Home-based elderly care health monitoring management system and method based on Internet of Things technology

Abstract

The present application discloses a home-based elderly care health monitoring management system and method based on the Internet of Things technology, which relates to the field of intelligent monitoring. It uses monitoring equipment to monitor and collect the physiological data of the elderly in real time, such as heart rate, blood pressure, blood oxygen and body temperature, and sends these physiological data to a cloud server for processing through a wireless network, so as to use data processing and analysis algorithms based on the Internet of Things and artificial intelligence technology on the cloud server to perform time-series collaborative analysis of these physiological data, so as to judge the health status and safety risks of the elderly. In this way, the physiological state of the elderly at home can be monitored in real time, and alarm information can be sent to relevant personnel in a timely manner when safety risks are detected, thereby providing necessary health monitoring management support for the elderly's home-based elderly care life.

Description

Technical Field

[0001] The present application relates to the field of intelligent monitoring, and more specifically, to a home-based elderly care health monitoring management system and method based on Internet of Things technology. Background Art

[0002] Home-based elderly care health monitoring and management refers to real-time monitoring and management of the health status and safety risks of the elderly by monitoring their physiological data, so as to detect abnormal situations in time and take corresponding measures to ensure the health and safety of the elderly.

[0003] However, traditional home-based elderly care health monitoring and management systems can only monitor the physiological data of the elderly, such as heart rate, blood pressure, blood oxygen, body temperature, etc., and issue alarms when thresholds are exceeded. This approach often has limited processing and analysis capabilities for these physiological data, and can only provide simple data statistics and alarm functions, but cannot conduct in-depth analysis and mining of large amounts of physiological data to extract more meaningful health information. In addition, traditional systems often lack personalized services and cannot provide customized care recommendations and guidance based on the specific needs and health conditions of the elderly, which results in limited adaptability and practicality of the system.

[0004] Therefore, a home-based elderly care health monitoring management system and method based on Internet of Things technology is desired. Summary of the invention

[0005] In order to solve the above technical problems, this application is proposed.

[0006] According to one aspect of the present application, a home-based elderly care health monitoring and management system based on the Internet of Things technology is provided, which includes:

[0007] A physiological data acquisition module, used to acquire a time series of physiological data of the monitored elderly person collected by a monitoring device worn by the monitored elderly person, wherein the physiological data includes heart rate value, blood pressure value, blood oxygen value and body temperature value;

[0008] A physiological data transmission module, used for transmitting the physiological data time series to a cloud server via a wireless network;

[0009] A physiological data time series division module is used to divide the physiological data time series of the monitored elderly person based on a predetermined time scale in the cloud server to obtain a sequence of local time series of physiological data;

[0010] A physiological data multi-parameter local time series arrangement module, used for arranging the sequence of the physiological data local time series according to the time dimension and the physiological parameter sample dimension to obtain a sequence of the physiological data local time series input matrix;

[0011] A physiological data multi-parameter local time series correlation feature extraction module, used to extract features from the sequence of the physiological data local time series input matrix respectively through a physiological multi-parameter local time series correlation feature extractor based on a deep neural network model to obtain a sequence of physiological data local time series correlation feature vectors;

[0012] A physiological data global time series feature expression module, used for fusing the sequence of the physiological data local time series associated feature vectors to obtain a physiological data global time series feature vector;

[0013] A physiological pattern semantic difference analysis module, used for respectively calculating the physiological pattern semantic difference coefficient of each physiological data local time series association feature vector in the sequence of the physiological data local time series association feature vector relative to the physiological data global time series feature vector to obtain a physiological pattern semantic difference feature vector as a physiological pattern semantic difference feature;

[0014] The physiological state abnormality detection and warning module is used to determine whether there is an abnormality in the physiological state of the monitored elderly person based on the semantic difference characteristics of the physiological pattern, and determine whether to generate a physiological abnormality warning signal.

[0015] In the above-mentioned home-based elderly care health monitoring and management system based on Internet of Things technology, the deep neural network model is a convolutional neural network model.

[0016] In the above-mentioned home-based elderly care health monitoring and management system based on Internet of Things technology, the physiological data global time series feature expression module is used to: use the projection fusion layer to fuse the sequence of the local time series associated feature vectors of the physiological data to obtain the global time series feature vector of the physiological data.

[0017] In the above-mentioned home-based elderly care health monitoring and management system based on the Internet of Things technology, the physiological data global time series feature expression module is used to: use the projection layer to fuse the sequence of the physiological data local time series associated feature vectors with the following formula to obtain the physiological data global time series feature vector;

[0018] Wherein, the formula is:

[0019] ;

[0020] in is a sequence of local temporal correlation feature vectors of the physiological data, is the global time series feature vector of the physiological data, Represents the projection layer.

[0021] In the above-mentioned home-based elderly care health monitoring and management system based on the Internet of Things technology, the physiological pattern semantic difference analysis module includes:

[0022] A physiological data global time series feature optimization unit, used for performing feature optimization on the physiological data global time series feature vector to obtain an optimized physiological data global time series feature vector;

[0023] The physiological pattern semantic difference measurement unit is used to respectively calculate the physiological pattern semantic difference coefficient of each physiological data local time series association feature vector in the sequence of the physiological data local time series association feature vector relative to the optimized physiological data global time series feature vector to obtain the physiological pattern semantic difference feature vector.

[0024] In the above-mentioned home-based elderly care health monitoring and management system based on the Internet of Things technology, the physiological data global time series feature optimization unit includes:

[0025] A feature fusion correction subunit, used for fusing the sequence of the global time series feature vector of the physiological data and the local time series associated feature vector of the physiological data with the feature value as the granularity to obtain a fused correction feature vector;

[0026] The cascade fusion subunit is used to cascade fuse the fused correction feature vector with the global time series feature vector of physiological data to obtain the optimized global time series feature vector of physiological data.

[0027] In the above-mentioned home-based elderly care health monitoring and management system based on the Internet of Things technology, the physiological pattern semantic difference measurement unit is used to: calculate the physiological pattern semantic difference coefficient of each physiological data local time series association feature vector in the sequence of the physiological data local time series association feature vector relative to the optimized physiological data global time series feature vector by the following semantic association formula to obtain the physiological pattern semantic difference feature vector;

[0028] Wherein, the semantic association formula is:

[0029] ;

[0030] in, represents the feature value of each position of any local time series correlation feature vector of physiological data in the sequence of local time series correlation feature vectors of physiological data, represents the eigenvalues ​​of each position of the optimized global time series eigenvector of physiological data, is the scale of the local temporal correlation feature vector of the physiological data and the optimized global temporal feature vector of the physiological data, is the feature value of each position in the semantic difference feature vector of the physiological pattern.

[0031] In the above-mentioned home-based elderly care health monitoring and management system based on the Internet of Things technology, the physiological state abnormality detection and early warning module includes:

[0032] A physiological state abnormality detection unit, used for passing the physiological pattern semantic difference feature vector through a classifier to obtain a classification result, wherein the classification result is used to indicate whether the physiological state of the monitored elderly person is abnormal;

[0033] The warning signal generating unit is used to determine whether to generate a physiological abnormality warning signal based on the classification result.

[0034] According to another aspect of the present application, a home-based elderly care health monitoring management method based on the Internet of Things technology is provided, which includes:

[0035] Acquire a time series of physiological data of the monitored elderly person collected by a monitoring device worn by the monitored elderly person, wherein the physiological data includes heart rate value, blood pressure value, blood oxygen value and body temperature value;

[0036] Transmitting the physiological data time series to a cloud server via a wireless network;

[0037] In the cloud server, the physiological data time series of the monitored elderly person is divided based on a predetermined time scale to obtain a sequence of local time series of the physiological data;

[0038] Arranging the sequence of the local time series of physiological data according to the time dimension and the physiological parameter sample dimension to obtain a sequence of the local time series input matrix of physiological data;

[0039] The sequence of the physiological data local time series input matrix is ​​subjected to feature extraction by a physiological multi-parameter local time series correlation feature extractor based on a deep neural network model to obtain a sequence of physiological data local time series correlation feature vectors;

[0040] fusing the sequence of local temporal correlation feature vectors of the physiological data to obtain a global temporal feature vector of the physiological data;

[0041] Respectively calculating the physiological pattern semantic difference coefficient of each physiological data local time series association feature vector in the sequence of the physiological data local time series association feature vector relative to the physiological data global time series feature vector to obtain a physiological pattern semantic difference feature vector as a physiological pattern semantic difference feature;

[0042] Based on the semantic difference characteristics of the physiological pattern, it is determined whether there is any abnormality in the physiological state of the monitored elderly person, and whether a physiological abnormality warning signal is generated.

[0043] Compared with the prior art, the present application provides a home-based elderly care health monitoring management system and method based on the Internet of Things technology, which uses monitoring equipment to monitor and collect the physiological data of the elderly in real time, such as heart rate, blood pressure, blood oxygen and body temperature, and sends these physiological data to the cloud server for processing through a wireless network, so as to use the data processing and analysis algorithm based on the Internet of Things and artificial intelligence technology on the cloud server to perform time-series collaborative analysis of these physiological data, so as to judge the health status and safety risks of the elderly. In this way, the physiological state of the elderly at home can be monitored in real time, and alarm information can be sent to relevant personnel in a timely manner when safety risks are detected, thereby providing necessary health monitoring management support for the elderly's home-based elderly care life. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] By describing the embodiments of the present application in more detail in conjunction with the accompanying drawings, the above and other purposes, features and advantages of the present application will become more apparent. The accompanying drawings are used to provide a further understanding of the embodiments of the present application and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the present application and do not constitute a limitation of the present application. In the accompanying drawings, the same reference numerals generally represent the same components or steps.

[0045] Figure 1 A block diagram of a home-based elderly care health monitoring and management system based on Internet of Things technology according to an embodiment of the present application;

[0046] Figure 2 This is a system architecture diagram of a home-based elderly care health monitoring and management system based on the Internet of Things technology according to an embodiment of the present application;

[0047] Figure 3 It is a block diagram of a physiological pattern semantic difference analysis module in a home-based elderly care health monitoring and management system based on the Internet of Things technology according to an embodiment of the present application;

[0048] Figure 4 It is a block diagram of a global time series feature optimization unit of physiological data in a home-based elderly care health monitoring and management system based on the Internet of Things technology according to an embodiment of the present application;

[0049] Figure 5 It is a block diagram of a physiological state abnormality detection and early warning module in a home-based elderly care health monitoring and management system based on the Internet of Things technology according to an embodiment of the present application;

[0050] Figure 6 This is a flowchart of a home-based elderly care health monitoring and management method based on Internet of Things technology according to an embodiment of the present application. DETAILED DESCRIPTION

[0051] Below, the exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the exemplary embodiments described here.

[0052] As shown in this application and claims, unless the context clearly indicates an exception, the words "a", "an", "an" and / or "the" do not refer to the singular and may also include the plural. Generally speaking, the terms "include" and "comprise" only indicate the inclusion of the steps and elements that have been clearly identified, and these steps and elements do not constitute an exclusive list. The method or device may also include other steps or elements.

[0053] Although the present application makes various references to certain modules in the system according to the embodiments of the present application, any number of different modules can be used and run on the user terminal and / or server. The modules are only illustrative, and different aspects of the system and method can use different modules.

[0054] Flowcharts are used in the present application to illustrate the operations performed by the system according to the embodiments of the present application. It should be understood that the preceding or following operations are not necessarily performed accurately in order. On the contrary, various steps may be processed in reverse order or simultaneously as required. Meanwhile, other operations may also be added to these processes, or a certain step or several steps of operations may be removed from these processes.

[0055] Below, the exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the exemplary embodiments described here.

[0056] Traditional home-based elderly care health monitoring and management systems can often only monitor the physiological data of the elderly, such as heart rate, blood pressure, blood oxygen, body temperature, etc., and alarm when exceeding the threshold. This method often has limited processing and analysis capabilities for these physiological data, and can only provide simple data statistics and alarm functions, and cannot conduct in-depth analysis and mining of a large amount of physiological data to extract more meaningful health information. In addition, traditional systems often lack personalized services and cannot provide customized care advice and guidance based on the specific needs and health conditions of the elderly, which leads to limited adaptability and practicality of the system. Therefore, a home-based elderly care health monitoring and management system based on the Internet of Things technology is desired.

[0057] It should be understood that the Internet of Things technology refers to a new technology that uses the Internet, sensors, communications and other technologies to achieve information exchange and intelligent control between objects. The Internet of Things technology has broad application prospects in the field of medical health, especially for the elderly living at home. The Internet of Things technology can provide effective health monitoring and management services to improve the quality of life and safety of the elderly. Therefore, the development of the Internet of Things technology provides a new solution for the health monitoring and management system for home-based elderly care.

[0058] In the technical solution of the present application, a home-based elderly care health monitoring and management system based on Internet of Things technology is proposed. Figure 1 This is a block diagram of a home-based elderly care health monitoring and management system based on Internet of Things technology according to an embodiment of the present application. Figure 2 : is a system architecture diagram of a home-based elderly care health monitoring and management system based on the Internet of Things technology according to an embodiment of the present application. Figure 1 and Figure 2 As shown, according to the embodiment of the present application, a home-based elderly care health monitoring and management system 300 based on the Internet of Things technology includes: a physiological data acquisition module 310, which is used to obtain the physiological data time series of the monitored elderly person collected by the monitoring device worn by the monitored elderly person, wherein the physiological data includes heart rate value, blood pressure value, blood oxygen value and body temperature value; a physiological data transmission module 320, which is used to transmit the physiological data time series to a cloud server through a wireless network; a physiological data time series division module 330, which is used to divide the physiological data time series of the monitored elderly person based on a predetermined time scale in the cloud server to obtain a sequence of local time series of physiological data; a physiological data multi-parameter local time series arrangement module 340, which is used to arrange the sequence of the local time series of physiological data according to the time dimension and the physiological parameter sample dimension to obtain a sequence of the local time series input matrix of physiological data; a physiological data multi-parameter local time series association feature extraction module Block 350 is used to perform feature extraction on the sequence of the local time series input matrix of the physiological data respectively through a physiological multi-parameter local time series association feature extractor based on a deep neural network model to obtain a sequence of local time series association feature vectors of physiological data; a physiological data global time series feature expression module 360 ​​is used to fuse the sequence of local time series association feature vectors of physiological data to obtain a global time series feature vector of physiological data; a physiological pattern semantic difference analysis module 370 is used to respectively calculate the physiological pattern semantic difference coefficient of each local time series association feature vector of physiological data in the sequence of local time series association feature vectors of physiological data relative to the global time series feature vector of physiological data to obtain a physiological pattern semantic difference feature vector as a physiological pattern semantic difference feature; a physiological state abnormality detection and warning module 380 is used to determine whether there is an abnormality in the physiological state of the monitored elderly person based on the physiological pattern semantic difference feature, and determine whether to generate a physiological abnormality warning signal.

[0059] In particular, the physiological data acquisition module 310 is used to acquire the time series of physiological data of the monitored elderly person acquired by the monitoring device worn by the monitored elderly person, wherein the physiological data includes heart rate value, blood pressure value, blood oxygen value and body temperature value. It should be understood that by acquiring the time series of physiological data of the monitored elderly person, the health status of the monitored elderly person can be monitored in real time.

[0060] In particular, the physiological data transmission module 320 is used to transmit the physiological data time series to the cloud server via a wireless network. That is, after the physiological data of the elderly is detected, the physiological data time series is transmitted to the cloud server via a wireless network to facilitate the processing and analysis of the physiological data to determine whether there is an abnormality.

[0061] In particular, the physiological data time series division module 330 is used to divide the physiological data time series of the monitored elderly based on a predetermined time scale in the cloud server to obtain a sequence of local time series of physiological data. Considering that the physiological data of the monitored elderly include the heart rate value, blood pressure value, blood oxygen value and body temperature value of the elderly, these data have a dynamic change law of time series in the time dimension, and these data also have volatility and correlation in time series. In other words, each physiological parameter in the physiological data will present different time series change characteristics and correlation relationships in different local time periods. Therefore, in order to better analyze and understand the coordinated change of the physiological data time series of the elderly, so as to extract more meaningful features, in the technical solution of the present application, the physiological data time series of the monitored elderly is further divided based on a predetermined time scale in the cloud server to obtain a sequence of local time series of physiological data. By dividing the physiological data time series by a predetermined time scale, the entire time series can be decomposed into multiple local time series, each local time series represents physiological data within a specific time range. In this way, the physiological data of the elderly can be observed and analyzed more carefully, so as to capture the changing trends and periodicity of the physiological data and further understand the health status of the elderly.

[0062] In particular, the physiological data multi-parameter local time series arrangement module 340 is used to arrange the sequence of the local time series of the physiological data according to the time dimension and the physiological parameter sample dimension to obtain the sequence of the local time series input matrix of the physiological data. Considering that in the local time series of the physiological data, each time point corresponds to a group of physiological parameter samples, such as heart rate value, blood pressure value, blood oxygen value and body temperature value. Therefore, in order to facilitate the subsequent more comprehensive and comprehensive time series collaborative correlation analysis of these physiological parameters, it is necessary to integrate the physiological data parameters in each local time series. Specifically, in the technical solution of the present application, the sequence of the local time series of the physiological data is arranged according to the time dimension and the physiological parameter sample dimension to obtain the sequence of the local time series input matrix of the physiological data.

[0063] In particular, the physiological data multi-parameter local temporal correlation feature extraction module 350 is used to extract features from the sequence of the physiological data local temporal input matrix respectively through a physiological multi-parameter local temporal correlation feature extractor based on a deep neural network model to obtain a sequence of physiological data local temporal correlation feature vectors. In particular, the deep neural network model is a convolutional neural network model. That is, in the technical solution of the present application, the sequence of the physiological data local temporal input matrix is ​​subjected to feature mining in a physiological multi-parameter local temporal correlation feature extractor based on a convolutional neural network model respectively, so as to extract the temporal collaborative correlation feature information between each physiological parameter in the physiological data in each time period respectively, thereby obtaining a sequence of physiological data local temporal correlation feature vectors. Specifically, each layer of the physiological multi-parameter local temporal association feature extractor based on the convolutional neural network model performs the following operations on the input data in the forward pass of the layer: convolution processing on the input data to obtain a convolution feature map; pooling based on the feature matrix on the convolution feature map to obtain a pooled feature map; and nonlinear activation on the pooled feature map to obtain an activation feature map; wherein the output of the last layer of the physiological multi-parameter local temporal association feature extractor based on the convolutional neural network model is a sequence of local temporal association feature vectors of the physiological data, and the input of the first layer of the physiological multi-parameter local temporal association feature extractor based on the convolutional neural network model is a sequence of local temporal input matrices of the physiological data.

[0064] It is worth noting that the Convolutional Neural Network (CNN) is a deep learning model that is particularly suitable for processing tasks with grid-structured data (such as images, audio, text, etc.). CNN extracts features by applying convolution operations on input data, and gradually extracts higher-level features through multiple convolutional layers and pooling layers. The following is the basic structure of the convolutional neural network model: Convolutional layer: The convolutional layer is the core component of CNN. It extracts features by applying convolution operations on input data. The convolution operation uses a convolution kernel (also called a filter) to slide on the input data and calculate the convolution result at each position. The convolution operation can capture the local relationship and pattern of the data and generate a feature map as output; Activation function: The output of the convolution layer is usually transformed nonlinearly through an activation function. Common activation functions include ReLU, Sigmoid, and Tanh. The activation function introduces nonlinear characteristics and increases the expressive power of the model; Pooling layer: The pooling layer is used to reduce the size of the feature map and extract more important features. Common pooling operations include maximum pooling and average pooling. Pooling can reduce the number of model parameters, reduce computational complexity, and improve the robustness of the model; Fully connected layer: The fully connected layer flattens the output of the pooling layer and connects it to the output layer. The neurons in the fully connected layer are connected to all neurons in the previous layer, and can learn higher-level feature representations. The last fully connected layer is usually used for classification tasks and outputs the model's prediction results.

[0065] In particular, the physiological data global temporal feature expression module 360 ​​is used to fuse the sequence of local temporal correlation feature vectors of the physiological data to obtain the global temporal feature vector of the physiological data. It should be understood that, since in the sequence of local temporal correlation feature vectors of the physiological data, each feature vector represents the temporal collaborative correlation feature information between various physiological parameters in the physiological data within a local time window, only considering the local temporal features between the physiological parameters may not be able to capture the global pattern and trend of the entire time series of the physiological data. Therefore, in order to integrate the local temporal correlation features of the physiological parameters to capture a more global and comprehensive feature representation, in the technical solution of the present application, a projection fusion layer is further used to fuse the sequence of local temporal correlation feature vectors of the physiological data to obtain the global temporal feature vector of the physiological data. Specifically, using a projection fusion layer to fuse the sequence of local temporal correlation feature vectors of the physiological data to obtain the global temporal feature vector of the physiological data includes: using a projection layer to fuse the sequence of local temporal correlation feature vectors of the physiological data with the following formula to obtain the global temporal feature vector of the physiological data; wherein the formula is:

[0066] ;

[0067] in is a sequence of local temporal correlation feature vectors of the physiological data, is the global time series feature vector of the physiological data, Represents the projection layer.

[0068] In particular, the physiological pattern semantic difference analysis module 370 is used to respectively calculate the physiological pattern semantic difference coefficient of each physiological data local time series association feature vector in the sequence of the physiological data local time series association feature vector relative to the physiological data global time series feature vector to obtain the physiological pattern semantic difference feature vector as the physiological pattern semantic difference feature. In particular, in a specific example of the present application, Figure 3 As shown, the physiological pattern semantic difference analysis module 370 includes: a physiological data global temporal feature optimization unit 371, which is used to perform feature optimization on the global temporal feature vector of the physiological data to obtain an optimized global temporal feature vector of the physiological data; and a physiological pattern semantic difference measurement unit 372, which is used to respectively calculate the physiological pattern semantic difference coefficients of each local temporal association feature vector of the physiological data in the sequence of local temporal association feature vectors of the physiological data relative to the optimized global temporal feature vector of the physiological data to obtain the physiological pattern semantic difference feature vector.

[0069] Specifically, the physiological data global time series feature optimization unit 371 is used to optimize the physiological data global time series feature vector to obtain an optimized physiological data global time series feature vector. In particular, in a specific example of the present application, Figure 4 As shown, the physiological data global time series feature optimization unit 371 includes: a feature fusion correction subunit 3711, used to fuse the sequence of the physiological data global time series feature vector and the physiological data local time series associated feature vector with the feature value as the granularity to obtain a fused correction feature vector; a cascade fusion subunit 3712, used to cascade fuse the fused correction feature vector with the physiological data global time series feature vector to obtain the optimized physiological data global time series feature vector.

[0070] More specifically, the feature fusion correction subunit 3711 is used to fuse the sequence of the global time series feature vector of the physiological data and the local time series correlation feature vector of the physiological data with the feature value as the granularity to obtain a fused correction feature vector. In particular, in the above technical scheme, each local time series association feature vector of physiological data in the sequence of local time series association feature vectors of physiological data expresses the time series-sample cross-dimensional association characteristics in the local time domain of the physiological data of the monitored elderly person determined by time scale division in the global time domain, and after using the projection fusion layer to fuse the sequence of local time series association feature vectors of physiological data, the global time series feature vector of physiological data obtained further expresses the time series-sample cross-dimensional association characteristics of the local time domain projection fusion in the global time domain. Therefore, if the local time domain-global time domain fusion expression effect of the global time series feature vector of physiological data can be improved, the calculation accuracy of the physiological pattern semantic difference coefficient of each local time series association feature vector of physiological data in the sequence of local time series association feature vectors of physiological data relative to the global time series feature vector of physiological data can be improved, thereby improving the accuracy of the classification result obtained by the classifier of the physiological pattern semantic difference feature vector. Therefore, in order to improve the local time domain-global time domain fusion expression effect of the global time series feature vector of the physiological data, the applicant of the present application optimizes the global time series feature vector of the physiological data by fusing the sequence of the local time series associated feature vector of the physiological data. However, considering that the global time series feature vector of the physiological data has variability in the global-local time series association calculation dimension relative to the sequence of the local time series associated feature vector of the physiological data, in order to improve the fusion effect of the global time series feature vector of the physiological data and the sequence of the local time series associated feature vector of the physiological data under the category judgment based on the classifier, the applicant of the present application preferably fuses the sequence of the global time series feature vector of the physiological data and the sequence of the local time series associated feature vector of the physiological data with the feature value as the granularity to obtain a fused correction feature vector, which is specifically expressed as:

[0071] ;

[0072] in, is a concatenated feature vector obtained by concatenating the sequence of local temporal correlation feature vectors of the physiological data, is the global time series feature vector of the physiological data after the linear interpolation, and are the eigenvalues ​​of the cascade feature vector and the global temporal feature vector of the physiological data, respectively, and Respectively represent the square of the first norm of the feature vector and the square root of the second norm of the feature vector, and the cascade feature vector and the global time series feature vector of the physiological data have the same feature vector length ,and is the weight hyperparameter, represents exponential operation, It means adding by position. It means to make a difference by position. is the eigenvalue of the fused correction feature vector. Here, the above-mentioned optimization fusion is based on the correspondence under the eigenvalue granularity to divide the serialized fusion representation of the sequence of the global temporal feature vector of the physiological data and the sequence of the local temporal association feature vector of the physiological data into foreground manifold and background manifold based on the vector scale, so as to stack the dynamic eigenvalue channelized association of the sequence of the global temporal feature vector of the physiological data and the sequence of the local temporal association feature vector of the physiological data under the feature corresponding channel supermanifold aggregation mechanism, thereby marking the feature semantic information of the changes between the sequences of the global temporal feature vector of the physiological data and the sequence of the local temporal association feature vector of the physiological data, and realizing the quasi-fully connected stacking fusion according to the variability of the semantic content between the sequences of the global temporal feature vector of the physiological data and the sequence of the local temporal association feature vector of the physiological data in different computing dimensions. Then, the fused correction feature vector is cascade-fused with the global temporal feature vector of the physiological data to obtain an optimized global temporal feature vector of the physiological data, so as to improve the fusion effect of the sequence of the global temporal feature vector of the physiological data and the local temporal association feature vector of the physiological data, thereby improving the calculation accuracy of the physiological pattern semantic difference coefficient of each local temporal association feature vector of the physiological data in the sequence of the local temporal association feature vector of the physiological data relative to the global temporal feature vector of the physiological data, so as to improve the accuracy of the classification result obtained by the classifier of the semantic difference feature vector of the physiological pattern. In this way, the physiological state of the elderly at home can be monitored in real time, and alarm information can be sent to relevant personnel in a timely manner when safety risks are detected, thereby providing necessary health monitoring and management support for the elderly's home-based elderly care life.

[0073] More specifically, the cascade fusion subunit 3712 is used to cascade fuse the fused correction feature vector with the physiological data global time series feature vector to obtain the optimized physiological data global time series feature vector. It should be understood that the expression ability of the feature can be improved by cascading the fused correction feature vector and the physiological data global time series feature vector.

[0074] Specifically, the physiological pattern semantic difference measurement unit 372 is used to respectively calculate the physiological pattern semantic difference coefficient of each physiological data local time series association feature vector in the sequence of the physiological data local time series association feature vector relative to the optimized physiological data global time series feature vector to obtain the physiological pattern semantic difference feature vector. It should be understood that each feature vector in the sequence of the physiological data local time series association feature vector represents the physiological data time series association feature within a local time window. The physiological data global time series feature vector integrates the physiological data time series feature information of the entire time series, including local time series association and global time series trend. Therefore, in order to further improve the adequacy and accuracy of abnormal detection of the physiological state of the elderly, it is necessary to compare and analyze the differences between the physiological data time series association features in each local time series and the overall physiological data time series features, so as to judge and detect the degree of deviation and abnormality of the physiological data time series synergistic features in each local time series and the overall change. Based on this, in the technical solution of the present application, the physiological pattern semantic difference coefficients of each local time series association feature vector of physiological data in the sequence of local time series association feature vectors of physiological data relative to the global time series feature vector of physiological data are further calculated to obtain the physiological pattern semantic difference feature vector. In particular, by comparing the difference between the local feature vector and the global feature vector, not only the difference and abnormality between the physiological state temporal changes in the local time series and the physiological state temporal changes in the overall time series can be captured, but also the pattern changes and semantic information of the physiological data can be revealed, which is conducive to further analysis and understanding of the health status of the elderly. More specifically, the physiological pattern semantic difference coefficients of each local time series association feature vector of physiological data in the sequence of local time series association feature vectors of physiological data relative to the optimized global time series feature vector of physiological data are calculated respectively by the following semantic association formula to obtain the physiological pattern semantic difference feature vector; wherein, the semantic association formula is:

[0075] ;

[0076] in, represents the feature value of each position of any local time series correlation feature vector of physiological data in the sequence of local time series correlation feature vectors of physiological data, represents the eigenvalues ​​of each position of the optimized global time series eigenvector of physiological data, is the scale of the local temporal correlation feature vector of the physiological data and the optimized global temporal feature vector of the physiological data, is the feature value of each position in the semantic difference feature vector of the physiological pattern.

[0077] It is worth mentioning that in other specific examples of the present application, the physiological pattern semantic difference coefficient of each local time series associated feature vector of the physiological data in the sequence of the local time series associated feature vector of the physiological data relative to the global time series feature vector of the physiological data can also be calculated in other ways to obtain the physiological pattern semantic difference feature vector as the physiological pattern semantic difference feature, for example: preprocessing the physiological data, including noise removal, normalization, smoothing and other operations to ensure the quality and stability of the data; using appropriate methods (such as average value, peak value, spectrum analysis, etc.) to extract the global time series feature vector from the complete time series sequence of the physiological data; dividing the time series sequence of the physiological data into multiple local time series associated feature vector sequences; for each local time series associated feature vector sequence, calculating the physiological pattern semantic difference coefficient between it and the global time series feature vector. The physiological pattern semantic difference coefficient can be calculated using a suitable similarity measurement method, such as correlation coefficient, cosine similarity, etc.; the calculated physiological pattern semantic difference coefficients are combined in a certain order to obtain the physiological pattern semantic difference feature vector.

[0078] In particular, the physiological state abnormality detection and warning module 380 is used to determine whether the physiological state of the monitored elderly person is abnormal based on the physiological pattern semantic difference characteristics, and determine whether to generate a physiological abnormality warning signal. In particular, in a specific example of the present application, Figure 5 As shown, the physiological state abnormality detection and warning module 380 includes: a physiological state abnormality detection unit 381, which is used to pass the physiological pattern semantic difference feature vector through a classifier to obtain a classification result, and the classification result is used to indicate whether there is an abnormality in the physiological state of the monitored elderly person; and a warning signal generation unit 382, ​​which is used to determine whether to generate a physiological abnormality warning signal based on the classification result.

[0079] Specifically, the physiological state abnormality detection unit 381 is used to pass the physiological pattern semantic difference feature vector through a classifier to obtain a classification result, and the classification result is used to indicate whether the physiological state of the monitored elderly person is abnormal. That is, the local and global physiological pattern semantic difference feature information is used for classification processing to detect physiological state abnormalities, thereby judging the health status and safety risks of the elderly, and sending alarm information to relevant personnel in a timely manner when safety risks are detected. More specifically, passing the physiological pattern semantic difference feature vector through a classifier to obtain a classification result includes: using multiple fully connected layers of the classifier to fully connect encode the physiological pattern semantic difference feature vector to obtain an encoded classification feature vector; and passing the encoded classification feature vector through the Softmax classification function of the classifier to obtain the classification result.

[0080] That is, in the technical solution of the present application, the label of the classifier includes whether the physiological state of the monitored elderly is abnormal (first label), and whether the physiological state of the monitored elderly is not abnormal (second label), wherein the classifier determines which classification label the semantic difference feature vector of the physiological pattern belongs to through a soft maximum function. It is worth noting that the first label p1 and the second label p2 here do not contain artificially set concepts. In fact, during the training process, the computer model does not have the concept of "whether the physiological state of the monitored elderly is abnormal". It only has two classification labels and the probability of the output feature under these two classification labels, that is, the sum of p1 and p2 is one. Therefore, the classification result of whether the physiological state of the monitored elderly is abnormal is actually converted into a binary class probability distribution that conforms to the laws of nature through the classification label. In essence, the physical meaning of the natural probability distribution of the label is used, rather than the linguistic text meaning of "whether the physiological state of the monitored elderly is abnormal".

[0081] A classifier is a machine learning model or algorithm that is used to classify input data into different categories or labels. Classifiers are part of supervised learning and perform classification tasks by learning the mapping relationship from input data to output categories.

[0082] A fully connected layer is a common type of layer in neural networks. In a fully connected layer, each neuron is connected to all neurons in the previous layer, and each connection has a weight. This means that each neuron in the fully connected layer receives inputs from all neurons in the previous layer, and performs a weighted sum of these inputs through the weights, and then passes the result to the next layer.

[0083] The Softmax classification function is a commonly used activation function for multi-classification problems. It converts each element of the input vector into a probability value between 0 and 1, and the sum of these probability values ​​is equal to 1. The Softmax function is often used in the output layer of a neural network, especially for multi-classification problems, because it can map the network output to the probability distribution of each category. During the training process, the output of the Softmax function can be used to calculate the loss function and update the network parameters through the back propagation algorithm. It is worth noting that the output of the Softmax function does not change the relative size relationship between the elements, but only normalizes them. Therefore, the Softmax function does not change the characteristics of the input vector, but only converts it into a probability distribution form.

[0084] Specifically, the warning signal generating unit 382 is used to determine whether to generate a physiological abnormality warning signal based on the classification result. That is, in a specific example of the present application, after obtaining the classification result of whether the physiological state of the monitored elderly person is abnormal, in response to the classification result that the physiological state of the monitored elderly person is abnormal, a physiological abnormality warning signal is generated; otherwise, no physiological abnormality warning signal is generated.

[0085] It is worth mentioning that in other specific examples of the present application, it is also possible to determine whether there is an abnormality in the physiological state of the monitored elderly based on the semantic difference characteristics of the physiological pattern in other ways, and determine whether to generate a physiological abnormality warning signal, for example: collecting physiological data of the monitored elderly, such as electrocardiogram, blood pressure, body temperature, etc.; preprocessing the collected physiological data; extracting physiological pattern semantic difference characteristics from the preprocessed physiological data; constructing a model for abnormality detection based on the extracted physiological pattern semantic difference characteristics; using the constructed model to detect abnormalities in the physiological state of the monitored elderly. By inputting the semantic difference characteristics of the physiological pattern, the model can determine whether the current physiological state is normal; if an abnormal physiological state is detected, a physiological abnormality warning signal is generated.

[0086] Furthermore, a home-based elderly care health monitoring and management method based on the Internet of Things technology is also provided.

[0087] Figure 6 Flow chart of the home-based elderly care health monitoring management method based on the Internet of Things technology according to an embodiment of the present application. Figure 6As shown, according to the embodiment of the present application, the home-based elderly care health monitoring and management method based on the Internet of Things technology includes the following steps: S1, obtaining a time series of physiological data of the monitored elderly person collected by a monitoring device worn by the monitored elderly person, wherein the physiological data includes heart rate value, blood pressure value, blood oxygen value and body temperature value; S2, transmitting the time series of physiological data to a cloud server via a wireless network; S3, in the cloud server, dividing the time series of physiological data of the monitored elderly person based on a predetermined time scale to obtain a sequence of local time series of physiological data; S4, arranging the sequence of local time series of physiological data according to the time dimension and the physiological parameter sample dimension to obtain a sequence of local time series input matrices of physiological data; S5, through a deep neural network-based The physiological multi-parameter local time series association feature extractor of the network model respectively extracts features from the sequence of the local time series input matrix of the physiological data to obtain a sequence of local time series association feature vectors of the physiological data; S6, fuses the sequence of local time series association feature vectors of the physiological data to obtain a global time series feature vector of the physiological data; S7, respectively calculates the physiological pattern semantic difference coefficient of each local time series association feature vector of the physiological data in the sequence of local time series association feature vectors of the physiological data relative to the global time series feature vector of the physiological data to obtain a physiological pattern semantic difference feature vector as a physiological pattern semantic difference feature; S8, based on the physiological pattern semantic difference feature, determines whether there is any abnormality in the physiological state of the monitored elderly person, and determines whether to generate a physiological abnormality warning signal.

[0088] In summary, the home-based elderly care health monitoring and management method based on the Internet of Things technology according to the embodiment of the present application is explained, which uses monitoring equipment to monitor and collect the physiological data of the elderly in real time, such as heart rate, blood pressure, blood oxygen and body temperature, and sends these physiological data to the cloud server for processing through a wireless network, so as to use the data processing and analysis algorithm based on the Internet of Things and artificial intelligence technology on the cloud server to perform time-series collaborative analysis of these physiological data, so as to judge the health status and safety risks of the elderly. In this way, the physiological state of the elderly at home can be monitored in real time, and when a safety risk is detected, an alarm message can be sent to relevant personnel in a timely manner, thereby providing necessary health monitoring management support for the elderly's home-based elderly care life.

[0089] As described above, the home-based elderly care health monitoring and management system 300 based on the Internet of Things technology according to the embodiment of the present application can be implemented in various wireless terminals, such as a server with a home-based elderly care health monitoring and management algorithm based on the Internet of Things technology. In a possible implementation, the home-based elderly care health monitoring and management system 300 based on the Internet of Things technology according to the embodiment of the present application can be integrated into the wireless terminal as a software module and / or a hardware module. For example, the home-based elderly care health monitoring and management system 300 based on the Internet of Things technology can be a software module in the operating system of the wireless terminal, or it can be an application developed for the wireless terminal; of course, the home-based elderly care health monitoring and management system 300 based on the Internet of Things technology can also be one of the many hardware modules of the wireless terminal.

[0090] Alternatively, in another example, the home-based elderly care health monitoring and management system 300 based on the Internet of Things technology and the wireless terminal may also be separate devices, and the home-based elderly care health monitoring and management system 300 based on the Internet of Things technology may be connected to the wireless terminal via a wired and / or wireless network, and transmit interactive information in accordance with an agreed data format.

[0091] The embodiments of the present disclosure have been described above, and the above description is exemplary, not exhaustive, and is not limited to the disclosed embodiments. Many modifications and changes will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the described embodiments. The selection of terms used herein is intended to best explain the principles of the embodiments, practical applications, or improvements to the technology in the market, or to enable other persons of ordinary skill in the art to understand the embodiments disclosed herein.

Claims

1. A home-based elderly care health monitoring and management system based on Internet of Things technology, characterized in that: include: A physiological data acquisition module, used to acquire a time series of physiological data of the monitored elderly person collected by a monitoring device worn by the monitored elderly person, wherein the physiological data includes heart rate value, blood pressure value, blood oxygen value and body temperature value; A physiological data transmission module, used for transmitting the physiological data time series to a cloud server via a wireless network; A physiological data time series division module is used to divide the physiological data time series of the monitored elderly person based on a predetermined time scale in the cloud server to obtain a sequence of local time series of physiological data; A physiological data multi-parameter local time series arrangement module, used for arranging the sequence of the physiological data local time series according to the time dimension and the physiological parameter sample dimension to obtain a sequence of the physiological data local time series input matrix; A physiological data multi-parameter local time series correlation feature extraction module, used to extract features from the sequence of the physiological data local time series input matrix respectively through a physiological multi-parameter local time series correlation feature extractor based on a deep neural network model to obtain a sequence of physiological data local time series correlation feature vectors; A physiological data global time series feature expression module, used for fusing the sequence of the physiological data local time series associated feature vectors to obtain a physiological data global time series feature vector; A physiological pattern semantic difference analysis module, used for respectively calculating the physiological pattern semantic difference coefficient of each physiological data local time series association feature vector in the sequence of the physiological data local time series association feature vector relative to the physiological data global time series feature vector to obtain a physiological pattern semantic difference feature vector as a physiological pattern semantic difference feature; A physiological state abnormality detection and warning module is used to determine whether the physiological state of the monitored elderly person is abnormal based on the semantic difference characteristics of the physiological pattern, and determine whether to generate a physiological abnormality warning signal; Wherein, the physiological pattern semantic difference analysis module includes: A physiological data global time series feature optimization unit, used for performing feature optimization on the physiological data global time series feature vector to obtain an optimized physiological data global time series feature vector; A physiological pattern semantic difference measurement unit, used for respectively calculating the physiological pattern semantic difference coefficient of each physiological data local time series association feature vector in the sequence of the physiological data local time series association feature vector relative to the optimized physiological data global time series feature vector to obtain the physiological pattern semantic difference feature vector; Wherein, the physiological data global time series feature optimization unit includes: A feature fusion correction subunit, used for fusing the sequence of the global time series feature vector of the physiological data and the local time series associated feature vector of the physiological data with the feature value as the granularity to obtain a fused correction feature vector; A cascade fusion subunit, configured to cascade fuse the fused correction feature vector with the physiological data global time series feature vector to obtain the optimized physiological data global time series feature vector; The sequence of the global temporal feature vector of the physiological data and the local temporal correlation feature vector of the physiological data is fused with the feature value as the granularity to obtain the fused correction feature vector, which is specifically expressed as: ； in, is a concatenated feature vector obtained by concatenating the sequence of local temporal correlation feature vectors of the physiological data, is the global time series feature vector of physiological data after linear interpolation, and are the eigenvalues ​​of the cascade feature vector and the global temporal feature vector of the physiological data, respectively, and Respectively represent the square of the first norm of the feature vector and the square root of the second norm of the feature vector, and the cascade feature vector and the global time series feature vector of the physiological data have the same feature vector length ,and is the weight hyperparameter, represents exponential operation, It means adding by position. It means to make a difference by position. is the eigenvalue of the fused corrected eigenvector.

2. The home-based elderly care health monitoring and management system based on Internet of Things technology according to claim 1 is characterized in that: The deep neural network model is a convolutional neural network model.

3. The home-based elderly care health monitoring and management system based on Internet of Things technology according to claim 2 is characterized in that: The physiological data global temporal feature expression module is used to: use a projection fusion layer to fuse the sequence of the physiological data local temporal association feature vectors to obtain the physiological data global temporal feature vector.

4. The home-based elderly care health monitoring and management system based on Internet of Things technology according to claim 3 is characterized in that: The physiological data global temporal feature expression module is used to: use the projection fusion layer to fuse the sequence of the physiological data local temporal correlation feature vectors according to the following formula to obtain the physiological data global temporal feature vector; Wherein, the formula is: ； in is a sequence of local temporal correlation feature vectors of the physiological data, is the global time series feature vector of the physiological data, Represents the projection fusion layer.

5. The home-based elderly care health monitoring and management system based on Internet of Things technology according to claim 4 is characterized in that: The physiological pattern semantic difference measurement unit is used to: calculate the physiological pattern semantic difference coefficient of each physiological data local time series association feature vector in the sequence of the physiological data local time series association feature vector relative to the optimized physiological data global time series feature vector by using the following semantic association formula to obtain the physiological pattern semantic difference feature vector; Wherein, the semantic association formula is: ； in, represents the feature value of each position of any local time series correlation feature vector of physiological data in the sequence of local time series correlation feature vectors of physiological data, represents the eigenvalues ​​of each position of the optimized global time series eigenvector of physiological data, is the scale of the local temporal correlation feature vector of the physiological data and the optimized global temporal feature vector of the physiological data, is the feature value of each position in the semantic difference feature vector of the physiological pattern.

6. The home-based elderly care health monitoring and management system based on Internet of Things technology according to claim 5 is characterized in that: The physiological state abnormality detection and early warning module comprises: A physiological state abnormality detection unit, used for passing the physiological pattern semantic difference feature vector through a classifier to obtain a classification result, wherein the classification result is used to indicate whether the physiological state of the monitored elderly person is abnormal; The warning signal generating unit is used to determine whether to generate a physiological abnormality warning signal based on the classification result.

7. A home-based elderly care health monitoring and management method based on Internet of Things technology, characterized in that: include: Acquire a time series of physiological data of the monitored elderly person collected by a monitoring device worn by the monitored elderly person, wherein the physiological data includes heart rate value, blood pressure value, blood oxygen value and body temperature value; Transmitting the physiological data time series to a cloud server via a wireless network; In the cloud server, the physiological data time series of the monitored elderly person is divided based on a predetermined time scale to obtain a sequence of local time series of the physiological data; Arranging the sequence of the local time series of physiological data according to the time dimension and the physiological parameter sample dimension to obtain a sequence of the local time series input matrix of physiological data; The sequence of the physiological data local time series input matrix is ​​subjected to feature extraction by a physiological multi-parameter local time series correlation feature extractor based on a deep neural network model to obtain a sequence of physiological data local time series correlation feature vectors; fusing the sequence of local temporal correlation feature vectors of the physiological data to obtain a global temporal feature vector of the physiological data; Respectively calculating the physiological pattern semantic difference coefficient of each physiological data local time series association feature vector in the sequence of the physiological data local time series association feature vector relative to the physiological data global time series feature vector to obtain a physiological pattern semantic difference feature vector as a physiological pattern semantic difference feature; Based on the semantic difference characteristics of the physiological pattern, determining whether the physiological state of the monitored elderly person is abnormal, and determining whether to generate a physiological abnormality warning signal; The method of calculating the physiological pattern semantic difference coefficient of each physiological data local time series association feature vector in the sequence of the physiological data local time series association feature vector relative to the physiological data global time series feature vector to obtain the physiological pattern semantic difference feature vector as the physiological pattern semantic difference feature includes: Performing feature optimization on the physiological data global time series feature vector to obtain an optimized physiological data global time series feature vector; Respectively calculating the physiological pattern semantic difference coefficient of each physiological data local time series association feature vector in the sequence of the physiological data local time series association feature vector relative to the optimized physiological data global time series feature vector to obtain the physiological pattern semantic difference feature vector; The step of performing feature optimization on the physiological data global time series feature vector to obtain an optimized physiological data global time series feature vector includes: Fusion the sequence of the global time series feature vector of the physiological data and the local time series associated feature vector of the physiological data with the feature value as the granularity to obtain a fused correction feature vector; Cascadingly fusing the fused correction feature vector with the global time series feature vector of physiological data to obtain the optimized global time series feature vector of physiological data; The sequence of the global temporal feature vector of the physiological data and the local temporal correlation feature vector of the physiological data is fused with the feature value as the granularity to obtain the fused correction feature vector, which is specifically expressed as: ； in, is a concatenated feature vector obtained by concatenating the sequence of local temporal correlation feature vectors of the physiological data, is the global time series feature vector of physiological data after linear interpolation, and are the eigenvalues ​​of the cascade feature vector and the global temporal feature vector of the physiological data, respectively, and Respectively represent the square of the first norm of the feature vector and the square root of the second norm of the feature vector, and the cascade feature vector and the global time series feature vector of the physiological data have the same feature vector length ,and is the weight hyperparameter, represents exponential operation, It means adding by position. It means to make a difference by position. is the eigenvalue of the fused corrected eigenvector.

Images