An intelligent management system and method applied to a health management device

Abstract

This invention discloses an intelligent management system and method for health management devices, relating to the field of intelligent health monitoring technology. The system includes: collecting physiological micro-motion signals of a target object, extracting physiological rhythm feature parameters, and establishing an individualized physiological baseline model based on time periods; extracting physiological rhythm feature parameters in real time and calculating a deviation index relative to the individualized physiological baseline model; inputting the deviation index into a sensory compensation mapping model to generate physical stimulation parameters corresponding to the visual and tactile channels; calculating an effectiveness coefficient based on historical activation records of each sensory channel, simultaneously applying subthreshold detection stimuli to the target object, and calculating the normalized cross-correlation peak value with the detection stimulus template to obtain a perceptual accessibility coefficient; combining both to determine the delivery channel; and finally driving the corresponding sensory stimulation execution unit to output a physical stimulation signal. This invention effectively solves the problems of reminder habituation failure and ineffective reminder delivery due to obstructed perceptual channels in existing solutions.

Description

Technical Field

[0001] This invention relates to the field of intelligent health monitoring technology, specifically an intelligent management system and method applied to health management equipment. Background Technology

[0002] With the increasing aging of the population, the need for home-based health monitoring for elderly people living alone is becoming increasingly prominent. When elderly people experience abnormal health conditions, the abnormalities often go undetected due to the lack of caregivers, missing the optimal window for intervention.

[0003] For the aforementioned scenarios, existing technologies mainly fall into two categories. The first category is physiological parameter monitoring solutions based on wearable devices, which continuously collect physiological indicators and set threshold alarms using devices such as heart rate monitors and pulse oximeter clips. This type of solution has two fundamental drawbacks: first, elderly people living alone generally have poor adherence to wearing the devices, leading to monitoring failure due to prolonged periods without wearing them; second, fixed threshold alarm methods lack individualized benchmarks, resulting in significant differences in health status among different elderly individuals for the same indicators, leading to a high false alarm rate. The second category is physiological detection solutions based on non-contact sensing, utilizing technologies such as millimeter-wave radar and ultra-wideband to achieve non-invasive vital sign collection and mapping the detection results to physical stimuli such as changes in ambient light or vibration alerts. However, existing solutions of this type suffer from two deeply overlooked problems: First, the existing scheme sets a fixed mapping relationship between physical stimulation parameters and health deviation levels. Long-term repetitive fixed stimulation will lead to habituation of the nervous system. The elderly will gradually stop perceiving the same pattern of stimulation. The system will continue to output reminder signals, but the elderly will actually be completely unaware, forming a blind spot in monitoring.

[0004] Second, existing solutions all assume that the target object will necessarily perceive the physical signal output by the sensory stimulus execution unit, neglecting two actual situations that can lead to sensory failure: First, the elderly may be preoccupied with their current task, causing their corresponding sensory channels to be occupied; second, health abnormalities are often accompanied by a simultaneous decline in sensory abilities, such as decreased visual sensitivity when blood pressure is abnormally high, or sluggish tactile response when blood sugar is low. These situations occur precisely when sensory reminders are most needed, yet the receiving ability of the sensory channels weakens simultaneously. These circumstances result in the system outputting a reminder, but the reminder may not actually reach the elderly person's perception level, creating a false state of monitoring completion. Summary of the Invention

[0005] The purpose of this invention is to provide an intelligent management system and method for health management equipment, so as to solve the problems raised in the prior art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: an intelligent management method for health management devices, comprising: Step S100: Collect physiological micro-motion signals of the target object within a preset time period through a non-contact sensing unit, extract physiological rhythm feature parameters, and establish an individualized physiological benchmark model corresponding to the time period; Step S200: Real-time acquisition of physiological micro-motion signals of the target object, extraction of real-time physiological rhythm feature parameters, and calculation of the deviation index of the real-time physiological rhythm feature parameters relative to the individualized physiological benchmark model; Step S300: Input the deviation index into the sensory compensation mapping model to generate corresponding physical stimulus parameters, wherein the physical stimulus parameters include stimulation features for at least two sensory channels; Step S400: Based on the historical activation records of each sensory channel, calculate the current effectiveness coefficient of each sensory channel; apply a detection stimulus to the target object, and collect the micro-motion response of the target object to the detection stimulus through the non-contact sensing unit, and calculate the current perceptual accessibility coefficient of each sensory channel; determine the delivery channel from the at least two sensory channels based on the effectiveness coefficient and the perceptual accessibility coefficient. Step S500: Based on the determined delivery channel and its corresponding physical stimulation parameters, drive the corresponding sensory stimulation execution unit in the target space to output physical stimulation signals, so as to realize the reminder of abnormal health status of the target object.

[0007] Furthermore, in step S100, the physiological micro-motion signals of the target object within a preset time period are collected through a non-contact sensing unit, physiological rhythm feature parameters are extracted, and an individualized physiological benchmark model corresponding to the time period is established, specifically as follows: Step S101: The non-contact sensing unit transmits a frequency-modulated continuous wave signal to the target object, receives the echo signal reflected from the surface of the target object, performs frequency mixing processing on the echo signal and the transmitted signal, extracts the baseband complex signal, performs unwinding processing on the phase component of the baseband complex signal, and obtains a continuous phase sequence reflecting the periodic displacement of the target object's surface. The continuous phase sequence is converted into a surface displacement time sequence according to the correspondence between radar wavelength and phase displacement. The surface displacement time sequence is the physiological micro-motion signal. Step S102: The physiological rhythm characteristic parameters include three dimensions: respiratory rate, respiratory amplitude, and respiratory rhythm regularity; power spectral density estimation is performed on the physiological micro-motion signal of a preset time length, and the frequency value corresponding to the main peak of the power spectrum is extracted as the respiratory rate within the respiratory rate range of 0.1Hz to 0.6Hz, the square root of the power value of the main peak is extracted as the respiratory amplitude, and the reciprocal of the standard deviation of the time interval between adjacent respiratory periods is extracted as the respiratory rhythm regularity; Step S103: Divide 24 hours into several preset time periods; within each preset time period, continuously collect physiological rhythm characteristic parameters of the target object for no less than a preset number of days; calculate the mean and variance of each dimension of characteristic parameters collected within each preset time period, and use the normal distribution represented by the mean and variance as the individualized physiological baseline model corresponding to that time period; the individualized physiological baseline models of each preset time period together constitute a set of time-segmented baseline models covering the entire day.

[0008] Furthermore, in step S200, the real-time acquisition of the physiological micro-motion signals of the target object, the extraction of real-time physiological rhythm feature parameters, and the calculation of the deviation index of the real-time physiological rhythm feature parameters relative to the individualized physiological benchmark model are specifically as follows: Based on the preset time period to which the current moment belongs, the corresponding individualized physiological baseline model is invoked; the standardized deviation value relative to the corresponding baseline mean is calculated for each dimension of the real-time physiological rhythm characteristic parameters. The standardized deviation value is the quotient obtained by dividing the difference between the real-time parameter of that dimension and the baseline mean by the baseline standard deviation; the absolute values ​​of the standardized deviation values ​​of each dimension are weighted and summed according to preset weights, and the result is used as the deviation index. Among them, the weight corresponding to the respiratory rate dimension is greater than the weight corresponding to the respiratory amplitude dimension and the respiratory rhythm regularity dimension.

[0009] Furthermore, in step S300, the deviation index is input into the sensory compensation mapping model to generate corresponding physical stimulus parameters, specifically: Step S301: Pre-set at least three grading thresholds θ1<θ2<θ3, and divide the deviation index D(t) into the first level, the second level, the third level and the fourth level; Step S302: For each sensory channel, the deviation level is mapped to physical stimulus parameters according to the Weber-Fechner perceptual law. The correspondence between the set value of the physical stimulus quantity and the deviation degree satisfies an exponential function relationship, so that the subjective perception intensity of the physical stimulus and the deviation degree have a linear correspondence. For the visual channel, the color temperature offset ΔTv satisfies: For the tactile channel, the pulse amplitude Ah satisfies: Where Dmax is the upper limit of the deviation range, ΔTmax and Amin are hardware calibration parameters, and λ is the perception calibration coefficient; the perception calibration coefficient is individually calibrated through subjective perception scoring during the initial use of the target object.

[0010] Furthermore, in step S400, the calculation of the current effectiveness coefficient of each sensory channel based on the historical activation records of each sensory channel specifically involves: Step S401: For each sensory channel, record the timestamps of each activation within a preset history window; using the time interval between each activation timestamp and the current time as the independent variable, perform a weighted summation of the effects of each historical activation event according to an exponential decay function to obtain the time-decayed weighted activation density Ac(t) of that channel. Where ρ is the habituation rate coefficient, ti,c is the timestamp of the i-th activation, and nc is the number of activations within the historical window; substituting the weighted activation density into the exponential decay function, the effectiveness coefficient Ec(t) is obtained: Where E0 is the initial effectiveness coefficient and ρE is the effectiveness decay coefficient; when the effectiveness coefficient is lower than the preset minimum effectiveness threshold, random perturbations within a preset range are introduced into the temporal interval of the physical stimulus signal output by the channel in this instance, so as to maintain the novelty of the stimulus to the target object.

[0011] Furthermore, in step S400, a probing stimulus is applied to the target object, and the non-contact sensing unit collects the target object's micro-motion response to the probing stimulus, calculating the current perceptual accessibility coefficient of each sensory channel, specifically: According to the effectiveness coefficients from high to low, subthreshold detection stimuli with an intensity of a preset ratio to the corresponding formal physical stimulus parameters are generated sequentially for each sensory channel, and the subthreshold detection stimuli are applied to the target object through the corresponding sensory stimulus execution unit. Simultaneously with the application of the detection stimuli, the physiological micro-motion signals of the target object are collected in real time through the non-contact sensing unit. After removing the physiological baseline within the respiratory frequency range, the micro-motion residual signal is obtained. The normalized cross-correlation function between the micro-motion residual signal and the time-domain template of the detection stimulus is calculated, and the peak value of the cross-correlation within a preset time delay range is taken as the perceptual accessibility coefficient Cc(t) of that sensory channel. ; Where δx(tj+g) is the sampled value of the time-shifted micro-motion residual signal, and j represents the sampling identifier; s probe,c (tj) represents the value of the temporal template of the probe stimulus at the j-th sampling time; gmax is the maximum search delay covering the individual reaction time differences of the target object, and g represents the search delay identifier.

[0012] The higher the perceptual accessibility coefficient, the more the target object has generated a micromotion response related to the stimulus pattern to the channel's detection stimulus, meaning that the current perceptual accessibility of the channel is higher.

[0013] In step S400, based on the effectiveness coefficient and the perceived accessibility coefficient, a delivery channel is determined from the at least two sensory channels, specifically as follows: For each sensory channel, the effectiveness coefficient and the perceived accessibility coefficient are weighted and summed according to preset weights to obtain the comprehensive delivery priority score of each channel, where the weight corresponding to the perceived accessibility coefficient is greater than the weight corresponding to the effectiveness coefficient; the sensory channel with the highest comprehensive delivery priority score is selected as the delivery channel; if the perceived accessibility coefficient of all sensory channels is lower than the preset minimum accessibility threshold, it is determined that the target object is currently unreachable by all sensory channels, the physical stimulus output is skipped, and instead an abnormal notification is sent to the preset remote monitoring terminal, and the event is recorded in the abnormal log.

[0014] Furthermore, in step S500, based on the determined delivery channel and its corresponding physical stimulus parameters, the corresponding sensory stimulus execution unit within the target space is driven to output a physical stimulus signal to alert the target object to an abnormal health status. Specifically: After completing a physical stimulus signal output, the delivery channel identifier and corresponding output timestamp used in this operation are written into the historical activation record of that channel. When the number of historical activation records exceeds the preset historical window length, the earliest record is removed. The system then enters the next acquisition cycle at a preset step interval and repeats steps S100 to S400 to implement continuous closed-loop monitoring of the target object.

[0015] An intelligent management system for health management devices includes: A non-contact sensing unit is used to transmit electromagnetic wave signals to the target object and receive echo signals; The signal processing module, connected to the non-contact sensing unit, is used to perform phase unwinding processing on the physiological micro-motion signal and extract physiological rhythm feature parameters, and to output the micro-motion residual signal after removing the physiological baseline from the physiological micro-motion signal when a detection timing command is received. A benchmark modeling cloud platform, connected to the signal processing module, is used to establish and maintain individualized physiological benchmark models according to the physiological rhythm characteristic parameters over time periods. The deviation cloud computing module is connected to the signal processing module and the benchmark modeling cloud platform respectively, and is used to calculate the deviation index based on the real-time physiological rhythm characteristic parameters and the individualized physiological benchmark model of the corresponding time period. A sensory compensation mapping module, connected to the deviation cloud computing module, is used to map the deviation index into physical stimulation parameters for at least two sensory channels. The cloud platform management module is connected to the sensory compensation mapping module, the signal processing module, and the execution transmission control module, respectively. The execution transmission control module is connected to the cloud platform management module and includes at least two types of sensory stimulation execution units, used to drive the corresponding sensory stimulation execution unit to output detection stimuli in the detection mode according to the subthreshold intensity, and to drive the corresponding sensory stimulation execution unit to output physical stimulation signals in the delivery mode according to the determined delivery channel and corresponding physical stimulation parameters. The remote notification module, connected to the cloud platform management module, is used to push an exception notification to a preset remote terminal when the trigger signal is received.

[0016] Furthermore, the cloud platform management module is used to calculate the effectiveness coefficient based on the historical activation records of each sensory channel, send a detection stimulus command to the execution transmission control module and a detection timing command to the signal processing module, calculate the perceptual accessibility coefficient of each sensory channel based on the micro-motion residual signal, determine the delivery channel based on the effectiveness coefficient and the perceptual accessibility coefficient, and send a trigger signal to the remote notification module when the perceptual accessibility coefficient of all sensory channels is lower than a preset threshold.

[0017] Compared with the prior art, the beneficial effects of the present invention are: 1. This invention introduces an effectiveness coefficient model based on time decay weighted activation density to quantify the degree of habituation caused by repeated activation of each sensory channel. When the effectiveness coefficient is lower than the threshold, it automatically introduces controlled random perturbation into the stimulus temporal structure to maintain stimulus novelty, fundamentally counteract the neural habituation mechanism, and ensure the continuous effectiveness of reminders in long-term use scenarios. 2. This invention utilizes existing non-contact sensing units to apply subthreshold detection stimuli to the target object before the formal reminder. It then uses the normalized cross-correlation peak value of the micro-motion residual signal and the time-domain template of the detection stimulus to assess the perceptual accessibility of each sensory channel in real time. Formal reminders are only delivered to channels whose accessibility coefficients meet the requirements, upgrading reminder delivery from open-loop output to closed-loop verification, fundamentally eliminating false monitoring completion statuses caused by invalid delivery. When all channels become unreachable, it automatically escalates to remote notification, ensuring the monitoring link remains uninterrupted even in the most dangerous situations. 3. The perception and accessibility detection mechanism reuses existing non-contact sensing units to complete micro-motion response acquisition without increasing hardware costs. The overall system architecture has good compatibility with existing non-contact health monitoring platforms. Attached Figure Description

[0018] Figure 1 This is a flowchart illustrating an intelligent management method for health management devices according to the present invention. Detailed Implementation

[0019] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0020] Example: Figure 1 As shown, the present invention provides a technical solution, step S100: collecting physiological micro-motion signals of the target object within a preset time period through a non-contact sensing unit, extracting physiological rhythm feature parameters, and establishing an individualized physiological benchmark model corresponding to the time period; Step S200: Real-time acquisition of physiological micro-motion signals of the target object, extraction of real-time physiological rhythm feature parameters, and calculation of the deviation index of the real-time physiological rhythm feature parameters relative to the individualized physiological benchmark model; Step S300: Input the deviation index into the sensory compensation mapping model to generate corresponding physical stimulus parameters, wherein the physical stimulus parameters include stimulation features for at least two sensory channels; Step S400: Based on the historical activation records of each sensory channel, calculate the current effectiveness coefficient of each sensory channel; apply a detection stimulus to the target object, and collect the micro-motion response of the target object to the detection stimulus through the non-contact sensing unit, and calculate the current perceptual accessibility coefficient of each sensory channel; determine the delivery channel from the at least two sensory channels based on the effectiveness coefficient and the perceptual accessibility coefficient. Step S500: Based on the determined delivery channel and its corresponding physical stimulation parameters, drive the corresponding sensory stimulation execution unit in the target space to output physical stimulation signals, so as to realize the reminder of abnormal health status of the target object.

[0021] Furthermore, in step S100, the physiological micro-motion signals of the target object within a preset time period are collected through a non-contact sensing unit, physiological rhythm feature parameters are extracted, and an individualized physiological benchmark model corresponding to the time period is established, specifically as follows: Step S101: The non-contact sensing unit transmits a frequency-modulated continuous wave signal to the target object, receives the echo signal reflected from the surface of the target object, performs frequency mixing processing on the echo signal and the transmitted signal, extracts the baseband complex signal, performs unwinding processing on the phase component of the baseband complex signal, and obtains a continuous phase sequence reflecting the periodic displacement of the target object's surface. The continuous phase sequence is converted into a surface displacement time sequence according to the correspondence between radar wavelength and phase displacement. The surface displacement time sequence is the physiological micro-motion signal. The non-contact sensing unit can be configured in smart home devices such as smart canes and smart TVs. It transmits a continuous wave signal with a frequency that changes linearly with time to the target object via a frequency-modulated continuous wave radar. The echo signal reflected from the target object's surface is received and mixed with the transmitted signal to obtain a difference frequency signal corresponding to the distance to the target. The periodic, minute displacements of the target object's surface caused by respiratory movements result in continuous phase changes in the echo signal. By unwinding the phase of the difference frequency signal, the time series of surface displacement can be reconstructed, obtaining the micro-motion envelope signal.

[0022] Step S102: The physiological rhythm characteristic parameters include three dimensions: respiratory rate, respiratory amplitude, and respiratory rhythm regularity; power spectral density estimation is performed on the physiological micro-motion signal of a preset time length, and the frequency value corresponding to the main peak of the power spectrum is extracted as the respiratory rate within the respiratory rate range of 0.1Hz to 0.6Hz, the square root of the power value of the main peak is extracted as the respiratory amplitude, and the reciprocal of the standard deviation of the time interval between adjacent respiratory periods is extracted as the respiratory rhythm regularity; Step S103: Divide 24 hours into several preset time periods; within each preset time period, continuously collect physiological rhythm characteristic parameters of the target object for no less than a preset number of days; calculate the mean and variance of each dimension of characteristic parameters collected within each preset time period, and use the normal distribution represented by the mean and variance as the individualized physiological baseline model corresponding to that time period; the individualized physiological baseline models of each preset time period together constitute a set of time-segmented baseline models covering the entire day.

[0023] Furthermore, in step S200, the real-time acquisition of the physiological micro-motion signals of the target object, the extraction of real-time physiological rhythm feature parameters, and the calculation of the deviation index of the real-time physiological rhythm feature parameters relative to the individualized physiological benchmark model are specifically as follows: Based on the preset time period to which the current moment belongs, the corresponding individualized physiological baseline model is invoked; the standardized deviation value relative to the corresponding baseline mean is calculated for each dimension of the real-time physiological rhythm characteristic parameters. The standardized deviation value is the quotient obtained by dividing the difference between the real-time parameter of that dimension and the baseline mean by the baseline standard deviation; the absolute values ​​of the standardized deviation values ​​of each dimension are weighted and summed according to preset weights, and the result is used as the deviation index. Among them, the weight corresponding to the respiratory rate dimension is greater than the weight corresponding to the respiratory amplitude dimension and the respiratory rhythm regularity dimension.

[0024] Furthermore, in step S300, the deviation index is input into the sensory compensation mapping model to generate corresponding physical stimulus parameters, specifically: Step S301: Pre-set at least three grading thresholds θ1<θ2<θ3, and divide the deviation index D(t) into the first level, the second level, the third level and the fourth level; Step S302: For each sensory channel, the deviation level is mapped to physical stimulus parameters according to the Weber-Fechner perceptual law. The correspondence between the set value of the physical stimulus quantity and the deviation degree satisfies an exponential function relationship, so that the subjective perception intensity of the physical stimulus and the deviation degree have a linear correspondence. For the visual channel, the color temperature offset ΔTv satisfies: For the tactile channel, the pulse amplitude Ah satisfies: Where Dmax is the upper limit of the deviation range, ΔTmax and Amin are hardware calibration parameters, and λ is the perception calibration coefficient; the perception calibration coefficient is individually calibrated through subjective perception scoring during the initial use of the target object.

[0025] Furthermore, in step S400, the calculation of the current effectiveness coefficient of each sensory channel based on the historical activation records of each sensory channel specifically involves: Step S401: For each sensory channel, record the timestamps of each activation within a preset history window; using the time interval between each activation timestamp and the current time as the independent variable, perform a weighted summation of the effects of each historical activation event according to an exponential decay function to obtain the time-decayed weighted activation density Ac(t) of that channel. Where ρ is the habituation rate coefficient, ti,c is the timestamp of the i-th activation, and nc is the number of activations within the historical window; substituting the weighted activation density into the exponential decay function, the effectiveness coefficient Ec(t) is obtained: Where E0 is the initial effectiveness coefficient and ρE is the effectiveness decay coefficient; when the effectiveness coefficient is lower than the preset minimum effectiveness threshold, random perturbations within a preset range are introduced into the temporal interval of the physical stimulus signal output by the channel in this instance, so as to maintain the novelty of the stimulus to the target object.

[0026] Furthermore, in step S400, a probing stimulus is applied to the target object, and the non-contact sensing unit collects the target object's micro-motion response to the probing stimulus, calculating the current perceptual accessibility coefficient of each sensory channel, specifically: According to the effectiveness coefficients from high to low, subthreshold detection stimuli with an intensity of a preset ratio to the corresponding formal physical stimulus parameters are generated sequentially for each sensory channel, and the subthreshold detection stimuli are applied to the target object through the corresponding sensory stimulus execution unit. Simultaneously with the application of the detection stimuli, the physiological micro-motion signals of the target object are collected in real time through the non-contact sensing unit. After removing the physiological baseline within the respiratory frequency range, the micro-motion residual signal is obtained. The normalized cross-correlation function between the micro-motion residual signal and the time-domain template of the detection stimulus is calculated, and the peak value of the cross-correlation within a preset time delay range is taken as the perceptual accessibility coefficient Cc(t) of that sensory channel. ; Where δx(tj+g) is the sampled value of the time-shifted micro-motion residual signal, and j represents the sampling identifier; s probe,c (tj) represents the value of the temporal template of the probe stimulus at the j-th sampling time; gmax is the maximum search delay covering the individual reaction time differences of the target object, and g represents the search delay identifier.

[0027] The higher the perceptual accessibility coefficient, the more the target object has generated a micromotion response related to the stimulus pattern to the channel's detection stimulus, meaning that the current perceptual accessibility of the channel is higher.

[0028] In this embodiment, the preset ratio αsub is set based on the principle that the intensity of the detection stimulus should be near the target object's perception threshold, so that a measurable micro-motion response can be triggered when the perception channel is unobstructed, and no interference will occur when the perception channel is blocked. Specifically, during the system initialization phase, a perception threshold calibration process is performed on the target object. The visual channel color temperature offset and tactile channel pulse amplitude are increased step by step from zero, with each level applied for 2 seconds and the micro-motion residual signal collected. When the normalized cross-correlation peak first exceeds the noise baseline mean plus 3 times the standard deviation, the corresponding stimulus intensity is recorded as the individual perception threshold Sthreshold,c. Taking the preset ratio αsub = 0.6, the subthreshold detection stimulus intensity is: ; The color temperature shift step for the visual channel is typically ΔTstep = 50K; the pulse amplitude step for the tactile channel is typically ΔAstep = 5%Amax. The reason αsub is chosen to be 0.6 rather than a lower percentage is that if the percentage is too low, the detection stimulus is too weak, and even if the sensory channel is fully open, it is difficult to elicit a detectable micro-motion response, resulting in a falsely low reachability coefficient. If the percentage is too high, the detection stimulus itself is already close to the intensity of the formal reminder, losing the meaning of separating detection from reminder. 0.6 is an empirically optimal value that balances measurability and non-interference. Furthermore, for the visual channel, a color temperature shift of 0.6 times the sensory threshold subjectively manifests only as a very slight change in hue, which will not attract the active attention of the elderly.

[0029] Visual channel detection stimulus is continuous T probe A rectangular color temperature offset pulse of 2 seconds, with the temporal template defined as: ; where 1[·] is a rectangular window function, S probe,v This represents the subthreshold color temperature offset for the visual channel, measured in Kelvin. The start and end edges of the rectangular pulse correspond to the most likely blinking or slight head movement responses of the target object when the lighting environment changes abruptly. The time delay search interval [-gmax, gmax] in the cross-correlation calculation is typically set to gmax = 0.8 seconds, covering the typical reaction time range from visual stimulus to micro-motion response.

[0030] The tactile channel detection stimulus is a single decaying sine pulse, and the time-domain template is defined as: Where f is the center frequency of the vibratory motor, in Hz; γ is the attenuation coefficient, typically taken as γ = 15s. -1 ; TP = 0.15 seconds is the duration of a single pulse, S probe,h The amplitude is the subthreshold pulse. The typical micromotion response evoked by tactile stimulation is a slight contraction of the hand or upper limb, with a reaction time shorter than that of the visual channel. The time delay search interval corresponds to gmax = 0.5 seconds.

[0031] The time-domain templates for both channels are calculated and stored in the channel management module during system initialization, and can be directly called for each subsequent probe without needing to be regenerated in real time.

[0032] For the current acquisition window, synchronized with the detection stimulus, the length is T. probe The micro-motion envelope signal x(t) within +gmax is used to extract the respiratory baseline using a zero-phase bandpass filter with a passband of [0.1Hz, 0.6Hz]. Zero-phase filtering is used to avoid introducing phase delay and to ensure that the temporal alignment between the micro-motion residual signal and the probe stimulus time-domain template is not affected by the filter phase. .

[0033] In step S400, based on the effectiveness coefficient and the perceived accessibility coefficient, a delivery channel is determined from the at least two sensory channels, specifically as follows: For each sensory channel, the effectiveness coefficient and the perceived accessibility coefficient are weighted and summed according to preset weights to obtain the comprehensive delivery priority score of each channel, where the weight corresponding to the perceived accessibility coefficient is greater than the weight corresponding to the effectiveness coefficient; the sensory channel with the highest comprehensive delivery priority score is selected as the delivery channel; if the perceived accessibility coefficient of all sensory channels is lower than the preset minimum accessibility threshold, it is determined that the target object is currently unreachable by all sensory channels, the physical stimulus output is skipped, and instead an abnormal notification is sent to the preset remote monitoring terminal, and the event is recorded in the abnormal log.

[0034] In one embodiment of this application, the system calibrates the tactile channel perception threshold S of the target object during the initialization phase. threshold,h =0.4Amax, then the subthreshold detection stimulus amplitude S probe,h =0.6×0.4Amax=0.24Amax; The noise baseline statistics are μnoise=0.05, σnoise=0.02, so the minimum reachability threshold Cmin=0.05+3×0.02=0.11.

[0035] In one detection, the elderly person was sitting still and holding a cane. The calculated peak cross-correlation value of the tactile channel was Cc(t) = 0.38 > 0.11, indicating that the tactile channel was currently perceptible. At the same time, the elderly person was looking at a television screen, and the peak cross-correlation value of the visual channel was Cc(t) = 0.07 < 0.11, indicating that the visual channel was currently inaccessible. The channel management module selected the tactile channel as the delivery channel based on the comprehensive delivery priority score, driving the cane vibration execution unit to output a formal reminder pulse sequence.

[0036] Furthermore, in step S500, based on the determined delivery channel and its corresponding physical stimulus parameters, the corresponding sensory stimulus execution unit within the target space is driven to output a physical stimulus signal to alert the target object to an abnormal health status. Specifically: After completing a physical stimulus signal output, the delivery channel identifier and corresponding output timestamp used in this operation are written into the historical activation record of that channel. When the number of historical activation records exceeds the preset historical window length, the earliest record is removed. The system then enters the next acquisition cycle at a preset step interval and repeats steps S100 to S400 to implement continuous closed-loop monitoring of the target object.

[0037] An intelligent management system for health management devices includes: A non-contact sensing unit is used to transmit electromagnetic wave signals to the target object and receive echo signals; The signal processing module, connected to the non-contact sensing unit, is used to perform phase unwinding processing on the physiological micro-motion signal and extract physiological rhythm feature parameters, and to output the micro-motion residual signal after removing the physiological baseline from the physiological micro-motion signal when a detection timing command is received. A benchmark modeling cloud platform, connected to the signal processing module, is used to establish and maintain individualized physiological benchmark models according to the physiological rhythm characteristic parameters over time periods. The deviation cloud computing module is connected to the signal processing module and the benchmark modeling cloud platform respectively, and is used to calculate the deviation index based on the real-time physiological rhythm characteristic parameters and the individualized physiological benchmark model of the corresponding time period. A sensory compensation mapping module, connected to the deviation cloud computing module, is used to map the deviation index into physical stimulation parameters for at least two sensory channels. The cloud platform management module is connected to the sensory compensation mapping module, the signal processing module, and the execution transmission control module, respectively. The execution transmission control module is connected to the cloud platform management module and includes at least two types of sensory stimulation execution units, used to drive the corresponding sensory stimulation execution unit to output detection stimuli in the detection mode according to the subthreshold intensity, and to drive the corresponding sensory stimulation execution unit to output physical stimulation signals in the delivery mode according to the determined delivery channel and corresponding physical stimulation parameters. The remote notification module, connected to the cloud platform management module, is used to push an exception notification to a preset remote terminal when the trigger signal is received.

[0038] Furthermore, the cloud platform management module is used to calculate the effectiveness coefficient based on the historical activation records of each sensory channel, send a detection stimulus command to the execution transmission control module and a detection timing command to the signal processing module, calculate the perceptual accessibility coefficient of each sensory channel based on the micro-motion residual signal, determine the delivery channel based on the effectiveness coefficient and the perceptual accessibility coefficient, and send a trigger signal to the remote notification module when the perceptual accessibility coefficient of all sensory channels is lower than a preset threshold.

[0039] It also includes a computer storage medium configured to programmatically store Weber-Fechner sensing laws, time-domain templates, and cross-correlation peaks, enabling cloud computing and improving computing speed.

[0040] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

Claims

1. An intelligent management method for health management equipment, characterized in that: Specifically, the steps include the following: Step S100: Collect physiological micro-motion signals of the target object within a preset time period through a non-contact sensing unit, extract physiological rhythm feature parameters, and establish an individualized physiological benchmark model corresponding to the time period; Step S200: Real-time acquisition of physiological micro-motion signals of the target object, extraction of real-time physiological rhythm feature parameters, and calculation of the deviation index of the real-time physiological rhythm feature parameters relative to the individualized physiological benchmark model; Step S300: Input the deviation index into the sensory compensation mapping model to generate corresponding physical stimulus parameters, wherein the physical stimulus parameters include stimulation features of at least two sensory channels; Step S400: Calculate the current effectiveness coefficient of each sensory channel based on the historical activation records of each sensory channel; A probing stimulus is applied to the target object, and the micro-motion response of the target object to the probing stimulus is collected through the non-contact sensing unit. The current perceptual accessibility coefficient of each sensory channel is calculated. Based on the effectiveness coefficient and the perceptual accessibility coefficient, a delivery channel is determined from the at least two sensory channels. Step S500: Based on the determined delivery channel and its corresponding physical stimulation parameters, drive the corresponding sensory stimulation execution unit in the target space to output physical stimulation signals, so as to realize the reminder of abnormal health status of the target object.

2. The intelligent management method for health management equipment according to claim 1, characterized in that: In step S100, physiological micro-motion signals of the target object within a preset time period are collected through a non-contact sensing unit, physiological rhythm feature parameters are extracted, and an individualized physiological baseline model corresponding to the time period is established, specifically as follows: Step S101: The non-contact sensing unit transmits a frequency-modulated continuous wave signal to the target object, receives the echo signal reflected from the surface of the target object, performs frequency mixing processing on the echo signal and the transmitted signal, extracts the baseband complex signal, performs unwinding processing on the phase component of the baseband complex signal, and obtains a continuous phase sequence reflecting the periodic displacement of the target object's surface. The continuous phase sequence is converted into a surface displacement time sequence according to the correspondence between radar wavelength and phase displacement. The surface displacement time sequence is the physiological micro-motion signal. Step S102: The physiological rhythm characteristic parameters include three dimensions: respiratory rate, respiratory amplitude, and respiratory rhythm regularity; power spectral density estimation is performed on the physiological micro-motion signal of a preset time length, and the frequency value corresponding to the main peak of the power spectrum is extracted as the respiratory rate within the respiratory rate range of 0.1Hz to 0.6Hz, the square root of the power value of the main peak is extracted as the respiratory amplitude, and the reciprocal of the standard deviation of the time interval between adjacent respiratory periods is extracted as the respiratory rhythm regularity; Step S103: Divide 24 hours into several preset time periods; within each preset time period, continuously collect physiological rhythm characteristic parameters of the target object for no less than a preset number of days; calculate the mean and variance of each dimension of characteristic parameters collected within each preset time period, and use the normal distribution represented by the mean and variance as the individualized physiological baseline model corresponding to that time period; the individualized physiological baseline models of each preset time period together constitute a set of time-segmented baseline models covering the entire day.

3. The intelligent management method for health management equipment according to claim 2, characterized in that: In step S200, the real-time acquisition of the physiological micro-motion signals of the target object, the extraction of real-time physiological rhythm feature parameters, and the calculation of the deviation index of the real-time physiological rhythm feature parameters relative to the individualized physiological baseline model are specifically as follows: Based on the preset time period to which the current moment belongs, the corresponding individualized physiological baseline model is invoked; the standardized deviation value relative to the corresponding baseline mean is calculated for each dimension of the real-time physiological rhythm characteristic parameters. The standardized deviation value is the quotient obtained by dividing the difference between the real-time parameter of that dimension and the baseline mean by the baseline standard deviation; the absolute values ​​of the standardized deviation values ​​of each dimension are weighted and summed according to preset weights, and the result is used as the deviation index.

4. The intelligent management method for health management equipment according to claim 1, characterized in that: In step S300, the deviation index is input into the sensory compensation mapping model to generate corresponding physical stimulus parameters, specifically: Step S301: Pre-set at least three grading thresholds θ1<θ2<θ3, and divide the deviation index D(t) into the first level, the second level, the third level and the fourth level; Step S302: For each sensory channel, the deviation level is mapped to physical stimulus parameters according to the Weber-Fechner perceptual law.

5. The intelligent management method for health management equipment according to claim 4, characterized in that: In step S400, the calculation of the current effectiveness coefficient of each sensory channel based on the historical activation records of each sensory channel is specifically as follows: For each sensory channel, record the timestamp of each activation of that channel within a preset history window; Using the time interval between each activation timestamp and the current time as the independent variable, the influence of each historical activation event is weighted and summed according to the exponential decay function to obtain the time decay weighted activation density Ac(t) of this channel: Where ρ is the habituation rate coefficient, ti,c is the timestamp of the i-th activation, nc is the number of activations within the historical window; t represents the current timestamp; substituting the weighted activation density into the exponential decay function, the effectiveness coefficient Ec(t) is obtained: Where E0 is the initial effectiveness coefficient and ρE is the effectiveness decay coefficient; when the effectiveness coefficient is lower than the preset minimum effectiveness threshold, random perturbations within a preset range are introduced into the temporal interval of the physical stimulus signal output by the channel in this instance, so as to maintain the novelty of the stimulus to the target object.

6. The intelligent management method for health management equipment according to claim 5, characterized in that: In step S400, a probing stimulus is applied to the target object, and the non-contact sensing unit collects the target object's micro-motion response to the probing stimulus. The current perceptual accessibility coefficient of each sensory channel is calculated, specifically as follows: According to the effectiveness coefficients from high to low, subthreshold detection stimuli with an intensity of a preset ratio to the corresponding formal physical stimulus parameters are generated for each sensory channel in sequence, and the subthreshold detection stimuli are applied to the target object through the corresponding sensory stimulus execution unit; while applying the detection stimuli, the physiological micro-motion signals of the target object are collected in real time through the non-contact sensing unit, and the micro-motion residual signal is obtained after removing the physiological baseline within the respiratory frequency range; the normalized cross-correlation function between the micro-motion residual signal and the time-domain template of the detection stimulus is calculated, and the peak value of the cross-correlation within the preset time delay range is taken as the perceptual accessibility coefficient Cc(t) of the sensory channel: ； Wherein, δx(tj+g) is the micro-motion residual signal sample value after time shift, j represents the sample identifier; s probe,c (tj) represents the value of the detection stimulation time domain template at the jth sample time; gmax is the maximum search time delay covering the individual reaction time difference of the target object, and g represents the search time delay identifier.

7. The intelligent management method for health management equipment according to claim 6, characterized in that: In step S400, based on the effectiveness coefficient and the perceived accessibility coefficient, a delivery channel is determined from the at least two sensory channels, specifically as follows: For each sensory channel, the effectiveness coefficient and the perceived accessibility coefficient are weighted and summed according to preset weights to obtain the comprehensive delivery priority score for each channel, where the weight corresponding to the perceived accessibility coefficient is greater than the weight corresponding to the effectiveness coefficient. The sensory channel with the highest overall delivery priority score is selected as the delivery channel. If the perceptual accessibility coefficients of all sensory channels are lower than the preset minimum accessibility threshold, it is determined that the target object is currently unreachable by all sensory channels. The physical stimulus output is skipped, and instead, an abnormal notification is sent to the preset remote monitoring terminal, and the event is recorded in the abnormal log.

8. The intelligent management method for health management equipment according to claim 1, characterized in that: In step S500, based on the determined delivery channel and its corresponding physical stimulus parameters, the corresponding sensory stimulus execution unit in the target space is driven to output a physical stimulus signal to alert the target object to an abnormal health status. Specifically: After completing a physical stimulus signal output, the delivery channel identifier and corresponding output timestamp used in this operation are written into the historical activation record of that channel. When the number of historical activation records exceeds the preset historical window length, the earliest record is removed. The system then enters the next acquisition cycle at a preset step interval and repeats steps S100 to S400 to implement continuous closed-loop monitoring of the target object.

9. An intelligent management system for health management equipment, employing the intelligent management method for health management equipment as described in any one of claims 1-8, characterized in that: include: A non-contact sensing unit is used to transmit electromagnetic wave signals to the target object and receive echo signals; The signal processing module, connected to the non-contact sensing unit, is used to perform phase unwinding processing on the physiological micro-motion signal and extract physiological rhythm feature parameters, and to output the micro-motion residual signal after removing the physiological baseline from the physiological micro-motion signal when a detection timing command is received. A benchmark modeling cloud platform, connected to the signal processing module, is used to establish and maintain individualized physiological benchmark models according to time periods based on the physiological rhythm characteristic parameters. The deviation cloud computing module is connected to the signal processing module and the benchmark modeling cloud platform respectively, and is used to calculate the deviation index based on the real-time physiological rhythm characteristic parameters and the individualized physiological benchmark model of the corresponding time period. A sensory compensation mapping module, connected to the deviation cloud computing module, is used to map the deviation index into physical stimulation parameters for at least two sensory channels. The cloud platform management module is connected to the sensory compensation mapping module, the signal processing module, and the execution transmission control module, respectively. The execution transmission control module is connected to the cloud platform management module and includes at least two types of sensory stimulation execution units, used to drive the corresponding sensory stimulation execution unit to output detection stimuli in the detection mode according to the subthreshold intensity, and to drive the corresponding sensory stimulation execution unit to output physical stimulation signals in the delivery mode according to the determined delivery channel and corresponding physical stimulation parameters. The remote notification module, connected to the cloud platform management module, is used to push an exception notification to a preset remote terminal when the trigger signal is received.

10. The intelligent management system for health management equipment according to claim 9, characterized in that: The cloud platform management module is used to calculate the effectiveness coefficient based on the historical activation records of each sensory channel, send a detection stimulus command to the execution transmission control module and a detection timing command to the signal processing module, calculate the perceptual accessibility coefficient of each sensory channel based on the micro-motion residual signal, determine the delivery channel based on the effectiveness coefficient and the perceptual accessibility coefficient, and send a trigger signal to the remote notification module when the perceptual accessibility coefficient of all sensory channels is lower than a preset threshold.

Images