Intelligent internet-of-things-based health monitoring data processing method and system for physical exercise of the elderly

Abstract

The application discloses an old-age physical exercise health monitoring data processing method and system based on a smart Internet of Things, relates to the technical field of health monitoring data processing, and comprises the following steps: collecting light intensity data, surface reflectivity data and acceleration data in the health monitoring data processing link, performing time alignment processing on the collected light intensity data, surface reflectivity data and acceleration data, and generating exercise scene time sequence features. The application constructs exercise scene time sequence features through multi-source data synchronous collection and time alignment, realizes the differentiation of light and motion interference, and combines inverse sampling and position marking to maintain signal continuity. Through intermittent shading rhythm linkage sampling rhythm and brightness threshold dynamic regulation of light sources, blood flow signals are corrected in real time, and the health monitoring data is stable and reliable.

Description

Technical Field

[0001] This invention relates to the field of health monitoring data processing technology, specifically to a method and system for processing health monitoring data of elderly people's physical exercise based on the Internet of Things. Background Technology

[0002] Data processing for health monitoring of elderly people's physical exercise based on the Internet of Things (IoT) refers to the entire process of continuously collecting, stably transmitting, and fusing multidimensional health data of the elderly during physical exercise in an IoT environment, relying on various sensing devices such as wearable sensors, smart bracelets, posture recognition devices, and environmental monitoring nodes. The collected data covers heart rate, blood pressure, body temperature, cadence, exercise posture, and environmental temperature and humidity, and is aggregated to a cloud platform via a wireless communication network. On the cloud side, combined with big data processing technology, the raw data undergoes cleaning, time alignment, feature extraction, and correlation analysis to construct monitoring curves and behavioral profiles reflecting individual exercise status and health trends. Based on this, historical health data and group statistical characteristics are further integrated to continuously assess and warn of abnormal vital signs, fatigue accumulation, and potential risks during exercise, forming a closed-loop management process integrating collection, analysis, judgment, and intervention, thereby providing safer and more scientific support for physical exercise for the elderly.

[0003] The existing technology has the following shortcomings:

[0004] In existing technologies, skin blood flow monitoring based on optical sensors in the health monitoring data processing of elderly people during physical exercise typically determines changes in blood flow by detecting the intensity of light reflected from the skin surface. However, in environments with sudden changes in light, specular reflection, or strong light interference, optical sensors are easily affected by external light, generating random high-frequency flicker signals. These signals are not actual physiological changes but can easily be misinterpreted as a sharp increase in blood flow, leading to distorted judgments of blood flow trends. If these optical false peaks are not identified and eliminated in time, it can trigger false cardiovascular risk warnings or physical overload alarms, causing abnormal interruptions to the elderly person's exercise process and affecting the accuracy and continuity of health monitoring data processing.

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

[0006] The purpose of this invention is to provide a method and system for processing health monitoring data of elderly people's physical exercise based on the Internet of Things, so as to solve the problems in the background art mentioned above.

[0007] To achieve the above objectives, the present invention provides the following technical solution: a method for processing health monitoring data on physical exercise of the elderly based on the Internet of Things, comprising the following steps:

[0008] In the health monitoring data processing stage, light intensity data, surface reflectivity data, and acceleration data are collected. The collected light intensity data, surface reflectivity data, and acceleration data are time-aligned to generate time series features of the exercise scene that include light intensity change features, reflectivity change features, and acceleration change features.

[0009] Based on the time series features of the exercise scene, light intensity variation segments are extracted, and the motion factors and light factors in the light intensity variation segments are distinguished by combining posture parameters and gait parameters, resulting in a set of interference false peak features that includes light interference features and motion interference features.

[0010] The inverse sampling window is set according to the interference false peak feature set. Inverse sampling operation is performed on the optical monitoring signal within the inverse sampling window to suppress co-frequency flicker interference and generate traceable position markers for time positioning.

[0011] Based on the location markers, time resampling and synchronization alignment are performed on the heart rate curve data and cadence curve data. During the time resampling process, short-term shading control commands are generated, and an intermittent shading rhythm with periodic rhythm is formed based on the short-term shading control commands.

[0012] Based on the intermittent shading rhythm linked to the optical sensor sampling rhythm and the ambient brightness threshold, the system performs early backsampling, staggered sampling, and segmented light source shutdown operations according to the intermittent shading rhythm to correct blood flow trend data online, thereby achieving dynamic and stable processing of health monitoring data for elderly people's physical exercise.

[0013] Preferably, the steps for generating time-series features for training scenarios are as follows:

[0014] During the health monitoring data collection phase, optical sensing units and acceleration sensing units were set up at key movement sites of the elderly to acquire light intensity data, surface reflectivity data, and acceleration data, respectively.

[0015] A time index table is generated based on the collected light intensity data, surface reflectivity data and acceleration data, and the time correspondence of the three types of data is established with the sampling time of optical data as the time reference.

[0016] Time alignment processing is performed on light intensity data, surface reflectivity data, and acceleration data based on the time index table, so that each time node contains light intensity change characteristics, reflectivity change characteristics, and acceleration change characteristics.

[0017] By integrating time-aligned light intensity data, surface reflectivity data, and acceleration data, time-series features of the exercise scene are generated, reflecting the temporal correlation between motion state and light changes.

[0018] Preferably, the time alignment processing uses the sampling time of optical data as a unified time reference. By comparing the timestamps of surface reflectivity data and acceleration data, a set of matching time points is generated. The time distribution of acceleration data is adjusted according to the sampling rhythm of optical signals, so that each light intensity change point corresponds to a reflectivity change point and an acceleration change point, thereby forming a continuous training scene time series feature on a unified time axis.

[0019] Preferably, the steps for generating the interference pseudo-peak feature set are as follows:

[0020] The light intensity variation features in the time series features of the exercise scene are continuously scanned. The initial range of the light intensity variation segment is determined based on the amplitude trend of the light intensity variation. The reflection response accompanied by the light intensity variation is confirmed by combining the surface reflectivity variation features.

[0021] After the light intensity variation range is determined, the acceleration variation characteristics within the same time period are extracted synchronously to form time-corresponding data of attitude parameters and step frequency parameters.

[0022] Based on the characteristics of light intensity change, reflectivity change, and the time correlation between posture parameters and step frequency parameters, the factors in the light intensity change range are identified to distinguish between motion factors and light factors.

[0023] The feature data of the illumination factor segment and the motion factor segment are structured and integrated to generate an interference false peak feature set that includes illumination interference features and motion interference features.

[0024] Preferably, when generating the interference false peak feature set, the illumination interference features are identified according to the amplitude of light intensity change, the degree of reflectivity synchronization response and the duration of interference, and the motion interference features are classified according to the peak value of acceleration change, the range of attitude angle change and the step frequency synchronization deviation. The correspondence between illumination interference features and motion interference features is established with time index as the core, so as to achieve accurate positioning and classification recording of interference types on the time axis.

[0025] Preferably, the steps for generating location markers are as follows:

[0026] The time range and duration of the inverted sampling window are determined based on the interference pseudo-peak feature set. The duration of interference in the illumination interference feature is used as the basic boundary of the time window, and buffers are added at both ends of the time axis to enclose the variation of interference peak value.

[0027] Within the inverted sampling window, sampling points and phase pairing are performed on the optical monitoring signal so that each pair of sampling points is located in the rising and falling segments of the interference peak change, thereby achieving reverse cancellation of the signal in the time dimension.

[0028] Time positioning is performed on the optical monitoring signal after inversion sampling within the inversion sampling window to generate a traceable location marker carrying time index and interference type information;

[0029] The optical monitoring signal within the inverted sampling window is smoothed at the boundary, and the location markers are incorporated into the time series features to form the optical signal timeline.

[0030] Preferably, when the time range is determined, the inverted sampling window is based on the start and end times of the light interference characteristics. Buffers are set at both ends of the interference duration to cover the rising and attenuating phases of the interference peak change. Position markers are generated in the middle of the light interference segment to mark the equilibrium time of the optical monitoring signal after inverted sampling and serve as the reference node for subsequent time resampling.

[0031] Preferably, the steps for forming the intermittent shading rhythm are as follows:

[0032] Using location markers as time anchors, the heart rate curve data and cadence curve data are matched in terms of time range and node correspondence, so that the two data sequences have a unified time reference at each marker position.

[0033] Starting from the location marker, time resampling is performed on the heart rate curve data and cadence curve data to keep the time distribution of the two types of signals consistent and obtain traceable time node information.

[0034] During time resampling, a short-term shading control command is generated based on the synchronization status of the heart rate curve and the cadence curve to control the illumination period and exposure rhythm during the optical sensing acquisition process.

[0035] By sequencing and integrating short-term shading control commands into a time-based rhythm, an intermittent shading rhythm with periodicity is formed, achieving dynamic coordination of the optical acquisition process.

[0036] Preferably, the steps for linking the intermittent shading rhythm with the optical sensor sampling rhythm and the ambient brightness threshold, and performing sampling and light source control operations are as follows:

[0037] A time-linked relationship is established based on the intermittent shading rhythm and the optical sensor sampling rhythm, and an ambient brightness threshold is introduced to determine the illumination status, so that the optical signal sampling and illumination control are synchronously correlated in time.

[0038] Perform early backsampling operation according to the intermittent shading rhythm to ensure that the sampling window completes signal acquisition before the sudden change in light intensity, so as to maintain sampling stability;

[0039] After backing up the sampling in advance, a staggered sampling operation is performed. By adjusting the sampling trigger interval, the optical sampling point is time-shifted to the ambient light change cycle in order to suppress the effect of co-frequency flicker.

[0040] The light source is turned off in segments according to the intermittent shading rhythm, and the blood flow trend data is corrected online using time anchors obtained by early backsampling and staggered sampling to maintain time continuity and data stability.

[0041] The data processing system for monitoring the health of elderly people's physical exercise based on the Internet of Things includes a multi-source data synchronous acquisition module, an interference feature identification module, an anti-phase sampling suppression module, a time resampling and shading control module, and a rhythm linkage correction module.

[0042] The multi-source data synchronous acquisition module collects light intensity data, surface reflectivity data, and acceleration data in the health monitoring data processing stage. It performs time alignment processing on the collected light intensity data, surface reflectivity data, and acceleration data to generate time series features of the exercise scene that include light intensity change features, reflectivity change features, and acceleration change features.

[0043] The interference feature identification module extracts light intensity change segments based on the time series features of the exercise scene, and distinguishes motion factors and light factors in the light intensity change segments by combining posture parameters and gait frequency parameters, thus obtaining an interference false peak feature set containing light interference features and motion interference features.

[0044] The inverted sampling suppression module sets an inverted sampling window based on the interference false peak feature set, performs inverted sampling operation on the optical monitoring signal within the inverted sampling window to suppress co-frequency flicker interference, and generates traceable position markers for time positioning.

[0045] The time resampling and shading control module performs time resampling and synchronization alignment on heart rate curve data and cadence curve data based on location markers. During the time resampling process, it generates short-term shading control commands and forms an intermittent shading rhythm with periodic rhythm based on the short-term shading control commands.

[0046] The rhythm linkage correction module, based on the intermittent shading rhythm linkage optical sensor sampling rhythm and ambient brightness threshold, performs early backsampling, staggered sampling and segmented light source shutdown operations according to the intermittent shading rhythm to perform online correction of blood flow trend data, so as to achieve dynamic and stable processing of health monitoring data of elderly people's physical exercise.

[0047] The technical effects and advantages provided by the present invention in the above technical solution are as follows:

[0048] This invention constructs a time-series feature of an exercise scene that reflects the relationship between motion state and illumination changes by introducing synchronous acquisition and time alignment of light intensity data, surface reflectivity data, and acceleration data during health monitoring data processing. This feature distinguishes between illumination interference and motion interference in subsequent processing and maintains temporal continuity of the optical signal under complex motion and illumination conditions through the linkage of an inverted sampling window and location markers. Therefore, the optical monitoring signal can maintain stable output in environments with changes in motion posture, sudden changes in illumination, and fluctuations in skin reflection, improving the temporal consistency and detection accuracy of blood flow trend data.

[0049] This invention achieves dynamic linkage between the optical sensing sampling rhythm and the ambient brightness threshold by establishing an intermittent shading rhythm. It actively adjusts the light source state and sampling sequence when illumination fluctuates, enabling the optical acquisition to have adaptive interference suppression capabilities. Through a combination of pre-sampling backsampling, staggered sampling, and segmented light source shutdown control, the blood flow signal is corrected in real time during sampling, avoiding abnormal peak interference caused by illumination flicker. Therefore, the blood flow trend curve remains smooth and continuous in motion scenarios, ensuring the dynamic stability and reliability of health monitoring data. Attached Figure Description

[0050] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0051] Figure 1 This is a flowchart of the method for processing health monitoring data of elderly people's physical exercise based on the Internet of Things according to the present invention.

[0052] Figure 2 This is a schematic diagram of the modules of the smart Internet of Things-based elderly physical exercise health monitoring data processing system of the present invention. Detailed Implementation

[0053] Exemplary embodiments will now be described more fully with reference to the accompanying drawings. However, these exemplary embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, they are provided so that the description of this disclosure will be more complete and fully convey the concept of the exemplary embodiments to those skilled in the art.

[0054] This invention provides, for example Figure 1 The data processing method for health monitoring of elderly people's physical exercise based on smart Internet of Things, as shown, includes the following steps:

[0055] In the health monitoring data processing stage, light intensity data, surface reflectivity data, and acceleration data are collected. The collected light intensity data, surface reflectivity data, and acceleration data are time-aligned to generate time series features of the exercise scene that include light intensity change features, reflectivity change features, and acceleration change features.

[0056] In the health monitoring data processing, by jointly acquiring and synchronously processing light intensity data, surface reflectivity data, and acceleration data, accurate acquisition and temporal consistency of multidimensional sensory information of elderly people during physical exercise are achieved. The specific implementation steps are as follows:

[0057] During the health monitoring data collection phase, optical sensing units and accelerometers distributed at key movement sites of the elderly were used to acquire light intensity data, surface reflectivity data, and acceleration data, respectively. Light intensity data reflects changes in ambient light intensity on the skin surface; surface reflectivity data reflects the dynamic response of the skin to incident light over time; and acceleration data describes the amplitude, rhythm, and posture changes of the elderly's limb movements. During data acquisition, the sampling frequencies of the optical sensing units and the accelerometers were kept within the same order of magnitude to ensure accurate time alignment. When the elderly engaged in physical exercise, these three types of data were collected simultaneously, covering a full range of information from changes in external light intensity and skin surface light reflection to dynamic changes in body posture, ensuring sufficient temporal continuity and physical correlation for subsequent data fusion.

[0058] After acquiring multi-source raw data, a preliminary time index is generated. Light intensity data, surface reflectivity data, and acceleration data are affected by their respective sampling trigger signals during acquisition, resulting in time distribution shifts typically on the order of microseconds or milliseconds. To ensure comparability of the three types of data on the time axis, each type needs to be sorted and indexed according to its sampling timestamp. The time index generation process establishes a time correspondence between light intensity, surface reflectivity, and acceleration data by recording the trigger time and data packet sequence number of each sampling, and uses the sampling time of the optical data as the time base to form a unified time index table. This index table ensures that in subsequent time alignment, each set of light intensity, surface reflectivity, and acceleration data can be precisely matched in the form of time keys, thereby avoiding time mismatches caused by trigger delays or sampling drift between different sensor data.

[0059] Based on a time index table, time alignment processing is performed on light intensity data, surface reflectivity data, and acceleration data. During time alignment, the timestamps of surface reflectivity and acceleration data are compared point-by-point using the optical sampling time as a baseline. A matching set of time points is generated through interpolation, ensuring that the three types of data have a one-to-one correspondence of numerical records at the same time node. To ensure the continuity and stability of the time-aligned data, the time distribution of acceleration data is adjusted according to the sampling rhythm of the optical signal during alignment, so that each light intensity change point corresponds to an acceleration change point and a reflectivity change point. After this processing, all data are mapped onto a unified time axis, with each time node containing light intensity change characteristics, surface reflectivity change characteristics, and acceleration change characteristics. This time alignment method can reflect the instantaneous correlation between optical and motion signals during the movement of elderly individuals, enabling subsequent feature analysis to perform multidimensional inference under time-consistent conditions.

[0060] After time alignment, the light intensity data, surface reflectivity data, and acceleration data are integrated to generate time-series features of the exercise scene. This process takes the time-aligned 3D data as input and outputs a continuous data stream containing light intensity variation features, reflectivity variation features, and acceleration variation features in time node order. Light intensity variation features describe the changing trend of ambient light over time, including light enhancement, attenuation, and abrupt changes. Surface reflectivity variation features characterize the subtle fluctuations in skin reflectivity, reflecting the dynamic changes in light reflection on the skin surface under the influence of factors such as sweat, sebum, and posture adjustments. Acceleration variation features characterize the intensity and rhythmic changes of limb movements in the elderly during exercise, revealing the coupling relationship between movement and light together with optical signals. In generating the time-series features, the three types of features are arranged sequentially along the time axis, ensuring a logical one-to-one correspondence between optical changes and movement changes at each moment, forming a continuous feature trajectory reflecting the entire exercise process of the elderly.

[0061] Based on the time series features of the exercise scene, light intensity variation segments are extracted, and the motion factors and light factors in the light intensity variation segments are distinguished by combining posture parameters and gait parameters, resulting in a set of interference false peak features that includes light interference features and motion interference features.

[0062] In the subsequent analysis of the time-series features of the exercise scene, in order to extract the difference information between illumination interference and motion interference from the joint data of light intensity change features, reflectivity change features, and acceleration change features, it is necessary to extract and discriminate the light intensity change segments to identify the different effects of illumination factors caused by sudden changes in ambient light and motion factors caused by human movement, thereby generating an interference false peak feature set containing both illumination interference features and motion interference features. The specific implementation steps are as follows:

[0063] Continuous scanning of light intensity variation features within the time series characteristics of exercise scenarios is performed to determine the initial range of light intensity variation segments. This process is based on the fluctuation amplitude of light intensity variation features over time. By monitoring the increasing and decreasing trends of amplitude over continuous time periods of light intensity data, the trend of light intensity variation over time is determined. When light intensity changes show a sudden increase or decrease within a short period, it indicates a possible abrupt change in ambient lighting conditions within that segment. When determining light intensity variation segments, surface reflectance variation features are also used as an auxiliary criterion to confirm whether light intensity changes are accompanied by synchronous changes in skin reflectance. When both light intensity and reflectance changes show significant fluctuations, it can be preliminarily determined that the time segment contains potential light interference events. Through this process, a set of light intensity variation segments covering the entire exercise process for older adults is formed. Each segment is identified by a time index, marking its start and end times, providing a basic time reference for subsequent correlation analysis of posture and cadence.

[0064] After determining the light intensity variation segments, acceleration variation features within the same time period are extracted synchronously to form time-corresponding data for posture parameters and gait frequency parameters. Posture parameters describe the tilt angle, rotation direction, and spatial movement trends of the elderly person's limbs during this time period, while gait frequency parameters describe the movement rhythm and frequency. Both posture and gait frequency parameters are derived from the time series of acceleration variation features. By matching the time index of acceleration variation features within the light intensity variation segments, motion state data synchronized with light intensity change events can be obtained. In this way, the actual situation of elderly person's posture and gait frequency changes can be accurately obtained within each light intensity variation segment. Since the acceleration variation features, light intensity variation features, and reflectivity variation features have already been time-aligned in the previous stage, the posture and gait frequency parameters here correspond perfectly with the light intensity variation segments on the same time axis, ensuring temporal consistency in the discrimination of light and motion factors.

[0065] After establishing the time correspondence between light intensity variation segments and posture and gait parameters, factor discrimination is performed on each light intensity variation segment to distinguish between motion and illumination factors. This discrimination process is based on a comprehensive analysis of the changing trends of light intensity variation characteristics, the response characteristics of reflectivity variation characteristics, and the correlation between the dynamic amplitudes of posture and gait parameters. When the light intensity and reflectivity variation characteristics fluctuate synchronously within a certain segment, while the amplitude and frequency of posture and gait parameters are small and stable, it indicates that the light intensity variation in this segment is mainly caused by changes in ambient lighting conditions, belonging to the influence of illumination factors. Conversely, when posture and gait parameters exhibit drastic fluctuations or rhythmic peaks, while there is an asynchronous relationship between light intensity and reflectivity changes, it indicates that the optical signal fluctuations in this segment mainly originate from posture changes or local occlusion caused by body movement, belonging to the influence of motion factors. By comparing the optical and motion characteristics within the light intensity variation segments, the segment types of different interference sources can be accurately identified, providing a classification basis for the subsequent generation of feature sets.

[0066] After distinguishing between light and motion factors, the feature data from various segments are structurally integrated to generate a set of interference pseudo-peak features that includes both light and motion interference characteristics. Light interference features describe abnormal fluctuations in optical signals caused by sudden changes in external light intensity; their data includes the amplitude of light intensity changes, the degree of reflectivity synchronization response, and the duration of the interference. Motion interference features describe optical reflection anomalies caused by changes in posture or sudden changes in gait frequency; their data includes the peak value of acceleration changes, the range of posture angle changes, and gait frequency synchronization deviation. Each interference pseudo-peak feature is based on a time index, maintaining a consistent correspondence with the time series features of the original exercise scenario, allowing for precise location of light and motion interference on the time axis. The generated interference pseudo-peak feature set not only fully reflects the temporal distribution and amplitude characteristics of various interference events but also records the occurrence patterns and mutual influences of different interference types during elderly individuals' exercise. Through the establishment of this feature set, health monitoring data can be processed in subsequent stages to perform targeted signal correction and trend optimization based on the characteristic differences between light and motion interference, thereby achieving stable processing of optical signals in complex exercise scenarios.

[0067] The inverse sampling window is set according to the interference false peak feature set. Inverse sampling operation is performed on the optical monitoring signal within the inverse sampling window to suppress co-frequency flicker interference and generate traceable position markers for time positioning.

[0068] In processing the interference spurious peak feature set, to effectively suppress co-frequency interference caused by ambient light flicker and ensure the traceability and stability of the optical monitoring signal in the time dimension, it is necessary to set an inverse sampling window based on the interference spurious peak feature set and perform inverse sampling on the optical monitoring signal within this window. This cancels high-frequency flicker interference at the signal level and generates traceable position markers for time positioning. The specific implementation steps are as follows:

[0069] The time range and duration of the inverse sampling window are determined based on the interference pseudo-peak feature set. This feature set includes illumination interference features and motion interference features. Illumination interference features reflect information such as the amplitude of light intensity changes, reflectivity response delay, and interference duration. When setting the inverse sampling window, the interference duration in the illumination interference features is used as the basic boundary of the time window, while the start and end times of the illumination interference are combined to determine the window's time position. To ensure the inverse sampling window covers the complete cycle of illumination intensity fluctuations, buffer zones are added at both ends of the interference duration to guarantee that the inverse sampling window completely encompasses the rising and falling phases of the interference peak change on the time axis. In this process, the window's time length and position are not fixed but dynamically generated based on the time markers in the interference pseudo-peak feature set, ensuring that each illumination interference event receives an inverse sampling window matching its fluctuation period. Through this time mapping method, the inverse sampling window and the interference period of the optical signal are synchronously correlated, providing a time boundary for subsequent inverse sampling operations.

[0070] After determining the inverse sampling window, sampling points and phase pairing are performed on the optical monitoring signal within the window to form inverse sampling pairs. Optical monitoring signals typically exhibit periodic flickering or abrupt changes within interference zones. The basic principle of inverse sampling is to utilize the phase difference between adjacent time points to form complementary samples, thus canceling high-frequency fluctuations. In this embodiment, the sampling points of the optical signal within the inverse sampling window are divided into time-mirror pairs using a time index within the interference false peak feature set. Each pair of sampling points is located in the rising and falling segments of the interference peak change, resulting in opposite signal offset directions under the influence of illumination interference. Through this pairing relationship, the inverse sampling operation can reverse the interference signal at the time level, weakening the high-frequency components of the optical signal under the influence of co-frequency flicker. Because the setting of the inverse sampling window matches the time distribution of the interference false peak feature set, the inverse sampling operation is always performed within the interference zone, effectively avoiding over-correction of the normal optical signal.

[0071] After completing the inversion sampling operation, the processed optical monitoring signal within the inversion sampling window needs to be time-localized to generate traceable position markers. These position markers mark the reference time on the time axis after the optical signal has undergone inversion sampling, ensuring accurate time reference for the subsequent synchronization of the heart rate and cadence curves. When generating position markers, the moment with the most stable fluctuation amplitude in the inversion-sampled signal is selected as the marker location based on the start and end times of the inversion sampling window and the corresponding start and end points of the interference false peak feature set. This location is typically in the middle of the light interference segment, representing the equilibrium state of the signal after inversion sampling. Each position marker carries a time index, window number, and interference type information to enable cross-signal time mapping in subsequent processing. Since the inversion sampling window and the interference false peak feature set are one-to-one in time, the generation of position markers not only provides end-to-end traceability but also allows for the establishment of classification indexes based on different types of interference events, enabling independent identification of light interference and motion interference in the time dimension.

[0072] After generating location markers, the optical monitoring signals within the inverted sampling window undergo boundary smoothing, and the location markers are incorporated into the time series features, forming a traceable optical signal timeline. This timeline, with the location markers as core nodes, remaps the processed light intensity variation, reflectivity variation, and acceleration variation features onto a unified time axis. After inverted sampling and boundary smoothing, the high-frequency flicker components in the illumination interference region are effectively weakened, while the temporal continuity of the signal is restored. By introducing location markers into the time series, these time anchors can be directly referenced during subsequent resampling of heart rate and cadence curves, achieving precise synchronization of multi-source signals. The location markers not only record the time nodes of the inverted sampling operation but also establish a unified time tracking system for the entire health monitoring data processing process, ensuring temporal consistency and comparability between optical and motion signals in dynamic motion scenarios.

[0073] Based on the location markers, time resampling and synchronization alignment are performed on the heart rate curve data and cadence curve data. During the time resampling process, short-term shading control commands are generated, and an intermittent shading rhythm with periodic rhythm is formed based on the short-term shading control commands.

[0074] After inverting the sampling of the optical monitoring signal and generating location markers, the heart rate curve data and cadence curve data need to be re-matched and aligned in the time dimension to ensure that they reflect the dynamic relationship between the elderly's exercise state and physiological changes under the same time reference. In this process, time resampling not only compensates for the time offset between different signals but also actively adjusts the optical acquisition process by generating short-term shading control commands, thereby forming a periodic intermittent shading rhythm that coordinates the sampling light source with the exercise state in the time dimension. The specific implementation steps are as follows:

[0075] Using previously generated location markers as time anchors, time range matching and node mapping are performed on heart rate and cadence curve data. The location markers originate from the time positioning information within the optical signal inversion sampling window and have precise time indices corresponding to light interference segments. During time range matching, the corresponding sampling segment in the heart rate and cadence curve data is first determined based on the time index of each location marker, ensuring that both heart rate and cadence changes can be synchronously extended along the time axis with that marker as the reference. In this way, the sampling windows of the heart rate and cadence curve data are unified in time, giving both data sequences the same time reference at each marker position. When elderly individuals are in a continuous state of motion, the time interval between different markers reflects the cycle of motion changes; therefore, time range matching not only provides a unified time baseline between signals but also provides a rhythmic framework for the subsequent resampling process.

[0076] After completing time range matching, the heart rate and cadence curve data are resampled to eliminate sampling frequency differences and time drift between different signals. Time resampling uses location markers as the starting point and the time interval between preceding and following markers as the sampling period to redistribute and insert the heart rate and cadence curve data, ensuring consistency in their temporal distribution. During resampling, the time intervals in the heart rate curve are remapped to the time scale of the cadence curve, so that each time point corresponds to both a heart rate change and a cadence change. In this way, heart rate and cadence changes are strictly aligned on the time axis, enabling accurate analysis of the synchronization between motion state and physiological response in subsequent optical signal processing. Since the time resampling reference is determined by the location markers, the resampled time series not only maintains consistency with the optical signal but also possesses traceable time node information, providing a time reference for generating shading control commands.

[0077] While performing time resampling, short-term shading control commands are generated based on the resampled time nodes to control the illumination cycle and exposure rhythm during optical sensing acquisition. The generation of these short-term shading control commands is based on the synchronization of heart rate and cadence curves. When the rates of change of the heart rate and cadence curves tend to be consistent within a certain time period, it indicates that the elderly person is in a stable exercise rhythm, and normal sampling light intensity can be maintained during that period. However, when a phase shift or abrupt change occurs between the heart rate and cadence curves, it indicates possible adjustment of movement posture or light interference. In this case, the short-term shading control command triggers a momentary interruption or reduction of brightness of the light source, thereby avoiding the impact of sudden light reflection or rapid posture changes on optical acquisition. Each short-term shading control command corresponds to a time resampling node, with a clear trigger time and duration, enabling dynamic coordination between optical acquisition and physiological motion signals at the temporal level. Through this process, the sampling time of the optical signal can be adaptively adjusted according to the movement rhythm, thereby reducing the impact of ambient light fluctuations on the sampled data.

[0078] After generating short-term shading control commands, these commands are time-sequentially ordered and rhythmically integrated to form an intermittent shading rhythm with a periodic pattern. The intermittent shading rhythm uses the synchronous variation cycle of heart rate and cadence as the time baseline, arranging the short-term shading control commands in chronological order, so that the light interruption and recovery operations exhibit a regular, intermittent distribution on the time axis. This rhythm not only reflects the coordination between optical sampling and motion state but also demonstrates the active control effect of the light collection rhythm on the stability of signal acquisition. In the process of forming the intermittent shading rhythm, each shading cycle includes a sampling preparation phase, a light suppression phase, and a recovery phase, enabling the optical sensing unit to sample under the most suitable lighting conditions in different motion states. The intermittent shading rhythm is synchronized with the heart rate and cadence variation cycles in time, allowing the optical acquisition process to dynamically adjust with the movement rhythm, achieving a coordinated unity between physiological signal acquisition and motion state perception. In this way, the optical monitoring signal avoids the superimposed effects of light interference in continuous motion scenarios, while preserving the temporal accuracy and sampling integrity of physiological parameters.

[0079] Based on the intermittent shading rhythm linked to the optical sensor sampling rhythm and the ambient brightness threshold, the system performs early backsampling, staggered sampling, and segmented light source shutdown operations according to the intermittent shading rhythm to correct blood flow trend data online, so as to achieve dynamic and stable processing of health monitoring data of elderly people's physical exercise.

[0080] After establishing the intermittent shading rhythm, it needs to be dynamically linked with the optical sensor sampling rhythm and ambient brightness threshold to achieve real-time coordination between optical signal acquisition and ambient lighting conditions, thereby maintaining the continuity and stability of blood flow trend data under different motion and lighting scenarios. In this process, the intermittent shading rhythm not only serves as a time control reference for optical signal acquisition but also forms an active intervention mechanism through its interaction with the optical sampling rhythm and ambient brightness threshold, used to dynamically adjust the sampling method and light source control strategy. The specific implementation steps are as follows:

[0081] A time-linked relationship is established between the intermittent shading rhythm and the optical sensor sampling rhythm, and an ambient brightness threshold is introduced to determine the illumination state. The intermittent shading rhythm reflects the periodic changes in illumination control over time, while the optical sensor sampling rhythm reflects the temporal distribution of signal acquisition. The two need to be synchronized or coordinated on the time axis. By comparing the time points of the two rhythms, the shading period and illumination recovery period in the intermittent shading rhythm are paired with the sampling point distribution in the optical sensor sampling rhythm. This ensures that optical signal sampling can complete interference-free sampling under shading conditions and normal signal sampling during the illumination recovery phase. The ambient brightness threshold is introduced at this stage as a benchmark for external illumination determination, used to determine whether the external light intensity exceeds the tolerable range. When the ambient light intensity exceeds the preset threshold, even during the illumination recovery phase, a short-term shading state will automatically be entered according to the rhythm linkage mechanism. In this way, the intermittent shading rhythm, the optical sensor sampling rhythm, and the ambient brightness threshold are synchronously correlated at the time and state levels, providing a real-time reference for subsequent dynamic sampling adjustments.

[0082] After establishing the rhythmic linkage, the optical sensor sampling is pre-back-sampling based on the intermittent shading rhythm to acquire stable signals before sudden changes in illumination. Pre-back-sampling means that the system enters the sampling window in advance when the illumination state is about to change. By comparing the time period before the shading node in the rhythm cycle, the sampling trigger time is shifted forward, allowing the sampling operation to avoid the strong light interference phase. Since the intermittent shading rhythm and the optical sensor sampling rhythm are linked, pre-back-sampling can proactively execute data acquisition in advance based on the time prediction characteristics of the illumination cycle. This operation effectively prevents overexposure signals from being sampled in the early stages of enhanced illumination, thus ensuring the stability of the sampled data. In scenarios involving elderly people in motion, since changes in external illumination are often related to the direction of body movement, the time shift of pre-back-sampling ensures that the sampling is completed before environmental changes, thus keeping the sampled signal in a steady state before interference occurs. This processing method not only improves the temporal accuracy of the optical signal but also provides a transition buffer space for subsequent sampling rhythm adjustments.

[0083] After the early back-up sampling is completed, staggered sampling is performed on the sampling sequence to avoid co-frequency interference when the ambient light fluctuation frequency is close to the sampling frequency. Staggered sampling adjusts the sampling trigger interval to create a time offset between the optical sampling points and the ambient light change cycle, thereby reducing the impact of co-frequency flicker. This operation relies on the rhythmic linkage information from the previous stage. When the ambient brightness threshold is detected to fluctuate continuously within a short period, the system fine-tunes the sampling sequence according to the rhythmic period of the intermittent shading rhythm, distributing the sampling times at unequal intervals within each shading cycle. In this way, the optical sampling points are staggered from the peak of light fluctuation, resulting in a more stable reflected light signal received by the optical sensor. The implementation of staggered sampling makes the sampling rhythm variable in the time dimension, automatically adjusting the sampling interval according to the light change frequency. In elderly exercise scenarios, this mechanism can maintain the continuity of the optical monitoring signal under environments such as sunlight flicker, light reflection, or equipment obstruction, avoiding the interference of co-frequency flicker on blood flow trends, and laying the time distribution foundation for the next stage of light source control.

[0084] After staggered sampling, segmented light source shutdown is performed according to the intermittent shading rhythm, and online correction is applied to the blood flow trend data. Segmented light source shutdown refers to dividing the light source control into multiple time periods within a complete shading rhythm cycle, using different illumination states at different stages to achieve active distributed suppression of light interference. During periods of strong light interference, the light source is completely off to avoid the combined effects of strong external light; in the middle stage of stable illumination, the light source operates in low-brightness mode to maintain signal detection sensitivity; at the end of the period when ambient light is dim, the light source gradually restores normal brightness to enhance signal sampling quality. Each segmented operation corresponds to the optical sensing sampling rhythm, ensuring that the sampling operation is synchronized with changes in illumination. Through this dynamic illumination control, blood flow trend data can obtain multi-stage signals during sampling, allowing for correction based on signal amplitude differences in different illumination segments during data processing. During online correction, the time anchor points obtained from advance backsampling and staggered sampling are used to reconstruct the blood flow trend curve, ensuring it reflects the true physiological trend without being affected by sudden changes in illumination. This process ensures the coordination and unity of blood flow trends in terms of temporal continuity and data smoothness, so that the change curve of health monitoring data maintains a natural transition and stable output under dynamic lighting conditions.

[0085] This invention constructs a time-series feature of an exercise scene that reflects the relationship between motion state and illumination changes by introducing synchronous acquisition and time alignment of light intensity data, surface reflectivity data, and acceleration data during health monitoring data processing. This feature distinguishes between illumination interference and motion interference in subsequent processing and maintains temporal continuity of the optical signal under complex motion and illumination conditions through the linkage of an inverted sampling window and location markers. Therefore, the optical monitoring signal can maintain stable output in environments with changes in motion posture, sudden changes in illumination, and fluctuations in skin reflection, improving the temporal consistency and detection accuracy of blood flow trend data.

[0086] This invention achieves dynamic linkage between the optical sensing sampling rhythm and the ambient brightness threshold by establishing an intermittent shading rhythm. It actively adjusts the light source state and sampling sequence when illumination fluctuates, enabling the optical acquisition to have adaptive interference suppression capabilities. Through a combination of pre-sampling backsampling, staggered sampling, and segmented light source shutdown control, the blood flow signal is corrected in real time during sampling, avoiding abnormal peak interference caused by illumination flicker. Therefore, the blood flow trend curve remains smooth and continuous in motion scenarios, ensuring the dynamic stability and reliability of health monitoring data.

[0087] This invention provides, for example Figure 2 The data processing system for monitoring the health of elderly people's physical exercise based on the Internet of Things (IoT) includes a multi-source data synchronous acquisition module, an interference feature identification module, an anti-phase sampling suppression module, a time resampling and shading control module, and a rhythm linkage correction module.

[0088] The multi-source data synchronous acquisition module collects light intensity data, surface reflectivity data, and acceleration data in the health monitoring data processing stage. It performs time alignment processing on the collected light intensity data, surface reflectivity data, and acceleration data to generate time series features of the exercise scene that include light intensity change features, reflectivity change features, and acceleration change features.

[0089] The interference feature identification module extracts light intensity change segments based on the time series features of the exercise scene, and distinguishes motion factors and light factors in the light intensity change segments by combining posture parameters and gait frequency parameters, thus obtaining an interference false peak feature set containing light interference features and motion interference features.

[0090] The inverted sampling suppression module sets an inverted sampling window based on the interference false peak feature set, performs inverted sampling operation on the optical monitoring signal within the inverted sampling window to suppress co-frequency flicker interference, and generates traceable position markers for time positioning.

[0091] The time resampling and shading control module performs time resampling and synchronization alignment on heart rate curve data and cadence curve data based on location markers. During the time resampling process, it generates short-term shading control commands and forms an intermittent shading rhythm with periodic rhythm based on the short-term shading control commands.

[0092] The rhythm linkage correction module, based on the intermittent shading rhythm linkage optical sensor sampling rhythm and ambient brightness threshold, performs early backsampling, staggered sampling and segmented light source shutdown operations according to the intermittent shading rhythm to perform online correction of blood flow trend data, so as to achieve dynamic and stable processing of health monitoring data of elderly people's physical exercise.

[0093] The method for processing health monitoring data of elderly people's physical exercise based on smart Internet of Things provided in this embodiment of the invention is implemented through the above-mentioned data processing system for monitoring health monitoring data of elderly people's physical exercise based on smart Internet of Things. For details of the specific methods and processes of the data processing system for monitoring health monitoring data of elderly people's physical exercise based on smart Internet of Things, please refer to the above-mentioned embodiment of the method for processing health monitoring data of elderly people's physical exercise based on smart Internet of Things, which will not be repeated here.

[0094] The foregoing has only described certain exemplary embodiments of the present invention by way of illustration. Undoubtedly, those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the foregoing drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. A method for processing health monitoring data of elderly people's physical exercise based on smart Internet of Things, characterized in that, Includes the following steps: In the health monitoring data processing stage, light intensity data, surface reflectivity data, and acceleration data are collected. The collected light intensity data, surface reflectivity data, and acceleration data are time-aligned to generate time series features of the exercise scene. Based on the time series features of the exercise scene, light intensity variation segments are extracted, and the motion factors and light factors in the light intensity variation segments are distinguished by combining posture parameters and gait parameters, thus obtaining a set of interference false peak features. The inverted sampling window is set according to the interference false peak feature set. Inverted sampling operation is performed on the optical monitoring signal within the inverted sampling window to suppress co-frequency flicker interference and generate position marker points. The steps for generating location markers are as follows: The time range and duration of the inverted sampling window are determined based on the interference pseudo-peak feature set. The duration of interference in the illumination interference feature is used as the basic boundary of the time window, and buffers are added at both ends of the time axis to enclose the variation of interference peak value. Within the inverted sampling window, sampling points and phase pairing are performed on the optical monitoring signal so that each pair of sampling points is located in the rising and falling segments of the interference peak change, thereby achieving reverse cancellation of the signal in the time dimension. Time positioning is performed on the optical monitoring signal after inversion sampling within the inversion sampling window to generate a traceable location marker carrying time index and interference type information; The optical monitoring signal within the inverted sampling window is smoothed at the boundary, and the position markers are incorporated into the time series features to form an optical signal timeline. Based on location markers, time resampling and synchronization alignment are performed on heart rate curve data and cadence curve data. During the time resampling process, short-term shading control commands are generated, and intermittent shading rhythms are formed based on the short-term shading control commands. Based on the intermittent shading rhythm linked to the optical sensor sampling rhythm and the ambient brightness threshold, the system performs early backsampling, staggered sampling, and segmented light source shutdown operations according to the intermittent shading rhythm to perform online correction of blood flow trend data.

2. The data processing method for health monitoring of elderly people's physical exercise based on smart Internet of Things according to claim 1, characterized in that, The steps for generating time-series features for training scenarios are as follows: During the health monitoring data collection phase, optical sensing units and acceleration sensing units distributed at key movement sites of the elderly were set up to acquire light intensity data, surface reflectivity data, and acceleration data, respectively. A time index table is generated based on the collected light intensity data, surface reflectivity data and acceleration data, and the time correspondence of the three types of data is established with the sampling time of optical data as the time reference. Time alignment processing is performed on light intensity data, surface reflectivity data, and acceleration data based on the time index table, so that each time node contains light intensity change characteristics, reflectivity change characteristics, and acceleration change characteristics. By integrating time-aligned light intensity data, surface reflectivity data, and acceleration data, time-series features of the exercise scene are generated, reflecting the temporal correlation between motion state and light changes.

3. The data processing method for health monitoring of elderly people's physical exercise based on smart Internet of Things according to claim 2, characterized in that, The time alignment process uses the sampling time of optical data as a unified time reference. It generates a set of matching time points by comparing the timestamps of surface reflectivity data and acceleration data, and adjusts the time distribution of acceleration data according to the sampling rhythm of optical signals, so that each light intensity change point corresponds to a reflectivity change point and an acceleration change point, thereby forming a continuous training scene time series feature on a unified time axis.

4. The data processing method for health monitoring of elderly people's physical exercise based on smart Internet of Things according to claim 2, characterized in that, The steps for generating the feature set of interference pseudo-peaks are as follows: The light intensity variation features in the time series features of the exercise scene are continuously scanned. The initial range of the light intensity variation segment is determined based on the amplitude trend of the light intensity variation. The reflection response accompanied by the light intensity variation is confirmed by combining the surface reflectivity variation features. After the light intensity variation range is determined, the acceleration variation characteristics within the same time period are extracted synchronously to form time-corresponding data of attitude parameters and step frequency parameters. Based on the characteristics of light intensity change, reflectivity change, and the time correlation between posture parameters and step frequency parameters, the factors in the light intensity change range are identified to distinguish between motion factors and light factors. The feature data of the illumination factor segment and the motion factor segment are structured and integrated to generate a feature set of interference false peaks.

5. The data processing method for health monitoring of elderly people's physical exercise based on smart Internet of Things according to claim 4, characterized in that, When generating the interference false peak feature set, the illumination interference features are identified according to the amplitude of light intensity change, the degree of reflectivity synchronization response and the duration of interference. The motion interference features are classified according to the peak value of acceleration change, the range of attitude angle change and the step frequency synchronization deviation. The correspondence between illumination interference features and motion interference features is established with time index as the core.

6. The data processing method for health monitoring of elderly people's physical exercise based on smart Internet of Things according to claim 1, characterized in that, When the time range is determined, the inverse sampling window is based on the start and end times of the light interference characteristics. Buffers are set at both ends of the interference duration to cover the rising and attenuation phases of the interference peak change. Position markers are generated in the middle of the light interference segment to mark the equilibrium time of the optical monitoring signal after inverse sampling and serve as the reference node for subsequent time resampling.

7. The data processing method for health monitoring of elderly people's physical exercise based on smart Internet of Things according to claim 1, characterized in that, The steps involved in the formation of intermittent shading rhythms are as follows: Using location markers as time anchors, the heart rate curve data and cadence curve data are matched in terms of time range and node correspondence, so that the two data sequences have a unified time reference at each marker position. Starting from the location marker, time resampling is performed on the heart rate curve data and cadence curve data to keep the time distribution of the two types of signals consistent and obtain traceable time node information. During the time resampling process, a short-term shading control command is generated based on the synchronization status of the heart rate curve and the cadence curve, which is used to control the illumination period and exposure rhythm during the optical sensing acquisition process. By sequencing and integrating short-term shading control commands into a time-based rhythm, an intermittent shading rhythm with periodicity is formed, achieving dynamic coordination of the optical acquisition process.

8. The method for processing health monitoring data of elderly people's physical exercise based on smart Internet of Things according to claim 7, characterized in that, The steps for performing sampling and light source control operations based on the intermittent shading rhythm, the linkage between the optical sensor sampling rhythm and the ambient brightness threshold, are as follows: A time-linked relationship is established based on the intermittent shading rhythm and the optical sensor sampling rhythm, and an ambient brightness threshold is introduced to determine the illumination status, so that the optical signal sampling and illumination control are synchronously correlated in time. Perform early backsampling operation according to the intermittent shading rhythm to ensure that the sampling window completes signal acquisition before the sudden change in light intensity, so as to maintain sampling stability; After backing up the sampling in advance, a staggered sampling operation is performed. By adjusting the sampling trigger interval, the optical sampling point is time-shifted to the ambient light change cycle in order to suppress the effect of co-frequency flicker. The light source is turned off in segments according to the intermittent shading rhythm, and the blood flow trend data is corrected online using time anchors obtained by early backsampling and staggered sampling to maintain time continuity and data stability.

9. A data processing system for monitoring the health of elderly people's physical exercise based on the Internet of Things, used to implement the data processing method for monitoring the health of elderly people's physical exercise based on the Internet of Things as described in any one of claims 1-8, characterized in that, It includes a multi-source data synchronous acquisition module, an interference feature identification module, an anti-phase sampling suppression module, a time resampling and shading control module, and a rhythm linkage correction module: The multi-source data synchronous acquisition module collects light intensity data, surface reflectivity data, and acceleration data in the health monitoring data processing stage. It performs time alignment processing on the collected light intensity data, surface reflectivity data, and acceleration data to generate time series features of the exercise scene. The interference feature identification module extracts light intensity change segments based on the time series features of the exercise scene, and combines posture parameters and gait frequency parameters to distinguish motion factors and light factors in the light intensity change segments, thereby obtaining a set of interference false peak features. The inverted sampling suppression module sets an inverted sampling window based on the interference false peak feature set, performs inverted sampling operation on the optical monitoring signal within the inverted sampling window to suppress co-frequency flicker interference, and generates position marker points; The time resampling and shading control module performs time resampling and synchronization alignment on heart rate curve data and cadence curve data based on location markers. During the time resampling process, it generates short-term shading control commands and forms an intermittent shading rhythm based on the short-term shading control commands. The rhythm linkage correction module, based on the intermittent shading rhythm linkage optical sensor sampling rhythm and ambient brightness threshold, performs early back-up sampling, staggered sampling and segmented light source shutdown operations according to the intermittent shading rhythm to correct blood flow trend data online.

Images