A multi-device linkage method, system and device for monitoring and early warning of the elderly data

Abstract

The application provides a multi-device linkage pension data monitoring and early warning method, system and device, relates to the technical field of data processing, and the method comprises the following steps: linkage monitoring is carried out on home-based pension users, a pension monitoring data set is obtained to identify activity scene features, multi-dimensional abnormal analysis attention optimization is carried out according to user activity scene labels, risk accident type fitting is carried out according to the user activity scene labels, a multi-dimensional risk accident deduction is carried out on a monitoring abnormal attention coupling matrix according to a dynamic disturbance distillation mechanism, a risk accident deduction graph is obtained, the risk accident deduction graph is optimized based on accompanying user monitoring data, a risk deduction optimization graph is obtained, and a pension monitoring and early warning signal is synchronously generated. The technical problem that the existing technology has one-sidedness in pension monitoring, resulting in insufficient accuracy and reliability of pension accident monitoring and prediction is solved. The technical effect of improving the accuracy and reliability of home-based pension user accident monitoring and early warning is achieved.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, specifically to a method, system, and device for monitoring and early warning of elderly care data through multi-device linkage. Background Technology

[0002] Traditional home-based elderly care monitoring typically relies on a single data source, such as using wearable devices to monitor the elderly's physiological data for risk identification, or simply using indoor cameras to collect user behavior information. While a single data source can provide basic monitoring functions, it cannot comprehensively assess the overall health status of the elderly. Furthermore, traditional monitoring methods often ignore the influence and intervention of accompanying users, resulting in a lack of comprehensiveness and accuracy in risk identification and early warning, thus affecting the accuracy and reliability of elderly care accident monitoring and early warning.

[0003] Existing technologies for elderly care monitoring suffer from limitations, leading to insufficient accuracy and reliability in the monitoring and prediction of elderly care incidents. Summary of the Invention

[0004] The purpose of this application is to provide a method, system, and device for monitoring and early warning of elderly care data through multi-device linkage, in order to solve the technical problem that the existing elderly care monitoring is one-sided, resulting in insufficient accuracy and reliability of elderly care accident monitoring and prediction.

[0005] In view of the above problems, this application provides a method, system and equipment for monitoring and early warning of elderly care data through multi-device linkage.

[0006] The first aspect of this application provides a multi-device linkage method for monitoring and early warning of elderly care data. The method includes: conducting linked monitoring of home-based elderly care users through a multi-device monitoring array to obtain an elderly care monitoring dataset; identifying activity scene features based on the elderly care monitoring dataset to generate user activity scene tags; performing multi-dimensional anomaly analysis and attention optimization on the elderly care monitoring dataset based on the user activity scene tags to establish a monitoring anomaly attention coupling matrix; fitting risk accident types based on the user activity scene tags to obtain N fitted risk accident types, where N is a positive integer greater than 1; performing multi-dimensional risk accident deduction on the monitoring anomaly attention coupling matrix based on the N fitted risk accident types using a dynamic perturbation distillation mechanism to obtain a risk accident deduction map; reading real-time accompanying user monitoring data corresponding to the accompanying user; and performing accompanying accident interference analysis optimization on the risk accident deduction map based on the accompanying user monitoring data to obtain a risk deduction optimization map, and simultaneously generating an elderly care monitoring early warning signal.

[0007] Optionally, data cleaning and feature recognition are performed on the elderly care monitoring dataset to obtain physiological monitoring feature sequences, environmental monitoring feature sequences, and behavioral monitoring feature sequences; anomaly analysis and attention optimization are performed on the physiological monitoring feature sequences based on the user activity scene labels to generate a physiological anomaly attention coupling matrix; anomaly analysis and attention optimization are performed on the environmental monitoring feature sequences based on the user activity scene labels to generate an environmental anomaly attention coupling matrix; anomaly analysis and attention optimization are performed on the behavioral monitoring feature sequences based on the user activity scene labels to generate a behavioral anomaly attention coupling matrix; the physiological anomaly attention coupling matrix, the environmental anomaly attention coupling matrix, and the behavioral anomaly attention coupling matrix are aggregated to generate the monitoring anomaly attention coupling matrix.

[0008] Optionally, based on the user activity scenario tags, normal physiological monitoring samples are retrieved for the home-based elderly care users to obtain a scenario-based normal physiological sample set; multidimensional physiological feature central trend analysis is performed on the scenario-based normal physiological sample set to construct a scenario-based normal physiological monitoring space; anomaly identification is performed on the physiological monitoring feature sequence based on the scenario-based normal physiological monitoring space to obtain physiological anomaly identification results; the degree of anomaly is evaluated based on the physiological anomaly identification results based on the scenario-based normal physiological monitoring space to obtain physiological anomaly evaluation results; and anomaly attention coupling optimization is performed on the physiological monitoring feature sequence based on the physiological anomaly identification results and the physiological anomaly evaluation results to obtain the physiological anomaly attention coupling matrix.

[0009] Optionally, risk incident retrieval is performed based on the user activity scenario tags to obtain a scenario-matching historical risk incident set; incident type identification is performed based on the scenario-matching historical risk incident set to obtain an incident type sample set; support is calculated for each incident type sample in the incident type sample set based on the scenario-matching historical risk incident set to obtain the support of each incident type; confidence is calculated for each incident type sample based on the support of each incident type to obtain the confidence of each incident type; and based on the predetermined type confidence, the incident type sample set is optimized and identified to generate the N fitted risk incident types.

[0010] Optionally, risk incident simulation records are retrieved based on the N fitted risk incident types to obtain N risk incident simulation record sets; incident tree training is performed based on the N risk incident simulation record sets to obtain N risk incident simulation models; dynamic perturbation distillation enhancement is performed on the N risk incident simulation models according to the dynamic perturbation distillation mechanism to establish N risk incident simulation channels; the monitoring anomaly attention coupling matrix is ​​input into the N risk incident simulation channels to obtain N risk incident simulation paths; the N risk incident simulation paths are organized to generate the risk incident simulation map.

[0011] Optionally, based on the N risk incident simulation models, an nth risk incident simulation model is extracted, where n is a positive integer, 1 ≤ n ≤ N; based on the N risk incident simulation record sets, an nth risk incident simulation record set corresponding to the nth risk incident simulation model is extracted; the dynamic perturbation distillation mechanism is activated, which includes a multimodal perturbation factor, comprising temporal perturbation, feature perturbation, and semantic perturbation; the nth risk incident simulation record set is perturbed multiple times according to the multimodal perturbation factor to obtain a temporal perturbation risk simulation. The system comprises a record set, a feature-based perturbation risk inference record set, and a semantic-based perturbation risk inference record set. Incremental learning is performed on the nth risk event inference model based on these three sets, respectively, to establish a first enhanced risk inference model, a second enhanced risk inference model, and a third enhanced risk inference model. Hierarchical distillation enhancement is then performed on these three enhanced models to generate the nth risk event inference channel.

[0012] Optionally, based on the accompanying user monitoring data, accident prediction is performed on the real-time accompanying users to obtain accompanying accident prediction results; based on the accompanying accident prediction results, accident triggering interference analysis is performed on the risk accident projection map to obtain accident triggering interference analysis results; based on the accompanying accident prediction results, accident propagation interference analysis is performed on the risk accident projection map to obtain accident propagation interference analysis results; based on the accompanying accident prediction results, accident accumulation interference analysis is performed on the risk accident projection map to obtain accident accumulation interference analysis results; based on the accident triggering interference analysis results, the accident propagation interference analysis results, and the accident accumulation interference analysis results, the risk accident projection map is corrected to obtain the optimized risk projection map.

[0013] Optionally, an accompanying early warning signal is generated based on the accompanying accident prediction results.

[0014] In a second aspect of the present application, an elderly care data monitoring and early warning system with multi-device linkage is provided, which comprises: a scene label generation module configured to perform linkage monitoring on home-based elderly care users through a multi-device monitoring array, obtain an elderly care monitoring data set, perform activity scene feature recognition based on the elderly care monitoring data set, and generate user activity scene labels; a matrix establishment module configured to perform multi-dimensional abnormal analysis attention optimization on the elderly care monitoring data set according to the user activity scene labels, and establish a monitoring abnormal attention coupling matrix; a risk accident type fitting module configured to perform risk accident type fitting according to the user activity scene labels, and obtain N fitting risk accident types, wherein N is a positive integer greater than 1; a risk accident deduction module configured to perform multi-dimensional risk accident deduction on the monitoring abnormal attention coupling matrix according to a dynamic perturbation distillation mechanism based on the N fitting risk accident types, and obtain a risk accident deduction graph; a monitoring and early warning signal generation module configured to read accompanying user monitoring data corresponding to real-time accompanying users, perform accompanying accident interference analysis and optimization on the risk accident deduction graph based on the accompanying user monitoring data, obtain an optimized risk deduction graph, and synchronously generate elderly care monitoring and early warning signals.

[0015] In a third aspect of the present application, an electronic device is provided, comprising: at least one processor; a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor, so that the at least one processor can execute the steps of the above-mentioned elderly care data monitoring and early warning method with multi-device linkage.

[0016] One or more technical solutions provided in the present application have at least the following technical effects or advantages: The method provided in this application embodiment uses a multi-device monitoring array to perform coordinated monitoring of home-based elderly care users, obtaining an elderly care monitoring dataset. Based on this dataset, activity scene features are identified to generate user activity scene labels. Multi-dimensional anomaly analysis and attention optimization are performed on the elderly care monitoring dataset according to the user activity scene labels to establish a monitoring anomaly attention coupling matrix. Risk accident types are fitted based on the user activity scene labels to obtain N fitted risk accident types, where N is a positive integer greater than 1. Based on the N fitted risk accident types, a dynamic perturbation distillation mechanism is used to perform multi-dimensional risk accident deduction on the monitoring anomaly attention coupling matrix to obtain a risk accident deduction map. Real-time accompanying user monitoring data is read, and accompanying user interference analysis is performed on the risk accident deduction map based on the accompanying user monitoring data to obtain an optimized risk deduction map, simultaneously generating an elderly care monitoring early warning signal. Through multi-device coordinated monitoring and multi-dimensional risk deduction, while considering the influence of accompanying users, the method achieves the technical effect of improving the accuracy and reliability of accident monitoring and early warning for home-based elderly care users.

[0017] The above description is merely an overview of the technical solution of this application. To enable a clearer understanding of the technical means of this application and to facilitate its implementation according to the description, and to make the above and other objects, features, and advantages of this application more apparent, specific embodiments of this application are described below. It should be understood that the content described in this section is not intended to identify key or important features of the embodiments of this application, nor is it intended to limit the scope of this application. Other features of this application will become readily apparent through the following description. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely exemplary. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0019] Figure 1 This is a flowchart illustrating a multi-device-linked elderly care data monitoring and early warning method provided in this application.

[0020] Figure 2 This is a schematic diagram of a multi-device linkage elderly care data monitoring and early warning system provided in this application.

[0021] Figure 3 A schematic diagram of the structure of an exemplary electronic device provided in this application.

[0022] Description of Reference Numerals: scene label generation module 11, matrix establishment module 12, risk accident type fitting module 13, risk accident deduction module 14, monitoring and early warning signal generation module 15, bus 300, receiver 301, processor 302, transmitter 303, memory 304, bus interface 305. DETAILED DESCRIPTION

[0023] The present application provides a pension data monitoring and early warning method, system and device with multi-device linkage, which is used to solve the technical problem in the prior art that the elderly care monitoring is one-sided, resulting in insufficient accuracy and reliability of elderly care accident monitoring and prediction. Through multi-device linkage monitoring and multi-dimensional risk deduction, and simultaneous consideration of user-related influences, the technical effect of improving the accuracy and reliability of accident monitoring and early warning for home-based elderly care users is achieved.

[0024] Hereinafter, the technical solution of the present invention will be described clearly and completely with reference to the accompanying drawings. Obviously, the described embodiments are only part, not all, of the embodiments of the present invention. It should be understood that the present invention is not limited to the exemplary embodiments described herein. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention. In addition, it should be noted that for the convenience of description, only the parts related to the present invention are shown in the accompanying drawings, not all of them.

[0025] Example 1, as Figure 1 shown, the present application provides a multi-device linkage elderly care data monitoring and early warning method, and the multi-device linkage elderly care data monitoring and early warning method comprises: Performing linkage monitoring on home-based elderly care users through a multi-device monitoring array to obtain an elderly care monitoring data set, and performing activity scene feature recognition based on the elderly care monitoring data set to generate user activity scene labels.

[0026] Specifically, a multi-device monitoring array is constructed, comprising various types of monitoring devices, including but not limited to wearable physiological monitoring devices, environmental sensors, and monitoring cameras. Wearable physiological monitoring devices, such as smart bracelets, can collect real-time physiological indicators such as heart rate, blood pressure, body temperature, and blood oxygen saturation. Environmental sensors, such as temperature and humidity sensors and light intensity sensors, are used to collect information such as temperature, humidity, and light intensity in the home environment. Monitoring cameras are used to collect users' daily behaviors and actions, reflecting their living activity status, such as activity trajectories, types of movements (sitting, standing, walking, lying down, etc.), and activity frequency. Multiple monitoring devices are interconnected via wireless communication technology, such as Wi-Fi. This multi-device monitoring array enables comprehensive and coordinated monitoring of home-based elderly care users, collecting multiple data points, including physiological monitoring data, environmental monitoring data, and behavioral monitoring data, to form an elderly care monitoring dataset.

[0027] The elderly care monitoring dataset is analyzed to identify user activity scenarios and generate user activity scenario labels based on the identified scenario types. For example, machine learning algorithms are used to analyze and identify the dataset. By collecting a large amount of home-based elderly care monitoring data, including physiological, environmental, and behavioral monitoring data under different activity scenarios, noise and outliers are removed from the data. This includes physiological data values ​​exceeding human limits and obviously unreasonable data in the environmental data. Each data sample is then labeled with a corresponding activity scenario label, such as resting, eating, or exercising. Behavioral data is processed by converting dynamic information acquired by cameras into data features, including the frequency, amplitude, and type of movements. The preprocessed dataset is divided into training and testing sets in a 7:3 or 8:2 ratio. The training set is used to train machine learning algorithms, such as decision tree algorithms. During training, information gain or Gini coefficient is calculated to determine the splitting features and split points of nodes. The trained decision tree model is then evaluated using a test set, with metrics including precision, recall, and F1 score. Once the model meets performance requirements, the completed model is obtained. The elderly care monitoring dataset is then input into the decision tree model, which classifies the input data based on learned rules and patterns, identifies user activity scene characteristics, and outputs corresponding user activity scene labels. Activity scene characteristics refer to different types of activities performed by the elderly in their home environment, such as resting, preparing food, eating, and exercising. After identifying the activity scene characteristics, each activity scene is labeled accordingly, such as preparing food in the kitchen or resting in the bedroom.

[0028] Users' physiological indicators, environmental needs, and behavioral patterns vary across different activity scenarios. For example, in a resting scenario, a user's heart rate, blood pressure, and other physiological indicators should be at relatively stable levels; abnormal fluctuations may indicate health problems. In contrast, elevated physiological indicators during exercise may be normal. By generating accurate user activity scenario tags, we can more accurately match users' actual conditions, improve the accuracy and reliability of elderly care data monitoring and early warning, and ensure the safety and health of home-based elderly care users.

[0029] Based on the user activity scenario labels, multidimensional anomaly analysis and attention optimization are performed on the elderly care monitoring dataset to establish a monitoring anomaly attention coupling matrix.

[0030] Furthermore, based on the user activity scene tags, multi-dimensional anomaly analysis attention optimization is performed on the elderly care monitoring dataset to establish a monitoring anomaly attention coupling matrix. This includes: performing data cleaning and feature recognition on the elderly care monitoring dataset to obtain physiological monitoring feature sequences, environmental monitoring feature sequences, and behavioral monitoring feature sequences; performing anomaly analysis attention optimization on the physiological monitoring feature sequences based on the user activity scene tags to generate a physiological anomaly attention coupling matrix; performing anomaly analysis attention optimization on the environmental monitoring feature sequences based on the user activity scene tags to generate an environmental anomaly attention coupling matrix; performing anomaly analysis attention optimization on the behavioral monitoring feature sequences based on the user activity scene tags to generate a behavioral anomaly attention coupling matrix; and aggregating the physiological anomaly attention coupling matrix, the environmental anomaly attention coupling matrix, and the behavioral anomaly attention coupling matrix to generate the monitoring anomaly attention coupling matrix.

[0031] Furthermore, based on the user activity scene tags, anomaly analysis and attention optimization are performed on the physiological monitoring feature sequence to generate a physiological anomaly attention coupling matrix. This includes: retrieving normal physiological monitoring samples from the home-based elderly care user based on the user activity scene tags to obtain a scene-based normal physiological sample set; performing multi-dimensional physiological feature central trend analysis on the scene-based normal physiological sample set to construct a scene-based normal physiological monitoring space; identifying anomalies in the physiological monitoring feature sequence based on the scene-based normal physiological monitoring space to obtain physiological anomaly identification results; evaluating the degree of anomaly in the physiological anomaly identification results based on the scene-based normal physiological monitoring space to obtain physiological anomaly evaluation results; and optimizing the physiological monitoring feature sequence for anomaly attention coupling based on the physiological anomaly identification results and the physiological anomaly evaluation results to obtain the physiological anomaly attention coupling matrix.

[0032] Specifically, the elderly care monitoring dataset undergoes data cleaning to remove noise and outliers, such as erroneous values ​​exceeding normal physiological limits in physiological monitoring data and obviously unreasonable data in environmental monitoring data. This data cleaning ensures data quality. Then, the cleaned data is used for feature identification based on time series data, forming physiological monitoring feature sequences, environmental monitoring feature sequences, and behavioral monitoring feature sequences. The physiological monitoring feature sequences are sequences of real-time collected user physiological data arranged chronologically, including heart rate, blood pressure, body temperature, and blood oxygen saturation, reflecting the user's health status. The environmental monitoring feature sequences include indoor temperature, humidity, and light intensity, reflecting the user's living environment. The behavioral monitoring feature sequences include behavioral information such as the user's action types, activity frequency, and activity trajectories, reflecting the user's activity patterns and lifestyle.

[0033] Anomaly analysis and attention optimization are performed on physiological monitoring feature sequences, environmental monitoring feature sequences, and behavioral monitoring feature sequences based on user activity scenario tags. In the process of anomaly analysis and attention optimization of physiological monitoring feature sequences based on user activity scenario tags, the physiological data of home-based elderly care users is first retrieved based on the user activity scenario tags. Activity scenarios already labeled in historical data, such as rest, meals, and exercise, are used to retrieve normal physiological monitoring samples that match the user's activity scenario tags. For example, when retrieving normal physiological samples for a rest scenario, data records showing that the user's heart rate, blood pressure, body temperature, and other physiological indicators are within the normal range and relatively stable under rest conditions are extracted, forming a normal physiological sample set for that scenario.

[0034] The central tendency analysis is performed on the multidimensional physiological features of the normal physiological sample set of a scenario, and statistical measures such as the mean and median of the multidimensional physiological samples are calculated. Taking heart rate in the physiological sample set as an example, the mean heart rate of all normal samples under the user activity scenario label is calculated as the central tendency value of heart rate in that scenario. By analyzing the central tendency of each physiological feature in the normal physiological sample set of the scenario, the normal fluctuation range of user indicators in the scenario can be reflected. The central tendency values ​​of multidimensional physiological features are integrated to construct a normal physiological monitoring space for the scenario, which is an integration of the normal range of multidimensional physiological features. For example, in the three-dimensional normal physiological monitoring space of the scenario, the three dimensions represent heart rate, blood pressure and body temperature, respectively, and the data points in the space correspond to the normal combination of physiological features in different scenarios. Then, the real-time collected physiological monitoring feature sequence is compared with the normal physiological monitoring space of the scenario. If the physiological feature at a certain time point exceeds the range of the normal monitoring space in the corresponding scenario, it is judged as abnormal, and the physiological abnormality identification result is obtained. For example, in a resting context, a user's normal blood pressure range is 90-120 mmHg systolic and 60-80 mmHg diastolic. If a systolic blood pressure of 130 mmHg and a diastolic blood pressure of 85 mmHg are monitored at a certain moment, which exceed the normal range, it will be identified as abnormal.

[0035] Based on the normal physiological monitoring space of the scenario, an abnormality deviation index is obtained for the physiological abnormality identification results. For each physiological feature identified as abnormal, the deviation of its actual value from the normal range in the normal physiological monitoring space is calculated. The abnormality is evaluated by weighting the abnormality duration. If the abnormality duration is long, it indicates a significant impact on the user's health, and a time weighting coefficient is set. For example, in a resting scenario, the normal blood pressure range for a user is 90-120 mmHg systolic and 60-80 mmHg diastolic. If, at a certain moment, the systolic blood pressure is 130 mmHg and the diastolic blood pressure is 85 mmHg, then the systolic blood pressure deviates from the upper limit of normal by 10 mmHg, and the diastolic blood pressure deviates from the upper limit of normal by 5 mmHg. If the blood pressure abnormality lasts for 5 minutes, the time weighting coefficient is 0.1, then the weighted deviation of systolic blood pressure is 10 × (1 + 0.1 × 5) = 15, and the weighted deviation of diastolic blood pressure is 5 × (1 + 0.1 × 5) = 7.5. As the abnormality duration increases, the time weighting coefficient also increases accordingly. Different physiological characteristics have varying degrees of impact on health. For example, abnormal heart rate is more urgent than abnormal body temperature. Based on medical knowledge and clinical experience, each physiological characteristic is assigned an importance weight, such as heart rate weight 0.4, blood pressure weight 0.4, and body temperature weight 0.2. Multiplying the weighted deviation by the importance weight yields the physiological abnormality assessment result corresponding to the physiological abnormality identification result. For example, for systolic blood pressure, the abnormality assessment score = 15 × 0.4 = 6, and the abnormality score for diastolic blood pressure = 7.5 × 0.4 = 3.

[0036] Anomaly attention coupling optimization is performed on the physiological monitoring sequence based on the results of physiological anomaly identification and evaluation. This optimization involves assigning different attention weights to different physiological characteristics in the physiological monitoring feature sequence according to the type and severity of the anomaly. For anomalies with high severity, higher attention weights are assigned. An attention mechanism is used to dynamically adjust the level of attention given to each anomaly based on its identification and evaluation results, thus obtaining anomaly attention weights. These anomaly attention weights are then sorted according to their order in the physiological monitoring feature sequence to form a physiological anomaly attention coupling matrix. Each element in this matrix represents the attention weight of the corresponding physiological characteristic in anomaly monitoring; a higher weight indicates a higher level of attention to that characteristic.

[0037] Similarly, based on the user activity scene tags, normal environmental monitoring samples and normal behavior monitoring samples are retrieved from the environmental monitoring feature sequences and behavior monitoring feature sequences, respectively. Central trend analysis is performed to construct the scene normal environmental monitoring space and the scene normal behavior monitoring space. Anomaly identification and anomaly degree evaluation are performed on the environmental monitoring feature sequences and behavior monitoring feature sequences, and anomaly attention coupling optimization is performed to obtain the environmental anomaly attention coupling matrix and the behavior anomaly attention coupling matrix, respectively.

[0038] Based on historical data statistics or expert experience, a weight is assigned to each abnormal attention coupling matrix to reflect the importance of different types of abnormalities in monitoring. For example, physiological abnormalities have a greater impact on the health of elderly users, so the weight of the physiological abnormality attention coupling matrix is ​​set higher, such as 0.4, while environmental abnormalities and behavioral abnormalities are each set to 0.3. Based on the assigned weights, the physiological abnormality attention coupling matrix, the environmental abnormality attention coupling matrix, and the behavioral abnormality attention coupling matrix are weighted and summed to generate the monitoring abnormality attention coupling matrix. This monitoring abnormality attention coupling matrix integrates abnormal information from three dimensions: physiological, environmental, and behavioral, and can comprehensively and accurately reflect the user's abnormal status in different scenarios.

[0039] By analyzing and integrating multidimensional anomalies, we can avoid misjudgments or omissions caused by single-dimensional analysis, and more accurately identify abnormal situations of users in different activity scenarios. This will improve the comprehensiveness and accuracy of user health and safety assessments, and safeguard the health and safety of home-based elderly care users.

[0040] Based on the user activity scenario tags, risk incident types are fitted to obtain N fitted risk incident types, where N is a positive integer greater than 1.

[0041] Furthermore, risk incident types are fitted based on the user activity scenario tags to obtain N fitted risk incident types, including: risk incident retrieval based on the user activity scenario tags to obtain a scenario-matching historical risk incident set; incident type identification based on the scenario-matching historical risk incident set to obtain an incident type sample set; support calculation for each incident type sample in the incident type sample set based on the scenario-matching historical risk incident set to obtain the support of each incident type; confidence calculation for each incident type sample based on the support of each incident type to obtain the confidence of each incident type; and optimization identification of the incident type sample set based on the predetermined type confidence and the confidence of each incident type to generate the N fitted risk incident types.

[0042] Specifically, based on user activity scenario tags, database retrieval technology is used to search for historical risk accident records that match the user activity scenario tags in a pre-built database containing a large amount of historical risk accident data, thus obtaining a scenario-matched historical risk accident set. For example, in a kitchen scenario, accident data related to falls and burns is retrieved. The scenario-matched historical risk accident set is then categorized to extract different accident types, such as falls, strokes, and heart attacks. These extracted accident types are then classified to generate an accident type sample set, where each sample represents a specific type of accident.

[0043] Support is calculated for each accident type sample in the historical risk accident set for scene matching. Support refers to the number of times a specific accident type appears in the historical risk accident set. The support for each accident type is obtained by statistically analyzing the occurrence counts of multiple accident types in the historical risk accident set for scene matching. For example, in the historical risk accident set for the cooking scene, a fall accident occurred 5 times, so the support for the fall accident type is 5. The confidence score is calculated using the formula: Accident Type Confidence Score = Accident Type Support Score ÷ Sum of Supports for All Accident Types. The confidence score for each accident type sample is calculated based on the support scores of each accident type. For example, in the historical risk accident set for the cooking scene, which includes fall accidents and burn accidents, the support for fall accidents is 5, and the support for burn accidents is 3. Therefore, the confidence score for fall accidents is 5 ÷ (5 + 3) = 0.625, and the confidence score for burn accidents is 3 ÷ (5 + 3) = 0.375.

[0044] The support levels of various accident types in historical accident data are statistically analyzed to calculate the probability of each type of accident occurring in different scenarios. Based on probability and safety risk tolerance requirements, a predetermined type confidence level is set to distinguish between high-risk and low-risk accident types. The accident type sample set is then screened and optimized based on the predetermined type confidence level. Specifically, the confidence level of each accident type is compared with the predetermined type confidence level, and accident types with confidence levels greater than or equal to the predetermined type confidence level are selected as N fitted risk accident types, where N is a positive integer greater than 1. These N fitted risk accident types represent the accidents most likely to occur under specific user activity scenarios.

[0045] By calculating support, confidence, and optimization, high-risk accident types related to specific activity scenarios are extracted from historical data. This enables accurate identification of potential risk factors in different user activity scenarios, providing users with real-time and accurate safety warnings and more precise safety protection for home-based elderly care users.

[0046] Based on the N fitted risk accident types, a multi-dimensional risk accident deduction is performed on the monitoring anomaly attention coupling matrix according to the dynamic perturbation distillation mechanism to obtain a risk accident deduction map.

[0047] Furthermore, based on the N fitted risk incident types, a multi-dimensional risk incident deduction is performed on the monitoring anomaly attention coupling matrix according to the dynamic perturbation distillation mechanism to obtain a risk incident deduction map, including: retrieving risk incident deduction records based on the N fitted risk incident types to obtain N risk incident deduction record sets; training an incident tree based on the N risk incident deduction record sets to obtain N risk incident deduction models; performing dynamic perturbation distillation enhancement on the N risk incident deduction models according to the dynamic perturbation distillation mechanism to establish N risk incident deduction channels; inputting the monitoring anomaly attention coupling matrix into the N risk incident deduction channels to obtain N risk incident deduction paths; and organizing the N risk incident deduction paths to generate the risk incident deduction map.

[0048] Specifically, based on N fitted risk incident types, risk incident simulation records are retrieved from a pre-stored database of numerous risk incident simulation records to obtain N risk incident simulation record sets. These sets contain detailed information about historical incident types related to the fitted risk incident types, including occurrence time, environment, and user health status. These N risk incident simulation records are used as a training set to train a fault tree analysis (FTA) method. FTA analyzes the probability and consequences of various events through logical relationships, identifying the root causes and triggering conditions of incidents in specific scenarios. During training, key information is first extracted from each risk incident simulation record set, including incident type, occurrence conditions, triggering factors, and results, to construct a preliminary fault tree framework. Then, the Fault Tree Analysis (FTA) method is used to analyze the high-risk nodes of each incident by logically associating the various conditions and events involved. Based on N risk incident simulation records, machine learning algorithms, such as decision trees or random forests, are used to train N incident tree models, adjust the weights and connections of each node, and form N risk incident simulation models. These N risk incident simulation models are used to simulate the occurrence paths of different types of accidents in specific scenarios and environments.

[0049] A dynamic perturbation distillation enhancement mechanism is used to enhance N risk event prediction models. This mechanism involves introducing random perturbations to optimize the N risk event prediction models and improve their generalization ability. During the enhancement process, random perturbations of different intensities and types are added to each risk event prediction model. The perturbed models are then fused using distillation techniques to obtain a more stable and accurate prediction model, which serves as the N risk event prediction channels. The monitoring anomaly attention coupling matrix is ​​input into these N prediction channels for analysis, outputting N risk event prediction paths. Each path represents the possible development process of a certain risk event under specific monitoring data. For example, if a user has an abnormal heart rate in a kitchen setting, a fall risk path might be predicted. The obtained N risk event prediction paths are then organized according to their importance and correlation, and a risk event prediction graph is generated using graphical tools. Each path in the risk incident projection map represents a potential risk evolution process, comprehensively and intuitively reflecting the possible risk incidents, their sequence, and correlations under different monitoring data and scenarios. This provides comprehensive and effective data support for user risk warning and intervention, thereby ensuring the safety of home-based elderly care users.

[0050] Furthermore, based on the dynamic perturbation distillation mechanism, the N risk incident simulation models are enhanced by dynamic perturbation distillation to establish N risk incident simulation channels, including: extracting the nth risk incident simulation model from the N risk incident simulation models, where n is a positive integer, 1≤n≤N; extracting the nth risk incident simulation record set corresponding to the nth risk incident simulation model from the N risk incident simulation record sets; activating the dynamic perturbation distillation mechanism, which includes a multimodal perturbation factor, including temporal perturbation, feature perturbation, and semantic perturbation; and enhancing the N risk incident simulation models with dynamic perturbation distillation based on the multimodal perturbation factor. The nth risk event simulation record set is subjected to multiple rounds of perturbation to obtain a temporal perturbation risk simulation record set, a feature perturbation risk simulation record set, and a semantic perturbation risk simulation record set. Based on the temporal perturbation risk simulation record set, the feature perturbation risk simulation record set, and the semantic perturbation risk simulation record set, the nth risk event simulation model is incrementally learned to establish a first risk simulation enhancement model, a second risk simulation enhancement model, and a third risk simulation enhancement model. Based on the first risk simulation enhancement model, the second risk simulation enhancement model, and the third risk simulation enhancement model, hierarchical distillation enhancement is performed to generate the nth risk event simulation channel.

[0051] Specifically, the nth risk event simulation model is extracted sequentially from N risk event simulation models, where n is a positive integer, 1 ≤ n ≤ N. Based on the extracted nth risk event simulation model, the corresponding nth risk event simulation record set is extracted from the N risk event simulation record sets. A dynamic perturbation distillation mechanism is activated, which includes multimodal perturbation factors, including temporal perturbation, feature perturbation, and semantic perturbation. Temporal perturbation refers to altering the time-series information in the risk event simulation records; the risk changes within different time periods of the model can be perturbed through preset time change rules. Feature perturbation refers to altering the monitoring features in the risk event simulation records, such as physiological data, environmental data, or behavioral data, simulating the impact of physiological, environmental, or behavioral changes on the occurrence of accidents; feature modification can be performed according to preset feature adjustment strategies. Semantic perturbation refers to interfering with the semantic information in the risk event simulation records, such as modifying the semantic expression of the accident description, improving the simulation model's ability to perform simulations under different semantic understandings; natural language processing techniques can be used to transform the semantics.

[0052] Based on the temporal, feature, and semantic perturbations in the multimodal perturbation factors, the nth risk event simulation record set is perturbed multiple times. The perturbed data is then organized to form temporal perturbation risk simulation record sets, feature perturbation risk simulation record sets, and semantic perturbation risk simulation record sets. An incremental learning algorithm is then used to incrementally learn the nth risk event simulation model using these record sets. Incremental learning refers to updating and optimizing the model by inputting new data based on the nth risk event simulation model. During incremental learning, the nth risk event simulation model processes the new information in the perturbed temporal, feature, and semantic perturbation risk simulation record sets to optimize the model and establish a first enhanced risk simulation model, a second enhanced risk simulation model, and a third enhanced risk simulation model. Among them, the first enhanced risk simulation model, the second enhanced risk simulation model, and the third enhanced risk simulation model correspond to the optimization of the nth risk event simulation model by the time-series perturbation risk simulation record set, the feature perturbation risk simulation record set, and the semantic perturbation risk simulation record set, respectively.

[0053] Knowledge distillation technology is employed to perform hierarchical distillation enhancement on the first, second, and third enhanced risk simulation models. Knowledge distillation is a technique that fuses knowledge from multiple models, extracting useful simulation knowledge from these models, including the independent simulation paths of each model and their outputs under different perturbation factors, and defining soft labels. Using a teacher-student model approach, the first, second, and third enhanced risk simulation models are used as teacher models, and the target simulation model is used as a learning model. The learning model improves its simulation capabilities by learning the soft labels of the teacher models and minimizing the differences between the soft labels output by the student models and the teacher models. Through hierarchical distillation enhancement, the knowledge from the first, second, and third enhanced risk simulation models is compressed and integrated to generate an nth risk incident simulation channel with higher simulation efficiency and accuracy. Repeating the above steps yields N risk incident simulation channels.

[0054] By enhancing dynamic perturbation and knowledge distillation, the risk incident simulation channel is improved to adapt to changing environments and risk scenarios, thereby increasing the accuracy of risk prediction. Furthermore, by processing elderly care monitoring data, it can quickly and accurately assess risks and provide early warnings, effectively preventing and responding to various potential safety risks and ensuring the safety of elderly users at home.

[0055] Read the real-time accompanying user monitoring data corresponding to the accompanying user, and perform accompanying accident interference analysis and optimization on the risk accident projection map based on the accompanying user monitoring data to obtain the risk projection optimization map, and simultaneously generate elderly care monitoring and early warning signals.

[0056] Furthermore, the risk incident projection map is optimized by performing accompanying incident interference analysis based on the accompanying user monitoring data to obtain an optimized risk projection map. This includes: performing incident prediction on the real-time accompanying users based on the accompanying user monitoring data to obtain accompanying incident prediction results; performing incident triggering interference analysis on the risk incident projection map based on the accompanying incident prediction results to obtain incident triggering interference analysis results; performing incident propagation interference analysis on the risk incident projection map based on the accompanying incident prediction results to obtain incident propagation interference analysis results; performing incident accumulation interference analysis on the risk incident projection map based on the accompanying incident prediction results to obtain incident accumulation interference analysis results; and correcting the risk incident projection map based on the incident triggering interference analysis results, the incident propagation interference analysis results, and the incident accumulation interference analysis results to obtain the optimized risk projection map.

[0057] Specifically, real-time monitoring data of accompanying users is read. Accompanying users refer to the group of people living or associated with home-based elderly care users. The monitoring data includes physiological indicators, environmental data, and behavioral activity data. Accident prediction is performed on accompanying users based on this monitoring data, yielding prediction results. For example, historical monitoring data and corresponding accident tags are acquired, and the historical data is divided into training and validation sets. A neural network model is trained using the training set and the corresponding accident tags. During training, the weights and biases in the network are continuously adjusted using a backpropagation algorithm to minimize the error between the predicted accident results and the actual labels. The trained neural network is then evaluated and adjusted using a validation set. Training stops when the performance of the neural network model on the validation set no longer improves. The accompanying user monitoring data is then input into the trained neural network model for accident prediction, yielding prediction results. These prediction results represent the types and probabilities of abnormal events that may occur to accompanying users in the future, such as falls.

[0058] Based on the accompanying accident prediction results, an accident triggering interference analysis is performed on the risk accident projection map to analyze whether changes in the behavior or health of accompanying users trigger accident risks for home-based elderly care users. For example, a fall by an accompanying user may cause severe fluctuations in the heart rate of the home-based elderly care user, affecting their health status. Using time series analysis, the accompanying accident prediction results are matched with the risk accident projection map to calculate the risk level and temporal relationship of the accompanying accident predictions, assess whether it will trigger accidents for elderly care users, and obtain the accident triggering interference analysis results. Simultaneously, by analyzing historical accident data and the risk accident projection map, a propagation path of accompanying accidents to elderly care users is established. The propagation path refers to the gradual expansion from the abnormal behavior of accompanying users to the health or behavioral status of elderly care users. Based on the accompanying accident prediction results, an accident propagation interference analysis is performed on the risk accident projection map to calculate the probability of different risk propagation paths and assess the impact of accompanying accidents on the propagation of accidents for elderly care users. For example, accompanying accidents may increase the risk of falls for elderly care users. Further analysis of the cumulative risk effect of accompanying accidents on elderly care users shows that long-term or repeated accompanying accidents may have a cumulative effect, increasing the risk for elderly care users. Based on the prediction results of accompanying accidents, the cumulative interference analysis of accidents in the risk accident projection map is performed. Multiple related nodes in the projection map are cumulatively weighted to identify the cumulative effect of accompanying accidents and obtain the results of the accident cumulative interference analysis.

[0059] The risk event projection map is revised based on the results of accident triggering interference analysis, accident propagation interference analysis, and accident accumulation interference analysis. Accident triggering interference analysis identifies which accompanying accidents may directly or indirectly cause risks to elderly users, and adjusts the risk nodes in the risk event projection map accordingly. The propagation weights of relevant nodes in the risk event projection map are updated based on the propagation paths identified in the accident propagation interference analysis. Accident accumulation interference analysis analyzes the cumulative impact of accompanying accidents, increasing the risk weights of relevant risk nodes based on the cumulative effect. By revising the accident nodes, paths, and weights in the risk event projection map using these three types of triggering interference analysis results, an optimized risk projection map is generated. When the optimized risk projection map shows that the accident risk exceeds a threshold, a relevant elderly care monitoring and early warning signal is generated and notified to relevant personnel via SMS, app push notifications, email, etc.

[0060] By analyzing accompanying user monitoring data and optimizing the risk incident projection map, the map becomes more realistic, improving the comprehensiveness, accuracy, and timeliness of risk incident prediction in the elderly care process. Simultaneously generating elderly care monitoring and early warning signals allows for timely notification of relevant personnel to take measures, effectively reducing the harm caused by risk incidents to users, ensuring their safety and health, and improving the quality and safety of elderly care services.

[0061] Further, an accompanying early warning signal is generated according to the accompanying accident prediction result.

[0062] Specifically, accident prediction is performed on the real-time accompanying user according to the accompanying user monitoring data, and after obtaining the accompanying accident prediction result, the occurrence probabilities of multiple risk events in the accompanying accident prediction result are compared with a preset threshold. When the occurrence probability of a risk event is greater than or equal to the prediction threshold, an accompanying early warning signal is generated, so that preventive measures can be taken before the risk occurs, which improves the safety of the home care environment, reduces the occurrence of potential accidents, and ensures the safety and health of the elderly.

[0063] In Embodiment 2, based on the same inventive concept as the elderly care data monitoring and early warning method linked by multiple devices in the foregoing embodiments, as Figure 2 shows, the present application provides an elderly care data monitoring and early warning system linked by multiple devices, wherein the elderly care data monitoring and early warning system linked by multiple devices includes: a scene label generation module 11, configured to perform linked monitoring on home care users through a multi-device monitoring array to obtain an elderly care monitoring data set, perform activity scene feature recognition based on the elderly care monitoring data set, and generate user activity scene labels; a matrix establishment module 12, configured to perform multi-dimensional anomaly analysis attention optimization on the elderly care monitoring data set according to the user activity scene labels, and establish a monitoring anomaly attention coupling matrix; a risk accident type fitting module 13, configured to perform risk accident type fitting according to the user activity scene labels to obtain N fitted risk accident types, where N is a positive integer greater than 1; a risk accident deduction module 14, configured to perform multi-dimensional risk accident deduction on the monitoring anomaly attention coupling matrix according to a dynamic perturbation distillation mechanism based on the N fitted risk accident types, so as to obtain a risk accident deduction graph; a monitoring and early warning signal generation module 15, configured to read accompanying user monitoring data corresponding to a real-time accompanying user, perform accompanying accident interference analysis optimization on the risk accident deduction graph based on the accompanying user monitoring data to obtain an optimized risk deduction graph, and synchronously generate an elderly care monitoring and early warning signal.

[0064] Furthermore, the matrix building module 12 is also used for: performing data cleaning and feature recognition based on the elderly care monitoring dataset to obtain physiological monitoring feature sequences, environmental monitoring feature sequences, and behavioral monitoring feature sequences; performing anomaly analysis and attention optimization on the physiological monitoring feature sequences based on the user activity scene labels to generate a physiological anomaly attention coupling matrix; performing anomaly analysis and attention optimization on the environmental monitoring feature sequences based on the user activity scene labels to generate an environmental anomaly attention coupling matrix; performing anomaly analysis and attention optimization on the behavioral monitoring feature sequences based on the user activity scene labels to generate a behavioral anomaly attention coupling matrix; and aggregating the physiological anomaly attention coupling matrix, the environmental anomaly attention coupling matrix, and the behavioral anomaly attention coupling matrix to generate the monitoring anomaly attention coupling matrix.

[0065] Furthermore, the matrix establishment module 12 is also used for: retrieving normal physiological monitoring samples of the home-based elderly care user based on the user activity scene tags to obtain a scene normal physiological sample set; performing multi-dimensional physiological feature central trend analysis based on the scene normal physiological sample set to construct a scene normal physiological monitoring space; performing anomaly identification on the physiological monitoring feature sequence based on the scene normal physiological monitoring space to obtain physiological anomaly identification results; evaluating the degree of anomaly based on the physiological anomaly identification results based on the scene normal physiological monitoring space to obtain physiological anomaly evaluation results; and performing anomaly attention coupling optimization on the physiological monitoring feature sequence based on the physiological anomaly identification results and the physiological anomaly evaluation results to obtain the physiological anomaly attention coupling matrix.

[0066] Furthermore, the risk incident type fitting module 13 is also used for: performing risk incident retrieval based on the user activity scenario tags to obtain a scenario-matching historical risk incident set; performing incident type identification based on the scenario-matching historical risk incident set to obtain an incident type sample set; calculating the support of each incident type sample in the incident type sample set based on the scenario-matching historical risk incident set to obtain the support of each incident type; calculating the confidence of each incident type sample based on the support of each incident type to obtain the confidence of each incident type; and performing optimization identification on the incident type sample set based on the predetermined type confidence and the confidence of each incident type to generate the N fitted risk incident types.

[0067] Furthermore, the risk incident simulation module 14 is also used for: retrieving risk incident simulation records based on the N fitted risk incident types to obtain N risk incident simulation record sets; training an incident tree based on the N risk incident simulation record sets to obtain N risk incident simulation models; performing dynamic perturbation distillation enhancement on the N risk incident simulation models according to the dynamic perturbation distillation mechanism to establish N risk incident simulation channels; inputting the monitoring anomaly attention coupling matrix into the N risk incident simulation channels to obtain N risk incident simulation paths; and organizing the N risk incident simulation paths to generate the risk incident simulation map.

[0068] Furthermore, the risk incident simulation module 14 is also used for: extracting the nth risk incident simulation model based on the N risk incident simulation models, where n is a positive integer, 1≤n≤N; extracting the nth risk incident simulation record set corresponding to the nth risk incident simulation model based on the N risk incident simulation record sets; activating the dynamic perturbation distillation mechanism, which includes a multimodal perturbation factor, including temporal perturbation, feature perturbation, and semantic perturbation; and performing multiple rounds of perturbation on the nth risk incident simulation record set based on the multimodal perturbation factor to obtain... Obtain a time-series perturbation risk projection record set, a feature perturbation risk projection record set, and a semantic perturbation risk projection record set; perform incremental learning on the nth risk event projection model based on the time-series perturbation risk projection record set, the feature perturbation risk projection record set, and the semantic perturbation risk projection record set, respectively, to establish a first enhanced risk projection model, a second enhanced risk projection model, and a third enhanced risk projection model; perform hierarchical distillation enhancement based on the first enhanced risk projection model, the second enhanced risk projection model, and the third enhanced risk projection model to generate the nth risk event projection channel.

[0069] Furthermore, the monitoring and early warning signal generation module 15 is also used to: perform accident prediction on the real-time accompanying user based on the accompanying user monitoring data, and obtain accompanying accident prediction results; perform accident triggering interference analysis on the risk accident projection map based on the accompanying accident prediction results, and obtain accident triggering interference analysis results; perform accident propagation interference analysis on the risk accident projection map based on the accompanying accident prediction results, and obtain accident propagation interference analysis results; perform accident accumulation interference analysis on the risk accident projection map based on the accompanying accident prediction results, and obtain accident accumulation interference analysis results; and correct the risk accident projection map based on the accident triggering interference analysis results, the accident propagation interference analysis results, and the accident accumulation interference analysis results, to obtain the optimized risk projection map.

[0070] Furthermore, the monitoring and early warning signal generation module 15 is also used to generate an accompanying early warning signal based on the accompanying accident prediction results.

[0071] Example 3: Based on the same inventive concept as the multi-device linkage elderly care data monitoring and early warning method in the foregoing examples, this application also provides an electronic device, including: at least one processor; a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform the steps of the multi-device linkage elderly care data monitoring and early warning method described in any one of Examples 1 above.

[0072] Appendix Figure 3 This is a schematic diagram of the structure of an exemplary electronic device of this application. Figure 3 In this document, the bus architecture is represented by bus 300. Bus 300 may include any number of interconnected buses and bridges, and bus 300 connects various circuits including one or more processors represented by processor 302 and memory represented by memory 304. Bus 300 may also connect various other circuits such as peripheral devices, voltage regulators, and power management circuits, which are well known in the art and therefore will not be described further herein. Bus interface 305 provides an interface between bus 300 and receiver 301 and transmitter 303. Receiver 301 and transmitter 303 may be the same element, i.e., a transceiver, providing a unit for communicating with various other devices over a transmission medium. Processor 302 is responsible for managing bus 300 and general processing, while memory 304 can be used to store data used by processor 302 during operation.

[0073] The above description of the disclosed embodiments enables those skilled in the art to make or use this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

[0074] Obviously, those skilled in the art can make several improvements and modifications to this application without departing from the principles of this application, and these improvements and modifications also fall within the protection scope of this application.

Claims

1. A method for monitoring and early warning of elderly care data involving multiple devices, characterized in that, The method includes: By using a multi-device monitoring array to conduct joint monitoring of home-based elderly care users, an elderly care monitoring dataset is obtained, and activity scene feature recognition is performed based on the elderly care monitoring dataset to generate user activity scene tags. Based on the user activity scenario tags, multidimensional anomaly analysis and attention optimization are performed on the elderly care monitoring dataset to establish a monitoring anomaly attention coupling matrix. Based on the user activity scenario tags, risk incident types are fitted to obtain N fitted risk incident types, where N is a positive integer greater than 1; Based on the N fitted risk accident types, a multi-dimensional risk accident deduction is performed on the monitoring anomaly attention coupling matrix according to the dynamic perturbation distillation mechanism to obtain a risk accident deduction map. Read the real-time accompanying user monitoring data corresponding to the accompanying user, and perform accompanying accident interference analysis and optimization on the risk accident inference map based on the accompanying user monitoring data to obtain the risk inference optimization map, and simultaneously generate elderly care monitoring early warning signal; Specifically, based on the N fitted risk incident types, a multi-dimensional risk incident extrapolation is performed on the monitoring anomaly attention coupling matrix according to the dynamic perturbation distillation mechanism to obtain a risk incident extrapolation map, including: Based on the N fitted risk accident types, risk accident simulation records are retrieved to obtain N risk accident simulation record sets. Based on the N risk incident simulation record sets, an incident tree is trained to obtain N risk incident simulation models; The N risk incident simulation models are enhanced by dynamic perturbation distillation based on the dynamic perturbation distillation mechanism to establish N risk incident simulation channels. Input the monitoring anomaly attention coupling matrix into the N risk incident deduction channels to obtain N risk incident deduction paths; Organize the N risk incident simulation paths to generate the risk incident simulation map; Specifically, based on the dynamic perturbation distillation mechanism, the N risk incident simulation models are enhanced by dynamic perturbation distillation to establish N risk incident simulation channels, including: Based on the N risk incident simulation models, extract the nth risk incident simulation model, where n is a positive integer, 1≤n≤N; Based on the N risk incident simulation record sets, extract the nth risk incident simulation record set corresponding to the nth risk incident simulation model; The dynamic perturbation distillation mechanism is activated, which includes a multimodal perturbation factor, comprising temporal perturbation, feature perturbation, and semantic perturbation. The nth risk event simulation record set is perturbed multiple times according to the multimodal perturbation factor to obtain the temporal perturbation risk simulation record set, the feature perturbation risk simulation record set, and the semantic perturbation risk simulation record set; Based on the time-series perturbation risk simulation record set, the feature perturbation risk simulation record set, and the semantic perturbation risk simulation record set, incremental learning is performed on the nth risk accident simulation model to establish a first enhanced risk simulation model, a second enhanced risk simulation model, and a third enhanced risk simulation model. Based on the first enhanced risk simulation model, the second enhanced risk simulation model, and the third enhanced risk simulation model, hierarchical distillation enhancement is performed to generate the nth risk accident simulation channel.

2. The method for monitoring and early warning of elderly care data through multi-device linkage as described in claim 1, characterized in that, Based on the user activity scenario tags, multidimensional anomaly analysis and attention optimization are performed on the elderly care monitoring dataset to establish a monitoring anomaly attention coupling matrix, including: Data cleaning and feature recognition were performed on the elderly care monitoring dataset to obtain physiological monitoring feature sequences, environmental monitoring feature sequences, and behavioral monitoring feature sequences. Based on the user activity scenario tags, the physiological monitoring feature sequence is subjected to anomaly parsing and attention optimization to generate a physiological anomaly attention coupling matrix. Based on the user activity scene tags, the environmental monitoring feature sequence is subjected to anomaly parsing and attention optimization to generate an environmental anomaly attention coupling matrix. Based on the user activity scenario tags, the behavior monitoring feature sequence is subjected to anomaly parsing and attention optimization to generate a behavior anomaly attention coupling matrix. The physiological abnormality attention coupling matrix, the environmental abnormality attention coupling matrix, and the behavioral abnormality attention coupling matrix are aggregated to generate the monitoring abnormality attention coupling matrix.

3. The method for monitoring and early warning of elderly care data through multi-device linkage as described in claim 2, characterized in that, Based on the user activity scenario tags, anomaly analysis and attention optimization are performed on the physiological monitoring feature sequence to generate a physiological anomaly attention coupling matrix, including: Based on the user activity scenario tags, normal physiological monitoring samples of the home-based elderly care users are retrieved to obtain a scenario normal physiological sample set. Based on the normal physiological sample set of the scenario, perform multidimensional physiological feature central trend analysis to construct a normal physiological monitoring space for the scenario; Based on the normal physiological monitoring space of the scenario, anomaly identification is performed on the physiological monitoring feature sequence to obtain physiological anomaly identification results; The degree of abnormality of the physiological abnormality identification result is evaluated based on the normal physiological monitoring space of the scenario to obtain the physiological abnormality evaluation result. Based on the physiological abnormality identification results and the physiological abnormality evaluation results, the physiological monitoring feature sequence is optimized for abnormal attention coupling to obtain the physiological abnormality attention coupling matrix.

4. The method for monitoring and early warning of elderly care data through multi-device linkage as described in claim 1, characterized in that, Based on the user activity scenario tags, risk incident types are fitted to obtain N fitted risk incident types, including: Based on the user activity scenario tags, risk incident retrieval is performed to obtain a set of historical risk incidents matching the scenario. Based on the scenario, historical risk incident sets are matched to identify incident types and obtain an incident type sample set. Based on the historical risk accident set matched to the scenario, the support of each accident type sample in the accident type sample set is calculated to obtain the support of each accident type. Calculate the confidence score for each accident type sample based on the support score for each accident type to obtain the confidence score for each accident type; Based on the predetermined type confidence level, the accident type sample set is optimized and identified according to the confidence level of each accident type to generate the N fitted risk accident types.

5. The method for monitoring and early warning of elderly care data through multi-device linkage as described in claim 1, characterized in that, Based on the accompanying user monitoring data, the risk incident projection map is optimized by accompanying incident interference analysis to obtain an optimized risk projection map, including: Based on the accompanying user monitoring data, an accident prediction is performed on the real-time accompanying user to obtain the accompanying accident prediction result. Based on the accompanying accident prediction results, the risk accident projection map is subjected to accident triggering interference analysis to obtain the accident triggering interference analysis results. Based on the accompanying accident prediction results, the accident propagation interference analysis is performed on the risk accident projection map to obtain the accident propagation interference analysis results; Based on the accompanying accident prediction results, the risk accident projection map is subjected to accident cumulative interference analysis to obtain the accident cumulative interference analysis results; Based on the accident triggering interference analysis results, the accident propagation interference analysis results, and the accident accumulation interference analysis results, the risk accident projection map is revised to obtain the optimized risk projection map.

6. The method for monitoring and early warning of elderly care data through multi-device linkage as described in claim 5, characterized in that, Based on the predicted results of the accompanying accidents, an accompanying early warning signal is generated.

7. An elderly care data monitoring and early warning system based on multi-device linkage, characterized in that, The steps for implementing the multi-device linkage elderly care data monitoring and early warning method according to any one of claims 1 to 6 include: The scene tag generation module is used to conduct joint monitoring of home-based elderly care users through a multi-device monitoring array, obtain elderly care monitoring dataset, and generate user activity scene tags based on the elderly care monitoring dataset to identify activity scene features. The matrix building module is used to perform multi-dimensional anomaly analysis and attention optimization on the elderly care monitoring dataset based on the user activity scene labels, and to build a monitoring anomaly attention coupling matrix. The risk incident type fitting module is used to fit risk incident types based on the user activity scenario tags to obtain N fitted risk incident types, where N is a positive integer greater than 1. The risk incident simulation module is used to perform multi-dimensional risk incident simulation on the monitoring anomaly attention coupling matrix based on the N fitted risk incident types and according to the dynamic perturbation distillation mechanism, so as to obtain a risk incident simulation map. The monitoring and early warning signal generation module is used to read the real-time monitoring data of the accompanying users, and perform accompanying accident interference analysis and optimization on the risk accident projection map based on the accompanying user monitoring data to obtain the risk projection optimization map, and simultaneously generate elderly care monitoring and early warning signals.

8. An electronic device, characterized in that, include: At least one processor; A memory that is communicatively connected to the at least one processor; The memory stores instructions that can be executed by the at least one processor, which are executed by the at least one processor to enable the at least one processor to perform the steps of the multi-device linkage elderly care data monitoring and early warning method according to any one of claims 1 to 6.

Images