Remote monitoring method and system applied to children's watch

Abstract

The invention provides a remote monitoring method and system applied to children's watches, and relates to the technical field of children's watches, time axis synchronous alignment is carried out on collected multi-source heterogeneous data through an interpolation method, data time sequence chaos caused by sampling rate differences is prevented, and then the data time sequence chaos caused by sampling rate differences is prevented. Through dynamic coupling calculation of a light intensity mutation gradient and a heart rate stress coefficient, the cooperative influence of environment change on the physiological state of the child is quantified in real time, and the physiological state of the child is determined by further combining power spectrum centroid frequency and physiological phase difference analysis of the motion rhythm and logarithmic correlation of acoustic envelope frequency and GPS speed. In addition, an acoustic propagation delay compensation mechanism is introduced, a two-factor dynamic equation model is constructed through a time delay cross term, and a dynamic risk threshold generation mechanism based on a noise energy ratio is combined, so that the dynamic risk threshold of the motion intensity, the rhythm characteristics and the acoustic scene is effectively distinguished. Effective remote monitoring on child safety in different scenes is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of children's watches, and in particular to a remote monitoring method and system applied to children's watches. Background Art

[0002] With the popularization of smart wearable devices, children's watches have become an important tool for parents to monitor the safety of children. Existing children's watches often integrate a variety of functions, which are convenient for parents to contact children and monitor their whereabouts and behaviors. Existing children's watches generally integrate basic functions such as GPS positioning, electronic fences, and heart rate monitoring. The movement trajectory is restricted and alarmed through geographical location tracking and preset electronic fence boundaries, or the activity status of children is inferred by means of simple physiological data such as heart rate and steps. However, the existing technologies have significant limitations. On the one hand, relying solely on GPS positioning is vulnerable to signal occlusion and interference, and positioning drift often occurs indoors or in dense urban areas. The static rules of electronic fences cannot adapt to the dynamic activity scenarios of children, resulting in a high false alarm rate. On the other hand, single-dimensional physiological monitoring based on heart rate or accelerometer is difficult to distinguish normal movements from sudden risk behaviors. Especially in complex environments, factors such as noisy acoustic interference and sudden light changes will further reduce the detection reliability. In addition, the existing solutions lack in-depth fusion and spatio-temporal correlation analysis of multi-modal data. For example, the abnormal calls for help in acoustic signals, the synergistic effects of sudden changes in ambient light intensity on physiological stress, and the rhythmic coupling characteristics of movement acceleration and heart rate variability are not synchronously processed, resulting in the system having a lag in response and frequent misjudgments in actual applications with noise interference and multi-scenario interweaving, and it is difficult to perform real-time and effective remote monitoring of children under the interference and influence of the current complex external environment.

[0003] Therefore, it is necessary to provide a remote monitoring method and system applied to children's watches to solve the above technical problems. Summary of the Invention

[0004] To solve the above technical problems, the present invention provides a remote monitoring method and system applied to children's watches, achieving the beneficial effect of effectively remotely monitoring the behavior status of children.

[0005] The present invention provides a remote monitoring method applied to children's watches, including: S1: Real-time collect raw data, and perform a time-axis alignment operation on the raw data to generate a basic data set, where the basic data set includes three-axis motion signals, ambient sound pressure signals, ambient light intensity, heart rate PPG signals, and GPS geographical location coordinates; S2: Extract the heart rate RR interval sequence from the heart rate PPG signal, and calculate the ambient physiological synergy parameter based on the ambient light intensity and the heart rate RR interval sequence; A data processing method and system applied to a digital intelligent platform S3: Calculate the motion oscillation factor based on the three-axis motion signal and the heart rate RR interval sequence, and calculate the acoustic oscillation factor based on the environmental sound pressure signal and the GPS geographical location coordinates; S4: Based on the motion oscillation factor and the acoustic oscillation factor, and in combination with the environmental physiological cooperation parameter, construct a two-factor dynamic equation with a time delay cross term to calculate the risk index in real time; S5: Separate the noise component and the effective signal component from the environmental sound pressure signal, calculate the energy ratio factor of the noise component, generate a dynamic risk threshold based on the energy ratio factor, and trigger a remote monitoring response in combination with the risk index.

[0006] Preferably, in step S1, the time axis alignment operation aligns the original data to a preset unified time axis through cubic spline interpolation to generate a basic data set with time axis alignment.

[0007] Preferably, in step S2, the calculation steps of the environmental physiological cooperation parameter include: Calculate the mean value of the environmental light intensity data within a preset historical data window as the light intensity baseline value; Calculate the difference between the current environmental light intensity and the environmental light intensity before a preset time interval, and divide it by the preset time interval to generate the light intensity instantaneous change rate; Calculate the difference between the current environmental light intensity and the light intensity baseline value, obtain the light intensity mutation direction factor in combination with the sign function, and at the same time adjust the difference between the current environmental light intensity and the light intensity baseline value by using a preset Sigmoid function to obtain the light intensity mutation confidence; Multiply the light intensity instantaneous change rate by the light intensity mutation confidence to obtain the light intensity mutation gradient; Calculate the standard deviation of the heart rate RR interval sequence within a preset first historical window and the root mean square of the adjacent differences of the heart rate RR interval sequence within a preset second historical window, and multiply the ratio of the root mean square to the standard deviation by the light intensity mutation direction factor to obtain the heart rate stress coefficient; Multiply the light intensity mutation gradient by the heart rate stress coefficient, and then perform normalization processing through a preset hyperbolic tangent function to obtain the environmental physiological cooperation parameter.

[0008] Preferably, in step S3, the generation steps of the motion oscillation factor include: Separate the three-axis acceleration components from the three-axis motion signal, filter out noise and baseline drift through a preset band-pass filter, and perform square sum and square root operations on the filtered three-axis acceleration components to generate an acceleration magnitude signal; Perform spectral analysis on the acceleration magnitude signal, calculate its power spectral density distribution, and determine the power spectral centroid frequency through energy weighted averaging; Adjust the center frequency of a preset band - pass filter based on the centroid frequency of the power spectrum, perform secondary filtering on the acceleration magnitude signal, and perform Hilbert transform on the acceleration magnitude signal after secondary filtering to extract the instantaneous motion phase; Convert the heart rate RR - interval sequence into a heart rate variability signal by interpolation method, and extract the physiological instantaneous phase of the heart rate variability signal through Hilbert transform; Scale the centroid frequency of the power spectrum according to a preset scaling ratio, and then multiply it by the cosine value of the phase difference between the physiological instantaneous phase and the instantaneous motion phase to generate a motion oscillation factor.

[0009] Preferably, in step S3, the generation steps of the acoustic oscillation factor include: Perform full - wave rectification and low - pass filtering on the ambient sound pressure signal to generate a sound pressure envelope signal, and calculate the sound pressure envelope frequency through the zero - crossing detection method; Calculate the magnitude of the GPS speed based on the temporal variation of the GPS geographical location coordinates; Scale the sound pressure envelope frequency according to a preset normalization factor to generate an acoustic rhythm coefficient, and perform natural logarithm transformation on the magnitude of the GPS speed to generate a speed - related gain; Multiply the acoustic rhythm coefficient by the speed - related gain to generate an acoustic oscillation factor.

[0010] Preferably, in step S4, the construction of the two - factor dynamic equation includes the following steps: According to the acoustic signal propagation delay time, shift the motion oscillation factor and the acoustic oscillation factor backward along a preset unified time axis by a preset delay duration to generate a delayed motion oscillation factor and a delayed acoustic oscillation factor; Correct the motion oscillation factor, the acoustic oscillation factor, the delayed motion oscillation factor, and the delayed acoustic oscillation factor respectively through the environmental - physiological cooperation parameter; Subtract the product of the corrected acoustic oscillation factor and the corrected delayed motion oscillation factor from the product of the corrected motion oscillation factor and the corrected delayed acoustic oscillation factor to generate a two - factor dynamic equation.

[0011] Preferably, the corrected motion oscillation factor and the corrected delayed motion oscillation factor are generated by respectively superimposing the motion oscillation factor and the delayed motion oscillation factor with the environmental - physiological cooperation parameter according to a preset superimposing ratio.

[0012] Preferably, the corrected acoustic oscillation factor and the corrected delayed acoustic oscillation factor are generated by respectively adjusting the acoustic oscillation factor and the delayed acoustic oscillation factor with the environmental - physiological cooperation parameter according to a preset exponential decay function.

[0013] Preferably, in step S5, the dynamic risk threshold is generated by scaling the noise energy ratio factor by a preset scaling factor and superimposing it on a preset reference risk threshold.

[0014] The present invention also provides a remote monitoring system applied to a children's smartwatch, and a remote monitoring method applied to a children's smartwatch, including: A data acquisition and alignment module, configured to collect raw data in real time and perform a time-axis alignment operation on the raw data to generate a basic data set, where the basic data set includes triaxial motion signals, ambient sound pressure signals, ambient light intensity, heart rate PPG signals, and GPS geographical location coordinates; A collaborative parameter generation module, configured to extract a heart rate RR interval sequence from the heart rate PPG signal and calculate an ambient physiological collaborative parameter based on the ambient light intensity and the heart rate RR interval sequence; A coupling factor generation module, configured to calculate a motion oscillation factor based on the triaxial motion signal and the heart rate RR interval sequence, and calculate an acoustic oscillation factor based on the ambient sound pressure signal and the GPS geographical location coordinates; A dynamic equation construction module, configured to construct a two-factor dynamic equation with a time-delay cross term based on the motion oscillation factor and the acoustic oscillation factor, and in combination with the ambient physiological collaborative parameter, and calculate a risk index in real time; A monitoring response trigger module, configured to separate a noise component from the ambient sound pressure signal, calculate an energy ratio factor of the noise component, generate a dynamic risk threshold based on the energy ratio factor, and trigger a remote monitoring response in combination with the risk index.

[0015] Compared with related technologies, a remote monitoring method and system applied to a children's smartwatch provided by the present invention have the following beneficial effects: First of all, the present invention uses the cubic spline interpolation method to synchronize and align the collected multi-source heterogeneous data on the time axis, preventing the problem of data timing chaos caused by sampling rate differences, and ensuring the high-precision spatio-temporal matching of motion, acoustic, environmental and physiological signals. Secondly, through the dynamic coupling calculation of the light intensity mutation gradient and the heart rate stress coefficient, the collaborative impact of environmental changes on children's physiological states is quantified in real time, enhancing the perception sensitivity to abnormal scenarios. Further, by combining the power spectral centroid frequency of the motion rhythm and the physiological phase difference analysis, as well as the logarithmic correlation between the acoustic envelope frequency and the GPS speed, the deep fusion of the motion intensity, rhythm characteristics and acoustic scenarios is realized, effectively distinguishing normal activities from risk behaviors such as dragging and falling. In addition, an acoustic propagation delay compensation mechanism is introduced. By eliminating the spatio-temporal asynchronous interference between the motion and acoustic signals through the time delay cross term, and combining the environmental-physiological collaborative parameters to perform superposition enhancement and exponential decay adjustment on the dynamic coupling factor, a two-factor dynamic equation model is constructed, significantly reducing the false alarm rate in a noisy environment. Finally, based on the dynamic risk threshold generation mechanism of the noise energy ratio, the adaptive optimization of the risk determination logic is realized, and high robustness can be maintained in different scenarios such as noisy streets and enclosed spaces, achieving high-precision, low-latency, and strong anti-interference remote monitoring of children's safety. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 is a flowchart of a remote monitoring method applied to a children's watch according to the present invention; Figure 2 is a module structure diagram of a remote monitoring system applied to a children's watch according to the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0017] The present invention will be further described in detail below with reference to the drawings and embodiments. It can be understood that the specific embodiments described herein are only used to explain the present invention, rather than limiting the present invention. In addition, it should be noted that for the sake of description, only parts related to the present invention are shown in the drawings rather than all the structures. In addition, without conflict, the embodiments in the present invention and the features in the embodiments can be combined with each other.

[0018] In addition, it should be noted that for the sake of description, only parts related to the present invention are shown in the drawings rather than all the content. Before discussing the exemplary embodiments in more detail, it should be mentioned that some exemplary embodiments are described as processes or methods depicted as flowcharts. Although the flowcharts describe the operations (or steps) as being processed sequentially, many of the operations can be performed in parallel, concurrently or simultaneously. In addition, the order of the operations can be rearranged. When the operations are completed, the process can be terminated, but there can also be additional steps not included in the drawings. The process can correspond to a method, function, procedure, subroutine, subprogram, etc.

[0019] Example 1 A remote monitoring method applied to a children's watch. In the specific implementation process, as Figure 1 shown, it shows a flowchart of a remote monitoring method applied to a children's watch, including: Step S1: Collect raw data in real time and perform a time-axis alignment operation on the raw data to generate a basic data set. The basic data set includes three-axis motion signals, ambient sound pressure signals, ambient light intensity, heart rate PPG signals, and GPS geographical location coordinates.

[0020] Specifically, in step S1, the time-axis alignment operation aligns the raw data to a preset unified time axis through cubic spline interpolation to generate a time-axis aligned basic data set.

[0021] In the specific implementation process, the raw data is collected through the sensors and GPS module built in the watch. Exemplarily, the three-axis motion signals collect the raw acceleration data by the built-in high-precision accelerometer at a sampling rate of not less than 100 Hz, the ambient sound pressure signal captures the acoustic wave signal through the miniature digital microphone at a sampling rate of 16 kHz and enhances the high-frequency components through a pre-emphasis filter, the ambient light intensity obtains the light intensity value through the optical sensor at a sampling rate of 10 Hz, the heart rate PPG signal is collected by the photoelectric pulse sensor at a sampling rate of 128 Hz and the waveform is corrected by an adaptive motion artifact cancellation algorithm, and the GPS geographical location coordinates obtain the longitude and latitude information through the low-power GNSS module at a frequency of 1 Hz. All sensor data are marked with independent timestamps through hardware interrupt signals and transmitted to the processing unit for time-axis alignment operation, unified to the preset unified time axis. Exemplarily, the preset unified time axis is a time axis of a continuous time series starting from the system startup moment and at intervals of 10 ms. The target alignment time points are determined according to the preset unified time axis. For sensor data with different sampling rates, the cubic spline interpolation algorithm is used to reconstruct the continuous signal curve. Exemplarily, the GPS geographical location coordinate data of 1 Hz is interpolated into a trajectory sequence at intervals of 10 ms, the acceleration data of 100 Hz is resampled to the preset unified time axis, and at the same time the ambient sound pressure signal is downsampled to match the preset unified time axis. During the cubic spline interpolation process, the smoothness of the interpolation curve is ensured by setting natural boundary conditions, and the weights of the interpolation nodes are optimized by the minimum mean square error. Finally, a multi-modal basic data set with strictly aligned timestamps and uniform sampling intervals is generated, where the three-axis motion signals include the X / Y / Z axis acceleration components, the ambient sound pressure signal is the sound pressure level sequence after A-weighting filtering of the time-domain waveform, the ambient light intensity is the standardized lux value, the heart rate PPG signal is the denoised pulse wave peak-valley sequence, and the GPS geographical location coordinates include longitude, latitude, altitude, and positioning accuracy information, providing an accurate spatio-temporal reference for subsequent multi-modal data fusion.

[0022] Step S2: Extract the heart rate RR interval sequence from the heart rate PPG signal, and calculate the environmental physiological cooperation parameter based on the ambient light intensity and the heart rate RR interval sequence.

[0023] Specifically, in step S2, the calculation steps of the environmental physiological cooperation parameter include: Based on the ambient light intensity data within a preset historical data window, calculate the mean value of the ambient light intensity as the light intensity baseline value; Calculate the difference between the current ambient light intensity and the ambient light intensity before a preset time interval, and divide it by the preset time interval to generate the light intensity instantaneous change rate; Calculate the difference between the current ambient light intensity and the light intensity baseline value, and combine it with the sign function to obtain the light intensity mutation direction factor. At the same time, use the preset Sigmoid function to adjust the difference between the current ambient light intensity and the light intensity baseline value to obtain the light intensity mutation confidence; Multiply the light intensity instantaneous change rate by the light intensity mutation confidence to obtain the light intensity mutation gradient; Calculate the standard deviation of the heart rate RR interval sequence within a preset first historical window and the root mean square of the adjacent differences of the heart rate RR interval sequence within a preset second historical window, and multiply the ratio of the root mean square to the standard deviation by the light intensity mutation direction factor to obtain the heart rate stress coefficient; Multiply the light intensity mutation gradient by the heart rate stress coefficient, and then perform normalization processing through the preset hyperbolic tangent function to obtain the environmental physiological cooperation parameter.

[0024] In the specific implementation process, first, the heart rate RR interval sequence is extracted from the heart rate PPG signal. Exemplarily, time-domain peak detection is performed on the PPG signal through a 5-second sliding time window to identify the main peak of the pulse wave and calculate the time difference between adjacent peaks as the RR interval sequence. Median filtering is used to eliminate abnormal interval values caused by motion artifacts. At the same time, exemplarily, the optical sensor acquires the ambient light intensity at a sampling rate of 10 Hz. Based on a 30-second window length, the light intensity baseline value generated by sliding average filtering is calculated to represent the steady-state level of the ambient light. Then, the instantaneous change rate of the light intensity is calculated. Exemplarily, the difference between the current ambient light intensity value and the historical ambient light intensity value 5 seconds ago is taken and divided by the time interval of 5 seconds to generate the instantaneous change rate of the light intensity, which reflects the severity of the light intensity mutation. At the same time, the offset direction of the current light intensity relative to the baseline is judged through the sign function to generate the light intensity mutation direction factor. A light intensity mutation direction factor of +1 indicates an increase in light intensity, and a light intensity mutation direction factor of -1 indicates a decrease in light intensity. The light intensity difference is non-linearly mapped by combining the preset Sigmoid function. Exemplarily, the Sigmoid slope is set to 0.01 and the central threshold is set to 100 lux, and the light intensity difference is compressed to the [0,1] interval to generate the light intensity mutation confidence, which quantifies the credibility of the environmental mutation. The instantaneous change rate of the light intensity is multiplied by the light intensity mutation confidence to obtain the light intensity mutation gradient, which reflects the effective intensity of the light intensity change. In the heart rate stress analysis, exemplarily, the standard deviation of RR intervals SDNN within a preset first historical window of 30 seconds is calculated to reflect the overall level of heart rate variability, and the root mean square of the differences between adjacent RR intervals RMSSD within a preset second historical window of 5 minutes is calculated to represent the short-term heart rate regulation ability. By calculating the ratio of RMSSD / SDNN and multiplying it by the light intensity mutation direction factor, the heart rate stress coefficient is obtained, and its physical meaning is the stress influence intensity of the ambient light intensity mutation on the autonomic nervous system. Finally, the light intensity mutation gradient is multiplied by the heart rate stress coefficient and then normalized by the hyperbolic tangent function to generate the environmental-physiological coordination parameter. A positive environmental-physiological coordination parameter indicates a positive correlation between the increase in light intensity and heart rate stress, and a negative environmental-physiological coordination parameter indicates a negative correlation between the decrease in light intensity and heart rate inhibition. The absolute value reflects the coupling strength between the environment and physiology.

[0025] Step S3: Calculate the motion oscillation factor based on the three-axis motion signal and the heart rate RR interval sequence, and calculate the acoustic oscillation factor based on the ambient sound pressure signal and the GPS geographical location coordinates.

[0026] In the specific implementation process, the generation of the motion oscillation factor and the acoustic oscillation factor is achieved through multi-modal signal processing and feature fusion. The core logic is to quantify the coupling relationship between the motion rhythm and the physiological state and the spatio-temporal correlation between the acoustic scene and the moving speed.

[0027] Specifically, in step S3, the steps for generating the motion oscillation factor include: Separate the three-axis acceleration components from the three-axis motion signals, filter out the noise and baseline drift through a preset band-pass filter, perform square sum and square root operations on the filtered three-axis acceleration components to generate an acceleration magnitude signal; Perform spectral analysis on the acceleration magnitude signal, calculate its power spectral density distribution, and determine the power spectral centroid frequency through the energy weighted average method; Based on the power spectral centroid frequency, adjust the center frequency of the preset band-pass filter, perform secondary filtering on the acceleration magnitude signal, and perform Hilbert transform on the secondarily filtered acceleration magnitude signal to extract the instantaneous motion phase; Convert the heart rate RR interval sequence into a heart rate variability signal through interpolation method, and extract the physiological instantaneous phase of the heart rate variability signal through Hilbert transform; Scale the power spectral centroid frequency according to a preset scaling ratio, and then multiply it by the cosine value of the phase difference between the physiological instantaneous phase and the motion instantaneous phase to generate a motion oscillation factor.

[0028] In the specific implementation process, for the generation of the motion oscillation factor, first separate the X / Y / Z axis acceleration components from the three-axis motion signals, filter out high-frequency noise and low-frequency baseline drift through a preset band-pass filter. Exemplarily, the high-frequency noise includes but is not limited to vibration interference, and the low-frequency baseline drift includes but is not limited to gravity offset. Perform square sum and square root operations on the filtered three-axis acceleration components to generate an acceleration magnitude signal, a scalar value representing the motion intensity. Then perform fast Fourier transform spectral analysis on the acceleration magnitude signal, calculate its power spectral density distribution, and determine the power spectral centroid frequency through the energy weighted average method, that is, the energy concentration frequency band. Exemplarily, calculate the weighted average of the energy ratios of each frequency point in the range of 0.5Hz - 5Hz as the main frequency characteristic of the motion rhythm. Dynamically adjust the center frequency of the band-pass filter based on the centroid frequency. Exemplarily, adjust from the initial 5Hz to 3Hz, perform secondary filtering on the acceleration magnitude signal to extract the target frequency band signal, and then calculate the instantaneous motion phase through Hilbert transform to reflect the timing fluctuation characteristics of the motion rhythm. At the same time, exemplarily, convert the heart rate RR interval sequence into a continuous heart rate variability signal sampled at 4Hz through cubic spline interpolation method, and also extract the physiological instantaneous phase through Hilbert transform. Finally, after magnifying the power spectral centroid frequency by a preset ratio, exemplarily, the scaling ratio is 10, multiply it by the cosine value of the difference between the motion instantaneous phase and the physiological instantaneous phase to generate a motion oscillation factor, which characterizes the synchronization strength between the motion rhythm and the heart rate variability. A positive value indicates an enhanced rhythm synchronization, and a negative value indicates a rhythm desynchronization.

[0029] Specifically, in step S3, the generation steps of the acoustic oscillation factor include: Full-wave rectify and low-pass filter the ambient sound pressure signal to generate a sound pressure envelope signal, and calculate the sound pressure envelope frequency through the zero-crossing detection method; Calculate the GPS speed magnitude based on the time-series change of GPS geographical location coordinates; Scale the sound pressure envelope frequency by a preset normalization factor to generate an acoustic rhythm coefficient, and perform a natural logarithm transformation on the GPS speed magnitude to generate a speed correlation gain; Multiply the acoustic rhythm coefficient by the speed correlation gain to generate an acoustic oscillation factor.

[0030] In the specific implementation process, for the generation of the acoustic oscillation factor, first, full-wave rectify the ambient sound pressure signal, that is, invert the negative half-cycle signal to positive and perform low-pass filtering. Exemplarily, a cut-off frequency of 20 Hz is used to generate a sound pressure envelope signal, which characterizes the time-varying characteristics of acoustic energy; count the number of zero-crossings of the envelope signal per unit time through the zero-crossing detection method. Exemplarily, it is obtained by calculating the number of zero-crossings within a 0.1-second window divided by the window duration. Then, calculate the sound pressure envelope frequency, which reflects the rhythm characteristics of acoustic events; at the same time, calculate the speed magnitude based on the time-series change of GPS coordinates, that is, the difference value of the spherical distance between adjacent coordinate points, and generate a speed correlation gain through natural logarithm transformation to suppress the noise sensitivity at low speeds; scale the sound pressure envelope frequency by a normalization factor to an acoustic rhythm coefficient, and multiply it by the speed correlation gain to generate an acoustic oscillation factor. The larger the value of the acoustic oscillation factor, the stronger the spatio-temporal correlation between the acoustic rhythm and the moving speed. The acoustic oscillation factor provides multi-dimensional risk assessment input for the subsequent dynamic equation by quantifying the acoustic and moving correlation, and the motion oscillation factor provides multi-dimensional risk assessment input for the subsequent dynamic equation by quantifying the physiological motion synchronization.

[0031] Step S4: Based on the motion oscillation factor and the acoustic oscillation factor, and combined with the environmental-physiological cooperation parameter, construct a two-factor dynamic equation with a time-delay cross term to calculate the risk index in real time.

[0032] Specifically, in step S4, the construction of the two-factor dynamic equation includes the following steps: According to the acoustic signal propagation delay time, shift the motion oscillation factor and the acoustic oscillation factor backward along a preset unified time axis by a preset delay duration to generate a delayed motion oscillation factor and a delayed acoustic oscillation factor; Correct the motion oscillation factor, the acoustic oscillation factor, the delayed motion oscillation factor, and the delayed acoustic oscillation factor respectively through the environmental-physiological cooperation parameter; Multiply the corrected motion oscillation factor by the corrected delayed acoustic oscillation factor, and subtract the product of the corrected acoustic oscillation factor and the corrected delayed motion oscillation factor to generate a two-factor dynamic equation.

[0033] Specifically, the corrected motion oscillation factor and the corrected delayed motion oscillation factor are generated by superimposing the motion oscillation factor and the delayed motion oscillation factor with the environmental physiological cooperation parameter according to a preset superimposition ratio respectively.

[0034] Specifically, the corrected acoustic oscillation factor and the corrected delayed acoustic oscillation factor are generated by adjusting the acoustic oscillation factor and the delayed acoustic oscillation factor with the environmental physiological cooperation parameter according to a preset exponential decay function respectively.

[0035] In the specific implementation process, first, according to the physical delay time of the acoustic signal propagation, for example, by default, it is 20 ms, the motion oscillation factor and the acoustic oscillation factor are shifted backward along a preset unified time axis by the corresponding delay amount. For example, the delayed motion oscillation factor, denoted as , represents the current moment, represents the physical delay time of the acoustic signal propagation, and the delayed acoustic oscillation factor is denoted as , forming the delayed motion oscillation factor and the delayed acoustic oscillation factor. The motion oscillation factor at the current moment is , and the acoustic oscillation factor at the current moment is denoted as . Then, the motion oscillation factor, the acoustic oscillation factor, the delayed motion oscillation factor and the delayed acoustic oscillation factor are corrected respectively by the environmental physiological cooperation parameter k . For example, for the motion oscillation factor, the environmental physiological cooperation parameter k is used to linearly enhance the motion oscillation factor and the delayed motion oscillation factor with a superimposition ratio coefficient of 0.5. For example, , , and are the corrected motion oscillation factor and the corrected delayed motion oscillation factor respectively. For the acoustic oscillation factor, for example, the acoustic oscillation factor and the delayed acoustic oscillation factor are corrected by an exponential decay function with an attenuation coefficient of 1.0, denoted as , is the corrected acoustic oscillation factor, , is the corrected delayed acoustic factor. Finally, the product of the corrected motion oscillation factor and the corrected delayed acoustic oscillation factor is subtracted from the product of the corrected acoustic oscillation factor and the corrected delayed motion oscillation factor to generate a two-factor dynamic equation, denoted as , represents the risk index.

[0036] Step S5: Isolate the noise component from the ambient sound pressure signal, calculate the energy ratio factor of the noise component, generate a dynamic risk threshold based on the energy ratio factor, and trigger a remote monitoring response in combination with the risk index.

[0037] Specifically, in step S5, the dynamic risk threshold is generated by scaling the noise energy ratio factor by a preset scaling factor and superimposing it on a preset reference risk threshold.

[0038] In the specific implementation process, first, exemplarily, perform a short-time Fourier transform on the ambient sound pressure signal to generate a time-frequency spectrum, calculate the noise power spectral density using a noise estimation window, and perform noise reduction filtering on each frequency point through a Wiener gain function. After filtering, reconstruct the time-domain signal to obtain the effective acoustic component and the noise component. The noise energy ratio factor is obtained by calculating the ratio of the root mean square energy of the noise component to the total energy of the original ambient sound pressure signal. The generation of the dynamic risk threshold is based on the linear superposition of a preset reference risk threshold and the noise energy ratio factor. The formula is , where is the dynamic risk threshold, the scaling factor is determined by optimizing experimental data, is the noise energy ratio factor, is the preset reference risk threshold. When the ambient noise increases, is adaptively increased to reduce false alarms, and vice versa to increase sensitivity. The risk index triggers a hierarchical response after comparison with the dynamic risk threshold. Exemplarily, if ≥ continues to exceed the preset duration, a first-level alarm is activated. At the same time, the watch's recording and video functions are activated to obtain the image and sound information of the real-time environment of the watch wearer, the precise positioning function is enabled to obtain the precise geographical location coordinates of the watch wearer, and the image and sound information as well as the precise geographical location coordinates are transmitted to the cloud for storage, and a message is pushed to the parent's mobile phone; if accumulates more than times within 5 seconds and reaches the preset number of times, a second-level alarm is triggered and a call is automatically made to the parent's mobile phone. By means of the dynamic threshold mechanism, the false alarm rate is reduced in noisy streets and the detection sensitivity is improved in a quiet environment, realizing all-weather adaptive remote monitoring applied to children's watches.

[0039] The working principle of a remote monitoring method applied to a children's watch provided by the present invention is as follows: The present invention first synchronously collects multi-source sensor data such as motion, acoustics, light intensity, heart rate, and geographical location, and uses the cubic spline interpolation method to align the original data with different sampling rates to a unified time axis, eliminate the timing deviation, and generate a spatio-temporally consistent basic data set. Subsequently, in combination with the product of the environmental light intensity mutation gradient and the heart rate variability stress coefficient, after hyperbolic tangent function normalization processing, the dynamic synergy effect between the environmental change and the physiological state is quantified to generate an environmental-physiological synergy parameter. Further, a motion rhythm coupling factor, that is, a motion oscillation factor, is constructed through the spectral centroid frequency of the acceleration signal and the instantaneous phase difference of the heart rate variability. At the same time, an acoustic scene feature factor, that is, an acoustic oscillation factor, is generated based on the logarithmic correlation between the acoustic pressure envelope frequency and the GPS speed modulus length, which respectively characterizes the synchrony between the motion intensity and the physiological rhythm and the spatio-temporal correlation between the acoustic event and the moving state. Then, an acoustic signal propagation delay compensation mechanism is introduced, and the motion oscillation factor and the acoustic oscillation factor at the current and historical moments are superimposed and enhanced and exponentially decayed and corrected through the environmental-physiological synergy parameter. The time-delay cross term is used to cancel the common-mode noise and highlight the asynchronous risk characteristics, and a two-factor dynamic equation is constructed to generate a risk index in real time. Finally, the risk threshold is dynamically adjusted by the noise energy ratio, and a hierarchical response is achieved by combining the risk index calculated in real time, so as to balance the detection sensitivity and the false alarm rate in a complex environment, thereby completing the full-process high-precision remote monitoring target for children's watches from multi-dimensional data collection, feature fusion, dynamic modeling to adaptive decision-making.

[0040] Embodiment 2 A remote monitoring system applied to a children's watch, which is applied to a remote monitoring method for a children's watch. In the specific implementation process, as Figure 2 shown, it shows a module structure diagram of a remote monitoring system applied to a children's watch, including: A data acquisition and alignment module 100, configured to collect original data in real time and perform a time axis alignment operation on the original data to generate a basic data set, where the basic data set includes a three-axis motion signal, an environmental sound pressure signal, an environmental light intensity, a heart rate PPG signal, and a GPS geographical location coordinate; A synergy parameter generation module 200, configured to extract a heart rate RR interval sequence from the heart rate PPG signal and calculate an environmental-physiological synergy parameter based on the environmental light intensity and the heart rate RR interval sequence; A coupling factor generation module 300, configured to calculate a motion oscillation factor based on the three-axis motion signal and the heart rate RR interval sequence, and calculate an acoustic oscillation factor based on the environmental sound pressure signal and the GPS geographical location coordinate; A dynamic equation construction module 400, configured to construct a two-factor dynamic equation with a time delay cross term based on the motion oscillation factor and the acoustic oscillation factor, and calculate a risk index in real time; The monitoring response trigger module 500 is used to separate the noise component from the environmental sound pressure signal, calculate the energy proportion factor of the noise component, generate a dynamic risk threshold based on the energy proportion factor, and trigger a remote monitoring response in combination with the risk index.

[0041] The working principle of a remote monitoring system for children's smartwatches provided by the present invention is as follows: First, the data acquisition and alignment module 100 obtains multi-source heterogeneous raw signals in real time, aligns data with different sampling rates to a unified time axis by using the cubic spline interpolation method, and generates a spatio-temporally synchronized basic data set; the collaborative parameter generation module 200 analyzes the dynamic correlation between the light intensity mutation gradient and the heart rate variability stress coefficient, generates an environmental-physiological collaborative parameter through hyperbolic tangent function normalization, and quantifies the real-time impact of environmental changes on the physiological state; the coupling factor generation module 300 calculates the motion rhythm synchronization factor, i.e., the motion oscillation factor, by combining the centroid frequency of the acceleration spectrum and the heart rate phase difference, and simultaneously generates an acoustic scene feature factor, i.e., the acoustic oscillation factor, through the logarithmic correlation between the sound pressure envelope frequency and the GPS speed modulus, respectively characterizing the coupling relationship between the motion intensity and the physiological rhythm and the spatio-temporal correlation between the acoustic event and the moving state; the dynamic equation construction module 400 introduces an acoustic propagation delay compensation mechanism, uses the environmental-physiological collaborative parameter to enhance and attenuate the current and historical motion and acoustic factors, constructs a two-factor dynamic equation and calculates the risk index, cancels the environmental noise interference through the cross term and highlights the asynchronous risk characteristics, and calculates and generates the risk index in real time; the monitoring response trigger module 500 separates the acoustic noise through a Wiener filter and calculates the energy proportion, dynamically adjusts the risk determination threshold, and triggers a hierarchical alarm strategy in combination with the risk index, speed mutation, and effective acoustic features, realizing an adaptive remote monitoring with high sensitivity and low false alarm rate in a complex environment. Finally, a warning message is sent to the guardian through the wireless communication module, forming a full-process closed-loop system from multi-source data acquisition, feature fusion, dynamic modeling to intelligent decision-making.

[0042] This application is described with reference to the flowcharts and / or block diagrams of methods, apparatuses (systems), and computer program products according to embodiments of the present application. It should be understood that each process and / or block in the flowchart and / or block diagram can be implemented by computer program instructions, and the combination of the processes and / or blocks in the flowchart and / or block diagram can also be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate a device for realizing the functions specified in Figure 1 one process or multiple processes and / or blocks Figure 1 one block or multiple blocks.

[0043] Those of ordinary skill in the art can understand that all or part of the steps in the various methods of the above embodiments can be completed by instructing relevant hardware through a program, and the program can be stored in a computer-readable storage medium. The storage medium includes read-only memory (ROM), random access memory (RAM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), one-time programmable read-only memory (OTPROM), electrically-erasable programmable read-only memory (EEPROM), compact disc read-only memory (CD-ROM) or other optical disc memories, magnetic disk memories, tape memories, or any other medium that can be used to carry or store data and is computer-readable.

[0044] It should also be noted that the term "including", "comprising" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements not only includes those elements, but also includes other elements not expressly listed, or also includes elements inherent in such a process, method, commodity or device. Without further limitation, an element defined by the statement "including one..." does not exclude the existence of another identical element in the process, method, commodity or device including the element.

Claims

1. A remote monitoring method applied to a children's watch, characterized in that, The remote monitoring method includes the following steps: S1: Collect raw data in real time, and perform a time-axis alignment operation on the raw data to generate a basic data set, where the basic data set includes triaxial motion signals, ambient sound pressure signals, ambient light intensity, heart rate PPG signals, and GPS geographical location coordinates; S2: Extract the heart rate RR interval sequence from the heart rate PPG signal, and calculate the ambient physiological coordination parameter based on the ambient light intensity and the heart rate RR interval sequence; S3: Calculate the motion oscillation factor based on the triaxial motion signal and the heart rate RR interval sequence, and calculate the acoustic oscillation factor based on the ambient sound pressure signal and the GPS geographical location coordinates; S4: Based on the motion oscillation factor and the acoustic oscillation factor, and combined with the ambient physiological coordination parameter, construct a two-factor dynamic equation with time-delay cross terms, and calculate the risk index in real time; S5: Separate the noise component from the ambient sound pressure signal, calculate the energy ratio factor of the noise component, generate a dynamic risk threshold based on the energy ratio factor, and trigger a remote monitoring response in combination with the risk index.

2. The remote monitoring method for a children's watch according to claim 1, characterized in that In step S1, the time-axis alignment operation aligns the raw data to a preset unified time axis through cubic spline interpolation to generate a time-axis aligned basic data set.

3. The remote monitoring method for a children's watch according to claim 2, characterized in that, In step S2, the calculation steps of the ambient physiological coordination parameter include: Based on the ambient light intensity data within a preset historical data window, calculate the mean value of the ambient light intensity as the light intensity baseline value; Calculate the difference between the current ambient light intensity and the ambient light intensity before a preset time interval, and divide it by the preset time interval to generate the light intensity instantaneous change rate; Calculate the difference between the current ambient light intensity and the light intensity baseline value, and obtain the light intensity mutation direction factor in combination with the sign function. At the same time, use a preset Sigmoid function to adjust the difference between the current ambient light intensity and the light intensity baseline value to obtain the light intensity mutation confidence; Multiply the light intensity instantaneous change rate by the light intensity mutation confidence to obtain the light intensity mutation gradient; Calculate the standard deviation of the heart rate RR interval sequence within a preset first historical window and the root mean square of the adjacent differences of the heart rate RR interval sequence within a preset second historical window, and multiply the ratio of the root mean square to the standard deviation by the light intensity mutation direction factor to obtain the heart rate stress coefficient; Multiply the light intensity mutation gradient by the heart rate stress coefficient, and then perform normalization processing through a preset hyperbolic tangent function to obtain the ambient physiological coordination parameter.

4. The remote monitoring method for a children's watch according to claim 3, wherein In step S3, the generation steps of the motion oscillation factor include: Separate the triaxial acceleration components from the triaxial motion signal, filter out noise and baseline drift through a preset band-pass filter, and perform a square sum and square root operation on the filtered triaxial acceleration components to generate an acceleration magnitude signal; Perform spectral analysis on the acceleration magnitude signal, calculate its power spectral density distribution, and determine the power spectral centroid frequency through energy weighted averaging; Adjust the center frequency of the preset band-pass filter based on the power spectral centroid frequency, perform secondary filtering on the acceleration magnitude signal, and perform Hilbert transform on the secondary filtered acceleration magnitude signal to extract the motion instantaneous phase; Convert the heart rate RR interval sequence into a heart rate variability signal through interpolation, and extract the physiological instantaneous phase of the heart rate variability signal through Hilbert transform; Scale the centroid frequency of the power spectrum by a preset scaling ratio, and then multiply it by the cosine value of the phase difference between the physiological instantaneous phase and the motion instantaneous phase to generate a motion oscillation factor.

5. The remote monitoring method for a children's watch according to claim 4, characterized in that, In step S3, the generation steps of the acoustic oscillation factor include: Perform full-wave rectification and low-pass filtering on the ambient sound pressure signal to generate a sound pressure envelope signal, and calculate the sound pressure envelope frequency by the zero-crossing detection method; Calculate the GPS speed magnitude based on the temporal variation of the GPS geographical location coordinates; Scale the sound pressure envelope frequency by a preset normalization factor to generate an acoustic rhythm coefficient, and perform a natural logarithm transformation on the GPS speed magnitude to generate a speed correlation gain; Multiply the acoustic rhythm coefficient by the speed correlation gain to generate an acoustic oscillation factor.

6. The remote monitoring method for a children's watch according to claim 5, wherein, In step S4, the construction of the two-factor dynamic equation includes the following steps: According to the acoustic signal propagation delay time, shift the motion oscillation factor and the acoustic oscillation factor backward along a preset unified time axis by a preset delay duration to generate a delayed motion oscillation factor and a delayed acoustic oscillation factor; Correct the motion oscillation factor, the acoustic oscillation factor, the delayed motion oscillation factor, and the delayed acoustic oscillation factor respectively through the environmental physiological cooperation parameter; Subtract the product of the corrected acoustic oscillation factor and the corrected delayed motion oscillation factor from the product of the corrected motion oscillation factor and the corrected delayed acoustic oscillation factor to generate a two-factor dynamic equation.

7. The remote monitoring method for a children's watch according to claim 6, wherein The corrected motion oscillation factor and the corrected delayed motion oscillation factor are generated by respectively superimposing the motion oscillation factor and the delayed motion oscillation factor with the environmental physiological cooperation parameter according to a preset superimposition ratio.

8. The remote monitoring method for a children's watch according to claim 7, characterized in that, The corrected acoustic oscillation factor and the corrected delayed acoustic oscillation factor are generated by respectively adjusting the acoustic oscillation factor and the delayed acoustic oscillation factor with the environmental physiological cooperation parameter according to a preset exponential decay function.

9. The remote monitoring method for a children's watch according to claim 8, characterized in that In step S5, the dynamic risk threshold is generated by scaling the noise energy proportion factor by a preset proportional factor and superimposing it with a preset reference risk threshold.

10. A remote monitoring system applied to a children's watch, characterized in that, Applied to a remote monitoring method for a children's watch as described in any one of claims 1-9, the remote monitoring system includes: A data acquisition and alignment module for real-time collecting raw data and performing a time-axis alignment operation on the raw data to generate a basic data set, the basic data set including a three-axis motion signal, an ambient sound pressure signal, ambient light intensity, a heart rate PPG signal, and GPS geographical location coordinates; A cooperation parameter generation module for extracting a heart rate RR interval sequence from the heart rate PPG signal and calculating an environmental physiological cooperation parameter based on the ambient light intensity and the heart rate RR interval sequence; A coupling factor generation module for calculating a motion oscillation factor based on the three-axis motion signal and the heart rate RR interval sequence, and calculating an acoustic oscillation factor based on the ambient sound pressure signal and the GPS geographical location coordinates; A dynamic equation construction module for constructing a two-factor dynamic equation with a time-delay cross term based on the motion oscillation factor and the acoustic oscillation factor, and combining the environmental physiological cooperation parameter to calculate the risk index in real time; A monitoring response trigger module is used to separate the noise component from the environmental sound pressure signal, calculate the energy proportion factor of the noise component, generate a dynamic risk threshold based on the energy proportion factor, and trigger a remote monitoring response in combination with the risk index.

Images