Safety monitoring and evaluation method and system based on multidimensional data

Abstract

The present invention relates to the technical field of multimodal data processing and intelligent health management, and in particular to a safety monitoring and assessment method and system based on multidimensional data. The method comprises: mapping multimodal feature data into health status vectors, and combining time series analysis to generate health status deviations and change trends, identifying health risks and generating behavioral intervention paths; the system expands scarce scenario data by generating an adversarial network, verifies the effectiveness of the path in a virtual simulation environment, and optimizes the parameters and associated weights of key nodes in the path; the dynamic collection of user feedback data further supports the real-time adjustment of the path and model; ultimately, the present invention realizes dynamic monitoring of health status, accurate assessment and personalized behavioral intervention, and has strong adaptability and foresight.

Description

Technical Field

[0001] The present invention relates to the technical field of multimodal data processing and intelligent health management, and in particular to a safety monitoring and evaluation method and system based on multidimensional data. Background Art

[0002] The health status of the elderly is complex, often influenced by multiple dimensions including physiological changes, behavioral patterns, and environmental factors. Traditional safety monitoring technologies struggle to fully capture the complex relationships between these multimodal factors, and even more so to dynamically assess changing health trends. Existing technology (Chinese invention patent, publication number: CN118379173B, title: A method for comprehensive safety monitoring based on digital elderly care) primarily relies on dynamic Bayesian networks (DBNs) to analyze the relationship between behavioral and physiological data and optimize behavioral recommendations by mapping weights. However, existing technology suffers from the following drawbacks:

[0003] While existing technologies combine behavioral and physiological data for analysis, they lack comprehensive consideration of environmental factors and long-term cumulative effects, limiting the accuracy and comprehensiveness of health assessments and behavioral interventions. Existing technologies tend to prioritize behavioral data based on impact weights, but are unable to dynamically capture the complex causal relationships underlying changes in health status across multiple dimensions, resulting in insufficient predictive and forward-looking capabilities. Model parameters in existing technologies primarily rely on static training results and lack the ability to dynamically adjust based on real-time feedback, making them difficult to adapt to individual differences and changing scenarios. Summary of the Invention

[0004] To address the numerous issues with the aforementioned existing technologies, the present invention provides a multidimensional data-based safety monitoring and assessment method and system. By enhancing the causal chain model, the present invention maps multimodal feature data into health status vectors, dynamically capturing the causal relationships between physiological, behavioral, and environmental factors. Combining generative adversarial networks and virtual simulation technology, the system generates personalized behavioral intervention pathways and dynamically adjusts model weights and pathway parameters based on real-time user feedback data. This invention enables precise monitoring of health status and proactive intervention, providing a scientific and personalized health management solution.

[0005] A safety monitoring and assessment method based on multidimensional data includes the following steps:

[0006] Collect multimodal data through distributed sensor networks, including physiological data, behavioral data, environmental data, and social interaction data, preprocess and prioritize the collected data, and generate optimized sampling data and event cluster data;

[0007] Based on optimized sampling data and event cluster data, multimodal feature data is extracted and fused to generate cross-modal feature data; a causal chain model is constructed by combining the causal inference network to generate short-term causal chain data and long-term causal chain data, and an enhanced causal chain model is generated through an optimization algorithm;

[0008] Utilize enhanced causal chain models and cross-modal feature data to assess health status and generate health status vectors and health status deviation data; generate health risk path data, health trend prediction data, and trend risk assessment data through path optimization and time series modeling;

[0009] Based on health risk path data and health trend prediction data, behavioral intervention paths are generated, and the effectiveness of the intervention paths is verified by combining simulation and optimization techniques; user feedback data is collected, and the key node weights of the behavioral intervention path and the associated parameters of the causal chain model are dynamically adjusted to generate optimized behavioral intervention path data and an updated causal chain model.

[0010] Preferably, the step of extracting and fusing multimodal feature data includes: extracting heart rate fluctuation characteristics and blood pressure change characteristics in physiological feature data, gait stability characteristics and activity frequency characteristics in behavioral feature data, and temperature and humidity fluctuation characteristics and light intensity change characteristics in environmental feature data based on optimized sampling data and event cluster data; and fusing the above feature data using a tensor decomposition method to generate cross-modal feature data.

[0011] Preferably, the tensor decomposition method includes: dynamically adjusting the correlation weights between feature data by decomposing the feature matrix; specifically, constructing a third-order tensor for physiological feature data, behavioral feature data, and environmental feature data, extracting key feature dimensions through singular value decomposition, and dynamically adjusting weight parameters to optimize the fusion effect and generate cross-modal feature data.

[0012] Preferably, the step of constructing a causal chain model includes: mapping cross-modal feature data into causal nodes using a causal inference network; generating short-term causal chain data based on time series analysis, and generating long-term causal chain data by combining the cumulative effects of multiple events; optimizing the correlation relationship and global weight of nodes in the causal chain through a nonlinear collaborative inference algorithm to generate an enhanced causal chain model.

[0013] Preferably, the nonlinear collaborative inference algorithm includes: applying a dynamic weight allocation strategy to adjust the node connection strength of short-term causal chain data according to the characteristic distribution of causal nodes; correcting the cumulative effect of long-term causal chain data through a recursive optimization mechanism, and using a multi-objective optimization method to calculate the global influence of nodes in the causal chain to generate a dynamically enhanced causal chain model.

[0014] Preferably, the step of evaluating the health status includes: constructing a health status mapping model based on the enhanced causal chain model and cross-modal feature data to generate a health status vector; generating health status deviation data by measuring the difference between the health status vector and the ideal health state, and generating health status change trend data in combination with historical health status change data.

[0015] Preferably, the step of generating health status deviation data includes: combining the current health status vector with the historical health status vector, using time series modeling to analyze the change trend, and generating a health status deviation curve; calculating the state deviation change rate of key nodes to generate health status deviation data.

[0016] Preferably, the step of generating a behavioral intervention path includes: generating preliminary behavioral intervention path data based on health risk path data and health trend prediction data in combination with a generative adversarial game optimization algorithm; verifying the improvement effect of each node in the preliminary behavioral intervention path on the health status through virtual simulation technology, adjusting the node parameters and connection weights in the path, and generating a behavioral intervention path.

[0017] Preferably, the step of collecting user feedback data includes: collecting feedback data after the user executes the behavioral intervention path, including behavior completion rate, subjective feelings and health status change data; combining the generative adversarial network to expand scarce scenario data, dynamically adjusting the key node parameters of the behavioral intervention path and the associated weights of the causal chain model, and generating optimized behavioral intervention path data and an updated causal chain model.

[0018] A system for implementing the multi-dimensional data-based safety monitoring and assessment method, comprising:

[0019] A data acquisition module for collecting physiological data, behavioral data, environmental data, and social interaction data through a distributed sensor network;

[0020] The data processing module is used to pre-process and prioritize the collected multimodal data to generate optimized sampling data and event cluster data;

[0021] The feature extraction and fusion module is used to extract physiological feature data, behavioral feature data, and environmental feature data based on optimized sampling data and event cluster data, and fuse the above feature data through tensor decomposition method to generate cross-modal feature data;

[0022] The causal inference and chain building module is used to combine the causal inference network, map cross-modal feature data into causal nodes, generate short-term causal chain data and long-term causal chain data, and generate an enhanced causal chain model through optimization algorithms;

[0023] The health status assessment module is used to evaluate health status using an enhanced causal chain model and cross-modal feature data, generating health status vectors and health status deviation data; and to generate health risk path data, health trend prediction data, and trend risk assessment data through path optimization and time series modeling.

[0024] The behavioral intervention path generation module is used to generate preliminary behavioral intervention path data based on health risk path data and health trend prediction data by generating adversarial game optimization algorithms, and verify the effectiveness of the path by combining virtual simulation technology to generate behavioral intervention paths;

[0025] The feedback and optimization module is used to collect user feedback data, dynamically adjust the key node weights of the behavior intervention path and the associated parameters of the causal chain model, and generate optimized behavior intervention path data and an updated causal chain model.

[0026] Compared with the prior art, the advantages and beneficial effects of the present invention are:

[0027] This invention achieves accurate modeling and dynamic optimization of the multi-dimensional causal relationship of health status by enhancing the causal chain model technology;

[0028] This paper uses a generative adversarial network to expand scarce scene data, thereby improving the robustness and adaptability of the model under data scarcity conditions.

[0029] This invention uses dynamic programming and virtual simulation technology to achieve scientific optimization and verification of behavioral intervention paths, ensuring the effectiveness and personalization of intervention measures.

[0030] The present invention achieves real-time optimization and enhances the adaptability to complex changes in health status by dynamically adjusting behavioral intervention paths and causal chain weights. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 Schematic diagram of the process of the present invention;

[0032] Figure 2 Schematic diagram of the construction process of the enhanced causal chain model in the present invention;

[0033] Figure 3 Schematic diagram of the generation and optimization process of the behavior intervention path in the present invention;

[0034] Figure 4 Schematic diagram of path optimization driven by dynamic feedback in the present invention;

[0035] Figure 5 It is a structural block diagram of the system of the present invention. DETAILED DESCRIPTION

[0036] Hereinafter, embodiments of the present disclosure will be described with reference to the accompanying drawings. However, it should be understood that these descriptions are merely exemplary and are not intended to limit the scope of the present disclosure. In the following detailed description, for ease of explanation, many specific details are set forth to provide a comprehensive understanding of the embodiments of the present disclosure.

[0037] like Figure 1 As shown, a safety monitoring and assessment method based on multidimensional data includes the following steps:

[0038] Collect multimodal data through distributed sensor networks, including physiological data, behavioral data, environmental data, and social interaction data, preprocess and prioritize the collected data, and generate optimized sampling data and event cluster data;

[0039] This invention utilizes a distributed sensor network to collect multimodal data, providing data support for monitoring and assessing the health status of the elderly. The distributed sensor network is composed of various types of sensors, including physiological, behavioral, environmental, and social interaction sensors, installed on devices worn by the elderly, their living environment, and interactive terminals. Through network-based collaborative data collection, it can cover the main data sources of the elderly's daily lives, laying the foundation for multidimensional health analysis.

[0040] The collection of multimodal data is achieved through the following equipment and principles:

[0041] Physiological data collection devices, such as smart bracelets or wearable blood pressure monitors, record heart rate, blood oxygen levels, blood pressure, and other indicators in real time. This data is obtained through optical detection principles of sensors (such as PPG and ECG). For example, heart rate monitoring is based on changes in light absorption caused by blood flow.

[0042] Behavioral data collection devices, such as gait monitors, use inertial sensors (such as accelerometers and gyroscopes) to monitor gait smoothness, activity frequency, and other behavioral characteristics. These sensors can identify abnormal behavior, such as falls or prolonged inactivity.

[0043] Environmental data collection devices, such as temperature and humidity sensors and light detectors, monitor the temperature, humidity, and light intensity of the living environment. By analyzing this environmental data, it can be inferred whether there are potential risks in the elderly's environment.

[0044] Social interaction data collection equipment: such as voice assistants or home interactive terminals, used to record data such as call frequency and visitor records to assist in assessing the social activity of the elderly.

[0045] Collected data is often affected by noise, equipment errors, and other factors. To ensure data reliability and consistency, raw data preprocessing is required. Data from different sources is aligned using a unified timestamp to ensure temporal consistency. Noise reduction algorithms tailored to the characteristics of different modal data are employed. For example, a Kalman filter algorithm is used for heart rate and blood pressure data to smooth random fluctuations in physiological data. A median filter algorithm is used for gait data to eliminate errors in abnormal behavior data.

[0046] In order to improve the efficiency and accuracy of data analysis, it is necessary to prioritize the preprocessed data.

[0047] Data is marked as abnormal using a threshold-based method. For example, a heart rate exceeding 120 bpm or a sudden cessation of gait are considered abnormal events, and the relevant data is marked. The DBSCAN algorithm is used to perform density clustering on the features of the marked event data. This algorithm calculates the local density of data points to identify highly correlated abnormal data clusters and generate event cluster data.

[0048] The aforementioned data collection, preprocessing, and priority screening process significantly improves the accuracy and analysis efficiency of elderly health data. Noise reduction and time synchronization eliminate environmental interference and data bias, providing high-quality data for subsequent analysis. Priority screening retains only high-priority data with strong relevance to health monitoring, preventing non-critical data from interfering with subsequent calculations. Real-time labeling of abnormal event data and generation of event clusters support rapid health status assessments, enabling, for example, timely detection of potential risk events such as abnormal heart rates and falls.

[0049] Example: A 75-year-old man wears a smart bracelet and a gait monitor, and his living environment is equipped with temperature and humidity sensors. Through a distributed sensor network, the following process is achieved:

[0050] Data Collection: The smart bracelet collected heart rate and blood pressure data, recording a current heart rate of 135 bpm (high) and a blood pressure of 160 / 100 mmHg (high). The gait monitor recorded a sudden pause in the elderly person's gait and a significant tilt. Environmental sensors recorded an indoor temperature of 28°C and a humidity of 70%, indicating no abnormal environmental risk.

[0051] Data preprocessing: Heart rate data was processed through a Kalman filter to remove random jitter and confirm that abnormal heart rate lasted for more than 2 minutes. Gait data was processed through a median filter to identify posture tilt trends that coincided with gait cessation.

[0052] Prioritization screening and event cluster generation: Based on a threshold method, we flagged heart rates exceeding 120 bpm as abnormal events, and gait cessation exceeding 10 seconds as behavioral abnormalities. We applied the DBSCAN algorithm to cluster these flagged events, confirming that abnormal heart rate and gait cessation belonged to the same event cluster and generating event cluster data.

[0053] The system analyzes the possibility of a fall and, based on event cluster data, sends real-time warnings to caregivers, including timestamps and key feature values of heart rate and gait abnormalities. The data is also recorded for subsequent health assessments and behavioral analysis.

[0054] like Figure 2 As shown, based on the optimized sampling data and event cluster data, multimodal feature data is extracted and fused to generate cross-modal feature data; a causal chain model is constructed by combining the causal inference network to generate short-term causal chain data and long-term causal chain data, and an enhanced causal chain model is generated through the optimization algorithm;

[0055] In the elderly care scenario of this invention, multimodal data (such as physiological, behavioral, environmental, and social data) can provide comprehensive information support for health monitoring and risk assessment. However, due to the heterogeneity of these data in feature space and differences in time scale, feature extraction, fusion, and causal chain modeling are required to capture their inherent correlations and underlying patterns.

[0056] Based on optimized sampling data and event cluster data, feature extraction is performed on data from different modalities. This includes features such as heart rate fluctuations and blood pressure changes in physiological data, gait stability and activity frequency in behavioral data, and temperature, humidity fluctuations, and light intensity changes in environmental data. After extraction, these multimodal feature data are fused using tensor decomposition to construct a high-dimensional feature space and generate cross-modal feature data. Tensor decomposition captures the correlations between data from different modalities by decomposing the feature matrix.

[0057] Based on cross-modal feature data, a causal chain model is constructed using a causal inference network (CIN). The CIN generates short-term causal chain data through time series analysis and generates long-term causal chain data by combining the long-term cumulative effects of multiple events. The specific steps are as follows:

[0058] Generation of short-term causal chain data: Dynamically modeling time series in cross-modal feature data to analyze the causal relationships between key events. Quantifying the causal impact of event A on event B using Granger causality analysis.

[0059] Generation of long-term causal chain data: Based on the time cumulative effects of multiple events, a long-term causal chain is constructed, and the total causal strength of the long-term effect is calculated through the cumulative weight function;

[0060] Enhanced causal chain model optimization: Optimize the relationships and weights of nodes in the causal chain through a nonlinear collaborative inference algorithm. Nonlinear collaborative inference uses a multi-objective optimization method to dynamically adjust the weight balance between short-term and long-term causal chains, ensuring that the causal chain model can simultaneously reflect immediate and cumulative effects.

[0061] Preferably, the step of extracting and fusing multimodal feature data includes:

[0062] Based on optimized sampling data and event cluster data, heart rate fluctuation characteristics and blood pressure change characteristics in physiological feature data, gait stability characteristics and activity frequency characteristics in behavioral feature data, and temperature and humidity fluctuation characteristics and light intensity change characteristics in environmental feature data are extracted; the above feature data are fused using the tensor decomposition method to generate cross-modal feature data.

[0063] In the elderly care scenario, the multimodal characteristics of physiological, behavioral and environmental data lead to differences and complex interactions between different modal data. For example, fluctuations in heart rate may be directly related to gait stability, while changes in ambient temperature and humidity may indirectly affect behavioral patterns through long-term cumulative effects. Therefore, the purpose of extracting and fusing multimodal feature data is to convert these heterogeneous data into a unified high-dimensional feature representation to provide support for subsequent causal analysis and health status assessment. The present invention achieves structured unification of multidimensional data by extracting features based on optimized sampling data and event cluster data, and using tensor decomposition methods for multimodal data fusion.

[0064] Based on the optimized sampling data and event cluster data, feature extraction is performed on physiological data, behavioral data and environmental data respectively.

[0065] Physiological feature data extraction: Extract heart rate fluctuation characteristics and blood pressure change characteristics from heart rate and blood pressure data through time series analysis. Heart rate fluctuation characteristics reflect short-term abnormalities in the cardiovascular state of the elderly. For example, the standard deviation of the fluctuation amplitude can be used to quantify the degree of fluctuation:

[0066] ,

[0067] in, Indicates the characteristic value of heart rate fluctuation; Indicates the Heart rate value at the moment; Indicates the mean heart rate; Indicates the time series length of the heart rate data. Blood pressure change characteristics identify key points of long-term and short-term blood pressure changes through trend analysis, such as peak points and downward trends.

[0068] Behavioral feature data extraction is based on gait stability and activity frequency analysis. Gait stability is determined by analyzing the rate of change of acceleration and angular velocity. For example, the standard deviation of the rate of change of acceleration can be used to quantify the degree of instability. Activity frequency features are calculated by counting the number of movements within a fixed time period, generating characteristic data related to the daily activity patterns of the elderly.

[0069] Environmental feature data extraction: extracting temperature and humidity fluctuation characteristics and light intensity change characteristics from environmental data. Temperature and humidity fluctuation characteristics are calculated based on the magnitude of changes in the time series, for example, by using a sliding window method to segment data trends. Light intensity change characteristics are generated by identifying the rising and falling cycles of light and are used to infer the correlation between activity time and ambient light.

[0070] The core of multimodal data fusion is to map the extracted features into a unified high-dimensional feature space to construct cross-modal feature data. The present invention adopts tensor decomposition method to perform data fusion. The specific steps are as follows:

[0071] Constructing feature tensors ,in:

[0072] : Physiological characteristic dimension (such as heart rate fluctuation characteristics, blood pressure change characteristics);

[0073] : Behavioral characteristic dimensions (such as gait stability characteristics, activity frequency characteristics);

[0074] : Environmental characteristic dimensions (such as temperature and humidity fluctuation characteristics, light intensity change characteristics).

[0075] Decompose the tensor and extract the key feature dimensions:

[0076] ,

[0077] in, Represents a multimodal feature tensor; represents the rank of tensor decomposition; Indicates the The weight of each feature component; 、 、 Represents the component vectors of physiological, behavioral and environmental characteristics respectively. By dynamically adjusting the rank of the tensor decomposition , which can control the complexity of fused data and balance the correlation between different modalities. The decomposed features are reconstructed into high-dimensional feature vectors to generate cross-modal feature data, ensuring that the data of different modalities can reflect their relevance and importance under a unified representation.

[0078] Converting heterogeneous physiological, behavioral, and environmental feature data into a unified high-dimensional feature representation facilitates subsequent causal analysis and health status assessment. Fusion of multimodal data using tensor decomposition not only preserves feature associations within a single modality but also captures potential interactions between modalities, such as the indirect impact of insufficient ambient lighting on gait stability. Highly correlated cross-modal feature data can support more accurate health risk assessments, for example, by analyzing the combined characteristics of heart rate fluctuations and activity frequency to identify fall risks earlier.

[0079] In this example, an elderly person was outdoors in winter, and the sensor recorded the following data:

[0080] Data collection, heart rate recording showed large fluctuations in heart rate over a short period of time, with a standard deviation of 12 bpm; gait monitoring records showed a decrease in activity frequency and unstable gait was detected; environmental data recorded a temperature of 10°C, a humidity of 80%, and light intensity below the standard threshold.

[0081] Feature extraction, physiological feature data: the heart rate fluctuation feature extraction value is 12, and blood pressure changes show a downward trend in systolic blood pressure; behavioral feature data: the gait stability feature value is 0.6 (based on the angular velocity change rate), and the activity frequency feature is 5 times per minute (lower than the daily activity level); environmental feature data: the temperature and humidity fluctuation feature value is medium to high, and the light intensity change feature shows a continuous low-light environment during the activity period.

[0082] Data fusion, tensor construction : Contains heart rate fluctuation characteristics and blood pressure change characteristics; Contains gait stability characteristics and activity frequency characteristics; Contains temperature and humidity fluctuation characteristics and light intensity change characteristics.

[0083] Tensor decomposition extracts key dimensions, and after weight adjustment, cross-modal feature data is generated. The fused cross-modal feature data reveals a strong correlation between low-light environments and gait instability. Combined with heart rate fluctuations, this further suggests that activities in low light may increase the risk of falls for elderly individuals. The system uses this risk for subsequent health status assessment and behavioral intervention pathway development.

[0084] Preferably, the tensor decomposition method includes: dynamically adjusting the correlation weights between feature data by decomposing the feature matrix; specifically, constructing a third-order tensor for physiological feature data, behavioral feature data, and environmental feature data, extracting key feature dimensions through singular value decomposition, and dynamically adjusting weight parameters to optimize the fusion effect and generate cross-modal feature data.

[0085] In elderly care settings, monitoring and assessing health status requires integrating multimodal data (such as physiological, behavioral, and environmental data). Due to the diverse sources of this data, its characteristics and dimensions vary significantly. For example, heart rate fluctuations and gait stability reflect short-term changes, while ambient temperature and humidity have long-term effects. Therefore, using tensor decomposition to fuse multimodal feature data allows for dynamic adjustment of the relevance weights of different modal features, ensuring the scientific and accurate nature of feature fusion.

[0086] Construct a third-order tensor based on the extracted multimodal feature data , to represent the multidimensional relationship between physiological, behavioral and environmental characteristics. Each element of Indicates the Physiological characteristics, Behavioral characteristics and The interaction value between the environmental features.

[0087] The constructed third-order tensor Perform singular value decomposition (SVD) to extract key feature dimensions. The decomposition formula is: By selecting The component with the largest singular value can capture the main cross-modal feature associations while reducing the feature dimension and avoiding redundant calculations.

[0088] In order to further optimize the feature fusion effect, the decomposed weight parameters Make dynamic adjustments based on the following:

[0089] Dynamically balance weights based on the proportion of physiological, behavioral, and environmental characteristics in a specific scenario. For example, in fall risk assessment, behavioral characteristics may be more important than environmental characteristics.

[0090] Adjust the feature weight decay rate according to the temporal correlation strength of the feature data. The dynamic adjustment formula is:

[0091] ,

[0092] in, represents the adjusted weight parameter; represents the original weight parameter; Represents the adjustment factor, which is calculated based on the importance and time correlation of the modal features.

[0093] The adjusted tensor decomposition results are then used to reconstruct the decomposed weights and component vectors into high-dimensional feature vectors, generating cross-modal feature data. This data uniformly represents the multidimensional associations between physiological, behavioral, and environmental characteristics, supporting subsequent causal analysis and health assessment.

[0094] Tensor decomposition methods can capture deep interactions between multimodal data, such as the combined effects of heart rate fluctuations, gait instability, and changes in light intensity, supporting comprehensive health assessments. Dynamic adjustment of weight parameters optimizes feature fusion based on different scenarios. For example, in fall risk assessment, emphasizing the weight of behavioral features helps to more accurately identify risk events. Singular value decomposition extracts key feature dimensions, reducing the interference of redundant features and improving computational efficiency while preserving key feature information.

[0095] In this example, an elderly person wears a health monitoring device during nighttime activities and collects the following data:

[0096] Characteristic Data: Physiological data: Heart rate fluctuation characteristic value of 15 bpm, blood pressure trend shows a decrease in systolic pressure. Behavioral data: Gait stability characteristic value of 0.4 (low), activity frequency of 3 times per minute (significant decrease). Environmental data: Temperature and humidity fluctuation characteristic value shows high humidity (80%), and light intensity is dim (less than 50 lux).

[0097] Constructing Tensors: Physiological Feature Dimensions :Heart rate fluctuation characteristics, blood pressure change characteristics; behavioral characteristics dimension :Gait stability characteristics, activity frequency characteristics; environmental characteristic dimensions :Temperature and humidity fluctuation characteristics, light intensity change characteristics. Construct tensor , where the element values represent the interaction effects of different feature combinations.

[0098] Extract the first two main feature components through SVD , .

[0099] Dynamically adjust the weight parameters to give behavioral features higher priority. The adjustment factor is , The adjusted weight is , .

[0100] The adjusted tensor decomposition results are used to reconstruct cross-modal feature vectors, ultimately generating unified feature data for causal analysis of fall risk and health status assessment. The system identifies strong correlations between excessive humidity, insufficient light, and unstable gait. It further assesses fall risk in the elderly by combining heart rate fluctuations, prompting caregivers to adjust ambient lighting and optimize activity schedules.

[0101] Preferably, the step of constructing a causal chain model includes: mapping cross-modal feature data into causal nodes using a causal inference network; generating short-term causal chain data based on time series analysis, and generating long-term causal chain data by combining the cumulative effects of multiple events; optimizing the correlation relationship and global weight of nodes in the causal chain through a nonlinear collaborative inference algorithm to generate an enhanced causal chain model.

[0102] In the elderly care scenario, the health status is affected by a combination of multimodal factors such as physiology, behavior, and environment, and there are complex causal relationships between these factors. In order to identify and evaluate these causal relationships, it is necessary to capture the interaction effects and cumulative effects of various factors through a causal chain model. The present invention maps cross-modal feature data into causal nodes through a causal inference network, generates short-term causal chain data based on time series analysis, and generates long-term causal chain data based on the cumulative effects of multiple events. Subsequently, the causal chain is optimized through a nonlinear collaborative inference algorithm to generate an enhanced causal chain model, which has both immediate response capabilities and long-term trend analysis capabilities.

[0103] Causal nodes are the basic units of the causal chain model, representing key events with causal relationships in feature data. Cross-modal feature data are mapped to causal nodes through the Causal Inference Network (CIN). The mapping rules are based on the conditional probability and correlation strength between feature data. For example, for the heart rate fluctuation feature (event A) and the gait stability feature (event B), by calculating the conditional probability and the total probability The difference between and is used to determine the causal strength:

[0104] ,

[0105] If the causal strength exceeds the set threshold , then connect event A and event B as causal nodes. In the elderly care scenario, causal nodes include: heart rate fluctuation (physiological node), unstable gait (behavioral node), and abnormal ambient temperature and humidity (environmental node).

[0106] Short-term causal chains are used to capture the immediate associations between health events based on time series analysis of cross-modal feature data. The specific process is as follows:

[0107] Time series modeling uses the timestamps of events to construct time series data.

[0108] Causal strength calculation, using the Granger causality analysis method, quantifies the causal influence of event A on event B. The calculation formula is:

[0109] ,

[0110] in, Indicates the Granger causal strength of event A on event B; represents the variance of event B under the condition of event A; represents the total variance of event B.

[0111] Short-term causal chain data records the causal nodes and their strengths within the current time window, for example:

[0112] Heart rate fluctuation → unstable gait, causal strength is 0.85.

[0113] Changes in temperature and humidity → heart rate fluctuations, the causal strength is 0.6.

[0114] Long-term causal chains capture the cumulative effects of multiple events and are particularly suitable for assessing the long-term impact of environmental factors. For example, persistent humidity abnormalities may lead to respiratory health problems. The specific process includes:

[0115] Cumulative causal calculation, through the time-weighted cumulative model, calculates the long-term causal strength:

[0116] ,

[0117] in, Indicates time the long-term causal strength of Indicates the The weight of each event node; Represents a decay function that represents the effect of time intervals on causal strength. It dynamically adjusts the node relationships in long-term causal chains, optimizing weights based on event frequency and cumulative strength.

[0118] There may be contradictions in the weight distribution and correlation between short-term and long-term causal chains, so it is necessary to optimize the causal chain through nonlinear collaborative inference algorithms. The optimization process includes:

[0119] Error calculation, calculate the correlation error of short-term causal chain and long-term causal chain respectively:

[0120] ,

[0121] in, Errors in short-term causal chains; Errors that represent long-term causal chains; 、 Represents the weight parameter, which is used to balance the short-term and long-term correlations.

[0122] Optimize the target, reduce the overall error through recursive iteration, update the causal node weights and chain associations, and generate an enhanced causal chain model.

[0123] Short-term causal chains can instantly capture the dynamic causal relationships between health events in older adults, such as the immediate correlation between heart rate fluctuations and gait abnormalities. Long-term causal chains effectively capture the long-term impact of environmental factors, such as the potential cumulative risk of humidity changes on the respiratory system. Nonlinear collaborative inference optimizes the causal chain model, enabling it to reflect both immediate relationships and analyze long-term trends, providing a scientific basis for health assessment and behavioral intervention.

[0124] In this example, an elderly person experienced increased heart rate fluctuations and an unstable gait during nighttime activities. Simultaneously, the environment was low light and high humidity. The system analyzed the causal relationships between these events using a causal chain model.

[0125] Causal node mapping:

[0126] Heart rate fluctuation (physiological node): Event A;

[0127] Unstable gait (behavior node): event B;

[0128] Insufficient ambient light (environment node): event C;

[0129] Abnormal ambient humidity (environmental node): Event D.

[0130] Short-term causal chain generation:

[0131] Heart rate fluctuation → unstable gait, causal strength is 0.8;

[0132] Low light → unstable gait, causal strength is 0.6.

[0133] Long-term causal chain generation:

[0134] Abnormal humidity → unstable gait, cumulative causal strength is 0.7.

[0135] Nonlinear collaborative optimization: After optimization, the weight of heart rate fluctuation on gait instability was adjusted to 0.85, and the weight of insufficient light was adjusted to 0.65.

[0136] The analysis results show that the risk of falls is caused by the combined action of multiple factors, among which heart rate fluctuations are immediate triggering factors and abnormal humidity is a cumulative risk factor. The optimized causal chain model supports the formulation of intervention strategies such as lighting enhancement and humidity control.

[0137] Preferably, the nonlinear collaborative inference algorithm includes: applying a dynamic weight allocation strategy to adjust the node connection strength of short-term causal chain data according to the characteristic distribution of causal nodes; correcting the cumulative effect of long-term causal chain data through a recursive optimization mechanism, and using a multi-objective optimization method to calculate the global influence of nodes in the causal chain to generate a dynamically enhanced causal chain model.

[0138] In elderly care health monitoring, causal chain models must analyze both the immediate relationships between short-term events and the cumulative effects of long-term events. The dynamic nature and complex interactions of these causal relationships require optimization through a nonlinear collaborative inference algorithm to enhance the accuracy and adaptability of causal chain models. This algorithm optimizes the connection strength of short-term causal chain nodes through a dynamic weight allocation strategy, corrects the cumulative effects of long-term causal chains using a recursive optimization mechanism, and calculates the global influence of causal nodes through a multi-objective optimization method, thereby generating a dynamically enhanced causal chain model.

[0139] The goal of the dynamic weight allocation strategy is to optimize the connection strength between nodes in the short-term causal chain based on the characteristic distribution of causal nodes, so that the model can more accurately reflect the immediate causal relationship.

[0140] Each node in the short-term causal chain data (e.g., heart rate fluctuation feature node A, gait instability feature node B) is assigned an initial weight based on the distribution of its feature value. Weight adjustment follows the following rules: highly correlated feature nodes (i.e., those with conditional probabilities significantly higher than the background probability) are assigned higher weights; the weights of low-correlation feature nodes are suppressed to avoid noise amplification.

[0141] Dynamic weight adjustment is achieved through the following formula:

[0142] ,

[0143] in, Indicates the connection strength between node A and node B; The conditional probability represents the immediate causal strength of node A on node B; Represents the correlation measure between the feature distributions of node A and node B. Through the above weight adjustment, the short-term causal chain can more accurately quantify the immediate causal relationship and ensure that the weights of key nodes are fully reflected in the model.

[0144] Long-term causal chains are used to analyze the cumulative effects of events, but these cumulative effects are uneven across different nodes. For example, the cumulative impact of abnormal ambient humidity on gait instability may be far greater than that of changes in light intensity. Therefore, a recursive optimization mechanism uses multiple rounds of iterations to correct for deviations in cumulative effects, ensuring the model's balance and global accuracy. Based on historical event data, the cumulative effects of nodes in the long-term causal chain are corrected using a weight decay function:

[0145] ,

[0146] in, represents the corrected cumulative causal strength between node A and node B; represents the original cumulative causal strength; Represents the time decay function, which is used to express the decreasing relationship between the cumulative effect of nodes over time. In each iteration, the system updates the cumulative effect of the nodes and verifies the new weight distribution. The optimization termination condition is that the overall error is lower than the set threshold .

[0147] The overall optimization of the causal chain model requires finding a balance between short-term immediate response and long-term trend analysis. Through multi-objective optimization methods, the global influence of each causal node is calculated and the global weight of the node is adjusted. The optimization objective expression is:

[0148] ,

[0149] in, Representation node Short-term causal errors; Representation node long-term causal errors; 、 The equilibrium parameters representing the short-term and long-term causal chains; Represents the total number of causal nodes.

[0150] Global influence calculation:

[0151] ,

[0152] in, Representation node global influence; 、 Represents the weight of a node in the short-term and long-term causal chains. Through multi-objective optimization, the global influence of the node is dynamically adjusted to generate a more balanced and accurate causal chain model.

[0153] A dynamic weight allocation strategy enables short-term causal chains to accurately capture the immediate causal relationships between key events, such as the direct impact of heart rate fluctuations on gait instability. A recursive optimization mechanism corrects for biases in long-term causal chains, enabling more accurate analysis of cumulative effects, such as the long-term impact of humidity abnormalities on respiratory health in the elderly. Through multi-objective optimization, the causal chain model achieves a dynamic balance between short-term immediate response and long-term trend prediction, providing more comprehensive data support for health status assessment and behavioral intervention.

[0154] In this example, an elderly person experiences an unstable gait and increased heart rate fluctuations during nighttime activities. Simultaneously, the ambient light is low and the humidity is high. The system optimizes the causal chain model using a nonlinear collaborative inference algorithm. The specific process is as follows:

[0155] In the short-term causal chain, the immediate causal strength of heart rate fluctuation (node A) on gait instability (node B) is 0.85, and the dynamic weight allocation strategy adjusts the connection strength to 0.9. The weight of insufficient ambient light (node C) on gait instability is adjusted to 0.65.

[0156] In the long-term causal chain, the original value of the cumulative causal strength of humidity anomaly (node D) on gait instability was 0.7, which was corrected to 0.75 after recursive optimization.

[0157] The equilibrium parameters of the short-term causal chain and the long-term causal chain are set as 、 The global influence of the unstable gait node is calculated as: .

[0158] The optimized causal chain model shows that abnormal humidity has a significant long-term association with gait instability, while heart rate fluctuations are the immediate trigger for falls. The system recommends optimizing the nighttime activity environment (increasing lighting brightness and reducing humidity) and monitoring heart rate fluctuations.

[0159] Utilize enhanced causal chain models and cross-modal feature data to assess health status and generate health status vectors and health status deviation data; generate health risk path data, health trend prediction data, and trend risk assessment data through path optimization and time series modeling;

[0160] The health status of the elderly is the result of the integration of multimodal data, such as physiological data (e.g., heart rate fluctuations), behavioral data (e.g., gait stability), and environmental data (e.g., light intensity and humidity changes). This paper assesses health status by enhancing causal chain models and cross-modal feature data. Furthermore, it combines path optimization and time series modeling to generate health risk path data, health trend prediction data, and trend risk assessment data, providing a scientific basis for dynamic health management and risk intervention.

[0161] The core of health status assessment is to combine the augmented causal chain model with cross-modal feature data to construct a health status mapping model, generating a health status vector and health status deviation data. The health status mapping model utilizes the weights and associations of nodes in the augmented causal chain model to transform cross-modal feature data into a multidimensional health status vector. Each component of the health status vector corresponds to a health indicator (such as heart rate fluctuation or gait stability). Health status deviation measures the difference between the current health status and the ideal health status. Health status deviation data is generated by calculating the Euclidean distance between the health status vector and the ideal health status vector.

[0162] Health risk pathways are generated between nodes in the augmented causal chain model using a path optimization algorithm. These pathways describe the optimal intervention path from the current health state to the desired health state. Path optimization, based on dynamic programming, aims to minimize health state deviations and generate the optimal path from the current node to the desired health state. The total weight of the pathway represents the required intervention intensity.

[0163] Health trend prediction is based on time series modeling, capturing changing health status trends and incorporating trend characteristics into risk assessment. Time series modeling generates health trend prediction data by predicting the future values of health status vectors. Common methods include temporal convolutional networks (TCNs), whose core approach is to learn temporal characteristics from historical status data and predict future changes. Health trend prediction data is combined with weights in the enhanced causal chain model to generate trend risk assessment data, which is used to quantify health risks over a period of time.

[0164] Preferably, the step of evaluating the health status includes: constructing a health status mapping model based on the enhanced causal chain model and cross-modal feature data to generate a health status vector; generating health status deviation data by measuring the difference between the health status vector and the ideal health state, and generating health status change trend data in combination with historical health status change data.

[0165] In the elderly care scenario, health status is affected by a combination of physiological, behavioral, and environmental multimodal factors. The key to assessing health status is to build a model that can integrate multimodal data and compare it with the ideal health state. The present invention builds a health status mapping model by enhancing the causal chain model and cross-modal feature data, generates a health status vector to quantify the current health status, and combines health status deviation data and historical health status change data to generate health status change trend data. This method can capture changes in health status in real time and provide accurate risk assessment and intervention basis.

[0166] The health state mapping model uses the node weights and association relationships in the enhanced causal chain model to map cross-modal feature data into a multidimensional health state vector. Each component of represents a health indicator (such as heart rate fluctuation characteristics, gait stability characteristics).

[0167] Input data to enhance the causal chain model to provide the weight of the causal node , cross-modal feature data provides feature values Mapping formula, the calculation formula of the health state vector is:

[0168] ,

[0169] in, Represents the health status vector Component, indicating the health indicators; Indicates the The causal node The weight of each health indicator; Indicates the The characteristic value of each causal node; Represents the total number of causal nodes. Through the above formula, the health status mapping model integrates multimodal data into a unified health status vector, reflecting the current overall health status of the elderly.

[0170] Health status deviation is used to quantify the difference between the current health status and the ideal health status, and is a basic indicator for health risk assessment. Ideal health status definition: Based on health standards or health benchmark data of elderly individuals, an ideal health status vector is constructed. The health state deviation is calculated by the Euclidean distance between the health state vector and the ideal health state vector:

[0171] ,

[0172] in, Indicates health status deviation; Represents the current health status vector Quantity The ideal health state vector Quantity Represents the dimension of the health status vector. The greater the health status deviation, the greater the gap between the current health status and the ideal health status, and the more timely intervention measures are needed.

[0173] Health status change trend data is used to reflect the historical trajectory of health status changes and future trends. It is generated by combining historical health status data with time series modeling. Through time series analysis, the changing trend of the health status vector (such as increase, decrease, or fluctuation) is extracted. Linear fitting or polynomial fitting is performed on the historical health status vector to generate trend characteristic values, such as: the rate of increase: indicating the speed of health status improvement; the amplitude of fluctuation: reflecting the instability of health status. The current health status vector is compared with the historical mean, and the calculation expression is:

[0174] ,

[0175] in, Indicates the Trend change values of health indicators; Represents the current health status vector Quantity Represents the historical average of health status. Trend change data can provide a basis for subsequent health intervention strategies, such as recommending that older adults increase their exercise frequency or adjust their living environment.

[0176] Health status vectors and health status deviation data provide a comprehensive and quantitative description of the health status of older adults and their differences from their ideal state. Health status trend data captures the dynamic changes in health indicators, providing reliable support for early warning and intervention. Combining health status deviation data with trend change data enables the system to develop personalized health management plans, such as strengthening monitoring of abnormal heart rate fluctuations or optimizing gait stability.

[0177] In this example, an elderly person exhibits abnormal heart rate fluctuations and an unstable gait during a monitoring period, while the ambient light is low. The system generates the following results by evaluating their health status:

[0178] Health state mapping: cross-modal feature data, heart rate fluctuation feature value is 15 bpm, gait stability feature value is 0.6, and light intensity feature value is 40 lux; enhanced causal chain model weight, heart rate fluctuation weight is 0.7, gait stability weight is 0.6, and light intensity weight is 0.5.

[0179] Health state vector calculation: ;

[0180] Ideal health state vector: ;

[0181] Health status deviation: ;

[0182] Health status change trend, historical average: heart rate fluctuation average is 12 bpm, gait stability average is 0.7; trend change value: It shows that the heart rate fluctuation is abnormally increased and the gait stability is reduced.

[0183] The system identifies that abnormal heart rate fluctuations are the main factor in the deterioration of health status. Unstable gait and insufficient ambient light may further increase the risk. It is recommended that caregivers increase the frequency of heart rate fluctuation monitoring and improve indoor lighting conditions.

[0184] Preferably, the step of generating health status deviation data includes: combining the current health status vector with the historical health status vector, using time series modeling to analyze the change trend, and generating a health status deviation curve; calculating the state deviation change rate of key nodes to generate health status deviation data.

[0185] In elderly care scenarios, health status deviation data is used to quantify the difference between the elderly's current health status and historical health trends, providing in-depth analysis and early warning of health changes. By combining the current health status vector with the historical health status vector, time series modeling is used to generate a health status deviation curve. The rate of change of status deviation at key nodes is further calculated to support health risk assessment and intervention strategies.

[0186] The health status deviation curve reflects the trend of the health status of the elderly over time and can capture the dynamic characteristics of multidimensional health status. and the current health status vector As the input of time series modeling, the trend of health status changes over time is extracted. The time series model uses a sliding window method to fit and predict continuous health status data.

[0187] ,

[0188] in, Represents the predicted health state vector at the next time point; Represents the historical health status vector at the current time point and before; A mapping function representing a time series model, such as a Temporal Convolutional Network (TCN) or an LSTM.

[0189] By the current health status vector and historical predicted health status vector The Euclidean distance of , generates the health status deviation curve:

[0190] ,

[0191] in, Indicates a time point Health status deviation value; represents the Euclidean distance.

[0192] The health status deviation change rate is used to measure the extent of health status changes at key nodes (such as heart rate fluctuations and gait instability) and is a key indicator of health risk. By analyzing the gradient of the health status deviation curve, key nodes with significant changes are identified, such as the time points when heart rate fluctuations increase significantly or gait stability decreases significantly. The health status deviation value at the key node is first-order differencing to calculate the state deviation change rate:

[0193] ,

[0194] in, Indicates key nodes The rate of change of state deviation; Indicates a time point Health status deviation value; Represents the time interval between adjacent key nodes. By calculating the state deviation change rate, the fluctuation intensity of the health state can be quantified, further supporting the quantitative analysis of health risks.

[0195] The health status deviation curve can reflect the dynamic changes in the health status of the elderly in real time, such as increased heart rate fluctuations or decreased gait stability, providing a basis for early risk warning.

[0196] The state deviation rate of change quantifies the rate of change in health status, enabling rapid identification of high-risk events, such as increased heart rate fluctuations caused by acceleration. By analyzing health state deviation data, the system can develop targeted interventions, such as optimizing ambient light intensity when insufficient lighting causes gait instability.

[0197] In this example, an elderly person exhibited increased heart rate fluctuations and unstable gait during the monitoring period, while also experiencing high humidity. The system generated health status deviation data to perform the following analysis:

[0198] Health state deviation curve generation: Historical health state vector: , corresponding to heart rate fluctuation, gait stability, and light intensity respectively; current health status vector: ; Predicting health status through time series modeling: ; Health status deviation curve: .

[0199] State deviation change rate calculation: Key node 1 (heart rate fluctuation): The state deviation increases from 2.0 to 4.69, and the change rate is: Key Node 2 (Gait Stability): The state deviation decreases from 0.2 to 0.1, and the rate of change is: .

[0200] The system identified a significant increase in heart rate fluctuations and a decrease in gait stability. Analysis of the rate of change indicated that heart rate fluctuations were high-risk, while changes in gait stability were more moderate. The system recommended strengthening real-time monitoring of heart rate fluctuations and gradually optimizing gait stability.

[0201] Based on health risk path data and health trend prediction data, behavioral intervention paths are generated, and the effectiveness of the intervention paths is verified by combining simulation and optimization techniques; user feedback data is collected, and the key node weights of the behavioral intervention path and the associated parameters of the causal chain model are dynamically adjusted to generate optimized behavioral intervention path data and an updated causal chain model.

[0202] Behavioral intervention pathways are generated using a generative adversarial game optimization algorithm based on health risk pathway data and health trend prediction data. Input data includes: health risk pathway data, which represents the optimal adjustment path from current health status to desired health status and its cost; and health trend prediction data, which describes the trend of health status over time, indicating risk points and optimization directions.

[0203] The behavioral intervention path generation process uses dynamic programming to find the least-cost adjustment path along the health risk path. Output: The generated preliminary behavioral intervention path includes multiple key nodes, each corresponding to an intervention action (e.g., increasing walking time, optimizing indoor lighting).

[0204] The generated preliminary behavioral intervention pathway is validated through simulation, and key nodes within the pathway are optimized using reinforcement learning algorithms. Each node in the intervention pathway is simulated and tested in a virtual simulation environment. For example, the effect of increasing indoor light levels on gait stability and heart rate fluctuations is simulated. The node intervention effect is calculated based on the simulation results. Based on the simulation results, the reinforcement learning algorithm optimizes the parameters of key nodes in the pathway, including adjusting the connection weights between nodes and the priority of actions. This optimized pathway is more adaptable and effective.

[0205] Optimizing behavioral intervention pathways requires dynamically adjusting key node weights and associated parameters of the causal chain model based on user feedback. The system collects real-time data on user completion rates, subjective experiences (such as comfort), and changes in health status after executing the intervention pathway, serving as the basis for dynamic adjustments. Node weights and associated parameters of the causal chain model are updated based on this feedback data. The resulting dynamically adjusted behavioral intervention pathway data and updated causal chain model are more responsive to the individual needs of older adults, enhancing intervention effectiveness.

[0206] Preferably, Figure 3As shown, the steps of generating a behavioral intervention path include: generating preliminary behavioral intervention path data based on health risk path data and health trend prediction data in combination with a generative adversarial game optimization algorithm; verifying the improvement effect of each node in the preliminary behavioral intervention path on the health status through virtual simulation technology, adjusting the node parameters and connection weights in the path, and generating a behavioral intervention path.

[0207] In elderly care settings, behavioral intervention pathway generation aims to plan effective intervention measures to improve the health of the elderly by analyzing health risk pathway data and health trend forecast data. This paper uses a generative adversarial game optimization algorithm to generate preliminary behavioral intervention pathway data and uses virtual simulation technology to verify the health improvement effects of key nodes in the pathway. Subsequently, by adjusting node parameters and connection weights in the pathway, an optimized behavioral intervention pathway is generated.

[0208] The generative adversarial game optimization algorithm dynamically plans the initial version of the intervention path through interactive analysis of health risk path data and health trend prediction data.

[0209] Input data: health risk path data, indicating the adjustment path from the current health status to the ideal health status, including risk nodes and adjustment costs; health trend prediction data, describing the future change trend of health status and identifying nodes that require priority intervention.

[0210] The generative adversarial game optimization algorithm is based on the principle of multi-objective optimization, and considers both health improvement effects and implementation costs in intervention path planning. The generation of preliminary behavioral intervention paths follows the following optimization objectives:

[0211] ,

[0212] in, Representation node To Node The weight of , which represents the intervention effect; Represents a slave node To Node implementation costs; Represents a node in the intervention path health benefits; Indicates the total number of nodes in the path.

[0213] Output data: The preliminary behavioral intervention path consists of several key nodes, each of which corresponds to a specific intervention measure (such as increasing activity time, improving light intensity, etc.).

[0214] The initially generated behavioral intervention path needs to be verified in a virtual simulation environment to evaluate the improvement effect of each node in the path on health status and optimize the parameters and weights of the path. The simulation environment simulates the actual situation of the elderly executing the intervention path, including the execution effect and the dynamic changes in health status. For example, in the node where the light intensity is increased, the simulation environment calculates the impact of the light change on gait stability and heart rate fluctuations. The health improvement effect of each node is quantitatively evaluated and its comprehensive effect value is calculated:

[0215] ,

[0216] in, Representation node The health improvement effect value; Representation node Health characteristics The weight of Representation characteristics The improvement value.

[0217] Based on the simulation results, the parameters and connection weights of the nodes in the pathway are adjusted to optimize the adaptability and implementation effectiveness of the intervention pathway. Node parameter optimization includes adjusting the intervention intensity within a node, for example, increasing the daily activity time from 5 minutes to 8 minutes. Based on the simulation results, node priorities are modified to ensure that efficient interventions are prioritized. Reinforcement learning algorithms are used to update the connection weights between nodes, balancing the health benefits of the intervention pathway with its implementation costs.

[0218] Initial behavioral intervention pathways are dynamically generated using a game optimization algorithm, fully considering both health improvement outcomes and implementation costs to develop scientific intervention plans for seniors. Virtual simulation technology is used to evaluate the health improvement outcomes of each node in the pathway, ensuring the feasibility and practical adaptability of the intervention pathway. Node parameters and weights are adjusted based on simulation validation results to better align the intervention pathway with the health needs of seniors and enhance intervention effectiveness.

[0219] In this example, an elderly person experienced unstable gait and abnormal heart rate fluctuations due to reduced indoor activity time and insufficient ambient light. The system generated a behavioral intervention path and performed simulation and optimization to complete the following steps:

[0220] Behavioral Intervention Path Generation: Input Data: The health risk path indicates that increasing activity time and optimizing light intensity are the top priorities; the health trend forecast indicates that heart rate fluctuations may further increase. Output Path: Node 1: Increase daily activity time to 5 minutes; Node 2: Increase indoor light brightness to 300 lux.

[0221] Virtual simulation verification: Simulation testing showed that increasing activity time at node 1 improved gait stability by 0.4 and heart rate fluctuation by 0.3. At node 2, increasing light brightness improved gait stability by 0.2 and heart rate fluctuation by 0.1. Overall effect: Node 1: E1 = 0.7; Node 2: E2 = 0.3.

[0222] Path optimization and adjustment: Node 1's activity time was adjusted to 8 minutes per day; Node 2's priority was increased to allow for earlier lighting improvements. The optimized behavioral intervention pathway demonstrated significant improvements in heart rate fluctuations and gait stability, significantly increasing the amount of time older adults spent indoors, and effectively reducing health risks.

[0223] Preferably, Figure 4 As shown, the step of collecting user feedback data includes: collecting feedback data after the user executes the behavioral intervention path, including behavior completion rate, subjective feelings and health status change data; combining the generative adversarial network to expand scarce scenario data, dynamically adjusting the key node parameters of the behavioral intervention path and the associated weights of the causal chain model, and generating optimized behavioral intervention path data and an updated causal chain model.

[0224] The purpose of collecting user feedback data is to evaluate the effectiveness of the user's implementation of the intervention path and its actual correlation with health status. The collection content includes:

[0225] Behavior completion rate: This records the ratio of the intervention behavior actually completed by the user to the intended plan. For example, if a user plans to walk for 10 minutes but actually completes 7 minutes, the behavior completion rate is 70%. Subjective experience: This records user feedback on comfort, fatigue, or other subjective evaluations after the behavior is performed. Health status change data: This records changes in multimodal data recorded by sensors, such as the improvement in gait stability characteristic values and the reduction in heart rate fluctuations.

[0226] After standardizing the collected feedback data, calculate the specific impact value of the feedback on each node in the intervention path. For example:

[0227] ,

[0228] in, Representation node Feedback impact value; Indicates the impact value of behavior completion rate; Indicates the subjective impact value; Indicates the impact value of health status change; , , Represents the weight parameter, which is dynamically adjusted according to the scenario.

[0229] Because actual feedback data may be scarce for certain scenarios (such as extreme environments and low-frequency events), the system utilizes a generative adversarial network (GAN) to augment the data and improve the model's robustness and adaptability. GANs generate high-fidelity simulated data through an adversarial generation mechanism. The generator generates new possible feedback scenarios based on known feedback data, while the discriminator verifies the authenticity of the generated data and optimizes the generator's output. For example, an augmented scenario might include the impact of varying humidity levels on gait stability. The augmented simulated data is integrated with actual feedback data to enrich the diversity of the feedback data and make the dynamic adjustment process more widely applicable.

[0230] Based on actual feedback and expanded data, the system dynamically adjusts the parameters of key nodes in the behavioral intervention path and the associated weights of the causal chain model. For nodes with low completion rates, the intervention intensity is reduced or the intervention method is adjusted, such as reducing daily walking time or switching to low-intensity exercise. For nodes with high feedback satisfaction, the priority is increased and the intervention intensity is increased. For example, optimizing the brightness of indoor lights has a significant effect on gait stability, and the brightness can be increased to a more optimal level. The associated weights of the nodes in the causal chain are reallocated based on the feedback data:

[0231] ,

[0232] in, Represents the updated node association weight; represents the original node association weight; Indicates the node feedback impact value; Represents the adjustment coefficient, which is used to control the impact of feedback data on weight updates.

[0233] By collecting and analyzing user feedback, behavioral intervention pathways can be dynamically adjusted based on actual implementation outcomes, ensuring personalized and adaptable intervention plans. Generative adversarial networks expand feedback data for low-frequency scenarios, enhancing the system's resilience and robustness in scarce scenarios. Dynamically adjusting the correlation weights within the causal chain model improves the model's adaptability to complex health status changes, providing more precise support for subsequent health assessments and intervention strategy development.

[0234] In this example, an elderly person actually completed some of the intervention behaviors while following the behavioral intervention path, and feedback data showed that their health status had improved. The system collected user feedback data and dynamically adjusted and optimized the intervention path: Planned path: Walk for 10 minutes every day and increase indoor light brightness to 300 lux. Feedback data:

[0235] Behavior completion rate: walking completion rate is 60%, light brightness adjustment completion rate is 90%;

[0236] Subjective feeling: The subjective fatigue score of walking behavior was 3 (out of 5, high fatigue);

[0237] Health status changes: Gait stability feature improvement value was 0.3, and heart rate fluctuation feature improvement value was 0.4.

[0238] Combined with the feedback, the feedback influence value of the walking node is calculated to be 0.25, and the feedback influence value of the light node is 0.4.

[0239] Generative Adversarial Network Data Augmentation: Expand the scene to generate data on the effects of different light levels (250 lux and 350 lux) on gait stability; and generate data on the improvement of heart rate fluctuations with different walking durations (5 minutes and 8 minutes). Output: The expanded data shows that an 8-minute walk improves heart rate fluctuations by 0.45, exceeding the current performance.

[0240] Dynamic Adjustment and Optimization: Path adjustments were made, with walking time adjusted to 8 minutes per day and increased in priority. Light nodes maintained at 300 lux, but with additional monitoring of ambient brightness stability. Causal chain model updates were made: the weight of the walking node was increased to 0.8, and the weight of the light node was updated to 0.7. The optimized behavioral intervention pathway better aligns with users' actual health needs. The system predicts further improvement in heart rate fluctuations and gait stability over subsequent monitoring cycles.

[0241] like Figure 5 As shown, a system for implementing the multi-dimensional data-based security monitoring and assessment method includes:

[0242] A data acquisition module for collecting physiological data, behavioral data, environmental data, and social interaction data through a distributed sensor network;

[0243] The data processing module is used to pre-process and prioritize the collected multimodal data to generate optimized sampling data and event cluster data;

[0244] The feature extraction and fusion module is used to extract physiological feature data, behavioral feature data, and environmental feature data based on optimized sampling data and event cluster data, and fuse the above feature data through tensor decomposition method to generate cross-modal feature data;

[0245] The causal inference and chain building module is used to combine the causal inference network, map cross-modal feature data into causal nodes, generate short-term causal chain data and long-term causal chain data, and generate an enhanced causal chain model through optimization algorithms;

[0246] The health status assessment module is used to evaluate health status using an enhanced causal chain model and cross-modal feature data, generating health status vectors and health status deviation data; and to generate health risk path data, health trend prediction data, and trend risk assessment data through path optimization and time series modeling.

[0247] The behavioral intervention path generation module is used to generate preliminary behavioral intervention path data based on health risk path data and health trend prediction data by generating adversarial game optimization algorithms, and verify the effectiveness of the path by combining virtual simulation technology to generate behavioral intervention paths;

[0248] The feedback and optimization module is used to collect user feedback data, dynamically adjust the key node weights of the behavior intervention path and the associated parameters of the causal chain model, and generate optimized behavior intervention path data and an updated causal chain model.

[0249] The system uses a distributed sensor network to collect physiological data (such as heart rate and blood pressure), behavioral data (such as gait stability and activity frequency), environmental data (such as temperature, humidity, and light intensity), and social interaction data from the daily lives of elderly people. This multimodal data is preprocessed and prioritized to generate optimized sampling data and event cluster data, providing high-quality data input for subsequent analysis.

[0250] The feature extraction and fusion module uses tensor decomposition to unify the key features in multimodal data and generate cross-modal feature data. This data is mapped into causal nodes by the causal inference network. By constructing short-term and long-term causal chains, an enhanced causal chain model is generated to capture the complex causal relationships between physiological, behavioral, and environmental factors.

[0251] Through the health status assessment module, the system generates health status vectors and health status deviation data, and combines path optimization and time series modeling to generate health risk paths and trend prediction data, providing support for real-time monitoring and forward-looking analysis of health status.

[0252] During the behavioral intervention phase, the system generates intervention pathways based on health risk pathway data and trend prediction data, and verifies their effectiveness through virtual simulation technology. Simultaneously, the system collects user feedback and continuously optimizes the pathways and models by dynamically adjusting the weights of key nodes in the intervention pathways and the associated parameters of the causal chain model.

[0253] The above are merely embodiments of the present application and are not intended to limit the present application. For those skilled in the art, the present application may have various modifications and variations. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present application should be included within the scope of the claims of the present application.

Claims

1. A safety monitoring and assessment method based on multidimensional data, characterized in that: The following steps are involved: Collect multimodal data through distributed sensor networks, including physiological data, behavioral data, environmental data, and social interaction data, preprocess and prioritize the collected data, and generate optimized sampling data and event cluster data; Based on optimized sampling data and event cluster data, multimodal feature data is extracted and fused to generate cross-modal feature data; a causal chain model is constructed by combining the causal inference network to generate short-term causal chain data and long-term causal chain data, and an enhanced causal chain model is generated through an optimization algorithm; The steps of constructing the causal chain model include: mapping cross-modal feature data into causal nodes using a causal inference network; generating short-term causal chain data based on time series analysis, and generating long-term causal chain data by combining the cumulative effects of multiple events; optimizing the correlation relationship and global weight of nodes in the causal chain through a nonlinear collaborative inference algorithm to generate an enhanced causal chain model; The nonlinear collaborative inference algorithm includes: applying a dynamic weight allocation strategy to adjust the node connection strength of short-term causal chain data based on the characteristic distribution of causal nodes; correcting the cumulative effect of long-term causal chain data through a recursive optimization mechanism; and using a multi-objective optimization method to calculate the global influence of nodes in the causal chain to generate a dynamically enhanced causal chain model; Utilize enhanced causal chain models and cross-modal feature data to assess health status and generate health status vectors and health status deviation data; generate health risk path data, health trend prediction data, and trend risk assessment data through path optimization and time series modeling; Based on health risk path data and health trend prediction data, behavioral intervention paths are generated, and the effectiveness of the intervention paths is verified by combining simulation and optimization techniques; user feedback data is collected, and the key node weights of the behavioral intervention path and the associated parameters of the causal chain model are dynamically adjusted to generate optimized behavioral intervention path data and an updated causal chain model.

2. The multi-dimensional data-based safety monitoring and assessment method according to claim 1, characterized in that: The step of extracting and fusing multimodal feature data includes: Based on optimized sampling data and event cluster data, heart rate fluctuation characteristics and blood pressure change characteristics in physiological feature data, gait stability characteristics and activity frequency characteristics in behavioral feature data, and temperature and humidity fluctuation characteristics and light intensity change characteristics in environmental feature data are extracted; the above feature data are fused using the tensor decomposition method to generate cross-modal feature data.

3. The multi-dimensional data-based safety monitoring and assessment method according to claim 2, characterized in that: The tensor decomposition method includes: The correlation weights between feature data are dynamically adjusted by decomposing the feature matrix; specifically, a third-order tensor is constructed for physiological feature data, behavioral feature data, and environmental feature data, key feature dimensions are extracted through singular value decomposition, and weight parameters are dynamically adjusted to optimize the fusion effect and generate cross-modal feature data.

4. The multi-dimensional data-based safety monitoring and assessment method according to claim 1, characterized in that: The steps of assessing health status include: A health status mapping model is constructed based on the enhanced causal chain model and cross-modal feature data to generate a health status vector; health status deviation data is generated by measuring the difference between the health status vector and the ideal health state, and health status change trend data is generated by combining historical health status change data.

5. The multi-dimensional data-based safety monitoring and assessment method according to claim 4, characterized in that: The step of generating health status deviation data comprises: By combining the current health state vector with the historical health state vector, time series modeling is used to analyze the change trend and generate a health state deviation curve; the state deviation change rate of key nodes is calculated to generate health state deviation data.

6. The multi-dimensional data-based safety monitoring and assessment method according to claim 1, characterized in that: The steps of generating a behavioral intervention path include: Based on health risk path data and health trend prediction data, preliminary behavioral intervention path data is generated in combination with the generative adversarial game optimization algorithm. The improvement effect of each node in the preliminary behavioral intervention path on the health status is verified through virtual simulation technology, and the node parameters and connection weights in the path are adjusted to generate a behavioral intervention path.

7. The multi-dimensional data-based safety monitoring and assessment method according to claim 1, characterized in that: The step of collecting user feedback data includes: Collect feedback data after users execute the behavioral intervention path, including behavior completion rate, subjective feelings and health status change data; combine with the generative adversarial network to expand scarce scenario data, dynamically adjust the key node parameters of the behavioral intervention path and the associated weights of the causal chain model, and generate optimized behavioral intervention path data and updated causal chain model.

8. A system for implementing the multidimensional data-based safety monitoring and assessment method according to any one of claims 1 to 7, characterized in that: include: A data acquisition module for collecting physiological data, behavioral data, environmental data, and social interaction data through a distributed sensor network; The data processing module is used to pre-process and prioritize the collected multimodal data to generate optimized sampling data and event cluster data; The feature extraction and fusion module is used to extract physiological feature data, behavioral feature data, and environmental feature data based on optimized sampling data and event cluster data, and fuse the above feature data through tensor decomposition method to generate cross-modal feature data; The causal inference and chain building module is used to combine the causal inference network, map cross-modal feature data into causal nodes, generate short-term causal chain data and long-term causal chain data, and generate an enhanced causal chain model through optimization algorithms; A health status assessment module is used to assess the health status using an enhanced causal chain model and cross-modal feature data, generating a health status vector and health status deviation data; Generate health risk pathway data, health trend prediction data, and trend risk assessment data through pathway optimization and time series modeling; The behavioral intervention path generation module is used to generate preliminary behavioral intervention path data based on health risk path data and health trend prediction data by generating adversarial game optimization algorithms, and verify the effectiveness of the path by combining virtual simulation technology to generate behavioral intervention paths; The feedback and optimization module is used to collect user feedback data, dynamically adjust the key node weights of the behavior intervention path and the associated parameters of the causal chain model, and generate optimized behavior intervention path data and an updated causal chain model.

Images