A sleep system based on internet of things communication and a method thereof

Abstract

The application discloses a sleep system and method based on internet communication, and particularly relates to the field of smart home and health monitoring, which is used to solve the problems of sleep microenvironment regulation lag and physical intervention easily inducing micro-awakening. First, based on multi-source sensing, heart-lung waveform and temperature and humidity data are acquired, and microclimate sweat index is calculated to accurately lock physiological heat dissipation demand. Then, when the sweat threshold is met, the time domain tracking of the respiratory envelope is carried out, the next expiration trough and time window are extrapolated, and the intervention time is limited to the feeling gate delay period. Through dynamic calculation of pure network time consumption and motor pressure building time consumption, the predicted time point is reversely compensated on the time axis to generate an advance trigger time stamp, and the system response time difference is smoothed. Finally, the ventilation volume is equivalent to discrete single pulse air volume, the driving control node only outputs airflow in the expiration time window, the dehumidification action is hidden in the natural expiration stage, and a non-inductive closed-loop intervention system is constructed, thereby providing scientific support for sleep health management.

Description

Technical Field

[0001] This invention relates to the field of smart home and health monitoring technology, specifically to a sleep system and method based on Internet of Things communication. Background Technology

[0002] With the accelerating pace of modern life and the growing awareness of public health management, the refined management of sleep quality has become an important branch of the health industry. The integration of IoT communication technology and non-contact sensing technology is driving smart home devices to evolve from simple passive data monitoring to proactive sleep environment intervention. Modern people have increasingly higher demands for the comfort of their home sleep environment, urgently requiring systems that can dynamically improve microclimate issues such as stuffiness and sweating during sleep without interfering with normal routines, in order to maintain continuous, high-quality deep sleep.

[0003] Existing IoT-based sleep environment control systems typically employ macroscopic spatial regulation logic, such as directly controlling the operation of the entire room's air conditioning system when a user enters a specific sleep stage or experiences a change in body temperature. This macroscopic regulation method faces significant spatial thermal inertia resistance, making it difficult for hot and cold airflows to quickly penetrate the bedding boundary layer. This results in a severe lag effect in changes to the bed microclimate, easily leading to overcooling or discomfort for the bed partner. Furthermore, existing devices operate randomly when issuing physical intervention commands, completely detached from the body's instantaneous physiological rhythms. Because the human body has high nerve sensitivity during the inspiratory phase, the sudden activation of fans or mechanical adjustment components during this phase can easily exceed the user's physiological arousal threshold, triggering stress-induced micro-awakening or even directly waking the user. In addition, the inherent network transmission jitter between IoT nodes, coupled with the build-up time of the underlying mechanical components from activation to effective output of physical quantities, causes severe time deviations in the coordinated control of multiple devices, further exacerbating the disruption to sleep continuity. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides a sleep system and method based on Internet of Things (IoT) communication, which solves the problems mentioned in the background.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a sleep system based on Internet of Things communication, comprising the following modules: a multi-source state sensing module, used to acquire the cardiopulmonary coupling vibration waveform output by a piezoelectric sensor array arranged on the mattress support layer and separate the continuous respiratory envelope and heart rate variability feature sequence, simultaneously acquiring transient temperature and humidity data of the microenvironment air chamber under the mattress surface, and calculating the microclimate perspiration index based on the rate of change of the transient temperature and humidity data; a rhythm phase analysis module, used to perform time-domain tracking of the respiratory envelope when the microclimate perspiration index is greater than a preset perspiration threshold and the high-frequency energy proportion of the heart rate variability feature sequence increases, identify the expiratory trough time points and expiratory duration of multiple consecutive respiratory cycles, establish a respiratory rhythm model based on the time interval of the expiratory trough time points, extrapolate the predicted expiratory trough time point and expiratory time window of the next respiratory cycle, and generate a model containing the target phase phase analysis module. The system addresses the cooling intervention requirements for air volume; a cross-domain latency compensation module, used to send timestamp detection frames to microenvironment control nodes in the IoT network, extract the dwell time and network round-trip time in the response frames returned by the microenvironment control nodes, calculate the one-way network communication latency, and sequentially remove the one-way network communication latency and the motor pressure build-up time of the microenvironment control nodes from the next exhalation trough prediction time point for advance offset compensation, generating an advance trigger timestamp; a rhythm-following intervention module, used to analyze the target ventilation volume in the cooling intervention requirements, equivalently discretize the target ventilation volume to a preset number of continuous respiratory cycles to set the single pulse air volume, take the advance trigger timestamp as the first execution starting point, generate intermittent control timing based on the span of the exhalation time window, and send drive messages to the microenvironment control nodes according to the control timing through the IoT communication protocol, driving the microenvironment control nodes to output a single pulse air volume only within the exhalation time window to perform dehumidification intervention.

[0006] Furthermore, the specific process of acquiring the cardiopulmonary coupled vibration waveform output by the piezoelectric sensor array arranged in the mattress support layer and separating the continuous respiratory envelope and heart rate variability feature sequence is as follows: Bandpass filtering is performed on the cardiopulmonary coupled vibration waveform output by the piezoelectric sensor array to extract the net cardiopulmonary vibration signal; empirical mode decomposition is applied to the net cardiopulmonary vibration signal to separate the low-frequency intrinsic mode function and the high-frequency intrinsic mode function; Hilbert transform is performed on the low-frequency intrinsic mode function to extract the amplitude sequence of the analytic signal and generate a continuous respiratory envelope; local peak detection is performed on the high-frequency intrinsic mode function to extract the time interval between adjacent peaks to generate a heart rate interval sequence; resampling and fast Fourier transform are performed on the heart rate interval sequence to extract frequency domain features and generate a heart rate variability feature sequence.

[0007] Furthermore, transient temperature and humidity data of the microenvironment air chambers beneath the mattress surface are collected simultaneously. The specific process for calculating the microclimate perspiration index based on the rate of change of the transient temperature and humidity data is as follows: The relative humidity component and temperature component are extracted from the transient temperature and humidity data. The relative humidity component and temperature component are transformed into an absolute humidity sequence of the air chamber using the Antoine equation. The first derivative of the absolute humidity sequence of the air chamber is obtained by sliding window to obtain the absolute humidity rate of change. The first derivative of the temperature component is obtained by sliding window to obtain the temperature rate of change. The absolute humidity rate of change and the temperature rate of change are linearly weighted and fused to output the microclimate perspiration index.

[0008] Furthermore, when the microclimate perspiration index is greater than the preset perspiration threshold and the proportion of high-frequency energy in the heart rate variability characteristic sequence increases, the respiratory envelope is time-domain tracked to identify the expiratory trough time points and expiratory duration for multiple consecutive respiratory cycles. The specific process is as follows: When the microclimate perspiration index is greater than the preset perspiration threshold and the proportion of high-frequency energy in the heart rate variability characteristic sequence increases, a first-order difference operation is performed on the respiratory envelope. The zero-crossing point in the first-order difference sequence of the respiratory envelope that changes from positive to negative is identified as the exhalation start point, and the zero-crossing point that changes from negative to positive is identified as the inhalation start point. A minimum value search is performed between the exhalation start point and the inhalation start point, and the location of the minimum value is marked as the expiratory trough time point. The relative span between the time coordinates of the inhalation start point and the exhalation start point is calculated to extract the expiratory duration.

[0009] Furthermore, a respiratory rhythm model is established based on the time interval of the expiratory trough time points. The specific process of extrapolating the predicted expiratory trough time point and expiratory time window of the next respiratory cycle to generate the cooling intervention requirement containing the target ventilation volume is as follows: The backward difference sequence of the expiratory trough time points of multiple consecutive respiratory cycles is calculated to obtain the historical respiratory cycle sequence. Autoregressive smoothing is performed on the historical respiratory cycle sequence to obtain the predicted respiratory cycle. Using the current last expiratory trough time point as the time reference, forward time axis extrapolation is performed according to the predicted respiratory cycle to generate the predicted expiratory trough time point of the next respiratory cycle. The exponential moving average of the expiratory duration of multiple consecutive respiratory cycles is extracted and its time span is determined as the expiratory time window of the next respiratory cycle. The excess sweating deviation is obtained by extracting the overflow difference between the microclimate sweating index and the preset sweating threshold. The excess sweating deviation is converted into a volume equivalent based on the preset air chamber volume constant to generate the target ventilation volume and encapsulate it as a cooling intervention requirement.

[0010] Furthermore, the specific process of sending timestamp probe frames to micro-environment control nodes in the IoT network, extracting dwell time and network round-trip time from the response frames returned by the micro-environment control nodes, and calculating the one-way network communication latency is as follows: Encapsulate a timestamp probe frame carrying a local transmission timestamp, route it to the micro-environment control node via the local IoT communication protocol, receive the response frame returned by the micro-environment control node, parse the response frame to extract the node reception timestamp, the node processing completion timestamp, and the local reception completion timestamp; calculate the time difference between the node processing completion timestamp and the node reception timestamp to extract the dwell time; calculate the time span between the reception completion timestamp and the local transmission timestamp to obtain the network round-trip time; filter out the dwell time from the network round-trip time to establish the pure network transmission time; perform a halving equivalent conversion on the pure network transmission time to output the one-way network communication latency.

[0011] Furthermore, the specific process of generating an advance trigger timestamp by sequentially removing the one-way network communication delay and the motor pressure build-up time of the microenvironment control node from the next exhalation trough prediction time point is as follows: The physical operating parameter table of the exhaust fan motor in the microenvironment control node is retrieved, the mechanical response time constant of the exhaust fan motor accelerating from a stationary state to the rated exhaust speed is extracted, the current airflow conduction damping coefficient of the microenvironment air chamber is obtained, and the current airflow conduction damping coefficient is introduced to perform damping hysteresis weighted calculation on the mechanical response time constant to calculate the motor pressure build-up time; based on the one-way network communication delay, reverse compensation of network delay is performed on the next exhalation trough prediction time point to locate the network synchronization time coordinate, and the motor pressure build-up time is further removed from the network synchronization time coordinate to complete the mechanical response advance compensation, establishing the absolute physical trigger time coordinate, and encapsulating the absolute physical trigger time coordinate as an advance trigger timestamp.

[0012] Furthermore, the specific process of analyzing the target ventilation volume in the cooling intervention requirement and equivalently discretizing the target ventilation volume to a preset number of continuous breathing cycles to set the single pulse air volume is as follows: Analyze the cooling intervention requirement to extract the target ventilation volume, obtain the preset number of continuous breathing cycles, and perform discrete equal distribution processing on the target ventilation volume according to the preset number to map the basic exhaust volume of a single cycle; obtain the current static pressure attenuation coefficient in the microenvironment air chamber, perform pipeline fluid resistance conversion compensation on the basic exhaust volume of a single cycle based on the current static pressure attenuation coefficient, and calculate the single pulse air volume; according to the air volume and speed mapping curve of the dehumidification fan motor, map the single pulse air volume to the target motor speed.

[0013] Furthermore, taking the advance trigger timestamp as the first execution starting point, an intermittent control sequence is generated based on the span of the exhalation time window. Drive messages are sent to the microenvironment control node according to the control sequence via the IoT communication protocol, driving the microenvironment control node to output a single pulse airflow within the exhalation time window to perform dehumidification intervention. The specific process is as follows: Using the advance trigger timestamp as the duty cycle starting point, the span of the exhalation time window is used as the pulse width, and the predicted respiratory cycle is used as the pulse interval to construct a pulse width modulation control sequence; the target motor speed and the pulse width modulation control sequence are encoded into an IoT communication control load, with the physical layer media access control address of the microenvironment control node added, and a drive message is encapsulated; the drive message is sent via the IoT communication protocol, controlling the microenvironment control node to start the dehumidification fan motor and enter an accelerated pressure-building state when the advance trigger timestamp is reached. After the motor pressure-building time, the target motor speed is maintained within the corresponding exhalation time window span to output a single pulse airflow, and the motor electromagnetic brake is triggered at the end of the exhalation time window span to stop airflow.

[0014] A sleep method based on Internet of Things (IoT) communication includes the following steps: S1. Acquiring the cardiopulmonary coupling vibration waveform output by a piezoelectric sensor array arranged in the mattress support layer and separating the continuous respiratory envelope and heart rate variability feature sequence; simultaneously collecting transient temperature and humidity data of the microenvironment air chamber under the mattress surface; calculating the microclimate perspiration index based on the rate of change of the transient temperature and humidity data; S2. When the microclimate perspiration index is greater than a preset perspiration threshold and the proportion of high-frequency energy in the heart rate variability feature sequence increases, performing time-domain tracking of the respiratory envelope, identifying the expiratory trough time points and expiratory duration of multiple consecutive respiratory cycles, establishing a respiratory rhythm model based on the time interval of the expiratory trough time points, extrapolating the predicted expiratory trough time point and expiratory time window of the next respiratory cycle, and generating a cooling intervention including the target ventilation volume. Requirements; S3. Send timestamp probe frames to the microenvironment control nodes in the IoT network, extract the dwell time and network round-trip time in the response frames returned by the microenvironment control nodes, calculate the one-way network communication latency, and subtract the one-way network communication latency and the motor pressure build-up time of the microenvironment control nodes from the predicted time point of the next exhalation trough in sequence to generate an advance trigger timestamp; S4. Analyze the target ventilation volume in the cooling intervention requirements, distribute the target ventilation volume evenly to a preset number of consecutive breathing cycles to set the single pulse air volume, take the advance trigger timestamp as the first execution starting point, generate intermittent control timing according to the span of the exhalation time window, and send drive messages to the microenvironment control nodes according to the control timing through the IoT communication protocol, driving the microenvironment control nodes to output a single pulse air volume only within the exhalation time window to perform dehumidification intervention.

[0015] The present invention has the following beneficial effects:

[0016] (1) A sleep system based on Internet of Things communication achieves a technological leap from blind adjustment of the macro environment to targeted identification and rhythm locking of microclimate through the coordinated operation of a multi-source state sensing module and a rhythm phase analysis module. The system utilizes a piezoelectric sensor array arranged in the mattress support layer and an air chamber probe under the surface to accurately separate the cardiopulmonary coupling vibration waveform and calculate the microclimate perspiration index, thereby accurately locking the real physiological heat dissipation needs without contacting the human body. More importantly, the system accurately extrapolates the expiratory trough time point and expiratory time window of the next respiratory cycle by tracking the time domain and extracting the extreme values ​​of the continuous respiratory envelope. This mechanism strictly limits the timing of environmental intervention to the end of expiration when the human sensory gating is at its most insensitive state, eliminating the sudden sensory stimulation caused by the activation of the device at any time in traditional systems, and ensuring the continuity of the user's sleep.

[0017] (2) A sleep method based on IoT communication overcomes the control failure problem caused by the superposition of IoT communication network jitter and mechanical entity response lag. By sending timestamp probe frames to accurately measure the one-way network communication delay, and using it together with the motor pressure build-up time of the micro-environment control node as the offset compensation base, the system performs reverse time-axis deduction on the predicted time point of the exhalation trough to generate a highly accurate advance trigger timestamp. On this basis, the system equivalently discretizes the huge target ventilation volume into single pulse air volume and generates intermittent control timing according to the exhalation time window span. This pulse width modulation mechanism based on absolute physical trigger time coordinates ensures that the dehumidifying fan can still complete pressure build-up and pulse output within the user's exhalation time window without any error under the dual interference of network delay and physical inertia, realizing truly imperceptible rhythmic follow-up intervention at the physical level.

[0018] Of course, any product implementing this invention does not necessarily need to achieve all of the advantages described above at the same time. Attached Figure Description

[0019] Figure 1 This is a flowchart of a sleep system based on Internet of Things communication according to the present invention.

[0020] Figure 2 This is a flowchart of a sleep method based on Internet of Things communication according to the present invention. Detailed Implementation

[0021] This application provides a sleep system and method based on Internet of Things (IoT) communication, which solves the problem of micro-awakening caused by the lag in macro-environmental regulation, misalignment between physical actions and ideal biological positioning, and superposition of network and mechanical response delays in existing IoT sleep intervention systems.

[0022] The overall concept of the solution in this application embodiment is as follows:

[0023] First, the system acquires data from a piezoelectric sensor array and temperature / humidity probes placed within the mattress, separating the respiratory envelope and heart rate variability (HRV) characteristic sequences, and calculating the microclimate perspiration index. Then, upon confirming a perspiration demand and an increase in the energy proportion of the HRV characteristic frequency band, the system performs time-domain tracking of the respiratory envelope to identify the expiratory trough time point, thereby establishing a respiratory rhythm model to deduce the predicted expiratory trough time point and expiratory time window for the next respiratory cycle, and generating a cooling intervention requirement. Next, the system calculates network communication latency by sending timestamp probe frames to the control node, sequentially removing network communication latency and motor pressure build-up time from the predicted expiratory trough time point for advance offset compensation, resulting in an advance trigger timestamp. Finally, the system discretizes the target ventilation volume into single-pulse airflow within multiple consecutive respiratory cycles, generating an intermittent control sequence starting from the advance trigger timestamp, driving the microenvironment control node to strictly limit the output pulse airflow within the user's expiratory time window, thereby executing precise and imperceptible dehumidification and cooling intervention.

[0024] Please see Figure 1 This invention provides a technical solution: a sleep system based on Internet of Things (IoT) communication, comprising the following modules: a multi-source state sensing module, used to acquire the cardiopulmonary coupling vibration waveform output by a piezoelectric sensor array arranged in the mattress support layer and separate the continuous respiratory envelope and heart rate variability feature sequence, simultaneously acquiring transient temperature and humidity data of the microenvironment air chamber under the mattress surface, and calculating the microclimate perspiration index based on the rate of change of the transient temperature and humidity data; and a rhythm phase analysis module, used to perform time-domain tracking of the respiratory envelope when the microclimate perspiration index is greater than a preset perspiration threshold and the high-frequency energy proportion of the heart rate variability feature sequence increases, identify the expiratory trough time points and expiratory duration of multiple consecutive respiratory cycles, establish a respiratory rhythm model based on the time interval of the expiratory trough time points, extrapolate the predicted expiratory trough time point and expiratory time window of the next respiratory cycle, and generate a reduction in the target ventilation volume. Temperature intervention requirements; Cross-domain latency compensation module, used to send timestamp detection frames to microenvironment control nodes in the IoT network, extract the dwell time and network round-trip time in the response frames returned by the microenvironment control nodes, calculate the one-way network communication latency, and sequentially remove the one-way network communication latency and the motor pressure build-up time of the microenvironment control nodes from the next exhalation trough prediction time point for advance offset compensation, generating an advance trigger timestamp; Rhythm follow-up intervention module, used to analyze the target ventilation volume in the cooling intervention requirements, equivalently discretize the target ventilation volume to a preset number of continuous breathing cycles to set the single pulse air volume, take the advance trigger timestamp as the first execution starting point, generate intermittent control timing according to the span of the exhalation time window, and send drive messages to the microenvironment control nodes according to the control timing through the IoT communication protocol, driving the microenvironment control nodes to output a single pulse air volume only within the exhalation time window to perform dehumidification intervention.

[0025] In this implementation scheme, the multi-source state sensing module collects and separates user physiological signals from bed environment signals. The cardiopulmonary coupled vibration waveform refers to the physical signal transmitted to the mattress by the combined mechanical vibrations caused by heartbeats and lung respiration; the respiratory envelope is the extended contour extracted from the combined vibration signal, representing the macroscopic amplitude of respiratory fluctuations; the heart rate variability characteristic sequence refers to the minute changes in the time interval between adjacent heartbeat cycles; and the microclimate perspiration index is a calculated value representing the rate of humidity increase within the small, enclosed space inside the bedding. The technical role of this module is to eliminate the constraints of wearable devices, achieving non-contact, precise perception of physiological and local microenvironmental states, providing a high-fidelity data foundation for subsequent intervention decisions. The rhythm phase analysis module identifies the optimal physiological time window for environmental intervention. An increased proportion of high-frequency energy typically indicates enhanced parasympathetic nervous system activity and autonomous heat dissipation in heart rate variability analysis. The expiratory trough refers to the extreme moment when a single exhalation is about to end and inhalation has not yet begun; at this time, the body's sensory nerve gating is at its least sensitive, making it highly resistant to wind and noise. The expiratory time window is the duration of a single exhalation. This module's technical function is to avoid the common misconception that continuous airflow causes chills or awakening, accurately capturing the end of exhalation when the user is least sensitive to external stimuli, defining the absolute safety boundary of intervention actions from a timeline perspective, and generating targeted intervention needs based on the body's autonomous physiological rhythms. The cross-domain latency compensation module eliminates the time difference caused by local communication delays in the Internet of Things (IoT) and the inertia of underlying hardware mechanical movements. The timestamp detection frame is a dedicated communication data packet used to accurately measure network link time nodes; dwell time refers to the waiting processing time consumed by the data packet in the hardware chip of the microenvironment control node for protocol parsing; motor pressure build-up time refers to the mechanical and physical response time required for the dehumidifying fan to accelerate from power-on to output effective airflow to the air chamber; advance offset compensation artificially advances the time scale of the system's issued commands to fully offset the above-mentioned lag time. The technical function of this module is to solve the technical problem of the disconnect between digital network commands and physical entity actions, ensuring that the adjustment commands issued by the system can still take effect precisely at the predicted expiratory trough time point after experiencing network routing transmission and mechanical acceleration resistance. The function of the rhythm-following intervention module is to transform macroscopic ventilation requirements into microscopic physical execution actions that closely match the respiratory rhythm. Equivalent discrete means that the large volume of air released at once for cooling is evenly distributed and executed in multiple consecutive breathing cycles while maintaining the total amount of moisture removed; single pulse air volume refers to a quantitative weak airflow that is strictly matched with the length of a single exhalation; intermittent control timing refers to a low-level communication control plan consisting of regular hardware start-up and emergency stop actions.The technical function of this module is to transform abstract network instructions into pulsed physical air delivery actions that are highly synchronized with human breathing rhythms. The control and regulation nodes only output dehumidification power during the user's exhalation phase and implement physical braking before inhalation, thereby effectively replacing the humid and hot air in the bed while achieving completely imperceptible intervention without triggering any physical awakening of the body surface.

[0026] Specifically, the process of acquiring the cardiopulmonary coupled vibration waveform output by the piezoelectric sensor array arranged in the mattress support layer and separating the continuous respiratory envelope and heart rate variability feature sequence is as follows: Bandpass filtering is performed on the cardiopulmonary coupled vibration waveform output by the piezoelectric sensor array to extract the net cardiopulmonary vibration signal; empirical mode decomposition is applied to the net cardiopulmonary vibration signal to separate the low-frequency intrinsic mode function and the high-frequency intrinsic mode function; Hilbert transform is performed on the low-frequency intrinsic mode function to extract the amplitude sequence of the analytic signal and generate a continuous respiratory envelope; local peak detection is performed on the high-frequency intrinsic mode function to extract the time interval between adjacent peaks to generate a heart rate interval sequence; resampling and fast Fourier transform are performed on the heart rate interval sequence to extract frequency domain features and generate a heart rate variability feature sequence.

[0027] In this implementation scheme, the initial signal acquired by the piezoelectric sensor array is mixed with mechanical resonance noise from the bed frame and high-frequency electromagnetic interference. Therefore, bandpass filtering is first used to remove abnormal frequency bands and retain the net cardiopulmonary vibration signal. Since the micro-vibrations caused by heartbeat and respiration are highly coupled in the time domain and are nonlinear and non-stationary signals, the system uses an empirical mode decomposition algorithm to adaptively separate them into multiple mode functions with a single instantaneous frequency. The low-frequency intrinsic mode function represents the macroscopic fluctuations of chest cavity respiration, while the high-frequency intrinsic mode function represents the micro-impacts caused by cardiac pumping. To obtain the pure trend of respiratory depth changes without interference from respiratory phase fluctuations, the system performs a Hilbert transform on the low-frequency intrinsic mode function to extract the instantaneous analytic signal. The real part of the analytic signal is the original signal, and the imaginary part is the phase-shifted signal. Taking the complex modulus of this signal constructs a smooth outer envelope that filters out high-frequency phase oscillations, i.e., a continuous respiratory envelope. The core calculation logic is as follows: The parameters are explained as follows: : Continuous respiratory envelope; : First discrete time series; : Low-frequency intrinsic mode function sequence; : Integral independent variable. For high-frequency intrinsic mode functions, the system locates the peak timestamps of cardiac mechanical contraction using a local peak detection algorithm and calculates the time intervals between adjacent peaks to form a heartbeat interval sequence. Because the heartbeat intervals are not uniformly distributed on the time axis, direct frequency domain transformation would lead to spectral aliasing. Therefore, the system uses cubic spline interpolation to uniformly resample the heartbeat interval sequence. Subsequently, the time-domain sequence is converted into a complex frequency sequence in the frequency domain using a fast Fourier transform, and energy integration in a specific frequency band is performed to extract heart rate variability feature sequences, especially extracting high-frequency energy representing parasympathetic nerve activity. The integration calculation logic is as follows: The parameters are explained as follows: Heart rate variability high-frequency energy; Frequency independent variable; : Lower limit of high-frequency band integration; High-frequency band integration limit; : The complex spectrum sequence after Fast Fourier Transform.

[0028] Specifically, the transient temperature and humidity data of the microenvironment air chambers beneath the mattress surface are collected synchronously. The specific process for calculating the microclimate perspiration index based on the rate of change of the transient temperature and humidity data is as follows: The relative humidity component and temperature component are extracted from the transient temperature and humidity data. The relative humidity component and temperature component are transformed into an absolute humidity sequence of the air chamber using the Antoine equation. The first derivative of the absolute humidity sequence of the air chamber is obtained by sliding window to obtain the absolute humidity rate of change. The first derivative of the temperature component is obtained by sliding window to obtain the temperature rate of change. The absolute humidity rate of change and the temperature rate of change are linearly weighted and fused to output the microclimate perspiration index.

[0029] In this implementation scheme, the relative humidity directly collected by the sensor is highly susceptible to interference from ambient temperature fluctuations, and cannot accurately reflect the actual mass of water vapor accumulated in the air chamber. Therefore, the system must first use the Antoine equation to solve for the saturated vapor pressure at the current temperature, and then convert the relative humidity component and temperature component into an absolute humidity sequence of the air chamber, characterizing the absolute moisture density within the physical space. The conversion calculation logic is as follows: The parameters are explained as follows: : Absolute humidity sequence of air chambers; : Second discrete time series; : Absolute humidity conversion coefficient; The first empirical constant for saturated water vapor pressure; : The second empirical constant of saturated water vapor pressure; The third empirical constant for saturated water vapor pressure; Temperature component sequence of the microenvironment air chamber; The system defines the relative humidity component sequence of the microenvironment air chamber. After obtaining the absolute moisture density, to distinguish whether the user is in a dynamic perspiration process or merely in a static high-humidity environment, the system uses a sliding window to perform first-order difference differentiation on the absolute humidity sequence and temperature component sequence of the air chamber, extracting the absolute humidity change rate and temperature change rate respectively. A positive slope of the change rate indicates that the microenvironment is continuously deteriorating, accumulating heat and moisture. Subsequently, the system performs linear weighted fusion of the absolute humidity change rate and temperature change rate to output the final microclimate perspiration index. Perspiration is a complex thermodynamic process involving both sensible heat rise and latent heat evaporation; therefore, the weighting allocation cannot be a fixed constant. The system discloses a method for determining the weight coefficients based on the thermodynamic physical properties of the closed air chamber, which adaptively allocates the weights according to the physical ratio of the latent heat of water vapor evaporation to the specific heat capacity of air at constant pressure. The calculation logic is as follows: The parameters are explained as follows: Microclimate perspiration index; The latent heat of vaporization of water; : The air density constant of the microenvironment air chamber; The specific heat capacity constant of air at constant pressure in a microenvironment chamber; : Rate of change of absolute humidity; Temperature change rate. Through the above steps, the system can filter out environmental noise and resting baseline, accurately quantifying the actual microclimate perspiration and heat accumulation of the human body during sleep.

[0030] Specifically, when the microclimate perspiration index is greater than the preset perspiration threshold and the high-frequency energy proportion of the heart rate variability characteristic sequence increases, the respiratory envelope is time-domain tracked to identify the expiratory trough time points and expiratory duration for multiple consecutive respiratory cycles. The specific process is as follows: When the microclimate perspiration index is greater than the preset perspiration threshold and the high-frequency energy proportion of the heart rate variability characteristic sequence increases, a first-order difference operation is performed on the respiratory envelope. The zero-crossing point in the first-order difference sequence of the respiratory envelope that changes from positive to negative is identified as the exhalation start point, and the zero-crossing point that changes from negative to positive is identified as the inhalation start point. A minimum value search is performed between the exhalation start point and the inhalation start point, and the location of the minimum value is marked as the expiratory trough time point. The relative span between the time coordinates of the inhalation start point and the exhalation start point is calculated to extract the expiratory duration.

[0031] In this implementation plan, the system first performs a dual-condition judgment based on cross-physiological characteristics. The microclimate perspiration index represents the accumulation of heat and humidity in the external physical space, while an increase in the high-frequency energy proportion of the heart rate variability sequence indicates enhanced parasympathetic nerve activity within the body, meaning the body is undergoing autonomous vasodilation and heat dissipation through perspiration. Only when both external physical and internal physiological characteristics are simultaneously satisfied does the system recognize a genuine need for intervention, thereby effectively filtering out false triggers caused by natural fluctuations in room temperature. The preset perspiration threshold is not a fixed, rigid value; the system discloses an adaptive determination method that dynamically calibrates based on static baseline data from the user's initial sleep stage. The judgment conditions and threshold determination logic are as follows: The parameters are explained as follows: : Preset perspiration threshold; : Mean baseline perspiration during resting state; Standard deviation of baseline perspiration at rest; : Dynamic sensitivity adjustment coefficient. After confirming that the intervention conditions are met, the system performs a first-order difference operation on the respiratory envelope representing the macroscopic fluctuations of the chest cavity. Physically, the first derivative of the chest cavity volume signal represents the direction and rate of respiratory airflow. When the difference sequence changes from a positive value crossing the zero axis to a negative value, it indicates that chest cavity expansion has stopped and contraction has begun; this zero-crossing point precisely corresponds to the expiratory initiation point where airflow transitions from inhalation to exhalation. Conversely, a zero-crossing point changing from negative to positive corresponds to the initiation point of inhalation. Since the human sensory nerves are least sensitive during the expiratory phase, the system searches for the minimum value of the waveform amplitude within the interval formed by the two zero-crossing points, calibrating it as the time point of the deepest expiratory trough in a single respiratory action. Subsequently, by calculating the difference between the two zero-crossing points on the time axis, the expiratory duration of that breath can be extracted. The calculation logic is as follows: The parameters are explained as follows: Duration of exhalation; : Time coordinate of the inhalation start point; : The time coordinate of the exhalation start point; : The current respiratory cycle number being processed. Through the above steps, the system accurately segments the continuous simulated respiratory waveform into discrete-time phase features with clear physical meaning.

[0032] Specifically, the process of establishing a respiratory rhythm model based on the time interval of the expiratory trough time points, extrapolating the predicted expiratory trough time point and expiratory time window of the next respiratory cycle, and generating a cooling intervention requirement including the target ventilation volume is as follows: The backward difference sequence of the expiratory trough time points of multiple consecutive respiratory cycles is calculated to obtain a historical respiratory cycle sequence. Autoregressive smoothing is performed on the historical respiratory cycle sequence to obtain a predicted respiratory cycle. Using the current last expiratory trough time point as the time reference, forward time-axis extrapolation is performed according to the predicted respiratory cycle to generate the predicted expiratory trough time point of the next respiratory cycle. The exponential moving average of the expiratory duration of multiple consecutive respiratory cycles is extracted, and its time span is determined as the expiratory time window of the next respiratory cycle. The excess perspiration deviation is obtained by extracting the overflow difference between the microclimate perspiration index and the preset perspiration threshold. Based on the preset air chamber volume constant, the excess perspiration deviation is converted into a volume equivalent to generate the target ventilation volume and encapsulate it as a cooling intervention requirement.

[0033] In this implementation scheme, because the respiratory rhythm of the human body during sleep has the characteristic of being stable and slowly drifting, directly using the previous respiratory cycle as the basis for predicting the next one would lead to serious time lag errors. Therefore, the system extracts the historical respiratory cycle sequence by calculating the difference between adjacent expiratory trough time points and introduces an autoregressive smoothing extrapolation algorithm. This algorithm uses a linear combination of multiple past cycles to eliminate the fluctuation noise caused by a single abnormal breath, thereby smoothly and accurately predicting the duration of the next complete respiratory cycle. Its core calculation logic is as follows: The parameters are explained as follows: Predicting the respiratory cycle; : Autoregressive model order index; : The maximum lag order of autoregression; Autoregressive weight coefficients; Measured values ​​from historical respiratory cycle sequences; White noise error term. By adding the time of the last trough of the current cycle to the predicted cycle, an absolutely safe intervention trigger point can be anchored on the future timeline. Subsequently, the system performs an exponential moving average calculation on the expiratory duration of multiple consecutive respiratory cycles. By assigning greater weight to data closer to the current moment, it quickly responds to subtle changes in the user's breathing depth to generate the expiratory time window for the next respiratory cycle. The calculation logic is as follows: The parameters are explained as follows: The expiratory time window of the next respiratory cycle; Smoothing decay factor; : The measured duration of exhalation in the current z-th respiratory cycle; The predicted value of the exhalation time window at the previous moment. This time window defines the longest safe operating range for the bottom fan. Finally, the system must convert the abstract exponential deviation into physical parameters that can directly guide mechanical execution. The system calculates the excess perspiration deviation by measuring the portion of the microclimate perspiration index that exceeds the preset perspiration threshold. Combining this with the inherent spatial volume of the enclosed microenvironment inside the mattress, the dimensionless exponential deviation is equivalently transformed, and the actual air volume required to be removed to restore a comfortable climate is derived and calculated. The calculation logic is as follows: The parameters are explained as follows: Target ventilation rate; Preset air chamber volume constant; : Dehumidification efficiency conversion factor. Through this volume equivalent conversion, the system completely encapsulates physiological feedback data into cooling intervention requirements that include precise time points, time spans, and absolute physical volumes, thereby providing a rigorous input source for the coordinated scheduling of downstream IoT devices.

[0034] Specifically, the process of sending a timestamp probe frame to the micro-environment control node in the IoT network, extracting the dwell time and network round-trip time from the response frame returned by the micro-environment control node, and calculating the one-way network communication latency is as follows: Encapsulate a timestamp probe frame carrying the local transmission timestamp, route it to the micro-environment control node via the local IoT communication protocol, receive the response frame returned by the micro-environment control node, parse the response frame to extract the node's receiving timestamp, the node's processing completion timestamp, and the receiving completion timestamp arriving locally; calculate the time difference between the node's processing completion timestamp and the node's receiving timestamp to extract the dwell time; calculate the time span between the receiving completion timestamp and the local transmission timestamp to obtain the network round-trip time; filter out the dwell time from the network round-trip time to establish the pure network transmission time; perform a halving equivalent conversion on the pure network transmission time; and output the one-way network communication latency.

[0035] In this implementation scheme, wireless local area networks (WLANs) are highly susceptible to interference from channel congestion and routing node task scheduling when transmitting control commands, resulting in unpredictable random jitter in the time it takes for commands to reach the micro-environment control node. To eliminate the disruption of precise physiological intervention caused by this network uncertainty, the system must dynamically calculate the underlying communication latency. The system first encapsulates a timestamp probe frame carrying a local transmission timestamp and sends it to the control node. The micro-environment control node records the node's reception timestamp the moment the hardware network card captures the frame, and records the node's processing completion timestamp when the main control chip completes protocol parsing and data read / write preparation. It then packages this into a response frame and returns along the same path. Upon receiving the response frame, the main system immediately records the arrival and completion timestamp. By analyzing these four key time scales, the system can accurately extract the dwell time consumed by the data packet during queuing and processing within the target hardware chip using the time difference between the node's processing completion timestamp and the node's reception timestamp. This is because the internal processing speed of IoT nodes of different models or under different load conditions varies greatly; directly halving the total round-trip time would introduce serious system errors. The system filters out dwell time from the time span formed by the receive completion timestamp and the local send timestamp, thereby determining the pure network transmission time that occurs entirely in the spatial medium. Based on the physical assumption of symmetrical transmit and receive channel paths, the system performs a halving equivalent conversion on the pure network transmission time to output the one-way network communication delay. Its core calculation logic is as follows: The parameters are explained as follows: One-way network communication latency; : The timestamp indicating that the reception was completed at the local location; : Locally sent timestamp; : Node processing completion timestamp; Node receive timestamp. Through this step, the system obtains the pure media link delay parameter, completely excluding interference from end-side operating system task processing.

[0036] Specifically, the process of generating an advance trigger timestamp by sequentially removing the one-way network communication delay and the motor pressure build-up time of the microenvironment control node from the next exhalation trough prediction time point and performing advance offset compensation is as follows: The physical operating parameter table of the exhaust fan motor in the microenvironment control node is retrieved, the mechanical response time constant of the exhaust fan motor accelerating from a stationary state to the rated exhaust speed is extracted, the current airflow conduction damping coefficient of the microenvironment air chamber is obtained, and the current airflow conduction damping coefficient is introduced to perform damping hysteresis weighted calculation on the mechanical response time constant to calculate the motor pressure build-up time; based on the one-way network communication delay, reverse compensation of network delay is performed on the next exhalation trough prediction time point to locate the network synchronization time coordinate, and the motor pressure build-up time is further removed from the network synchronization time coordinate to complete the mechanical response advance compensation, establishing the absolute physical trigger time coordinate, and encapsulating the absolute physical trigger time coordinate as an advance trigger timestamp.

[0037] In this implementation scheme, after obtaining the accurate network layer latency, the system still needs to overcome the mechanical inertia barrier of the physical hardware layer. The dehumidifying fan motor must undergo a physical response ramp-up phase from being powered on from a standstill until its blades accelerate to output effective dehumidifying airflow into the mattress's microenvironment air chamber. If the system issues the command only when the predicted physiological trough time is reached, the physical lag in motor pressure build-up will cause the effective dehumidifying airflow to perfectly miss the exhalation time window when the user's senses are least sensitive. Therefore, the system retrieves the physical operating parameter table of the motor at this node to obtain the mechanical response time constant, but this only characterizes the acceleration time of the motor under ideal no-load conditions. In actual sleep scenarios, the user's body weight pressing on the mattress changes the effective cross-sectional area of ​​the ventilation support holes inside the mattress, causing additional exhaust fluid resistance. The system obtains the current airflow conduction damping coefficient of the microenvironment air chamber and introduces this coefficient to perform damping hysteresis weighted calculation on the mechanical response time constant to derive the motor pressure build-up time under real load physical conditions. The calculation logic is as follows: The parameters are explained as follows: : Motor voltage build-up time; Mechanical response time constant of the dehumidifier fan motor; : Weighting coefficient for airflow damping hysteresis compensation; The current airflow conduction damping coefficient of the microenvironment air chamber. A method for determining the airflow damping hysteresis compensation weight coefficient is disclosed here: It uses historical data on the time increment of the dehumidifying fan reaching its rated output air pressure when pre-calibrated standard test blocks of different weights are pressed onto the mattress. The least squares method is used to perform linear regression fitting on the weight deformation distribution and the time increment, and the slope of the fitted line is extracted as the coefficient. Subsequently, the system uses the two major delay quantities derived above—communication and physical—to perform reverse time-axis extrapolation of the physiological prediction time. First, based on the one-way network communication delay, reverse network delay compensation is performed on the predicted time point of the next exhalation trough to locate the network synchronization time coordinate required for the command to be delivered to the hardware port on time. Next, the motor pressure build-up time is further removed from this network synchronization time coordinate to complete the mechanical response pre-compensation. This process decouples the digital dimension delay from the physical dimension delay in series calculation, establishing the absolute physical trigger time coordinate at which the main control chip must transmit electromagnetic wave wireless signals. The calculation logic is as follows: The parameters are explained as follows: : The absolute physical trigger time coordinate corresponding to the advance trigger timestamp; The system predicts the next exhalation trough time. Through this cross-domain dual advance offset compensation mechanism, the system ensures that network data packets can fly in advance in the space medium, and that the motor blades can overcome damping and accelerate in advance, so that the effective dehumidification pulse airflow and the user's deepest exhalation trough intersect precisely on the time axis without any error.

[0038] Specifically, the process of analyzing the target ventilation volume in the cooling intervention requirement and discretizing the target ventilation volume into a preset number of continuous breathing cycles to set the single pulse air volume is as follows: Analyze the cooling intervention requirement to extract the target ventilation volume, obtain the preset number of continuous breathing cycles, and perform discrete equal distribution processing on the target ventilation volume according to the preset number to map the basic exhaust volume of a single cycle; obtain the current static pressure attenuation coefficient in the microenvironment air chamber, perform pipeline fluid resistance conversion compensation on the basic exhaust volume of a single cycle based on the current static pressure attenuation coefficient, and calculate the single pulse air volume; according to the air volume and speed mapping curve of the dehumidification fan motor, map the single pulse air volume to the target motor speed.

[0039] In this implementation plan, if a large amount of cold air is injected into the mattress at once for rapid cooling, the resulting strong wind and sudden temperature drop will instantly exceed the arousal threshold of the human skin's nerves, causing the system intervention to directly wake the user. To achieve imperceptible dehumidification, the system must break down the intervention intensity into micro-dosages. After analyzing the target ventilation volume generated in the previous stage, the system extracts the number of executions of the system's preset continuous breathing cycles. Through simple division, the total macroscopic exhaust demand is equivalently and discretely distributed into multiple continuous natural breathing cycles, thereby determining the theoretical basic exhaust volume per cycle. However, the theoretical exhaust volume is not directly equivalent to the air volume that the fan needs to output. The mattress is filled with porous support materials and narrow, flexible ventilation channels. When airflow travels a long distance within these channels, severe energy loss occurs due to air viscosity friction and collision with the channel walls, resulting in a significant decrease in airflow pressure reaching the user's skin. Therefore, the system must obtain the current static pressure attenuation coefficient within the microenvironment chamber and use a fluid dynamics compensation model to perform pipeline fluid resistance compensation on the single-cycle basic exhaust volume, ensuring that the fan outputs additional power to offset the resistance along the way. The compensation calculation logic is as follows: The parameters are explained as follows: Single pulse air volume; Target ventilation rate; The preset number of consecutive respiratory cycles; Pipeline friction loss conversion factor; : Current static pressure attenuation coefficient. A method for determining the pipeline friction loss conversion factor is disclosed here: Using multi-point wind pressure difference measurements of the mattress pipeline in a constant temperature test chamber before the system leaves the factory, the nonlinear correspondence between multiple sets of static pressure attenuation data and terminal wind speed characteristic values ​​is extracted to establish a fluid impedance attenuation matrix. The maximum eigenvalue of this matrix is ​​obtained, and the reciprocal of the maximum eigenvalue is used as the pipeline friction loss conversion factor. After obtaining the single-pulse airflow after fluid resistance compensation, the system needs to convert the aerodynamic volume parameters into electrical parameters in the electrical control dimension. Due to the design characteristics of the fan blades, there is a nonlinear relationship between the output airflow and the motor speed. Based on the preset airflow and speed quadratic mapping curve, the system directly calculates the target motor speed that the underlying hardware can recognize. The calculation logic is as follows: The parameters are explained as follows: Target speed of the motor; : Coefficients of the quadratic term of the mapping curve; : Coefficient of the first term of the mapping curve; : Mapping curve constant term. Through the above calculations, the system completes a rigorous equivalent conversion from macroscopic physiological needs to microscopic mechanical control commands, ensuring that each micro-ventilation precisely matches the user's perceived tolerance.

[0040] Specifically, taking the advance trigger timestamp as the first execution starting point, an intermittent control sequence is generated based on the span of the exhalation time window. Drive messages are sent to the microenvironment control node according to the control sequence via the IoT communication protocol, driving the microenvironment control node to output a single pulse airflow within the exhalation time window to perform dehumidification intervention. The specific process is as follows: Using the advance trigger timestamp as the duty cycle starting point, the span of the exhalation time window is used as the pulse width, and the predicted respiratory cycle is used as the pulse interval to construct a pulse width modulation control sequence; the target motor speed and the pulse width modulation control sequence are encoded into an IoT communication control load, with the physical layer media access control address of the microenvironment control node added, and a drive message is encapsulated; the drive message is sent via the IoT communication protocol, controlling the microenvironment control node to start the dehumidification fan motor and enter an accelerated pressure-building state when the advance trigger timestamp is reached. After the motor pressure-building time, the target motor speed is maintained within the corresponding exhalation time window span to output a single pulse airflow, and the motor electromagnetic brake is triggered at the end of the exhalation time window span to stop airflow.

[0041] In this implementation plan, to completely eliminate the interference of the expulsion action on the user's sleep structure, the intervention action not only needs to be minimal in intensity but also needs to be perfectly concealed in the exhalation phase, when the user's sensory nerves are least sensitive. The system uses an advance trigger timestamp, compensated for by both cross-domain network and mechanical delay, as the pulse start point. It uses the next physiologically predicted respiratory cycle as the interval between each pulse trigger and strictly establishes the span of the exhalation time window as the duration of a single expulsion, thereby constructing a pulse width modulation control timing sequence specific to the current user's physiological rhythm. The system extracts the motor energization start time coordinate and motor de-energization braking time coordinate for each subsequent continuously executed respiratory cycle. The timing boundary calculation logic is as follows: ; The parameters are explained as follows: : Coordinate of the start time of motor energization in the m-th cycle; : The time coordinate of the motor power-off braking in the m-th cycle; : Execution index number of the continuous respiratory cycle; : The absolute physical trigger time coordinate corresponding to the advance trigger timestamp; Predicting the respiratory cycle; : Motor voltage build-up time; Exhalation Time Window. After completing the timing sequence construction, the system packages and encodes the motor target speed with the time coordinate sequence, accurately addresses the physical layer hardware address of the microenvironment control node, and encapsulates it into a standard LAN driver message for distribution. When the control node receives the message and enters the execution phase, due to the early triggering mechanism at the power-on moment, the motor begins to accelerate before the human body begins exhaling, using this safe time to build up pressure inside the pipeline. When the physical time advances to the starting point of the actual exhalation time window, the motor has just passed the mechanical response ramp-up period and reached the rated motor target speed, and begins to output a stable and gentle single-pulse airflow to the mattress air chamber. Subsequently, throughout the entire exhalation process, the warm and humid air in the microenvironment is smoothly extracted. More importantly, in order to prevent residual airflow caused by the mechanical inertia of high-speed rotation after the fan is powered off, which would cause excess cold airflow to overflow into the user's highly sensitive inhalation phase, the system does not cut off the power and let it decelerate naturally at the moment the timing advances to the end of the exhalation time window span, but directly triggers the motor's physical electromagnetic brake. The electromagnetic brake instantly locks the motor rotor through reverse excitation, forcibly cutting off the air pressure output and ensuring that the physical airflow and the human body's exhalation stop simultaneously. The entire execution process utilizes time compensation and robust electrical control to perfectly embed the physical intervention into the gaps of the physiological rhythm, achieving truly imperceptible sleep regulation.

[0042] Please see Figure 2A sleep method based on Internet of Things communication includes the following steps: S1. Acquiring the cardiopulmonary coupling vibration waveform output by a piezoelectric sensor array arranged in the mattress support layer and separating the continuous respiratory envelope and heart rate variability feature sequence; simultaneously collecting transient temperature and humidity data of the microenvironment air chamber under the mattress surface; calculating the microclimate perspiration index based on the rate of change of the transient temperature and humidity data; S2. When the microclimate perspiration index is greater than a preset perspiration threshold and the high-frequency energy proportion of the heart rate variability feature sequence increases, performing time-domain tracking of the respiratory envelope, identifying the expiratory trough time points and expiratory duration of multiple consecutive respiratory cycles, establishing a respiratory rhythm model based on the time interval of the expiratory trough time points, extrapolating the predicted expiratory trough time point and expiratory time window of the next respiratory cycle, and generating a cooling intervention including the target ventilation volume. Requirements; S3. Send timestamp probe frames to the microenvironment control nodes in the IoT network, extract the dwell time and network round-trip time in the response frames returned by the microenvironment control nodes, calculate the one-way network communication latency, and subtract the one-way network communication latency and the motor pressure build-up time of the microenvironment control nodes from the predicted time point of the next exhalation trough in sequence to generate an advance trigger timestamp; S4. Analyze the target ventilation volume in the cooling intervention requirements, distribute the target ventilation volume evenly to a preset number of consecutive breathing cycles to set the single pulse air volume, take the advance trigger timestamp as the first execution starting point, generate intermittent control timing according to the span of the exhalation time window, and send drive messages to the microenvironment control nodes according to the control timing through the IoT communication protocol, driving the microenvironment control nodes to output a single pulse air volume only within the exhalation time window to perform dehumidification intervention.

[0043] In this implementation scheme, step S1 serves as the underlying data sensing hub of the entire control method. Its core function is to establish a seamless mapping relationship between the human physiological state and the microenvironment of the bed. Through synchronous acquisition by piezoelectric sensors and temperature and humidity probes, this step can transform chaotic physical vibration signals and slow environmental change data into structured respiratory waveforms, heart rate characteristics, and perspiration indices without contacting the user's body or increasing sleep burden. This process completely eliminates the invasiveness of traditional wearable devices, providing the most basic and highest-fidelity data source support for subsequent judgments on whether and when to initiate intervention. Step S2 plays a central role in intelligent decision-making and rhythm locking. This step first cross-validates the perspiration index and high-frequency energy ratio to ensure that subsequent mechanisms are triggered only when physiological heat and sweating actually occur, effectively avoiding misjudgments caused by natural fluctuations in room temperature. More importantly, it breaks through the limitations of blind intervention by traditional devices, using time-domain tracking to lock onto the exhalation phase where human sensory nerves are least sensitive, and proactively predicting the spatiotemporal scale of the next breath. This step successfully transformed the vague environmental regulation requirements into targeted intervention needs with strict time boundaries and specific air exchange volume indicators, establishing an absolutely safe execution benchmark for precise control. Step S3 establishes a dynamic time delay offsetting mechanism across physical domains, addressing the practical engineering pain point of lag between IoT control command transmission and underlying mechanical execution. Due to the uncertainty of wireless LAN routing and the physical ramp-up time required for the exhaust fan motor to build up pressure, issuing commands only when the predicted trough is reached, as per conventional logic, would inevitably lead to severe misalignment of the exhaust airflow or even intrusion into the intake phase. This step precisely extracts the pure network communication time by actively sending probe frames and superimposes it with the motor's mechanical pressure build-up time, shifting the predicted physiological time point in reverse along the time axis, ultimately generating an advance trigger timestamp. This mechanism ensures that digital commands can be transmitted in space in advance, and that the motor can overcome resistance and accelerate in advance, completely eliminating the system-level time lag in device response. Step S4 undertakes the mission of transforming abstract control algorithms into end-effector execution of physical, imperceptible actions. The core function of this step is to perform dual micro-dosage reshaping of the intervention intensity and timing. By equivalently discretizing the macroscopic ventilation demand across multiple consecutive respiratory cycles, the system transforms the strong, continuous cold air into a gentle, single-pulse breeze. Simultaneously, the system strictly uses the compensated timestamp as the starting point and the exhalation time window as the execution limit, forcing the mechanical actions of the microenvironment regulation nodes to completely follow the user's respiratory rhythm. This process ensures that the dehumidifying airflow is precisely delivered only during exhalation and abruptly stopped before inhalation, thus effectively improving the sleep microclimate while protecting the user from physical awakening.

[0044] In summary, this application has at least the following effects:

[0045] A sleep system and method based on Internet of Things (IoT) communication overcomes the serious lag and micro-arousal induction defects of traditional macroscopic sleep environment regulation. Through a multi-source state perception module, it accurately acquires cardiopulmonary coupling characteristics and microclimate perspiration index in a non-contact state. Combined with a rhythm phase analysis module, it proactively locks the expiratory time window where human sensory gating is least sensitive, thus strictly limiting intervention timing within an absolutely safe physiological time boundary. Simultaneously, a cross-domain latency compensation module dynamically separates and calculates the pure network communication time and the pressure build-up time of the underlying fan motor, performing dual pre-offset compensation of digital communication and physical mechanics at the predicted time point, completely eliminating the system-level time difference between command transmission and physical acceleration. Finally, through a rhythm-driven intervention module, the macroscopic target ventilation volume is equivalently discretized into a single micro-dose pulse air volume, strictly executing intermittent timing control according to the compensated advance trigger timestamp. This solution enables the underlying physical dehumidification action to closely match and precisely follow the body's instantaneous breathing rhythm. While efficiently replacing the stuffy microclimate inside the mattress, it ensures that the mechanical air supply is perfectly concealed during the natural exhalation stage and is abruptly stopped before inhalation, thus fundamentally avoiding stress-induced arousal caused by physical stimulation and achieving truly imperceptible closed-loop sleep intervention.

[0046] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0047] This invention is described with reference to flowchart illustrations and / or block diagrams of systems, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0048] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0049] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0050] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention.

[0051] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A sleep system based on Internet of Things (IoT) communication, characterized in that, Includes the following modules: The multi-source state sensing module is used to acquire the cardiopulmonary coupling vibration waveform output by the piezoelectric sensor array arranged in the mattress support layer and separate the continuous respiratory envelope and heart rate variability characteristic sequence. It also collects transient temperature and humidity data of the microenvironment air chamber under the mattress surface and calculates the microclimate perspiration index based on the rate of change of the transient temperature and humidity data. The rhythm phase analysis module is used to track the respiratory envelope in the time domain when the microclimate sweating index is greater than the preset sweating threshold and the high-frequency energy proportion of the heart rate variability characteristic sequence increases. It identifies the time points of the expiratory troughs and the duration of exhalation in multiple consecutive respiratory cycles, establishes a respiratory rhythm model based on the time interval of the expiratory troughs, extrapolates the predicted time points of the expiratory troughs and the expiratory time window for the next respiratory cycle, and generates cooling intervention requirements that include the target ventilation volume. The cross-domain latency compensation module is used to send timestamp detection frames to micro-environment control nodes in the Internet of Things network, extract the dwell time and network round-trip time in the response frames returned by the micro-environment control nodes, calculate the one-way network communication latency, and sequentially remove the one-way network communication latency and the motor pressure build-up time of the micro-environment control nodes from the next exhalation valley prediction time point for advance offset compensation, and generate an advance trigger timestamp. The rhythm-following intervention module is used to analyze the target ventilation volume in the cooling intervention requirements, and to discretize the target ventilation volume into a preset number of continuous breathing cycles to set the single pulse air volume. Taking the advance trigger timestamp as the first execution start point, it generates an intermittent control sequence based on the span of the exhalation time window. Through the Internet of Things communication protocol, it sends drive messages to the microenvironment control node according to the control sequence, driving the microenvironment control node to output a single pulse air volume to perform dehumidification intervention only within the exhalation time window.

2. The sleep system based on Internet of Things communication according to claim 1, characterized in that: The specific process of acquiring the cardiopulmonary coupled vibration waveform output by the piezoelectric sensor array arranged in the mattress support layer and separating the continuous respiratory envelope and heart rate variability characteristic sequence is as follows: The cardiopulmonary coupled vibration waveform output by the piezoelectric sensor array is bandpass filtered to extract the net cardiopulmonary vibration signal. Empirical mode decomposition is then applied to the net cardiopulmonary vibration signal to separate the low-frequency intrinsic mode function and the high-frequency intrinsic mode function. Perform Hilbert transform on the low-frequency intrinsic mode functions to extract the amplitude sequence of the analytic signal and generate a continuous breathing envelope; Local peak detection is performed on the high-frequency intrinsic mode function, the time interval between adjacent peaks is extracted to generate a heart rate interval sequence, and resampling and fast Fourier transform are performed on the heart rate interval sequence to extract frequency domain features and generate a heart rate variability feature sequence.

3. The sleep system based on Internet of Things communication according to claim 1, characterized in that: The specific process of simultaneously collecting transient temperature and humidity data from the microenvironment air chambers beneath the mattress surface and calculating the microclimate perspiration index based on the rate of change of these transient temperature and humidity data is as follows: The relative humidity component and temperature component are extracted from the transient temperature and humidity data. The relative humidity component and temperature component are transformed into the absolute humidity sequence of the air chamber using the Antoine equation. The absolute humidity sequence is then subjected to the first derivative of the sliding window to obtain the absolute humidity rate of change. The temperature component is then subjected to the first derivative of the sliding window to obtain the temperature rate of change. The absolute humidity change rate and temperature change rate are linearly weighted and fused to output the microclimate perspiration index.

4. A sleep system based on Internet of Things communication according to claim 1, characterized in that: When the microclimate perspiration index exceeds the preset perspiration threshold and the high-frequency energy proportion of the heart rate variability characteristic sequence increases, the specific process of time-domain tracking of the respiratory envelope to identify the expiratory trough time points and expiratory duration for multiple consecutive respiratory cycles is as follows: When the microclimate perspiration index is greater than the preset perspiration threshold and the high-frequency energy proportion of the heart rate variability characteristic sequence increases, the first-order difference operation is performed on the respiratory envelope. The zero-crossing point in the first-order difference sequence of the respiratory envelope that changes from positive to negative is taken as the exhalation start point, and the zero-crossing point that changes from negative to positive is taken as the inhalation start point. Perform a minimum search between the start of exhalation and the start of inhalation, and mark the location of the minimum as the trough of the exhalation wave. Calculate the relative span between the time coordinates of the inhalation start point and the exhalation start point to extract the exhalation duration.

5. A sleep system based on Internet of Things communication according to claim 4, characterized in that: The specific process of establishing a respiratory rhythm model based on the time interval of the expiratory trough, extrapolating the predicted expiratory trough time point and expiratory time window for the next respiratory cycle, and generating the cooling intervention requirements that include the target ventilation volume is as follows: The backward difference sequence of the expiratory trough time points of multiple consecutive respiratory cycles is calculated to obtain the historical respiratory cycle sequence. The historical respiratory cycle sequence is then subjected to autoregressive smoothing extrapolation to obtain the predicted respiratory cycle. Using the current last expiratory trough time point as the time reference, the predicted respiratory cycle is used to perform forward time axis extrapolation to generate the predicted expiratory trough time point of the next respiratory cycle. The expiration duration of multiple consecutive respiratory cycles is extracted to calculate the exponential moving average, and its time span is determined as the expiration time window of the next respiratory cycle. The excess perspiration deviation is obtained by extracting the overflow difference between the microclimate perspiration index and the preset perspiration threshold. The excess perspiration deviation is converted into a volume equivalent based on the preset air cell volume constant to generate the target ventilation volume and encapsulate it as a cooling intervention requirement.

6. A sleep system based on Internet of Things communication according to claim 1, characterized in that: The specific process of sending timestamp probe frames to micro-environment control nodes in the Internet of Things (IoT) network, extracting the dwell time and network round-trip time from the response frames returned by the micro-environment control nodes, and calculating the one-way network communication latency is as follows: Encapsulate a timestamp probe frame carrying a local transmission timestamp, route it to the micro-environment control node via the local IoT communication protocol, receive the response frame returned by the micro-environment control node, parse the response frame to extract the node's reception timestamp, the node's processing completion timestamp, and the local reception completion timestamp. The dwell time is extracted by calculating the time difference between the node processing completion timestamp and the node receiving timestamp. The network round-trip time is obtained by calculating the time span between the receiving completion timestamp and the local sending timestamp. The dwell time is filtered out from the network round-trip time to establish the pure network transmission time. The pure network transmission time is subjected to a half-equivalent conversion to output the one-way network communication delay.

7. A sleep system based on Internet of Things communication according to claim 1, characterized in that: The specific process of generating an advance trigger timestamp by sequentially removing one-way network communication delay and motor pressure build-up time of micro-environment control nodes from the next exhalation trough prediction time point and performing advance offset compensation is as follows: Retrieve the physical operating parameter table of the exhaust fan motor in the microenvironment control node, extract the mechanical response time constant of the exhaust fan motor from a stationary state to the rated exhaust speed, obtain the current airflow conduction damping coefficient of the microenvironment air chamber, introduce the current airflow conduction damping coefficient to perform damping hysteresis weighted calculation on the mechanical response time constant, and calculate the motor pressure build-up time. Based on the one-way network communication delay, reverse network delay compensation is performed on the predicted time point of the next exhalation trough. The network synchronization time coordinate is located, and the motor pressure build-up time is further removed from the network synchronization time coordinate to complete the mechanical response pre-compensation. The absolute physical trigger time coordinate is established and encapsulated as an advance trigger timestamp.

8. A sleep system based on Internet of Things communication according to claim 1, characterized in that: The specific process of analyzing the target ventilation volume in the cooling intervention requirement and discretizing the target ventilation volume into a preset number of consecutive respiratory cycles to set the single pulse air volume is as follows: The target ventilation volume is extracted by analyzing the cooling intervention requirements, and the preset number of consecutive respiratory cycles is obtained. The target ventilation volume is then processed by discrete equal distribution according to the preset number to map the basic exhaust volume of a single cycle. Obtain the current static pressure attenuation coefficient in the microenvironment chamber, perform pipeline fluid resistance conversion compensation on the single-cycle basic exhaust volume based on the current static pressure attenuation coefficient, and calculate the single pulse air volume. Based on the air volume and speed mapping curve of the dehumidifying fan motor, the air volume of a single pulse is mapped to the target speed of the motor.

9. A sleep system based on Internet of Things communication according to claim 1, characterized in that: Starting with the advance trigger timestamp as the first execution point, an intermittent control sequence is generated based on the span of the exhalation time window. Drive messages are sent to the microenvironment control node according to the control sequence via the Internet of Things communication protocol. The specific process of driving the microenvironment control node to output a single pulse airflow to perform dehumidification intervention within the exhalation time window is as follows: Using the advance trigger timestamp as the duty cycle starting point, the span of the expiratory time window as the pulse width, and the predicted respiratory cycle as the pulse interval, a pulse width modulation control timing sequence is constructed. The target speed of the motor and the pulse width modulation control timing are encoded into IoT communication control load, and the physical layer media access control address of the micro-environment control node is added to encapsulate and generate a drive message. The driver sends a drive message via the Internet of Things communication protocol to control the microenvironment control node to start the dehumidification fan motor to enter the accelerated pressure build-up state when the advance trigger time stamp is reached. After the motor pressure build-up time has elapsed, the motor maintains the target speed and outputs a single pulse air volume within the corresponding exhalation time window span. At the end of the exhalation time window span, the motor electromagnetic brake is triggered to stop the air delivery.

10. A sleep method based on Internet of Things (IoT) communication, applied to a sleep system based on IoT communication as described in any one of claims 1-9, characterized in that, Includes the following steps: S1. Acquire the cardiopulmonary coupling vibration waveform output by the piezoelectric sensor array arranged on the mattress support layer and separate the continuous respiratory envelope and heart rate variability characteristic sequence. Simultaneously collect the transient data of temperature and humidity in the microenvironment air chamber under the mattress surface and calculate the microclimate perspiration index based on the rate of change of the transient data of temperature and humidity. S2. When the microclimate sweating index is greater than the preset sweating threshold and the high-frequency energy proportion of the heart rate variability characteristic sequence increases, the respiratory envelope is tracked in the time domain to identify the expiratory trough time point and expiration duration of multiple consecutive respiratory cycles. A respiratory rhythm model is established based on the time interval of the expiratory trough time point, and the predicted expiratory trough time point and expiration time window of the next respiratory cycle are extrapolated to generate cooling intervention requirements that include the target ventilation volume. S3. Send a timestamp probe frame to the micro-environment control node in the Internet of Things network, extract the dwell time and network round-trip time in the response frame returned by the micro-environment control node, calculate the one-way network communication delay, and subtract the one-way network communication delay and the motor pressure build-up time of the micro-environment control node from the next exhalation valley prediction time point in turn to generate an advance trigger timestamp. S4. Analyze the target ventilation volume in the cooling intervention requirements, distribute the target ventilation volume evenly to a preset number of consecutive breathing cycles to set the single pulse air volume, take the advance trigger timestamp as the first execution start point, generate intermittent control timing according to the span of the exhalation time window, and send drive messages to the microenvironment control node according to the control timing through the Internet of Things communication protocol, drive the microenvironment control node to output a single pulse air volume to perform dehumidification intervention only within the exhalation time window.

Images