrum diagram as input baseline.The application realizes the dynamic interactive identification of air bag inflation and deflation and human microvibration by spectrum mapping and coupling measure, early warning and reduce resonance risk, effectively reduce local stress fluctuation and bed sore occurrence.At the same time, when coupling coefficient is close to set threshold, trigger adaptive frequency offset mechanism, adjust inflation and deflation rhythm in real time, ensure the effect of decompression and massage, while actively avoiding sensitive frequency band, improve the safety, comfort and long-term stability of nursing process.

Description

Technical Field

[0001] This invention relates to the field of massage and nursing technology, specifically to a method and system for intelligent inflation and deflation of grid-distributed massage airbags. Background Technology

[0002] The grid-distributed massage airbag intelligent inflation and deflation system refers to the placement of multiple independent airbag units in a grid pattern inside the mattress. Combined with real-time monitoring data from pressure-sensing membrane sensors and humidity sensors, the system precisely adjusts the inflation and deflation of airbags in each area, achieving dynamic pressure relief and intelligent massage functions. This method continuously collects information on the pressure distribution and duration on various parts of the bedridden person's body, identifying their posture and local pressure state. When excessive pressure or prolonged pressure is detected in a certain area, the control unit will inflate or deflate the corresponding airbag, effectively dispersing pressure and reducing the risk of pressure sores. Simultaneously, the alternating inflation and deflation of the airbag units creates a rhythmic, gentle massage, promoting local blood circulation and tissue activity.

[0003] The existing technology has the following shortcomings:

[0004] In the inflation and deflation process of grid-distributed airbags, existing technologies often only focus on the overall pressure distribution and single-point threshold control, lacking analysis of the dynamic micro-vibration response of users in scenarios involving frequent and subtle turning over. When the alternating inflation and deflation frequency of the airbags couples with the micro-vibration frequency of the human body, it is highly likely to trigger a local resonance effect, causing periodic abnormal stress fluctuations in specific areas of the airbag. This problem is difficult to detect in the short term, but if it persists for a long time, it can cause persistent deep tissue damage, significantly increasing the risk of pressure ulcers and seriously affecting the recovery process and nursing safety of bedridden patients.

[0005] The information disclosed in the background section is only intended to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0006] The purpose of this invention is to provide a method and system for intelligent inflation and deflation of grid-distributed massage airbags to solve the problems mentioned in the background art.

[0007] To achieve the above objectives, the present invention provides the following technical solution: a method for intelligent inflation and deflation of a grid-distributed massage airbag, comprising:

[0008] A spatiotemporal distribution matrix based on the air pressure fluctuation field of the mattress is constructed inside the mattress. The pressure changes of each airbag during the inflation and deflation cycle are transformed into a continuous air pressure disturbance sequence. The continuous air pressure disturbance sequence is then subjected to Fourier transform to extract the main frequency component, forming a complete airbag alternating frequency spectrum as the input baseline.

[0009] Based on the input baseline, a high-sensitivity inertial measurement unit matrix is introduced to obtain the body local posture micro-disturbance curve, and the instantaneous micro-vibration frequency in the lying position is extracted through multi-dimensional displacement vector fitting, and then the wavelet packet decomposition is used for multi-scale processing of the instantaneous micro-vibration frequency to obtain the human body micro-vibration spectrum diagram;

[0010] The human body micro-vibration spectrum diagram and the air bag alternating frequency spectrum diagram are mapped in the unified frequency domain, a frequency spectrum overlap weight matrix is constructed, and a weighted mapping is performed on the frequency peak in the phase synchronization window to extract the frequency band with time sequence overlap characteristics as the coupling candidate frequency band.

[0011] For the coupling candidate frequency band, the cross-power spectral density is calculated in the sliding time window, and the normalized coherence function is combined to reconstruct the time domain coupling sequence to generate a multi-parameter coupling degree vector composed of frequency domain coincidence rate and phase locking degree.

[0012] Based on the multi-parameter coupling degree vector, covariance analysis is performed on the coupling factors of each frequency band, and a correction function is established combined with the dynamic risk factor to calculate the coupling coefficient between the air bag alternating inflation and deflation frequency and the human body micro-vibration frequency.

[0013] When the coupling coefficient gradually approaches the set threshold, the correction function is called to trigger the adaptive frequency offset mechanism to perform micro-disturbance modulation on the air bag alternating inflation and deflation frequency, so that the air bag alternating inflation and deflation frequency actively avoids the sensitive frequency band.

[0014] Preferably, the step of obtaining the air bag alternating frequency spectrum diagram comprises:

[0015] A digital air pressure sensor is embedded in each air bag unit inside the mattress to obtain pressure data of the air bag during inflation and deflation, and combined with spatial coordinate information to form a time-space distribution matrix varying with time;

[0016] After obtaining the time-space distribution matrix, interpolation and difference processing are performed on the time-space distribution matrix to form a continuous air pressure disturbance sequence;

[0017] After obtaining the continuous air pressure disturbance sequence, fast Fourier transform is used to convert it into a frequency domain signal, and the main frequency and secondary frequency components are identified;

[0018] After identifying the main frequency and secondary frequency components, the frequency domain signal is integrated with the spatial coordinates and time sequence to construct a three-dimensional matrix containing time, spatial position and frequency, and expanded into a two-dimensional frequency spectrum diagram to form the air bag alternating frequency spectrum diagram.

[0019] Preferably, the step of obtaining the human body micro-vibration spectrum diagram comprises:

[0020] High-sensitivity inertial measurement units are arranged at different key positions on the surface of the mattress to synchronously collect three-dimensional acceleration and three-dimensional angular velocity signals and form a posture data stream;

[0021] After obtaining the posture data stream, filtering and calibration are performed to obtain a posture micro-disturbance curve, and a posture change trajectory conforming to the lying posture direction is formed through integration and coordinate transformation;

[0022] After obtaining the posture change trajectory, three-dimensional displacement components are combined into a multi-dimensional displacement vector, and instantaneous micro-vibration frequencies are extracted using polynomial fitting within a sliding time window;

[0023] After obtaining the instantaneous micro-vibration frequencies, wavelet packet decomposition is performed on the instantaneous micro-vibration frequency sequence, and a multi-dimensional frequency spectrum map is formed by combining energy distribution and spatial position to obtain a human micro-vibration spectrum map.

[0024] Preferably, the wavelet packet decomposition uses Daubechies wavelet as the basis function, and the sub-band signals obtained by decomposition are evaluated for energy distribution characteristics through energy entropy calculation to identify the main micro-vibration frequency bands corresponding to breathing, heartbeat and muscle twitch, thereby improving the resolution and integrity of the frequency spectrum map.

[0025] Preferably, the step of extracting frequency bands with time sequence overlap characteristics as coupling candidate frequency bands includes:

[0026] After obtaining the air bag alternating frequency spectrum map and the human micro-vibration frequency spectrum map, they are projected into a unified frequency domain, and missing points are filled by re-sampling and interpolation, while high-frequency noise is removed by a band-pass filter;

[0027] A frequency spectrum overlap weight matrix is constructed in the unified frequency domain, the product of the air bag spectrum amplitude and the human spectrum amplitude at each frequency point is taken as the initial weight value and normalized, and the local frequency spectrum of the human body corresponding to different air bag units is calculated to form the frequency spectrum overlap weight matrix;

[0028] After obtaining the frequency spectrum overlap weight matrix, a phase synchronization window is introduced to weight map the frequency peaks, and the weight values are corrected according to the phase difference size;

[0029] After phase weighting, a threshold interval is set to extract frequency points that continuously exceed the threshold interval and combine them into frequency bands, and spatial position information is used to mark the human body parts and air bag units, thereby forming coupling candidate frequency bands.

[0030] Preferably, the step of generating a multi-parameter coupling degree vector composed of frequency domain coincidence rate and phase locking degree includes:

[0031] After obtaining the coupling candidate frequency band, a sliding window is introduced on the time axis and the cross-power spectral density of the airbag alternating inflation and deflation pressure disturbance signal and the human body micro-vibration signal is calculated to reveal the frequency domain coincidence characteristics of the two signals in the candidate frequency band;

[0032] After obtaining the cross-power spectral density, the phase relationship between the airbag alternating inflation and deflation pressure disturbance signal and the human body micro-vibration signal is analyzed by combining the normalized coherence function, and the phase difference is extracted in the sliding window to reconstruct the time domain coupling sequence, thereby reflecting the phase locking condition of the airbag signal and the human body signal in the candidate frequency band;

[0033] After obtaining the time domain coupling sequence, parameterization processing is performed on it, the integral result of the cross-power spectral density is taken as the frequency domain coincidence rate, the mean and variance calculation results of the normalized coherence function are taken as the phase locking degree, and a multi-parameter coupling degree vector is formed by combining the two.

[0034] Preferably, the step of calculating the coupling coefficient between the airbag alternating inflation frequency and the human body micro-vibration frequency comprises:

[0035] After obtaining the multi-parameter coupling degree vector composed of the frequency domain coincidence rate and the phase locking degree, the multi-parameter coupling degree vector is classified and arranged according to the frequency segment, and the energy superposition degree and the phase synchronization stability are respectively represented in each sub-frequency segment;

[0036] After completing the frequency segment division and vector arrangement, covariance analysis is performed on the coupling factors in each sub-frequency segment, and the stability of the coupling relationship is identified;

[0037] After obtaining the covariance results of each sub-frequency segment, the covariance results are weighted and corrected in combination with the dynamic risk factor, a correction function is established, and the coupling factors are adjusted;

[0038] After the correction function is established, the corrected coupling factors are normalized and weighted averaged in the candidate frequency band range, thereby obtaining the final coupling coefficient which can reflect energy, phase, stability and risk factors.

[0039] Preferably, the step of performing perturbation modulation on the airbag alternating inflation frequency comprises:

[0040] After the coupling coefficient is calculated, the coupling coefficient is monitored in real time, and when the coupling coefficient output by the correction function gradually approaches the set threshold and continuously rises in multiple time windows, the correction function is called to trigger the adaptive frequency offset mechanism;

[0041] After triggering the adaptive frequency offset mechanism, the direction and amplitude of the frequency offset are determined based on the results of the correction function, and the offset is performed in a step-by-step manner when the coupling coefficient approaches different degrees until the airbag alternating inflation frequency moves out of the sensitive frequency band;

[0042] After the direction and amplitude of the frequency offset are determined, the offset result is applied to the inflation and deflation control strategy and the new coupling coefficient is monitored in real time. When the coupling coefficient decreases significantly, the current setting is maintained. When the coupling coefficient is still close to the set threshold, the offset is continued, and comfort constraints are introduced during the modulation process to keep the inflation and deflation rhythm within the range that the human body can tolerate, thereby achieving active avoidance of the air bag inflation and deflation frequency.

[0043] A grid-distributed intelligent massage air bag inflation and deflation system, comprising a gas pressure disturbance spectrum construction module, a human body micro-vibration spectrum acquisition module, a spectrum overlap weight extraction module, a multi-parameter coupling degree generation module, a coupling coefficient calculation module, and an adaptive frequency offset module;

[0044] The gas pressure disturbance spectrum construction module constructs a space-time distribution matrix based on the bed surface gas pressure fluctuation field inside the mattress, converts the pressure changes of each air bag during the inflation and deflation cycle into a continuous gas pressure disturbance sequence, and performs Fourier transform on the continuous gas pressure disturbance sequence to extract the main frequency component, forming an air bag alternating frequency spectrum graph as an input baseline;

[0045] The human body micro-vibration spectrum acquisition module, based on the input baseline, introduces a high-sensitivity inertial measurement unit matrix, acquires a body local posture micro-disturbance curve, and extracts the instantaneous micro-vibration frequency in the lying posture through multi-dimensional displacement vector fitting. Then, the instantaneous micro-vibration frequency is processed by wavelet packet decomposition for multi-scale processing to obtain a human body micro-vibration spectrum graph;

[0046] The spectrum overlap weight extraction module maps the human body micro-vibration spectrum graph and the air bag alternating frequency spectrum graph in a unified frequency domain, constructs a spectrum overlap weight matrix, and performs weighted mapping on the frequency peak in the phase synchronization window to extract the frequency band with time sequence overlap characteristics as the coupling candidate frequency band;

[0047] The multi-parameter coupling degree generation module calculates the cross-power spectral density in the sliding time window for the coupling candidate frequency band, and reconstructs the time-domain coupling sequence combined with the normalized coherence function to generate a multi-parameter coupling degree vector composed of frequency domain coincidence rate and phase locking degree;

[0048] The coupling coefficient calculation module, based on the multi-parameter coupling degree vector, performs covariance analysis on the coupling factors of each frequency band, and establishes a correction function combined with the dynamic risk factor to calculate the coupling coefficient between the air bag alternating inflation and deflation frequency and the human body micro-vibration frequency;

[0049] The adaptive frequency offset module, when the coupling coefficient gradually approaches the set threshold, calls the correction function to trigger the adaptive frequency offset mechanism, and performs micro-disturbance modulation on the air bag alternating inflation and deflation frequency to actively avoid the sensitive frequency band.

[0050] In the above technical solution, the present application provides technical effects and advantages:

[0051] The present application realizes real-time quantitative identification of the coupling relationship between the dynamic behavior of the air bag and the subtle posture changes of the human body by constructing a unified frequency domain mapping relationship of the air bag alternating frequency spectrum and the human body micro-vibration frequency spectrum, and introducing mutual power spectrum density and coherence function calculation means, effectively improving the early perception ability of the resonance risk. Compared with the traditional scheme which only relies on pressure threshold for control, the present application can accurately capture the dynamic interaction process in the frequency domain, identify the coupling trend and generate a multi-parameter coupling degree vector, so that the control strategy is upgraded from a static threshold reaction to a dynamic frequency domain perception, effectively reducing the formation of local periodic stress fluctuations and significantly reducing the risk of bedsores.

[0052] When the coupling coefficient is close to the set threshold, the present application triggers an adaptive frequency offset mechanism by calling a correction function, which can actively perturb and modulate the air bag inflation and deflation frequency, so as to avoid the sensitive frequency band of the human body while maintaining the massage and decompression functions, thereby realizing more intelligent and safer air bag inflation and deflation control. The regulation process is real-time and gradual, which can realize dynamic avoidance of resonance risk without affecting the overall mattress working rhythm, effectively improving the stability and comfort of intelligent nursing equipment in the long-term clinical use process, and providing more physiological adaptive personalized intervention strategies for bedridden patients. BRIEF DESCRIPTION OF DRAWINGS

[0053] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed in the embodiments will be briefly introduced as follows. Obviously, the drawings in the following description are only some embodiments described in the present application, and other drawings can also be obtained by those skilled in the art based on these drawings.

[0054] Figure 1 The method flow chart of the present application is a grid distributed massage air bag intelligent inflation and deflation method.

[0055] Figure 2 The module schematic diagram of the present application is a grid distributed massage air bag intelligent inflation and deflation system.

[0056] Figure 3 The flow chart of obtaining the human body micro-vibration frequency spectrum of the present application is shown in the following figure.

[0057] Figure 4 The flow chart of extracting the frequency band with time sequence overlap characteristics as the coupling candidate frequency band of the present application is shown in the following figure. DETAILED DESCRIPTION

[0058] Example implementations are now described with reference to the drawings. Example implementations can, however, be implemented in many different forms and should not be construed as limited to the examples set forth herein; rather, these example implementations are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the example implementations to those skilled in the art.

[0059] The present application provides a method for intelligent inflation and deflation of a grid-distributed massage air bag as shown in Figure 1 、 3 , 4, the method comprising the following steps:

[0060] A space-time distribution matrix based on the air pressure fluctuation field of the bed surface is constructed inside the mattress, the pressure change of each air bag in the inflation and deflation cycle is converted into a continuous air pressure disturbance sequence, and the Fourier transform is performed on the air pressure disturbance sequence to extract the main frequency component, forming a complete air bag alternating frequency spectrum as the input baseline;

[0061] The air bag alternating frequency spectrum is obtained, including the following continuous sub-steps:

[0062] A digital air pressure sensor is embedded in each air bag unit inside the mattress, the measurement accuracy of each sensor is preferably 0.1 Pa, and the sampling frequency is set to 100 Hz to ensure that the rapid pressure fluctuations formed during the inflation and deflation of the air bag can be captured. During the manufacturing process of the mattress, a unique two-dimensional coordinate label is assigned to each air bag unit, for example, taking the center of the mattress as the origin, the spatial position of each air bag is calibrated using the Cartesian coordinate system, so that each sampling point contains the air bag number, time stamp and corresponding spatial coordinates. The sensor records the gradual rise of the internal pressure of the air bag during inflation, and records the gradual decline of the internal pressure of the air bag during deflation, and through continuous collection, the original pressure data stream of the air bag changing with time is formed. Taking a complete inflation and deflation cycle as the sampling window, for example, setting each cycle to 60 seconds, then each air bag can generate 6000 continuous sampling points in a cycle. Aligning the original pressure data of all air bags according to the time dimension and arranging them in the order of spatial coordinates, a space-time distribution matrix can be obtained, the horizontal axis of the matrix represents the time sequence, the vertical axis represents the spatial position of the air bag, and each element in the matrix represents the air pressure value at a certain position at a certain time. The space-time distribution matrix obtained in this way directly reflects the air pressure distribution change of the bed surface in a cycle.

[0063] In order to transform the discrete data in the above spatiotemporal distribution matrix into a sequence of air pressure disturbances with continuity, interpolation and difference processing are needed. Specifically, in the time dimension, a cubic spline interpolation algorithm is used to compensate the data between adjacent sampling points, so that the originally discrete pressure curve appears as a smooth continuous function on the time axis, avoiding waveform gaps caused by sampling intervals. In the spatial dimension, the pressure values of adjacent airbags are linearly interpolated, so that the airbags at different positions in the matrix can form a continuous spatial distribution surface. After obtaining the continuous matrix, first-order difference calculation is performed on its time axis, that is, the pressure value at each time point is subtracted from the pressure value at the previous time point, thereby extracting the disturbance change rate of the airbag during the inflation and deflation process. The differentiated curve can highlight the dynamic changes of the airbag in a short period of time, such as the rapid rise of the airbag pressure at the beginning of inflation and the rapid drop of the airbag pressure at the end of deflation. Finally, the smoothed curve after interpolation processing is combined with the disturbance rate curve obtained by difference calculation to form a complete continuous air pressure disturbance sequence, which can truly reflect the dynamic disturbance process of the airbag during the entire inflation and deflation cycle.

[0064] After obtaining the continuous air pressure disturbance sequence, Fourier transform is needed to convert the time domain signal into the frequency domain signal to extract the main frequency components. The specific steps are as follows: first, discrete Fourier transform (DFT) is performed on the continuous air pressure disturbance sequence, and fast Fourier transform (FFT) algorithm is used to realize it. The FFT point number is set to an integer power of 2 according to the sampling length, for example, if the sampling point is 8192, the FFT point number is set to 8192. Through FFT operation, the frequency spectrum distribution of the air pressure disturbance signal in the frequency range of 0 Hz to 50 Hz can be obtained. Then, the amplitude and phase of each frequency component are calculated, where the amplitude represents the energy size of the frequency in the disturbance signal, and the phase represents the offset of the signal on the time axis. Further, the amplitude is normalized so that the maximum amplitude corresponds to 100%, thereby forming a standardized amplitude spectrum. By observing the amplitude spectrum, the main frequency peak value in the airbag inflation and deflation process can be identified, for example, the main frequency peak value between 0.2 Hz and 0.5 Hz often corresponds to the rhythm of the airbag alternating inflation and deflation. In addition, there will be several secondary peak values with low amplitude in the high frequency interval, which may be caused by the nonlinear elastic response of the airbag material, the vortex effect of air flow or sensor electrical noise. By retaining the main frequency and secondary frequency peak values at the same time, this method can fully reveal the complex frequency characteristics of the airbag in the dynamic process, rather than just focusing on the main frequency components.

[0065] After completing the Fourier transform and obtaining the primary and secondary frequency components, it is necessary to re-integrate the frequency domain results with spatial coordinates and time series to generate a complete air bag alternating frequency spectrum. The specific implementation is to introduce a frequency dimension on the basis of the original space-time distribution matrix to construct a three-dimensional matrix, where the first dimension is time, the second dimension is spatial position, and the third dimension is frequency. Each three-dimensional matrix unit represents the corresponding frequency component and its amplitude size at a specific time and spatial position. In order to display intuitively, the three-dimensional matrix is expanded into a two-dimensional frequency spectrum, where the horizontal axis is frequency, the vertical axis is spatial position, and the color depth represents the amplitude size. Through this frequency spectrum, it can be observed whether different air bags have similar frequency response patterns at the same time and whether there are obvious frequency differences between different regions. For example, air bags located below the body's bony parts may exhibit stronger primary frequency responses, while air bags located at the edge of the body exhibit more secondary frequency components. The final air bag alternating frequency spectrum not only serves as an input baseline for subsequent coupling analysis, but also serves as a visualization tool to help nursing staff intuitively understand the overall state and local differences of the mattress during dynamic operation.

[0066] Through the above specific steps, the whole process from the collection of original air bag pressure data to the generation of continuous disturbance sequence, to the extraction of frequency domain components and the construction of frequency spectrum is realized. Compared with the traditional method of relying only on pressure threshold for judgment, this method can more comprehensively and meticulously reveal the dynamic characteristics of air bags during the inflation and deflation process, and provide a basis for subsequent human micro-vibration coupling analysis in the form of frequency spectrum.

[0067] Based on the input baseline, a high-sensitivity inertial measurement unit matrix is introduced to obtain the local posture disturbance curve of the body, and the instantaneous micro-vibration frequency in the lying position is extracted through multi-dimensional displacement vector fitting, and then the instantaneous micro-vibration frequency is processed by wavelet packet decomposition to obtain the human micro-vibration spectrum;

[0068] As shown in Figure 3 , obtaining the human micro-vibration spectrum specifically includes the following sub-steps:

[0069] A plurality of high-sensitivity inertial measurement units are arranged at different key positions on the surface of the mattress, each of which preferably comprises a three-axis acceleration sensor and a three-axis gyroscope, and the sampling frequency is set to be above 200 Hz to ensure that the instantaneous micro-vibration of the human body in a lying position due to breathing, heartbeat, turning over or muscle twitching can be captured. The inertial measurement units are arranged in a matrix manner, and the arrangement interval is optimized according to the main stress area of the human body, for example, the sensor distribution is more dense in the sacrococcygeal part, scapular part, thigh and ankle part, so as to enhance the signal capture ability of the key bone protrusion area. In actual use, all the inertial measurement units synchronously collect three-dimensional acceleration and three-dimensional angular velocity signals, and form a six-dimensional attitude data stream through time stamp alignment. Through Kalman filtering algorithm, the original signal is denoised and drift compensated to obtain a more smooth and true attitude change curve. On this basis, the relative position of each measuring point in the spatial coordinate is calibrated, so that the output data of each sensor not only contains the attitude change itself, but also can realize one-to-one correspondence with the input baseline of the bed surface pressure fluctuation field in space, thereby ensuring the consistency of spatial resolution in subsequent spectrum fusion.

[0070] After obtaining the filtered and calibrated six-dimensional attitude data stream, the data is converted into an attitude micro-disturbance curve that can intuitively reflect the subtle dynamic characteristics in a lying position. Specifically, first, the three-axis acceleration signals are integrated to obtain a displacement curve, and the three-axis angular velocity signals are integrated to obtain an angle change curve, and then the displacement and angle curves are uniformly mapped into the human anatomy coordinate system through a coordinate transformation matrix, so as to obtain an attitude change trajectory conforming to the actual lying position direction. In this trajectory, the breathing motion appears as a low-frequency, periodic displacement fluctuation, the heartbeat motion appears as a high-frequency, micro-amplitude vibration, and the turning over or muscle twitching appears as an instantaneous large-amplitude displacement peak. In order to extract the instantaneous micro-vibration frequency in a lying position, the present application proposes a multi-dimensional displacement vector fitting method, that is, at each sampling time, the three-dimensional displacement components are combined into a multi-dimensional displacement vector, and then a polynomial fitting method within a sliding time window is used to fit the multi-dimensional displacement vector, so as to obtain a locally smooth displacement change trend. By performing a fast Fourier transform on the fitted curve, the instantaneous main vibration frequency can be extracted within a short time segment, and the corresponding relationship between the instantaneous main vibration frequency and the spatial position is established, and finally a set of instantaneous micro-vibration frequency data with resolution in time and space is formed.

[0071] After obtaining the instantaneous micro-vibration frequency data, it is necessary to further perform multi-scale decomposition on the data to reveal the complex dynamic components of the human body in a lying position. The specific implementation is to perform wavelet packet decomposition operation on the instantaneous micro-vibration frequency sequence, use Daubechies wavelet as the base function, and recursively decompose the frequency signal in multiple hierarchical frequency bands. For example, the frequency range of 0 Hz to 20 Hz is decomposed into several sub-bands, and the width of each sub-band is gradually reduced with the increase of the decomposition layer, so as to obtain higher frequency resolution. In the decomposition process, the energy distribution characteristics of each sub-band signal are evaluated by energy entropy calculation method, and the energy concentrated frequency band often corresponds to the main micro-vibration source of the human body, such as 0.1 Hz to 0.3 Hz for breathing, 1 Hz to 2 Hz for heartbeat, and more than 5 Hz for muscle twitch. By layer-by-layer reconstruction, the energy information of each sub-band is combined with its corresponding spatial position to finally form a multi-dimensional frequency spectrum. The horizontal axis of the frequency spectrum represents the frequency, the vertical axis represents the spatial position, and the color depth represents the frequency energy size, so as to intuitively display the micro-vibration distribution characteristics of the human body in a lying position. Compared with the traditional method which only relies on average displacement or single frequency analysis, the present application realizes multi-scale characterization of non-stationary signal through wavelet packet decomposition, which not only reveals stable rhythms such as breathing and heartbeat, but also captures random and instantaneous muscle tremor and turning disturbance, greatly improving the authenticity and integrity of the frequency spectrum. The finally obtained micro-vibration frequency spectrum will be coupled with the previously constructed airbag alternating frequency spectrum in a unified frequency domain for subsequent coupling analysis, providing data support for preventing resonance.

[0072] Through the implementation of the above sub-steps, the present application introduces a high-sensitivity inertial measurement unit matrix based on the input baseline, not only obtains the posture micro-disturbance curve of the local body in a lying position, but also accurately extracts the instantaneous micro-vibration frequency through a multi-dimensional displacement vector fitting method, and then uses wavelet packet decomposition to complete the multi-scale human micro-vibration spectrum construction. Compared with the existing method which only relies on a single sensor or average value processing, the present application has stronger data integrity and higher frequency resolution.

[0073] The human micro-vibration frequency spectrum and the airbag alternating frequency spectrum are mapped in a unified frequency domain to construct a frequency spectrum overlap weight matrix, and the frequency peaks are weighted and mapped in a phase synchronization window to extract the frequency band with time sequence overlap characteristics as the coupling candidate band.

[0074] As shown in Figure 4 , extracting the frequency band with time sequence overlap characteristics as the coupling candidate band includes the following sub-steps:

[0075] After obtaining the airbag alternating frequency spectrum and the human body micro-vibration frequency spectrum, the two are projected to a unified frequency domain to ensure that frequency data from different sources can be compared and integrated under the same benchmark. Specifically, the frequency range of the airbag alternating frequency spectrum is usually concentrated between 0.1 Hz and 2 Hz, while the human body micro-vibration frequency spectrum covers a wider range, from the low frequency of 0.1 Hz of respiration to the high frequency of 2 Hz or even higher of heartbeat. In order to ensure the consistency of frequency resolution, the present application re-samples the two types of spectral data to the same frequency scale, for example, to establish a unified frequency axis with a frequency interval of 0.01 Hz. In this process, linear interpolation is used to fill in the missing points, and a band-pass filter is used to remove high-frequency noise outside the range of interest (such as more than 10 Hz). After this processing, the airbag spectrum and the human body spectrum are unified to the same frequency domain, and each frequency point corresponds to two independent energy amplitudes of the airbag and the human body, laying a data foundation for subsequent construction of the spectral overlap weight matrix.

[0076] On the basis of the unified frequency domain, the spectral overlap weight matrix is constructed to quantify the energy overlap degree of the airbag spectrum and the human body spectrum at different frequency points. The specific implementation is as follows: at each frequency point, take the product of the airbag spectrum amplitude and the human body spectrum amplitude as the initial weight value of the point, and then normalize the weight value to fall within the interval of 0 to 1. The closer the weight value is to 1, the higher the energy overlap degree of the frequency point, i.e. the frequency point is more likely to be a potential coupling point. In order to further enhance the resolution of the spatial dimension, the local frequency spectrum of the human body corresponding to different airbag units is calculated separately, and finally a three-dimensional overlap weight matrix is formed, where the first dimension is frequency, the second dimension is spatial position, and the third dimension is weight value. In this way, not only can the overlap in frequency be identified, but also the spectral coupling strength between different human body parts and corresponding airbags can be revealed. For example, when the sacrococcygeal airbag and the human body frequency spectrum of the adjacent bony process region simultaneously appear a higher energy peak value at 0.25 Hz, the overlap weight matrix will show a high weight at the corresponding position, which means that there may be a potential resonance risk in this area.

[0077] After obtaining the spectral overlap weight matrix, a phase factor is further introduced to distinguish the case of energy superposition but not synchronization from the case of real time coupling. Specifically, a phase synchronization window is introduced in the unified frequency domain, and each frequency peak is weighted and mapped. The length of the phase synchronization window is set according to the analysis requirements, for example, taking a complete inflation and deflation cycle as the window length, and calculating the phase difference between the airbag spectrum and the human body spectrum at the corresponding frequency within the window. If the phase difference is within a preset threshold range, for example, less than 30 degrees, it is considered that the frequency point has synchronization in time sequence, and at this time the weight value in the overlap weight matrix is multiplied by a phase synchronization factor (such as 0.8 to 1.0), thereby forming a phase-corrected weight value; if the phase difference is too large, it means that although the two are close in frequency, they are not synchronized in time sequence, and at this time the weight value is attenuated. In this way, the spectral overlap weight matrix not only reflects the energy overlap in frequency, but also combines the phase matching degree in time, ensuring that the final selected frequency point really has coupling potential.

[0078] After phase weighting processing, the frequency band with time sequence overlap characteristics is extracted as a coupling candidate frequency band. Specifically, a threshold interval is set, for example, when the weight value after phase weighting continuously exceeds the threshold interval (i.e. the maximum value of the threshold interval) 0.7, it is considered that the frequency band has high overlap and synchronization. Multiple points that continuously meet the conditions on the frequency axis are combined into a complete frequency band, and the corresponding human body parts and airbag units are marked in the spatial dimension. The coupling candidate frequency band obtained in this way not only contains the frequency range, but also contains the spatial position information, so as to accurately describe which parts and which frequency bands may have resonance risk. For example, if a candidate frequency band is formed between 0.28 Hz and 0.32 Hz, and the candidate frequency band appears in the high weight interval at the same time in the shoulder blade and thigh parts, it means that the airbag frequency and the human body micro-vibration frequency in these areas have high overlap and phase synchronization, and it is extremely likely to trigger local stress amplification effect. Through such an extraction process, the final coupling candidate frequency band provides a clear input for subsequent mutual power spectrum density calculation and coupling quantification.

[0079] Through the above steps, the whole process from the establishment of the unified frequency domain to the construction of the spectral overlap weight matrix, the introduction of the phase synchronization window, and finally the extraction of the coupling candidate frequency band is realized. This method not only considers the energy overlap degree, but also combines the phase matching relationship, and realizes the partition analysis in the spatial dimension, breaking through the existing technical solutions which only take the pressure threshold or average frequency as the index. The present application can more comprehensively and accurately reveal the potential coupling characteristics between human body micro-vibration and airbag inflation and deflation frequency, and provides a technical basis for actively avoiding the resonance sensitive frequency band.

[0080] For the coupling candidate frequency band, the cross-power spectral density is calculated in the sliding time window, and the normalized coherence function is combined to reconstruct the time-domain coupling sequence to generate a multi-parameter coupling degree vector composed of frequency domain coincidence rate and phase locking degree;

[0081] A multi-parameter coupling degree vector composed of frequency domain coincidence rate and phase locking degree is generated, specifically including the following consecutive sub-steps:

[0082] After obtaining the coupling candidate frequency band, a sliding window needs to be introduced on the time axis to calculate the cross-power spectral density. Specifically, the original signal is divided into the airbag alternating inflation and deflation pressure disturbance signal and the human body micro-vibration signal, both of which extract the corresponding frequency components in the candidate frequency band range. In order to ensure the balance between time domain and frequency domain, the sliding window length is set to ten times the main frequency period of the candidate frequency band, for example, the candidate frequency is 0.3 Hz, then the window length is set to about 33 seconds, and a 50% overlap rate is used between each window to ensure the balance of time resolution and stability. The signals in each sliding window are subjected to fast Fourier transform to obtain the power spectrum of the airbag signal and the human body signal, respectively, and then the two are cross-correlated to obtain the cross-power spectral density function. The cross-power spectral density not only reflects the energy distribution of the two signals in the frequency domain, but also reveals the mutual coupling strength between them. If the cross-power spectral density shows obvious peak value in the candidate frequency band range in a certain window, it indicates that there is a strong frequency domain coincidence feature between the airbag frequency and the human body micro-vibration frequency in that time period.

[0083] After obtaining the cross-power spectral density, the normalized coherence function needs to be combined to further analyze the phase relationship between the signals, so as to reconstruct the time-domain coupling sequence. Specifically, the coherence function is the ratio of the cross-power spectral density to the two individual power spectrums, and its value range is between 0 and 1, the closer the value is to 1, the higher the correlation of the two signals at a certain frequency. The coherence function is calculated for each frequency point in the candidate frequency band, and the result is normalized to eliminate the bias caused by the difference in signal amplitude. Subsequently, by calculating the change of the coherence function over time in the sliding window, a dynamic coherence curve can be obtained, which can reflect the phase locking degree of the airbag signal and the human body signal in the candidate frequency band. In order to convert this frequency domain coherence relationship into time domain performance, a phase reconstruction method is proposed, that is, in each sliding window, the phase difference is extracted according to the peak position of the coherence function, and the phase difference is mapped into the synchronization offset of the time domain waveform, thereby generating a time-domain coupling sequence. The sequence can intuitively describe the phase matching of the airbag signal and the human body signal in different time periods, for example, when the breathing period and the airbag alternating period are synchronized, the time-domain coupling sequence will show high-amplitude continuous fluctuation; while the two are mismatched, it will show low-amplitude or intermittent fluctuation.

[0084] After reconstructing the time-domain coupling sequence, the time-domain coupling sequence is parameterized to form a multi-parameter coupling degree vector composed of a frequency domain overlap rate and a phase locking degree. Specifically, the frequency domain overlap rate is obtained by integrating the cross power spectral density, that is, the cross power spectral density is normalized and integrated in the candidate frequency band to obtain the proportion of energy overlap; the phase locking degree is obtained by comprehensive calculation of the mean and variance of the normalized coherence function to measure the stability of phase synchronization in the entire time window. The two parameters are combined in the form of a vector, that is, a multi-parameter coupling degree vector is formed, for example, represented as C= (R_f, L_p), where R_f is the frequency domain overlap rate and L_p is the phase locking degree. Further, in order to enhance the robustness of the vector, it is repeatedly calculated in multiple sliding windows, and the final coupling degree vector is obtained by weighted average, so that it can reflect both the short-term instantaneous coupling characteristics and the long-term overall trend. The finally generated multi-parameter coupling degree vector not only provides basic data for subsequent covariance analysis and dynamic correction function construction, but also can be used as a key indicator for real-time monitoring to prompt nursing staff when and where potential resonance risks may exist.

[0085] Through the above sub-steps, the energy overlap is revealed by using the cross power spectral density in the coupling candidate frequency band, the phase synchronization is extracted by combining the normalized coherence function, and the method of time-domain coupling sequence reconstruction and multi-parameter vectorization description is realized. Compared with the existing technology which only relies on single power spectrum or single point coherence analysis, the method realizes double coupling analysis from the frequency domain to the time domain, forms a multi-parameter coupling degree vector containing a frequency domain overlap rate and a phase locking degree, and thus provides a basis for accurately identifying the deep coupling relationship between the airbag alternating frequency and the human body micro-vibration frequency.

[0086] Based on the multi-parameter coupling degree vector, covariance analysis is performed on the coupling factors of each frequency band, and a correction function is established by combining a dynamic risk factor to calculate the coupling coefficient between the airbag alternating inflation and deflation frequency and the human body micro-vibration frequency.

[0087] The coupling coefficient between the airbag alternating inflation and deflation frequency and the human body micro-vibration frequency is calculated, including the following sub-steps:

[0088] After obtaining the multi-parameter coupling degree vector composed of the frequency domain coincidence rate and the phase locking degree, the vector is classified and arranged according to the frequency band. Specifically, the coupling candidate frequency band is divided into several sub-frequency bands, and each sub-frequency band corresponds to a group of multi-parameter coupling degree vectors. In the same sub-frequency band, the frequency domain coincidence rate represents the superposition degree of the energy of the airbag signal and the human body signal in the frequency range, and the phase locking degree represents the synchronization stability of the two in time sequence. In this way, each sub-frequency band can be characterized by a group of parameterized vectors. Unlike the traditional method of only focusing on a single peak or a single indicator, the present application considers both energy factors and phase factors in this sub-step, thereby ensuring the comprehensiveness and reliability of subsequent analysis.

[0089] After completing the frequency band division and multi-parameter vector arrangement, the coupling factors of each sub-frequency band need to be further explored by covariance analysis. Specifically, in each sub-frequency band, the multi-parameter coupling degree vector is usually composed of the results of several time windows, and there may be fluctuation differences between different time windows. In order to reveal the correlation between these fluctuations, the frequency domain coincidence rate and the phase locking degree are taken as two core variables, and the covariance in the time dimension is calculated. If in a certain sub-frequency band, the two variables show a highly correlated trend in different time windows, it means that the energy superposition and phase synchronization have consistency in that sub-frequency band, representing a real and stable coupling relationship; on the contrary, if the covariance result is low, it means that the coupling relationship in that frequency band is not stable enough, and it may be a short-term and occasional phenomenon. Through covariance analysis, not only can the stability of the coupling relationship in different frequency bands be quantitatively reflected, but also a statistical basis can be provided for the introduction of dynamic risk factors in the subsequent step.

[0090] After obtaining the covariance results of each sub-frequency band, the results need to be modified in combination with dynamic risk factors to establish a modified function that is more consistent with clinical and engineering practice. The so-called dynamic risk factor is a comprehensive parameter that affects the health of human tissues and the safety of nursing in actual use scenarios, such as the length of bed rest, whether the stressed part is a high-risk area (such as the sacrococcygeal part, the scapular part), individual physical differences, and environmental humidity and temperature. Different risk factors have different sensitivities to coupling relationships, for example, in the bone protrusion area, even if the covariance value is not particularly high, the risk of injury may still be significantly increased due to the vulnerability of the tissue under pressure. Therefore, in this sub-step, the results of covariance analysis are combined with dynamic risk factors to establish a modified function. The modified function adjusts the covariance results by weighting, so that in a high-risk scenario, the weight of the coupling relationship is increased, and in a low-risk scenario, the weight of the coupling relationship is appropriately reduced. In this way, the modified function can ensure that the final coupling measure not only reflects the statistical correlation between signals, but also takes into account the actual nursing risk.

[0091] After the correction function is established, the covariance results of each sub-frequency band are corrected using the function, and the final coupling coefficient is calculated accordingly. Specifically, for each sub-frequency band, the original covariance value is combined with the correction weight corresponding to the dynamic risk factor to obtain a corrected coupling factor; then, in the entire candidate frequency band range, all corrected coupling factors are normalized and weighted averaged to form a global coupling coefficient. The coupling coefficient integrates the frequency domain coincidence rate, phase locking degree, time stability and risk correction factor, and can comprehensively reflect the real coupling degree between the air bag alternating inflation and deflation frequency and the human body micro-vibration frequency. If the final coupling coefficient exceeds the set threshold, it can be determined that the current operating state has a resonance risk, and the adaptive frequency offset mechanism needs to be triggered for adjustment in the subsequent steps. Through this method, the invention not only realizes accurate quantification of the coupling relationship at the signal level, but also realizes forward-looking assessment of potential hazards at the risk management level, providing practical value for the safe operation of intelligent nursing mattresses.

[0092] Through the above steps, the whole process from the arrangement of the multi-parameter coupling degree vector, to the stability evaluation of the covariance analysis, to the introduction of the dynamic risk factor and the establishment of the correction function, to the final coupling coefficient reflecting the real risk level is realized. Compared with the prior art, this method not only considers the dual characteristics of energy and phase, but also reveals the time stability through the covariance statistical method, and further combines the clinical risk factor for correction, providing accurate trigger basis for subsequent adaptive frequency adjustment.

[0093] When the coupling coefficient gradually approaches the set threshold, the adaptive frequency offset mechanism is triggered by calling the correction function to perform perturbation modulation on the air bag alternating inflation and deflation frequency, so that the air bag alternating inflation and deflation frequency actively avoids the sensitive frequency band;

[0094] Performing perturbation modulation on the air bag alternating inflation and deflation frequency includes the following sub-steps:

[0095] After the completion of the coupling coefficient calculation, real-time monitoring is performed on the coupling coefficient. When the coupling coefficient output by the correction function gradually approaches the set threshold, for example, the set threshold is 0.75, and the calculation result continuously remains between 0.70 and 0.74 and shows an upward trend, it is judged that the current state has a triggering risk. At this stage, not only the coupling coefficient value at a single time point is considered, but also the change trend thereof in multiple time windows is analyzed. If the coupling coefficient remains rising in three consecutive time windows, it means that the air bag alternating inflation and deflation frequency and the human body micro-vibration frequency are gradually tending towards the sensitive frequency band, and it is extremely likely to break through the set threshold in a short time. In order to avoid the damage risk caused by lagging reaction, the present application calls the correction function to trigger the adaptive frequency offset mechanism at this time, ensuring that the frequency regulation can intervene in advance before the real resonance occurs. In this way, the monitoring of the coupling coefficient is not only a single point judgment, but also a double consideration of the numerical value and the change trend, thereby improving the accuracy and foresight of the triggering.

[0096] After triggering the adaptive frequency offset mechanism, the direction and amplitude of the frequency offset are determined based on the results of the correction function. Specifically, the correction function has been established in combination with the covariance result and the dynamic risk factor in the previous sub-step, so its output result not only contains the numerical value of the coupling coefficient, but also implicitly contains the sensitivity information of the risk distribution. By analyzing the output of the correction function, the risk distribution pattern of the current frequency band can be determined. If the risk is concentrated at the low end of the frequency axis, it means that the air bag alternating inflation and deflation frequency needs to be offset to the high frequency direction; on the contrary, if the risk is concentrated at the high end, it needs to be offset to the low frequency direction. The setting of the offset amplitude is based on the proximity of the coupling coefficient, for example, when the coupling coefficient is only slightly higher than the lower limit of safety, a small amplitude offset can be taken, such as adjusting within 0.05 Hz; when the coupling coefficient is close to the upper limit, a larger amplitude offset needs to be taken, such as 0.1 Hz or even higher. In specific implementation, the present application adopts a step-by-step adjustment strategy, that is, a small amplitude offset is first performed, and then the coupling coefficient is recalculated in the next time window. If the coupling coefficient after offset is still in the critical interval, a second offset is continued to be performed until the air bag alternating inflation and deflation frequency is completely moved out of the sensitive frequency band. Through this step-by-step adjustment method, the decline of comfort or hemodynamic disturbance caused by one-time large amplitude adjustment can be avoided, so as to ensure safety while maintaining the comfortable experience of the user.

[0097] After the direction and amplitude of the frequency offset are determined, perturbation modulation is performed on the alternating inflation and deflation frequency of the air bag, and feedback results are monitored in real time during the modulation. Specifically, the offset results are applied to the inflation and deflation control strategy of the air bag, such as adjusting the opening period of the electromagnetic valve or the output rhythm of the air pump to change the actual inflation and deflation frequency of the air bag. While performing the modulation, the synchronization monitoring of the human body micro-vibration frequency spectrum and the air bag alternating frequency spectrum is continuously maintained, and the new coupling coefficient is calculated in real time. If the coupling coefficient after the offset is significantly reduced and far away from the set threshold, it means that the modulation is effective, and the current frequency setting can be maintained; if the coupling coefficient after the offset is still close to the set threshold, or a new high-risk point appears in the new frequency band, the next round of frequency offset needs to be continued by calling the correction function. At the same time, during the entire modulation process, the comfort constraint is particularly introduced, that is, on the premise of avoiding sensitive frequency bands, the inflation and deflation rhythm after the modulation still needs to be maintained within the range that the human body can tolerate, for example, the frequency should not be higher than 1 Hz or lower than 0.1 Hz, to avoid affecting the normal decompression and massage effect. Finally, through this real-time feedback and iterative correction strategy, the active avoidance of the air bag inflation and deflation frequency is realized, which can dynamically move away from the sensitive frequency band of the human body micro-vibration, thereby effectively preventing the resonance effect and the risk of deep tissue damage caused thereby.

[0098] Through the above sub-steps, the present application proposes an adaptive frequency offset mechanism triggered by a correction function, realizes a complete closed-loop process from coupling coefficient monitoring, to risk-oriented offset direction and amplitude determination, to perturbation modulation under real-time feedback and comfort constraint, and the method can realize continuous adaptive adjustment under the condition of dynamic change of human body behavior, has higher sensitivity and safety, and provides technical support for intelligent nursing mattress in preventing pressure sores and deep tissue damage.

[0099] The present application realizes real-time quantitative identification of the coupling relationship between the dynamic behavior of the air bag and the subtle posture change of the human body by constructing a unified frequency domain mapping relationship between the air bag alternating frequency spectrum and the human body micro-vibration frequency spectrum, and introducing mutual power spectrum density and coherence function calculation means, effectively improving the early perception ability of the resonance risk. Compared with the traditional scheme which only relies on pressure threshold for control, the present application can accurately capture the dynamic interaction process in the frequency domain, identify the coupling trend and generate a multi-parameter coupling degree vector, so that the control strategy is upgraded from a static threshold reaction to a dynamic frequency domain perception, effectively reducing the formation of local periodic stress fluctuations, and significantly reducing the risk of pressure sores.

[0100] The application can actively perturb the air bag inflation and deflation frequency by calling the correction function when the coupling coefficient approaches the set threshold, so as to avoid the sensitive frequency band of the human body while maintaining the massage and decompression functions, thereby realizing more intelligent and safer air bag inflation and deflation control. The regulation process has real-time and gradual characteristics, can realize dynamic avoidance of resonance risk without affecting the overall mattress working rhythm, effectively improves the stability and comfort of intelligent nursing equipment in the long-term clinical use process, and provides more physiological adaptive personalized intervention strategies for bedridden patients.

[0101] The application provides a grid distributed massage air bag intelligent inflation and deflation system as shown in Figure 2 The grid distributed massage air bag intelligent inflation and deflation system comprises an air pressure disturbance frequency spectrum construction module, a human body micro-vibration frequency spectrum acquisition module, a spectrum overlap weight extraction module, a multi-parameter coupling degree generation module, a coupling coefficient calculation module and an adaptive frequency offset module.

[0102] The air pressure disturbance frequency spectrum construction module constructs a space-time distribution matrix based on the bed surface air pressure fluctuation field inside the mattress, converts the pressure change of each air bag in the inflation and deflation cycle into a continuous air pressure disturbance sequence, and performs Fourier transform on the air pressure disturbance sequence to extract the main frequency component, thereby forming an air bag alternating frequency spectrum graph as an input baseline.

[0103] The human body micro-vibration frequency spectrum acquisition module introduces a high-sensitivity inertial measurement unit matrix based on the input baseline, acquires a body local posture micro-disturbance curve, extracts an instantaneous micro-vibration frequency in a lying posture through multi-dimensional displacement vector fitting, and performs multi-scale processing on the instantaneous micro-vibration frequency by using wavelet packet decomposition to obtain a human body micro-vibration frequency spectrum graph.

[0104] The spectrum overlap weight extraction module maps the human body micro-vibration frequency spectrum graph and the air bag alternating frequency spectrum graph in a unified frequency domain, constructs a spectrum overlap weight matrix, and performs weighted mapping on the frequency peak in a phase synchronization window to extract a frequency band with time sequence overlap characteristics as a coupling candidate frequency band.

[0105] The multi-parameter coupling degree generation module calculates the cross-power spectral density in a sliding time window for the coupling candidate frequency band, and combines a normalized coherence function to reconstruct a time-domain coupling sequence, thereby generating a multi-parameter coupling degree vector composed of a frequency domain coincidence rate and a phase locking degree.

[0106] The coupling coefficient calculation module performs covariance analysis on the coupling factor of each frequency band based on the multi-parameter coupling degree vector, establishes a correction function combined with a dynamic risk factor, and calculates the coupling coefficient between the air bag alternating inflation and deflation frequency and the human body micro-vibration frequency.

[0107] The adaptive frequency offset module is used to call the correction function to trigger the adaptive frequency offset mechanism when the coupling coefficient gradually approaches the set threshold, and to perform perturbation modulation on the air bag alternating inflation and deflation frequency, so that the air bag alternating inflation and deflation frequency actively avoids the sensitive frequency band.

[0108] The embodiment of the present application provides a kind of grid distributed massage air bag intelligent inflation and deflation method, it is realized by the above-mentioned one kind of grid distributed massage air bag intelligent inflation and deflation system, and the specific method and process of a kind of grid distributed massage air bag intelligent inflation and deflation system are described in the above-mentioned embodiment of one kind of grid distributed massage air bag intelligent inflation and deflation method, and will not be repeated here.

[0109] The above only describes some exemplary embodiments of the present application by way of illustration, without doubt, for those skilled in the art, the described embodiments can be modified in various ways without departing from the spirit and scope of the present application. Therefore, the above drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of the claims of the present application.

Claims

1. A method for controlling the intelligent inflation and deflation of a grid-distributed massage airbag, characterized in that, Includes the following steps: A spatiotemporal distribution matrix based on the air pressure fluctuation field of the mattress is constructed inside the mattress. The pressure changes of each airbag during the inflation and deflation cycle are transformed into a continuous air pressure disturbance sequence. The continuous air pressure disturbance sequence is subjected to Fourier transform to extract the main frequency component and form an alternating frequency spectrum of the airbag as the input baseline. Based on the input baseline, a high-sensitivity inertial measurement unit matrix is ​​introduced to obtain the local posture micro-perturbation curve of the body. The instantaneous micro-vibration frequency under the lying position is extracted by multi-dimensional displacement vector fitting. Then, wavelet packet decomposition is used to process the instantaneous micro-vibration frequency on multiple scales to obtain the human body micro-vibration spectrum. The spectrum of human micro-vibration and the frequency spectrum of airbag alternation are mapped in a unified frequency domain to construct a spectrum overlap weight matrix. Then, the frequency peaks are weighted and mapped within the phase synchronization window to extract frequency bands with temporal overlap characteristics as coupling candidate bands. For the candidate coupling frequency band, the cross power spectral density is calculated within the sliding time window, and the time-domain coupling sequence is reconstructed by combining the normalized coherence function to generate a multi-parameter coupling degree vector composed of frequency domain coincidence rate and phase locking degree. Based on the multi-parameter coupling degree vector, covariance analysis is performed on the coupling factors of each frequency band, and a correction function is established in combination with dynamic risk factors to calculate the coupling coefficient between the alternating inflation and deflation frequency of the airbag and the micro-vibration frequency of the human body. When the coupling coefficient gradually approaches the set threshold, the correction function is called to trigger the adaptive frequency offset mechanism, which performs perturbation modulation on the alternating inflation and deflation frequency of the airbags, so that the alternating inflation and deflation frequency of the airbags actively avoids the sensitive frequency band.

2. The intelligent inflation and deflation control method for a grid-distributed massage airbag according to claim 1, characterized in that, The steps for obtaining the alternating frequency spectrum of the airbags include: A digital pressure sensor is embedded in each airbag unit inside the mattress to obtain pressure data during the inflation and deflation process of the airbag, and to form a spatiotemporal distribution matrix that changes over time by combining it with spatial coordinate information. After obtaining the spatiotemporal distribution matrix, interpolation and difference processing are performed on the spatiotemporal distribution matrix to form a continuous pressure disturbance sequence; After obtaining the continuous pressure disturbance sequence, a fast Fourier transform is used to convert it into a frequency domain signal, and the dominant and secondary frequency components are identified. After identifying the primary and secondary frequency components, the frequency domain signal is integrated with spatial coordinates and time series to construct a three-dimensional matrix containing time, spatial location and frequency, and then expanded into a two-dimensional frequency spectrum to form an airbag alternating frequency spectrum.

3. The intelligent inflation and deflation control method for a grid-distributed massage airbag according to claim 1, characterized in that, The steps to obtain a human micro-vibration spectrum include: High-sensitivity inertial measurement units are deployed at different key locations on the mattress surface to simultaneously acquire three-dimensional acceleration and three-dimensional angular velocity signals and form an attitude data stream; After obtaining the attitude data stream, filtering and calibration are performed to obtain the attitude micro-perturbation curve, and the attitude change trajectory conforming to the supine orientation is formed through integration and coordinate transformation. After obtaining the attitude change trajectory, the three-dimensional displacement components are synthesized into a multi-dimensional displacement vector, and the instantaneous micro-vibration frequency is extracted by polynomial fitting within the sliding time window. After obtaining the instantaneous micro-vibration frequency, wavelet packet decomposition is performed on the instantaneous micro-vibration frequency sequence, and a multidimensional spectrum is formed by combining the energy distribution and spatial location to obtain the human body micro-vibration spectrum.

4. The intelligent inflation and deflation control method for a grid-distributed massage airbag according to claim 3, characterized in that, Wavelet packet decomposition uses Daubechies wavelets as basis functions, and evaluates the energy distribution characteristics of the sub-band signals obtained by decomposition through energy entropy calculation.

5. The intelligent inflation and deflation control method for a grid-distributed massage airbag according to claim 4, characterized in that, The steps for extracting frequency bands with temporal overlap characteristics as candidate coupling bands include: After obtaining the frequency spectrum of the airbag alternation and the frequency spectrum of human micro-vibration, the two are projected into a unified frequency domain. In a unified frequency domain, a spectrum overlap weight matrix is ​​constructed. At each frequency point, the product of the airbag spectrum amplitude and the human body spectrum amplitude is taken as the initial weight value and normalized. The local human body spectrum corresponding to different airbag units is calculated separately to form a spectrum overlap weight matrix. After obtaining the spectrum overlap weight matrix, a phase synchronization window is introduced to perform weighted mapping on the frequency peaks, and the weight values ​​are corrected according to the phase difference. After phase weighting, a threshold range is set to extract frequency points that continuously exceed the threshold range and merge them into a frequency band. Combined with spatial location information, human body parts and airbag units are marked to form a coupling candidate frequency band.

6. The intelligent inflation and deflation control method for a grid-distributed massage airbag according to claim 1, characterized in that, The steps for generating a multi-parameter coupling vector consisting of frequency domain overlap rate and phase locking degree include: After obtaining the candidate frequency bands for coupling, a sliding window is introduced on the time axis and the cross power spectral density of the pressure disturbance signal of the alternating inflation and deflation of the airbag and the human body micro-vibration signal is calculated. After obtaining the cross-power spectral density, the phase relationship between the pressure disturbance signal of the alternating inflation and deflation of the airbag and the micro-vibration signal of the human body is analyzed by combining the normalized coherence function, and the phase difference is extracted within the sliding window to reconstruct the temporal coupling sequence. After obtaining the time-domain coupled sequence, it is parameterized. The integral result of the cross-power spectral density is used as the frequency domain coincidence rate, and the mean and variance of the normalized coherence function are used as the phase locking degree. The two are combined to form a multi-parameter coupling degree vector.

7. The intelligent inflation and deflation control method for a grid-distributed massage airbag according to claim 6, characterized in that, The steps for calculating the coupling coefficient between the alternating inflation and deflation frequency of the airbag and the micro-vibration frequency of the human body include: After obtaining the multi-parameter coupling vector composed of frequency domain overlap rate and phase locking degree, the multi-parameter coupling vector is classified and organized according to frequency band. After completing the frequency band division and vector arrangement, covariance analysis was performed on the coupling factors within each sub-frequency band. After obtaining the covariance results for each sub-frequency band, the covariance results are weighted and corrected in conjunction with dynamic risk factors, a correction function is established, and the coupling factor is adjusted. After the correction function is established, the corrected coupling factor is normalized and weighted averaged within the candidate frequency band to obtain the coupling coefficient.

8. The intelligent inflation and deflation control method for a grid-distributed massage airbag according to claim 7, characterized in that, The steps for perturbation modulation of the airbag alternating inflation / deflation frequency include: After the coupling coefficient is calculated, the coupling coefficient is monitored in real time. When the coupling coefficient output by the correction function gradually approaches the set threshold and continues to rise in multiple time windows, the correction function is called to trigger the adaptive frequency offset mechanism. After the adaptive frequency offset mechanism is triggered, the direction and magnitude of the frequency offset are determined based on the result of the correction function. When the coupling coefficient is close to different levels, the offset is performed by gradually adjusting until the airbag alternating inflation and deflation frequency moves out of the sensitive frequency band. After determining the direction and magnitude of the frequency offset, the offset result is applied to the inflation / deflation control strategy and the new coupling coefficient is monitored in real time. When the coupling coefficient decreases, the current setting is maintained, and when the coupling coefficient is still close to the set threshold, the offset continues.

9. A grid-distributed massage airbag intelligent inflation / deflation system, used to implement the control method for intelligent inflation / deflation of the grid-distributed massage airbag as described in any one of claims 1-8, characterized in that, It includes a barometric disturbance spectrum construction module, a human micro-vibration spectrum acquisition module, a spectrum overlap weight extraction module, a multi-parameter coupling degree generation module, a coupling coefficient calculation module, and an adaptive frequency offset module. The air pressure disturbance spectrum construction module constructs a spatiotemporal distribution matrix based on the air pressure fluctuation field of the mattress inside the mattress. It transforms the pressure changes of each airbag during the inflation and deflation cycle into a continuous air pressure disturbance sequence, and performs Fourier transform on the continuous air pressure disturbance sequence to extract the main frequency component, forming an alternating frequency spectrum of the airbags as the input baseline. The human body micro-vibration spectrum acquisition module, based on the input baseline, introduces a high-sensitivity inertial measurement unit matrix to acquire the local posture micro-perturbation curve of the body, and extracts the instantaneous micro-vibration frequency under the lying position by fitting multi-dimensional displacement vectors. Then, wavelet packet decomposition is used to process the instantaneous micro-vibration frequency on multiple scales to obtain the human body micro-vibration spectrum. The spectrum overlap weight extraction module maps the human micro-vibration spectrum and the airbag alternating frequency spectrum in a unified frequency domain, constructs a spectrum overlap weight matrix, and performs weighted mapping on the frequency peaks within the phase synchronization window to extract frequency bands with temporal overlap characteristics as coupling candidate frequency bands. The multi-parameter coupling degree generation module calculates the cross power spectral density within a sliding time window for the coupling candidate frequency band, and reconstructs the time-domain coupling sequence by combining the normalized coherence function, generating a multi-parameter coupling degree vector composed of frequency domain coincidence rate and phase locking degree. The coupling coefficient calculation module performs covariance analysis on the coupling factors of each frequency band based on the multi-parameter coupling degree vector, and establishes a correction function in combination with dynamic risk factors to calculate the coupling coefficient between the alternating inflation and deflation frequency of the airbag and the micro-vibration frequency of the human body. The adaptive frequency offset module calls a correction function to trigger the adaptive frequency offset mechanism when the coupling coefficient gradually approaches the set threshold. This mechanism performs perturbation modulation on the alternating inflation and deflation frequency of the airbags, allowing the alternating inflation and deflation frequency of the airbags to actively avoid sensitive frequency bands.

Images