Intelligent remote monitoring system for the elderly and intelligent state monitoring method

Abstract

This invention relates to the field of remote monitoring technology, and in particular provides a smart elderly care remote monitoring system and intelligent status monitoring method. The system includes a data sequence acquisition subsystem for generating heart rate, blood pressure, and biochemical data sequences; these three data sequences are timestamped within the device to form a synchronized raw physiological and biochemical data stream; a weighted fusion subsystem generates a multidimensional health vector containing cardiovascular load index, metabolic activity, and autonomic nervous system regulation capacity through a weighted fusion algorithm; and an early warning level output subsystem allows edge computing nodes to compare the real-time generated multidimensional health vector with a personalized health baseline for the elderly pre-stored in a local database, calculating the offset between the vector and the baseline; and automatically classifying three early warning levels based on the offset magnitude. This invention strengthens the privacy protection of the elderly's personal health data from a technical architecture perspective.

Description

Technical Field

[0001] This invention relates to the field of remote monitoring technology, and in particular to a smart elderly care remote monitoring system and a method for intelligent status monitoring. Background Technology

[0002] As the global population continues to age, the demand for efficient and intelligent elderly care monitoring solutions is becoming increasingly urgent. Traditional elderly care monitoring models mainly rely on manual care or simple alarm devices, which suffer from problems such as delayed response, limited monitoring dimensions, and inability to achieve personalized health management. Elderly people aging at home often face challenges such as delayed detection of sudden health events, difficulty in continuously monitoring chronic diseases, and a lack of effective intervention for psychological anxiety.

[0003] In recent years, the development of the Internet of Things (IoT), wearable devices, and edge computing technologies has provided new possibilities for remote monitoring. Existing technologies attempt to collect health data by integrating multiple sensors and analyze and issue alerts through cloud platforms. However, these solutions still have significant limitations in terms of data fusion depth, real-time analysis capabilities, privacy protection, and proactive intervention. Specifically, most systems only achieve simple data aggregation and threshold alarms, lacking the ability to deeply fuse and intelligently analyze multi-source heterogeneous data. They cannot build comprehensive health status assessment models from multiple dimensions such as physiological and biochemical aspects, and it is even more difficult to achieve dynamic risk warnings and graded responses based on individual baselines.

[0004] Therefore, there is an urgent need for an integrated system that can simultaneously collect multimodal physiological and biochemical data, perform real-time intelligent fusion analysis at the near end, and automatically execute graded early warning and environmental intervention based on the analysis results, so as to improve the accuracy, real-time nature and initiative of elderly care monitoring. Summary of the Invention

[0005] To achieve the above objectives, the present invention adopts the following technical solution: In one aspect, the present invention provides a smart elderly care remote monitoring system, comprising: The data sequence acquisition subsystem is used by the optical sensor integrated in the wearable device to continuously acquire the pulse wave signal of the elderly wrist and generate a heart rate data sequence through the change of photoplethysmography; the pressure sensor in the wearable device simultaneously measures the radial artery blood pressure waveform and generates a blood pressure data sequence; the micro-fluid channel at the bottom of the wearable device collects sweat from the skin surface and detects the concentration of sodium ions, chloride ions and lactic acid in the sweat through the electrode array to generate a biochemical data sequence; the three data sequences are timestamped and aligned in the device to form a synchronized physiological and biochemical raw data stream. The weighted fusion subsystem is used by edge computing nodes to receive raw data streams, perform waveform decomposition on heart rate data sequences, extract the main wave height and dicrotic wave depth, and calculate time-domain indices of heart rate variability; analyze blood pressure data sequences to extract the values ​​and fluctuation amplitudes of systolic, diastolic, and mean blood pressure; perform sliding window analysis on sweat ion concentration sequences to obtain the rate of change of sodium ion, chloride ion, and lactic acid concentrations; and generate a multidimensional health vector containing cardiovascular load index, metabolic activity, and autonomic nervous system regulation capacity through a weighted fusion algorithm. The early warning level output subsystem is used by edge computing nodes to compare the real-time generated multidimensional health vector with the personalized health baseline of the elderly pre-stored in the local database, and calculate the offset between the vector and the baseline. Based on the magnitude of the offset, three early warning levels are automatically divided: when the offset exceeds the first threshold, a reminder message containing the health vector deviation is sent to the family's mobile terminal; when the offset exceeds the second threshold, an early warning is sent to both the family's terminal and the community medical workstation, along with detailed deviation data for each component; when the offset exceeds the third threshold, in addition to sending an early warning, a command is also sent to the home environment control unit via wireless network to automatically adjust the indoor lighting color temperature, temperature and humidity, and activate the aroma diffusion device to help the elderly alleviate discomfort.

[0006] Another aspect of the present invention provides a method for intelligent status monitoring based on a smart elderly care remote monitoring system, comprising the following steps: The optical sensor in the wearable device continuously collects the pulse wave signal from the elderly person's wrist and generates a heart rate data sequence through changes in photoplethysmography; the pressure sensor in the wearable device simultaneously measures the radial artery blood pressure waveform and generates a blood pressure data sequence; the micro-fluid channel at the bottom of the wearable device collects sweat from the skin surface and detects the concentrations of sodium ions, chloride ions, and lactic acid in the sweat through an electrode array, generating a biochemical data sequence; the three data sequences are timestamped and aligned within the device to form a synchronized raw physiological and biochemical data stream. After receiving the raw data stream, the edge computing node performs waveform decomposition on the heart rate data sequence, extracts the main wave height and dicrotic wave depth, and calculates the time-domain index of heart rate variability; it analyzes the blood pressure data sequence, extracting the values ​​and fluctuation amplitudes of systolic blood pressure, diastolic blood pressure, and mean blood pressure; it performs sliding window analysis on the sweat ion concentration sequence to obtain the rate of change of sodium ion, chloride ion, and lactic acid concentrations; and it integrates the indicators through a weighted fusion algorithm to generate a multidimensional health vector that includes cardiovascular load index, metabolic activity, and autonomic nervous system regulation ability. Edge computing nodes compare the real-time generated multidimensional health vector with the personalized health baseline of the elderly pre-stored in the local database, and calculate the offset between the vector and the baseline. Based on the magnitude of the offset, three warning levels are automatically defined: when the offset exceeds the first threshold, a reminder message containing the health vector deviation is sent to the family member's mobile device; when the offset exceeds the second threshold, a warning is sent to both the family member's device and the community medical workstation, along with detailed deviation data for each component; when the offset exceeds the third threshold, in addition to sending a warning, a command is also sent to the home environment control unit via the wireless network to automatically adjust the indoor lighting color temperature, temperature and humidity, and activate the aroma diffusion device.

[0007] This invention optimizes system efficiency and privacy protection by employing an edge computing model. High-computation feature extraction, fusion analysis, and baseline comparison are completed locally on local nodes, uploading only critical early warning information and condensed health vectors, rather than continuously uploading raw data streams. This significantly reduces the amount of data transmitted wirelessly and the burden on cloud storage, minimizing network dependence and latency. Simultaneously, a large amount of sensitive raw physiological and biochemical data is processed locally without uploading to the cloud, thus strengthening the privacy protection of elderly individuals' personal health data from a technical architecture perspective. Attached Figure Description

[0008] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings: Figure 1 This is a block diagram of the smart elderly care remote monitoring system provided in Embodiment 1 of the present invention; Figure 2 This is a schematic diagram of the smart elderly care remote monitoring system provided in Embodiment 1 of the present invention; Figure 3 This is a block diagram of the data sequence acquisition subsystem provided in Embodiment 2 of the present invention; Figure 4 This is a block diagram of the weighted fusion subsystem provided in Embodiment 4 of the present invention; Figure 5 This is a block diagram of the early warning level output subsystem provided in Embodiment 10 of the present invention; Figure 6 This refers to the intelligent status monitoring method based on the smart elderly care remote monitoring system provided in Embodiment 11 of the present invention. Figure 7 A block diagram of the electronic device provided by the present invention; Figure 8 A block diagram of a computer-readable storage medium provided for this invention. Detailed Implementation

[0009] The technical solutions of the present invention will now be described with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0010] Hereinafter, the terms "first," "second," etc., are used for descriptive convenience only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined with "first," "second," etc., may explicitly or implicitly include one or more of that feature. In the description of this invention, unless otherwise stated, "a plurality of" means two or more.

[0011] In this invention, unless otherwise explicitly specified and limited, the term "connection" should be interpreted broadly. For example, "connection" can be a fixed mechanical connection, a detachable mechanical connection, or an integral part; or, "connection" can be a direct connection or an indirect connection through an intermediate medium. Furthermore, unless otherwise explicitly specified and limited, the term "coupling" should be interpreted broadly. For example, "coupling" can be a direct electrical connection, such as physical contact and electrical conduction between two components; it can also be understood as an electrical connection between different components in a circuit structure through physical lines capable of transmitting electrical signals, such as copper foil or wires on a printed circuit board (PCB), to transmit electrical signals; or, "coupling" can be an indirect electrical connection between two components through an intermediate medium; or, "coupling" can be an electrical connection between two components in a non-contact manner, such as an electrical connection between two components using capacitive coupling to transmit electrical signals.

[0012] In this embodiment of the invention, directional terms such as "up," "down," "left," and "right" may be defined relative to the orientation of the components shown in the accompanying drawings. It should be understood that these directional terms can be relative concepts, used for relative description and clarification, and can change accordingly depending on the orientation of the components in the accompanying drawings.

[0013] Example 1: As Figure 1 As shown, this embodiment of the invention provides a smart elderly care remote monitoring system, comprising: The data sequence acquisition subsystem is used by the optical sensor integrated in the wearable device to continuously acquire the pulse wave signal of the elderly wrist and generate a heart rate data sequence through the change of photoplethysmography; the pressure sensor in the wearable device simultaneously measures the radial artery blood pressure waveform and generates a blood pressure data sequence; the micro-fluid channel at the bottom of the wearable device collects sweat from the skin surface and detects the concentration of sodium ions, chloride ions and lactic acid in the sweat through the electrode array to generate a biochemical data sequence; the three data sequences are timestamped and aligned in the device to form a synchronized physiological and biochemical raw data stream. The weighted fusion subsystem is used by edge computing nodes to receive raw data streams, perform waveform decomposition on heart rate data sequences, extract the main wave height and dicrotic wave depth, and calculate time-domain indices of heart rate variability; analyze blood pressure data sequences to extract the values ​​and fluctuation amplitudes of systolic, diastolic, and mean blood pressure; perform sliding window analysis on sweat ion concentration sequences to obtain the rate of change of sodium ion, chloride ion, and lactic acid concentrations; and generate a multidimensional health vector containing cardiovascular load index, metabolic activity, and autonomic nervous system regulation capacity through a weighted fusion algorithm. The early warning level output subsystem is used by edge computing nodes to compare the real-time generated multidimensional health vector with the personalized health baseline of the elderly pre-stored in the local database, and calculate the offset between the vector and the baseline. Based on the magnitude of the offset, three early warning levels are automatically divided: when the offset exceeds the first threshold, a reminder message containing the health vector deviation is sent to the family's mobile terminal; when the offset exceeds the second threshold, an early warning is sent to both the family's terminal and the community medical workstation, along with detailed deviation data for each component; when the offset exceeds the third threshold, in addition to sending an early warning, a command is also sent to the home environment control unit via wireless network to automatically adjust the indoor lighting color temperature, temperature and humidity, and activate the aroma diffusion device to help the elderly alleviate discomfort.

[0014] The cardiovascular load index is calculated by fusing time-domain indicators of heart rate variability, such as the root mean square of the difference between adjacent RR intervals (in milliseconds), and blood pressure characteristic sequences, including systolic blood pressure, diastolic blood pressure, pulse pressure, and fluctuation amplitude (in mmHg). Its value is typically a dimensionless standardized score, such as 0-100 points, or a deviation value based on baseline normalization, used to quantify the current stress level on the cardiovascular system. Metabolic activity is mainly generated by combining the rate of change in sweat ion concentration, the linear fitting slope of sodium, chloride, and lactic acid concentrations (usually in mmol / (L·min)), and heart rate characteristics. Since the rate of change in sweat ion concentration reflects the metabolic activity level of sweat glands under sympathetic nerve control, the fusion result is usually a dimensionless relative activity value, representing the body's current metabolic intensity. Autonomic nervous system regulation is derived by fusing time-domain indicators of heart rate variability, reflecting sympathetic and parasympathetic nerve tension, with blood pressure fluctuation amplitude. Its value is also a dimensionless comprehensive score, reflecting the dynamic regulatory ability of the autonomic nervous system to maintain cardiovascular homeostasis. The higher the value, the better the regulatory ability or the closer it is to the personalized health baseline.

[0015] The pre-defined weighted coefficient matrix is ​​the core computational parameter for fusing multi-source physiological and biochemical data and generating multi-dimensional health vectors. It is a pre-calibrated linear transformation matrix used to map the multi-dimensional input feature vector, composed of heart rate feature sequences, blood pressure feature sequences, and sweat ion concentration change rate, onto three output dimensions: cardiovascular load index, metabolic activity, and autonomic nervous system regulation ability. Determination method: Based on physiological prior-guided machine learning, such as partial least squares regression, it is trained using a multi-sample dataset. Dimensionality: 3×N, where N is the total number of input features, typically 12~20, much larger than 3. Dynamics: The matrix parameters are fixed, serving as solidified fusion coefficients for edge computing nodes to ensure real-time computational efficiency and system stability. Universality guarantee: Through a triple mechanism of group prior training, personalized health baseline dynamic calibration, and multi-level thresholds based on individual statistical distribution, the fixed matrix effectively addresses individual differences and different physiological states, achieving accurate early warning. Each row of the matrix corresponds to a health indicator, and each column corresponds to an input feature. The elements in the matrix represent the contribution weight of the corresponding feature to that health indicator. The values ​​are not randomly set, but are obtained from sample data based on prior physiological knowledge or through machine learning methods such as multiple regression, principal component analysis, and partial least squares regression. The weights corresponding to each health indicator are usually normalized to eliminate the influence of dimensions. Once the matrix is ​​calibrated, it becomes a fixed fusion parameter of the system and is invoked when performing multiplication and summation operations in edge computing nodes. The first threshold is the lowest-level trigger threshold in the early warning level classification system, used to define the degree to which the Euclidean distance between the multidimensional health vector and the elderly person's personalized health baseline reaches a slight deviation. Based on the multidimensional health vector data collected by the elderly person during a stable health period, such as the initial seven to fourteen days of system deployment, the mean and standard deviation of the Euclidean distance between the vector and the baseline mean are calculated, and the first threshold is set as the mean plus 1.5 to 2 times the standard deviation. When the real-time Euclidean distance exceeds this threshold, the system determines that the elderly person's physiological state has experienced a slight abnormal fluctuation, triggering the lowest-level response—sending a reminder message only to the family's mobile device. The second threshold is the intermediate-level trigger threshold in the early warning level classification system, used to define the degree of moderate deviation in Euclidean distance. Its technical characteristics are similar to the first threshold, based on the same set of personalized health baseline statistical parameters, but the threshold is set as the mean plus 2.5 to 3 times the standard deviation, which is numerically higher than the first threshold. When the real-time Euclidean distance exceeds the second threshold, it indicates that the elderly person's physiological state has deviated beyond the daily fluctuation range and has potential clinical risks. Therefore, an intermediate-level response mechanism is triggered: an early warning is sent to both the family member and the community medical care workstation, along with detailed deviation data of each component in the multidimensional health vector. This level of intervention upgrades from a single information notification to a two-way early warning and provides more granular deviation information to assist medical staff in making judgments.The third threshold is the highest-level trigger threshold in the early warning level classification system. It is used to define the degree to which the Euclidean distance has reached a severe deviation. It is usually the mean plus four to five times the standard deviation, which means that the elderly person's physiological state has become significantly abnormal or has deviated acutely. The response mechanism triggered by this threshold has the characteristics of active intervention: in addition to sending the highest-level early warning to family members and community medical care workstations, it also sends control commands to the home environment control unit through the wireless network to automatically adjust the indoor lighting color temperature, temperature and humidity, and activate the aroma diffusion device. It goes beyond the scope of simple information early warning and realizes closed-loop environmental regulation based on physiological state monitoring. It aims to help the elderly relieve discomfort by creating a soothing environment.

[0016] The principles described in the above embodiments are referenced in the appendix. Figure 2This embodiment achieves precise, dynamic, and proactive monitoring of the health status of the elderly through the synchronous acquisition of multimodal physiological and biochemical data, real-time analysis and processing by edge computing nodes, and linkage with a hierarchical early warning and response mechanism. Its core effects are reflected in the following aspects: It enables in-depth, continuous, and integrated monitoring of health status. Traditional monitoring methods often focus on single or isolated vital signs. This embodiment, by synchronously acquiring pulse waves or cardiovascular function, radial artery blood pressure waveforms or hemodynamics, and sweat biochemical indicators or metabolic and electrolyte status, and completing time alignment, constructs a multi-dimensional data stream from macrophysiology to microbiochemistry, and from mechanical activity to body fluid composition. Integrated monitoring can capture complex physiological correlations that cannot be reflected by single indicators, such as the synchronous changes between cardiovascular load and metabolic product accumulation, providing a richer and more fundamental data foundation for assessing overall health status. To improve the early detection and accuracy of health risk identification, the system performs real-time processing of raw data at the edge, extracting derived features such as heart rate variability, blood pressure fluctuation, and ion concentration change rate. These derived features are more sensitive than the raw heart rate and blood pressure values ​​in reflecting early functional changes such as autonomic nervous system regulation, vascular elasticity, and metabolic stress. The weighted fusion generates a multidimensional health vector that comprehensively quantifies the synergistic state of the cardiovascular, metabolic, and nervous systems. By comparing with an individual's historical baseline, the system can identify subtle and gradual changes that deviate from the normal fluctuation range, thereby enabling early warning of potential health risks before the appearance of clinical symptoms or the occurrence of risk events. A tiered, precise, and proactive intervention response loop is constructed. The system dynamically activates a three-tiered response mechanism based on the degree or magnitude of health deviations. The first-tier alert aims to attract family attention and facilitate non-emergency communication. The second-tier warning links community healthcare, providing detailed data support for potential professional intervention. The third-tier response goes beyond traditional information notification, directly triggering pre-set home environment adjustments, such as soothing lighting and temperature / humidity settings, and relaxing aromatherapy, attempting to provide immediate, non-pharmacological physical relief for the elderly's discomfort before emergency medical assistance arrives. This forms a complete closed loop from risk perception to intelligent analysis, and then to tiered alerts and proactive intervention, improving the timeliness and effectiveness of emergency response measures. The system optimizes efficiency and privacy protection by adopting an edge computing model. High-computational-load feature extraction, fusion analysis, and baseline comparison are completed on local nodes, uploading only key warning information and condensed health vectors, rather than continuously uploading raw data streams. This significantly reduces the amount of data transmitted wirelessly and the burden on cloud storage, reducing network dependence and latency. Simultaneously, a large amount of sensitive raw physiological and biochemical data is processed locally without uploading to the cloud, strengthening the privacy protection of the elderly's personal health data from a technical architecture perspective.

[0017] In summary, this embodiment systematically improves the predictability, accuracy, and timeliness of remote elderly care monitoring through deep data fusion, real-time and accurate edge intelligent analysis, and proactive hierarchical response mechanisms, while also taking into account system operating efficiency and data security.

[0018] Example 2: As Figure 3 As shown, based on Embodiment 1, the early warning level output subsystem provided in this embodiment of the invention specifically includes: The first functional module is used to continuously collect pulse wave signals from the wrist of the elderly by the optical sensor in the wearable device, generate a heart rate data sequence through the change of photoplethysmography, and record the collection time point when each heart rate data point is generated, so as to obtain a heart rate data sequence with the collection time point mark. The second functional module is used to synchronously measure the radial artery blood pressure waveform by the pressure sensor in the wearable device, generate a blood pressure data sequence, and record the acquisition time point when each blood pressure data point is generated, so as to obtain a blood pressure data sequence with acquisition time point markers. The third functional module is used to collect sweat from the skin surface through the micro-fluid channels at the bottom of the wearable device, detect the concentrations of sodium ions, chloride ions and lactic acid in the sweat through the electrode array, generate a biochemical data sequence, and record the collection time point when each biochemical data point is generated, so as to obtain a biochemical data sequence with the collection time point mark. The data association module is used to associate and combine three data sequences marked with collection time points within the wearable device based on a unified clock reference, and to form a synchronized physiological and biochemical raw data stream by associating heart rate data, blood pressure data and biochemical data at the same collection time point.

[0019] In the above embodiments, the early warning level output subsystem of this embodiment synchronously collects pulse wave signals, radial artery blood pressure waveforms, and sodium, chloride, and lactic acid concentration data in sweat through optical sensors, pressure sensors, and microfluidic channels. It records the collection time point at each data point, forming time-stamped heart rate, blood pressure, and biochemical data sequences. The data association module associates and combines the three types of data at the same collection time point based on a unified clock reference, forming a time-synchronized physiological and biochemical raw data stream. The combination of the technical features of each module achieves high-precision time alignment of multimodal physiological and biochemical data, ensuring temporal consistency between different signals during analysis. This provides a reliable and synchronous multi-source data foundation for feature extraction, vector fusion, and health status assessment by edge computing nodes.

[0020] Example 3: Based on Example 2, the data association module provided in this embodiment of the invention specifically includes: The first sequence acquisition submodule is used to identify the pulse wave rising limb start point corresponding to each heartbeat cycle from the heart rate data sequence, take the acquisition time point of the pulse wave rising limb start point as the reference time point of the heartbeat cycle, arrange all reference time points in chronological order, and generate a time reference sequence composed of the start time of the heartbeat cycle. The second sequence acquisition submodule is used to receive the time reference sequence. For each reference time point, it locates the systolic blood pressure peak and diastolic blood pressure trough adjacent to the time before and after the time in the blood pressure data sequence, records the acquisition time point of the two waveforms of systolic blood pressure peak and diastolic blood pressure trough and the corresponding blood pressure value, and generates a systolic blood pressure peak value sequence and diastolic blood pressure trough value sequence that correspond one-to-one with the heartbeat cycle. The third sequence acquisition submodule is used to receive the time reference sequence. For each reference time point, it extracts the concentration data of sodium ions, chloride ions and lactic acid in sweat within a fixed time window before and after that time from the biochemical data sequence, calculates the average value of each ion concentration within the window, and generates the average sodium ion concentration sequence, average chloride ion concentration sequence and average lactic acid concentration sequence corresponding one-to-one with the heartbeat cycle. The sequence combination and merging submodule is used to combine the time base sequence, systolic blood pressure peak sequence, diastolic blood pressure trough sequence, average sodium ion concentration sequence, average chloride ion concentration sequence, and average lactate concentration sequence in the order of the base time point to form a raw physiological and biochemical data stream with the heartbeat cycle as the synchronization unit.

[0021] In the above embodiments, the data association module extracts the starting point of the pulse wave's upward limb from the heart rate data sequence as the reference time point of the heartbeat cycle, forming a time reference sequence. Using this sequence as a reference, it locates the systolic blood pressure peak and diastolic blood pressure trough corresponding to each reference time point in the blood pressure data sequence, generating a systolic blood pressure peak sequence and a diastolic blood pressure trough sequence aligned with the heartbeat cycle. Simultaneously, it extracts sweat ion concentration data within a fixed window before and after each reference time point from the biochemical data sequence, calculates the average values, and generates average concentration sequences of sodium ions, chloride ions, and lactic acid. The sequence combination and merging submodule combines all the above sequences in the order of the reference time points, forming a physiological and biochemical raw data stream synchronized with the heartbeat cycle. This embodiment achieves precise alignment and structured integration of multi-source heterogeneous physiological signals on the heartbeat cycle scale, providing a time-consistent and cycle-normalized data foundation for calculating composite indicators such as cardiovascular load index and metabolic activity, effectively supporting multi-dimensional collaborative analysis and early warning judgment of health status.

[0022] Example 4: Figure 4 As shown, based on Embodiment 1, the weighted fusion subsystem provided in this embodiment of the invention specifically includes: The first numerical processing module is used to identify the peak value of the pulse wave and the trough value of the diabetic wave in each cardiac cycle from the heart rate data sequence, measure the amplitude difference between the peak value and the trough value and the time interval between the peak value and the peak value of the adjacent cycle, arrange the amplitude difference and the time interval in the order of the cardiac cycle, and generate a heart rate feature sequence containing time-domain indicators of heart rate variability. The second numerical processing module is used to locate the systolic blood pressure peak and diastolic blood pressure trough in each cardiac cycle from the blood pressure data sequence, record the values ​​of systolic and diastolic blood pressure, and calculate the difference between systolic and diastolic blood pressure in each cycle, i.e., pulse pressure; at the same time, it calculates the systolic blood pressure fluctuation amplitude and diastolic blood pressure fluctuation amplitude of multiple consecutive cycles, arranges them in the order of cardiac cycle, and generates a blood pressure feature sequence containing systolic blood pressure, diastolic blood pressure, mean pressure and fluctuation amplitude. The third numerical processing module receives heart rate and blood pressure feature sequences, and extracts fixed-length time windows from the sweat ion concentration sequence. Within each window, it calculates the linear fitting slopes of sodium ion concentration, chloride ion concentration, and lactic acid concentration as rates of change. The three rates of change within the same window are combined with the corresponding time-time values ​​in the heart rate and blood pressure feature sequences. The combinations are multiplied and summed using a preset weighting coefficient matrix to generate a multidimensional health vector that includes cardiovascular load index, metabolic activity, and autonomic nervous system regulation ability.

[0023] In the above embodiments, the weighted fusion subsystem extracts the amplitude difference between the peak value of the pulse wave and the trough value of the dicrotic wave and the time interval between the peak values ​​from the heart rate data sequence to form a heart rate feature sequence reflecting heart rate variability; it extracts the systolic blood pressure, diastolic blood pressure, pulse pressure, and continuous cycle fluctuation amplitude from the blood pressure data sequence to generate a blood pressure feature sequence containing dynamic changes in blood pressure; simultaneously, it calculates the linear fitting slope of sodium ion, chloride ion, and lactic acid concentrations within a fixed window based on the sweat ion concentration sequence as the metabolic change rate. The third numerical processing module aligns the above heart rate features, blood pressure features, and metabolic change rate according to the time window, and performs multiplication and summation operations through a preset weighting coefficient matrix to generate a multidimensional health vector that integrates the cardiovascular load index, metabolic activity, and autonomic nervous system regulation ability. This realizes feature extraction, dynamic quantification, and cross-modal fusion of multi-source physiological and biochemical indicators in the time dimension, transforming discrete time-series signals into composite health vectors with clear physiological meaning, providing a calculable and comparable multidimensional health status representation for early warning level determination.

[0024] Example 5: Based on Example 4, the third numerical processing module provided in this embodiment of the invention specifically includes: The measurement point extraction submodule is used to extract continuous time segments from the sweat ion concentration sequence according to a preset fixed time length; within each time segment, sodium ion concentration measurement points, chloride ion concentration measurement points, and lactic acid concentration measurement points are extracted respectively, and all measurement points within the same time segment are arranged in chronological order of collection time to generate a set of concentration segments arranged in chronological order. The average concentration calculation submodule receives a set of concentration segments. For each concentration segment, it selects several measurement points at the beginning of the sequence according to a preset front ratio for sodium ion concentration measurement point sequences, chloride ion concentration measurement point sequences, and lactic acid concentration measurement point sequences, and calculates the average concentration of these measurement points as the front average concentration. It also selects several measurement points at the end of the sequence according to a preset back ratio, and calculates the average concentration of these measurement points as the back average concentration. This generates the front average concentration and back average concentration of the three ions in each concentration segment. The difference acquisition submodule is used to receive the first and last average concentrations of the three ions in each concentration segment. For each concentration segment, it calculates the difference between the last and first average concentrations of sodium ions, chloride ions, and lactic acid. Each difference is divided by the time length of the concentration segment to obtain the sodium ion concentration change rate, chloride ion concentration change rate, and lactic acid concentration change rate.

[0025] In the above embodiments, the third numerical processing module extracts continuous time segments from the sweat ion concentration sequence at fixed durations. Within each segment, it extracts measurement points for sodium ion, chloride ion, and lactic acid concentrations and arranges them in chronological order to form a concentration segment set. The average concentration calculation submodule calculates the average concentration at the beginning of several measurement points and the average concentration at the end of several measurement points in each segment for the three ion concentration sequences. The difference acquisition submodule calculates the difference between the average concentrations at the beginning and end of each segment and divides it by the segment duration to obtain the change rate of sodium ion, chloride ion, and lactic acid concentrations. This achieves the extraction of local dynamic features of sweat ion concentration changes over time. By segmented averaging and difference calculation, the continuous concentration sequence is transformed into a rate of change index reflecting short-term metabolic trends, providing standardized metabolic dynamic parameters for cross-modal weighted fusion with heart rate and blood pressure features.

[0026] Example 6: Based on Example 5, the measurement point extraction submodule provided in this embodiment of the invention specifically includes: The time determination component is used to determine the start and end times of each time segment from the sweat ion concentration sequence according to a preset fixed time length, and to arrange the continuous time intervals in order to generate a list of time segments composed of time intervals. The ion type classification component is used to receive a list of time segments. For each time segment, it filters out all measurement points whose collection time is between the start and end time of the time segment from the sweat ion concentration sequence. The measurement points are divided into three groups according to ion type: sodium ion measurement points, chloride ion measurement points and lactic acid measurement points. This generates a set of measurement points grouped by ion type within the time segment. The sequence generation component receives a set of measurement points grouped by ion type within each time segment. It extracts the concentration value and acquisition time of each measurement point from the sodium ion measurement point group, arranges them in chronological order of acquisition time, and generates a sodium ion concentration measurement point sequence. Similarly, it extracts the concentration value and acquisition time of each measurement point from the chloride ion measurement point group, arranges them in chronological order of acquisition time, and generates a chloride ion concentration measurement point sequence. Finally, it extracts the concentration value and acquisition time of each measurement point from the lactic acid measurement point group, arranges them in chronological order of acquisition time, and generates a lactic acid concentration measurement point sequence.

[0027] In the above embodiments, the measurement point extraction submodule divides the sweat ion concentration sequence into continuous time segments of fixed duration using a time determination component, generating a list of time segments defined by start and end times. The ion type segmentation component filters all measurement points falling within the time segment from the original sequence based on the start and end times of each segment, and groups them according to three ion types: sodium, chloride, and lactic acid, forming a set of measurement point groups distinguished by time segment and ion type. The sequence generation component extracts the concentration values ​​and collection times of the three ion measurement points from the measurement point group set of each time segment, sorts them according to the collection time, and generates a sequence of sodium, chloride, and lactic acid concentration measurement points within each time segment. This achieves structured segmentation and typological recombination of sweat multi-ion concentration data in the time dimension, transforming the continuous, mixed ion type original measurement sequence into a set of time series segmented by fixed duration and independently arranged by ion type, providing a standardized data organization form for calculating the rate of change of each ion concentration within a time segment.

[0028] Example 7: Based on Example 6, the ion type classification component provided in this embodiment of the invention specifically includes: The channel identifier reading sub-component is used to read the electrode array channel identifier corresponding to each measurement point from all measurement points whose acquisition time is between the start and end times of the time segment; based on the correspondence between the channel identifier and the ion type, it identifies the ion type to which each measurement point belongs and generates a set of measurement points with ion type labels attached to each measurement point; The ion tag screening sub-component is used to receive a set of measurement points with attached ion type tags, and filter out all measurement points with the ion type tag being sodium ion; at the same time, based on the electrode response intensity of each sodium ion measurement point, it filters out measurement points with response intensity higher than a preset skin contact threshold, and generates a group of valid sodium ion measurement points. The response intensity filtering sub-component receives a set of measurement points with attached ion type tags, filters out all measurement points with the ion type tag of chloride ion, and filters out measurement points with response intensity higher than a preset skin contact threshold based on the electrode response intensity of each chloride ion measurement point, generating an effective chloride ion measurement point group; it also filters out all measurement points with the ion type tag of lactic acid from the set of measurement points with attached ion type tags, and filters out measurement points with response intensity higher than a preset skin contact threshold based on the electrode response intensity of each lactic acid measurement point, generating an effective lactic acid measurement point group.

[0029] In the above embodiments, the ion type classification component reads the electrode array channel identifiers from all measurement points within a time segment through the channel representation reading sub-component. Based on the correspondence between the channel identifiers and ion types, an ion type tag is attached to each measurement point, forming a set of measurement points with ion type tags. The ion tag screening sub-component filters out measurement points tagged as sodium ions from this set, and further filters out effective sodium ion measurement point groups based on the condition that the electrode response intensity is higher than a preset skin contact threshold. The response intensity screening sub-component filters out chloride and lactic acid measurement points from the set of attached tags in the same way, and generates effective chloride ion measurement point groups and effective lactic acid measurement point groups based on the electrode response intensity condition. This achieves accurate classification and quality screening of multi-ion mixed measurement points within a time segment. Ion types are identified through channel identifiers, and low-reliability measurement points caused by poor skin contact are excluded using the electrode response intensity threshold. This ensures that the measurement point data used to calculate the concentration change rate simultaneously meets the requirements of type accuracy and signal validity, improving the reliability and consistency of metabolic dynamic parameter extraction.

[0030] Example 8: Based on Example 7, the channel identifier reading sub-component provided in this embodiment of the invention specifically includes: The mapping relationship processing unit is used to extract the physical position coordinates of the electrode array channel in the array from the electrode array channel identifier corresponding to each measurement point, determine the ion detection functional area corresponding to the measurement point according to the mapping relationship between the physical position coordinates and the electrode functional area, and generate a preliminary classification set with functional area identifiers for each measurement point. The numerical comparison unit receives a preliminary classification set with functional area identifiers. For each measurement point, it reads the corresponding electrode response intensity value. It compares the response intensity value with a preset skin contact intensity threshold, filters out measurement points with response intensities higher than the threshold, and generates a set of valid contact measurement points. The identification conversion unit receives a set of valid contact measurement points. For each valid contact measurement point, it converts the functional area identification into an ion type identification of sodium ion, chloride ion, or lactic acid by referring to a preset functional area and ion type lookup table based on the functional area identification attached to it, thereby generating a set of measurement points with ion type labels attached to each measurement point.

[0031] In the above embodiments, the channel identifier reading sub-component extracts the physical position coordinates from the electrode array channel identifier corresponding to each measurement point through the mapping relationship processing unit. Based on the mapping relationship between the coordinates and the electrode functional areas, it determines the ion detection functional area to which the measurement point belongs, generating a preliminary classification set with functional area identifiers. The numerical comparison unit receives this set and compares the electrode response intensity of each measurement point with a preset skin contact intensity threshold, filtering out measurement points with response intensities higher than the threshold to form a set of valid contact measurement points. The identifier conversion unit, based on the functional area identifiers of each point in the set of valid contact measurement points and referring to a preset functional area and ion type lookup table, converts the functional area identifiers into corresponding sodium ion, chloride ion, or lactate ion type identifiers, ultimately generating a set of measurement points with ion type labels. This achieves a reliable conversion from the original electrode array channel identifiers to ion type labels, determines the functional area affiliation through physical position coordinate mapping, and filters low-quality contact data using response intensity thresholds, ensuring that the accuracy of ion type identification is based on valid electrode contact. This provides a structurally complete and signal-reliable data foundation for subsequent grouping by ion type.

[0032] Example 9: Based on Example 8, the numerical comparison unit provided in this embodiment of the invention specifically includes: The voltage value calculation subunit is used to extract the original voltage signal value output by the electrode array channel at the measurement time from the electrode array channel identifier corresponding to each measurement point; at the same time, it extracts the reference voltage signal value of the electrode channel at the moment before the measurement, subtracts the reference voltage signal value from the original voltage signal value, and generates the electrode differential voltage value corresponding to each measurement point. The amplitude adjustment subunit is used to receive the electrode differential voltage value corresponding to each measurement point, and at the same time read the skin temperature value at the time of measurement point acquisition from the temperature sensor built into the wearable device. The amplitude of the electrode differential voltage value is adjusted according to the preset temperature-response compensation coefficient to generate the temperature-compensated voltage value corresponding to each measurement point. The temperature compensation subunit is used to receive the temperature-compensated voltage value corresponding to each measurement point, determine the corresponding ion type according to the functional area identifier attached to the measurement point, and convert the temperature-compensated voltage value into the concentration value of the ion according to the preset ion type voltage-concentration conversion curve, which is used as the electrode response intensity value corresponding to the measurement point.

[0033] In the above embodiments, the numerical comparison unit of this embodiment extracts the original voltage signal value at the measurement time and the reference voltage signal value at the moment before measurement from the electrode array channel identifier corresponding to each measurement point through the voltage value calculation subunit, and calculates the difference between the two to generate an electrode differential voltage value. The amplitude adjustment subunit receives the differential voltage value, combines it with the skin temperature value at the measurement point acquisition time, and adjusts the amplitude of the differential voltage value according to the preset temperature-response compensation coefficient to generate a temperature-compensated voltage value. The temperature compensation subunit determines the ion type according to the functional area identifier attached to the measurement point, and converts the temperature-compensated voltage value into the concentration value of the corresponding ion with reference to the preset ion type voltage-concentration conversion curve, which serves as the electrode response intensity value corresponding to the measurement point. This realizes the conversion process from the original electrode voltage signal to the standardized ion concentration value, eliminates the influence of baseline drift through differential calculation, corrects the interference of ambient temperature on the electrode response through temperature compensation, and converts the voltage signal into a physiologically meaningful concentration value based on the ion-specific conversion curve, ultimately generating reliable electrode response intensity data for quality screening, ensuring the accuracy of effective contact judgment and the physical consistency of ion concentration calculation.

[0034] Example 10: As Figure 5 As shown, based on Embodiment 1, the early warning level output subsystem provided in this embodiment of the invention includes: The timeliness adjustment component is used to retrieve the personalized health baseline vector of the elderly from the local database, and at the same time extract the historical data collection time point when the personalized health baseline vector of the elderly was established; according to the time interval between the current system time and the historical data collection time point, the time decay coefficient is calculated according to the preset time decay program; the time decay coefficient is multiplied by the values ​​of the cardiovascular load index, metabolic activity and autonomic nervous regulation ability in the baseline vector to generate the timeliness adjusted baseline vector. The activity level confirmation component receives multidimensional health vectors generated in real time by edge computing nodes and reads the current body movement intensity value of the elderly from the accelerometer built into the wearable device. Based on the body movement intensity value and a preset activity status classification table, it determines the elderly’s current level of rest, light activity, or moderate activity. According to the activity status level, it extracts the dynamic range compensation coefficients of cardiovascular load index, metabolic activity, and autonomic nervous regulation ability from the preset compensation coefficient storage area, and multiplies each component of the real-time health vector with the corresponding compensation coefficient to generate a real-time vector after activity status compensation. The weighted square component receives the baseline vector adjusted for timeliness and the real-time vector compensated for activity status. It reads the pre-set medical importance weight coefficients for cardiovascular load index, metabolic activity, and autonomic nervous system regulation ability from the local database. It calculates the differences in cardiovascular load index, metabolic activity, and autonomic nervous system regulation ability between the real-time vector and the baseline vector, respectively. It squares each difference and multiplies it by the corresponding weight coefficient to obtain three weighted squared differences. The three weighted squared differences are summed and the square root is taken to generate the final weighted offset.

[0035] In the above embodiments, the early warning level output subsystem of this embodiment retrieves the personalized health baseline vector of the elderly and its historical data collection time point from the local database through the timeliness adjustment component. Based on the time interval between the current system time and the historical collection time point, it calculates the time decay coefficient according to the preset time decay program, and multiplies the coefficient by the cardiovascular load index, metabolic activity and autonomic nervous regulation ability components of the baseline vector to generate the timeliness-adjusted baseline vector. The activity level confirmation component receives the multidimensional health vector generated in real time by the edge computing node, and reads the current body movement intensity value of the elderly from the wearable device's accelerometer. Based on the body movement intensity and the preset activity state classification table, it determines that the elderly are currently in a resting state. The system identifies the activity level as either cessation, mild activity, or moderate activity. Based on this level, it extracts dynamic range compensation coefficients for cardiovascular load index, metabolic activity, and autonomic nervous system regulation from a pre-defined compensation coefficient storage area. Each component of the real-time health vector is multiplied by its corresponding compensation coefficient to generate an activity-compensated real-time vector. A weighted squaring component receives both a time-adjusted baseline vector and the activity-compensated real-time vector. It reads pre-defined medical importance weight coefficients for cardiovascular load index, metabolic activity, and autonomic nervous system regulation from a local database. The differences between each component of the real-time and baseline vectors are calculated, squared, and multiplied by the corresponding weight coefficient to obtain three weighted squared differences. These differences are summed and the square root is taken to generate the final weighted offset. This system enables a quantitative assessment of the difference between the elderly's real-time health status and their personalized baseline. Time-adjustment reflects the dynamic changes in baseline data, activity-compensated measures correct for the interference of body movement on physiological parameters, and medical importance weights highlight the impact of key health indicators. The final weighted offset serves as the basis for determining the warning level, improving the accuracy and clinical relevance of health status deviation calculations.

[0036] Example 11: As Figure 6 As shown, based on Examples 1-10, the intelligent status monitoring method based on a smart elderly care remote monitoring system provided in this embodiment of the invention includes the following steps: Step S100: The optical sensor in the wearable device continuously collects the pulse wave signal from the elderly person's wrist and generates a heart rate data sequence through changes in photoplethysmography; the pressure sensor in the wearable device simultaneously measures the radial artery blood pressure waveform and generates a blood pressure data sequence; the micro-fluid channel at the bottom of the wearable device collects sweat from the skin surface and detects the concentrations of sodium ions, chloride ions, and lactic acid in the sweat through an electrode array, generating a biochemical data sequence; the three data sequences are time-stamped and aligned within the device to form a synchronized physiological and biochemical raw data stream; Step S200: After receiving the raw data stream, the edge computing node performs waveform decomposition on the heart rate data sequence, extracts the main wave height and dicrotic wave depth, and calculates the time-domain index of heart rate variability; it analyzes the blood pressure data sequence, extracts the values ​​and fluctuation amplitudes of systolic blood pressure, diastolic blood pressure, and mean blood pressure; it performs sliding window analysis on the sweat ion concentration sequence to obtain the rate of change of sodium ion, chloride ion, and lactic acid concentrations; and it integrates the above indicators through a weighted fusion algorithm to generate a multidimensional health vector that includes cardiovascular load index, metabolic activity, and autonomic nervous system regulation ability. Step S300: The edge computing node compares the real-time generated multidimensional health vector with the personalized health baseline of the elderly pre-stored in the local database, and calculates the offset between the vector and the baseline. Based on the magnitude of the offset, three warning levels are automatically defined: when the offset exceeds the first threshold, a reminder message containing the health vector deviation is sent to the family member's mobile terminal; when the offset exceeds the second threshold, a warning is sent to both the family member's terminal and the community medical workstation, along with detailed deviation data for each component; when the offset exceeds the third threshold, in addition to sending a warning, a command is also sent to the home environment control unit via the wireless network to automatically adjust the indoor lighting color temperature, temperature and humidity, and activate the aroma diffusion device to help the elderly alleviate discomfort.

[0037] In the above embodiments, this embodiment achieves simultaneous monitoring of the elderly's physiological and biochemical states, fusing multi-source data such as heart rate, blood pressure, and sweat ions into a multi-dimensional health vector, thereby comprehensively assessing the elderly's cardiovascular load, metabolic activity, and autonomic nervous system regulation capabilities. The system compares the real-time generated multi-dimensional health vector with a personalized health baseline, calculates the offset to quantitatively identify abnormal states, and automatically classifies warning levels based on the degree of deviation, sending differentiated information to family members and medical staff. At the highest warning level, the system connects to the home environment control unit via wireless network, automatically adjusting indoor lighting color temperature, humidity, and activating an aromatherapy diffuser to help alleviate the elderly's discomfort from an environmental perspective. This forms a complete closed loop from data acquisition, feature fusion, state recognition to environmental intervention, achieving accurate perception and timely response to the elderly's health status.

[0038] Figure 7 A block diagram of an exemplary electronic device suitable for implementing embodiments of the present invention is shown.

[0039] Electronic devices may include a central processing unit / microprocessor / main control chip; and a storage medium coupled to the central processing unit / microprocessor / main control chip, wherein computer-executable instructions are stored for performing the steps of various methods of embodiments of the present invention when executed by a processor.

[0040] The central processing unit / microprocessor / main control chip may include, but is not limited to, one or more processors or microprocessors.

[0041] Storage media may include, but are not limited to, random access memory (RAM), read-only memory (ROM), flash memory, EPROM memory, EEPROM memory, registers, and computer storage media (such as hard disks, floppy disks, solid-state drives, removable disks, CD-ROMs, DVD-ROMs, Blu-ray discs, etc.).

[0042] In addition, the electronic device may include (but is not limited to) a data bus, an input / output bus / external bus / device bus, a display, and input / output devices (e.g., keyboard, mouse, speaker, etc.).

[0043] The central processing unit / microprocessor / main control chip can communicate with external devices via wired or wireless networks (not shown) through input / output buses / external buses / device buses.

[0044] The storage medium may also store at least one computer-executable instruction for performing the steps of various functions and / or methods in the embodiments described herein when the central processing unit / microprocessor / main control chip is running.

[0045] In one embodiment, the at least one computer-executable instruction may also be compiled into or comprise a software product, wherein one or more computer-executable instructions are executed by a processor to perform the steps of the various functions and / or methods in the embodiments described herein.

[0046] Figure 8 A schematic diagram of a computer-readable storage medium according to an embodiment of the present invention is shown.

[0047] like Figure 8 As shown, instructions, such as computer-readable instructions, are stored on a non-transitory computer-readable storage medium. When the computer-readable instructions are executed by a processor, the various methods described above can be performed. The non-transitory computer-readable storage medium includes, but is not limited to, volatile memory and / or non-volatile memory. Volatile memory may include, for example, random access memory (RAM) and / or cache memory. Non-transitory non-volatile memory may include, for example, read-only memory (ROM), hard disk, flash memory, etc. For example, the non-transitory computer-readable storage medium can be connected to a computing device such as a computer, and then, when the computing device executes the computer-readable instructions stored on the non-transitory computer-readable storage medium, the various methods described above can be performed.

[0048] In the embodiments provided by this invention, it should be understood that the disclosed systems and methods can be implemented in other ways. For example, the system embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between systems or units may be electrical, mechanical, or other forms.

[0049] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0050] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0051] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for executing all or part of the steps of the methods of the various embodiments of this invention through a computer device (which may be a personal computer, server, or network device, etc.). The aforementioned storage medium includes: USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, optical disks, and other media capable of storing program code.

[0052] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A smart elderly care remote monitoring system, characterized in that, Include: The data sequence acquisition subsystem is used to acquire heart rate data sequences, blood pressure data sequences, and biochemical data sequences through wearable devices; the three data sequences are timestamped within the device to form a synchronized raw physiological and biochemical data stream; The weighted fusion subsystem is used by edge computing nodes to receive raw data streams, perform waveform decomposition on heart rate data sequences, extract the main wave height and dicrotic wave depth, calculate time-domain indicators of heart rate variability, extract the values ​​and fluctuation amplitudes of systolic blood pressure, diastolic blood pressure, and mean blood pressure, and obtain the rate of change of sodium ion, chloride ion, and lactic acid concentrations. The heart rate data sequence and blood pressure data sequence are processed to obtain the corresponding target feature sequence, which is combined with the rate of change of sodium ion, chloride ion, and lactic acid concentrations in sweat within the same time window. The combination is multiplied and accumulated through a preset weighting coefficient matrix to generate a multidimensional health vector containing cardiovascular load index, metabolic activity, and autonomic nervous system regulation ability. The early warning level output subsystem is used by edge computing nodes to compare the multi-dimensional health vector generated in real time with the personalized health baseline of the elderly that is pre-stored in the local database, and calculate the offset between the vector and the baseline. Based on the magnitude of the offset, three warning levels are automatically assigned.

2. The smart elderly care remote monitoring system as described in claim 1, characterized in that, The weighted fusion subsystem includes: The first numerical processing module is used to identify the peak value of the pulse wave and the trough value of the diabetic wave in each cardiac cycle from the heart rate data sequence, measure the amplitude difference between the peak value and the trough value and the time interval between the peak value and the peak value of the adjacent cycle, arrange the amplitude difference and the time interval in the order of the cardiac cycle, and generate a heart rate feature sequence containing time-domain indicators of heart rate variability. The second numerical processing module is used to locate the systolic blood pressure peak and diastolic blood pressure trough in each cardiac cycle from the blood pressure data sequence, record the values ​​of systolic and diastolic blood pressure, and calculate the difference between systolic and diastolic blood pressure in each cycle, i.e., pulse pressure; at the same time, it calculates the systolic blood pressure fluctuation amplitude and diastolic blood pressure fluctuation amplitude of multiple consecutive cycles, arranges them in the order of cardiac cycle, and generates a blood pressure feature sequence containing systolic blood pressure, diastolic blood pressure, mean pressure and fluctuation amplitude. The third numerical processing module receives heart rate and blood pressure feature sequences, and extracts fixed-length time windows from the sweat ion concentration sequence. Within each window, it calculates the linear fitting slopes of sodium ion concentration, chloride ion concentration, and lactic acid concentration as rates of change. The three rates of change within the same window are combined with the corresponding time-time values ​​in the heart rate and blood pressure feature sequences. The combinations are multiplied and summed using a preset weighting coefficient matrix to generate a multidimensional health vector that includes cardiovascular load index, metabolic activity, and autonomic nervous system regulation ability.

3. The smart elderly care remote monitoring system as described in claim 2, characterized in that, The third numerical processing module includes: The measurement point extraction submodule is used to extract continuous time segments from the sweat ion concentration sequence according to a preset fixed time length; within each time segment, sodium ion concentration measurement points, chloride ion concentration measurement points, and lactic acid concentration measurement points are extracted respectively, and all measurement points within the same time segment are arranged in chronological order of collection time to generate a set of concentration segments arranged in chronological order. The average concentration calculation submodule receives a set of concentration segments. For each concentration segment, it selects several measurement points at the beginning of the sequence according to a preset front ratio for sodium ion concentration measurement point sequences, chloride ion concentration measurement point sequences, and lactic acid concentration measurement point sequences, and calculates the average concentration of these measurement points as the front average concentration. It also selects several measurement points at the end of the sequence according to a preset back ratio, and calculates the average concentration of these measurement points as the back average concentration. This generates the front average concentration and back average concentration of the three ions in each concentration segment. The difference acquisition submodule is used to receive the first and last average concentrations of the three ions in each concentration segment, and calculates the slope of the concentration change over time within the concentration segment using a linear fitting method to obtain the change rate of sodium ion concentration, chloride ion concentration, and lactic acid concentration.

4. The smart elderly care remote monitoring system as described in claim 3, characterized in that, The measurement point extraction submodule includes: The time determination component is used to determine the start and end times of each time segment from the sweat ion concentration sequence according to a preset fixed time length, and to arrange the continuous time intervals in order to generate a list of time segments composed of time intervals. The ion type classification component is used to receive a list of time segments. For each time segment, it filters out all measurement points whose collection time is between the start and end time of the time segment from the sweat ion concentration sequence. The measurement points are divided into three groups according to ion type: sodium ion measurement points, chloride ion measurement points and lactic acid measurement points. This generates a set of measurement points grouped by ion type within the time segment. The sequence generation component receives a set of measurement points grouped by ion type within each time segment. It extracts the concentration value and acquisition time of each measurement point from the sodium ion measurement point group, arranges them in chronological order of acquisition time, and generates a sodium ion concentration measurement point sequence. Similarly, it extracts the concentration value and acquisition time of each measurement point from the chloride ion measurement point group, arranges them in chronological order of acquisition time, and generates a chloride ion concentration measurement point sequence. Finally, it extracts the concentration value and acquisition time of each measurement point from the lactic acid measurement point group, arranges them in chronological order of acquisition time, and generates a lactic acid concentration measurement point sequence.

5. The smart elderly care remote monitoring system as described in claim 4, characterized in that, Ion type classification component, including: The channel identifier reading sub-component is used to read the electrode array channel identifier corresponding to each measurement point from all measurement points whose acquisition time is between the start and end times of the time segment; based on the correspondence between the channel identifier and the ion type, it identifies the ion type to which each measurement point belongs and generates a set of measurement points with ion type labels attached to each measurement point; The ion tag screening sub-component is used to receive a set of measurement points with attached ion type tags, and filter out all measurement points with the ion type tag being sodium ion; at the same time, based on the electrode response intensity of each sodium ion measurement point, it filters out measurement points with response intensity higher than a preset skin contact threshold, and generates a group of valid sodium ion measurement points. The response intensity filtering sub-component receives a set of measurement points with attached ion type tags, filters out all measurement points with the ion type tag of chloride ion, and filters out measurement points with response intensity higher than a preset skin contact threshold based on the electrode response intensity of each chloride ion measurement point, generating an effective chloride ion measurement point group; it also filters out all measurement points with the ion type tag of lactic acid from the set of measurement points with attached ion type tags, and filters out measurement points with response intensity higher than a preset skin contact threshold based on the electrode response intensity of each lactic acid measurement point, generating an effective lactic acid measurement point group.

6. The smart elderly care remote monitoring system as described in claim 5, characterized in that, The channel identifier reading sub-component includes: The mapping relationship processing unit is used to extract the physical position coordinates of the electrode array channel in the array from the electrode array channel identifier corresponding to each measurement point, determine the ion detection functional area corresponding to the measurement point according to the mapping relationship between the physical position coordinates and the electrode functional area, and generate a preliminary classification set with functional area identifiers for each measurement point. The numerical comparison unit receives a preliminary classification set with functional area identifiers. For each measurement point, it reads the corresponding electrode response intensity value. It compares the response intensity value with a preset skin contact intensity threshold, filters out measurement points with response intensities higher than the threshold, and generates a set of valid contact measurement points. The identification conversion unit receives a set of valid contact measurement points. For each valid contact measurement point, it converts the functional area identification into an ion type identification of sodium ion, chloride ion, or lactic acid by referring to a preset functional area and ion type lookup table based on the functional area identification attached to it, thereby generating a set of measurement points with ion type labels attached to each measurement point.

7. The smart elderly care remote monitoring system as described in claim 6, characterized in that, Numerical comparison unit, including: The voltage value calculation subunit is used to extract the original voltage signal value output by the electrode array channel at the measurement time from the electrode array channel identifier corresponding to each measurement point; at the same time, it extracts the reference voltage signal value of the electrode channel at the moment before the measurement, subtracts the reference voltage signal value from the original voltage signal value, and generates the electrode differential voltage value corresponding to each measurement point. The amplitude adjustment subunit is used to receive the electrode differential voltage value corresponding to each measurement point, and at the same time read the skin temperature value at the time of measurement point acquisition from the temperature sensor built into the wearable device. The amplitude of the electrode differential voltage value is adjusted according to the preset temperature-response compensation coefficient to generate the temperature-compensated voltage value corresponding to each measurement point. The temperature compensation subunit is used to receive the temperature-compensated voltage value corresponding to each measurement point, determine the corresponding ion type according to the functional area identifier attached to the measurement point, and convert the temperature-compensated voltage value into the concentration value of the ion according to the preset ion type voltage-concentration conversion curve, which is used as the electrode response intensity value corresponding to the measurement point.

8. The smart elderly care remote monitoring system as described in claim 1, characterized in that, In the data sequence acquisition subsystem, the optical sensor integrated in the wearable device continuously collects the pulse wave signal from the elderly person's wrist and generates a heart rate data sequence through changes in photoplethysmography; the pressure sensor in the wearable device simultaneously measures the radial artery blood pressure waveform and generates a blood pressure data sequence; the micro-fluid channel at the bottom of the wearable device collects sweat from the skin surface and detects the concentrations of sodium ions, chloride ions, and lactic acid in the sweat through an electrode array, generating a biochemical data sequence; the three data sequences are timestamped and aligned within the device to form a synchronized raw physiological and biochemical data stream.

9. The smart elderly care remote monitoring system as described in claim 1, characterized in that, After receiving the raw data stream, the edge computing nodes in the weighted fusion subsystem perform waveform decomposition on the heart rate data sequence, extracting the main wave height and dicrotic wave depth, and calculating the time-domain index of heart rate variability; analyze the blood pressure data sequence, extracting the values ​​and fluctuation amplitudes of systolic blood pressure, diastolic blood pressure, and mean blood pressure; perform sliding window analysis on the sweat ion concentration sequence to obtain the rate of change of sodium ion, chloride ion, and lactic acid concentrations; and generate a multidimensional health vector containing cardiovascular load index, metabolic activity, and autonomic nervous system regulation capacity through a weighted fusion algorithm. Three warning levels: When the deviation exceeds the first threshold, a reminder message containing the health vector deviation is sent to the family's mobile device; When the offset exceeds the second threshold, an alert is sent to both the family member's terminal and the community medical care workstation, along with detailed offset data for each component. When the offset exceeds the third threshold, in addition to sending an alert, a command is sent to the home environment control unit via wireless network to automatically adjust the indoor lighting color temperature, temperature and humidity, and activate the aroma diffusion device to help the elderly alleviate discomfort.

10. A method for intelligent status monitoring based on a smart elderly care remote monitoring system, used to implement the smart elderly care remote monitoring system as described in any one of claims 1 to 9, characterized in that, Includes the following steps: The optical sensor in the wearable device continuously collects the pulse wave signal from the elderly person's wrist and generates a heart rate data sequence through changes in photoplethysmography; the pressure sensor in the wearable device simultaneously measures the radial artery blood pressure waveform and generates a blood pressure data sequence; the micro-fluid channel at the bottom of the wearable device collects sweat from the skin surface and detects the concentrations of sodium ions, chloride ions, and lactic acid in the sweat through an electrode array, generating a biochemical data sequence; the three data sequences are timestamped and aligned within the device to form a synchronized raw physiological and biochemical data stream. After receiving the raw data stream, the edge computing node performs waveform decomposition on the heart rate data sequence, extracts the main wave height and dicrotic wave depth, and calculates the time-domain index of heart rate variability; it analyzes the blood pressure data sequence, extracting the values ​​and fluctuation amplitudes of systolic blood pressure, diastolic blood pressure, and mean blood pressure; it performs sliding window analysis on the sweat ion concentration sequence to obtain the rate of change of sodium ion, chloride ion, and lactic acid concentrations; and it integrates the indicators through a weighted fusion algorithm to generate a multidimensional health vector that includes cardiovascular load index, metabolic activity, and autonomic nervous system regulation ability. Edge computing nodes compare the real-time generated multidimensional health vector with the personalized health baseline of the elderly pre-stored in the local database, and calculate the offset between the vector and the baseline. Based on the magnitude of the offset, three warning levels are automatically divided: when the offset exceeds the first threshold, a reminder message containing the health vector deviation is sent to the family's mobile device. When the offset exceeds the second threshold, an alert is sent to both the family member's terminal and the community medical care workstation, along with detailed offset data for each component. When the offset exceeds the third threshold, in addition to sending an alert, a command is sent to the home environment control unit via wireless network to automatically adjust the indoor lighting color temperature, temperature and humidity, and activate the aroma diffusion device.

Images