Data communication terminal and method with emergency rescue function
By connecting the data communication terminal with the smart bracelet, using artificial intelligence to analyze vital sign data and sending help information through satellite communication, the problems of insufficient communication coverage and false alarms in remote areas are solved, and more accurate health risk identification and Emergency rescue services.
Patent Information
- Application Number
- CN202411157974.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-22
- Publication Date
- 2024-11-01
- Estimated Expiration
- 2044-08-22
AI Technical Summary
Traditional data communication terminals with emergency rescue functions cannot guarantee communication coverage in remote or maritime areas, and are low in intelligence, which is prone to false alarm signals, affecting the effectiveness of emergency rescue services.
By connecting the data communication terminal with the smart bracelet, using artificial intelligence-based data analysis and processing algorithms to analyze the timing characteristics of vital sign data, judge whether there is an abnormality in the physiological state, and use the satellite communication and positioning module when abnormal situations are detected Send location information and help information.
It realizes more accurately identifying and predicting health risks in places that are not covered by traditional communication networks, providing effective communication and help services, and providing protection for emergency situations for outdoor personnel.
Smart Images

Figure CN118694427B_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of data analysis, and more specifically, to a data communication terminal and method with emergency rescue function. Background Art
[0002] Outdoor work is usually done in remote areas or areas with poor communication; emergency rescue, outdoor hiking, mountaineering and other activities may be accompanied by high risks. A data communication terminal with emergency rescue function is a device that integrates advanced communication technology and emergency response functions, designed to provide users with rapid medical assistance and rescue services. Through a data communication terminal with emergency rescue function, timely rescue and communication support can be provided for these outdoor personnel.
[0003] However, traditional data communication terminals with emergency rescue functions rely on specific communication networks and may not be able to guarantee communication coverage in remote areas or at sea. In addition, in emergency rescue, it is crucial to quickly and accurately assess the health status of the victim, and traditional terminals usually rely on monitoring the user's vital signs, such as heart rate and body temperature, and compare them with predetermined thresholds to determine whether to send an emergency rescue signal. This method has a low degree of intelligence and will be affected by external environmental factors or changes in user factors, resulting in data fluctuations within the normal range, resulting in frequent false alarm signals, thereby affecting the effectiveness of emergency rescue services.
[0004] Therefore, an optimized data communication terminal with emergency rescue function is needed to solve the above technical problems. Summary of the invention
[0005] In order to solve the above technical problems, the present application is proposed. The embodiment of the present application provides a data communication terminal and method with emergency rescue function, which connects the data communication terminal to a smart bracelet worn by the user, and the smart bracelet is used to collect vital signs data. The data communication terminal processes and analyzes the collected vital signs data to determine whether there is an abnormality in the physiological state. If there is an abnormality, the location information and distress message are sent through a satellite communication antenna. Specifically, an artificial intelligence-based data analysis and processing algorithm is used to perform a time series feature analysis of the vital signs data, and then to determine whether there is an abnormality in the vital signs of the monitored human subject, and send the location information and distress message through a satellite communication and positioning module when an abnormality is detected. The present application can more accurately identify and predict the health risks of the monitored person and can also achieve effective communication and distress in places where traditional communication networks are not covered.
[0006] According to one aspect of the present application, a data communication terminal with an emergency rescue function is provided, which includes:
[0007] A human vital signs data receiving module, used to receive a time series of vital signs data of a monitored human subject sent by a smart bracelet, wherein the vital signs data include body temperature, heart rate and blood pressure;
[0008] A vital sign data regularization module, used for regularizing the time series of the vital sign data according to the parameter sample dimension to obtain a time series of body temperature values, a time series of heart rate values, and a time series of blood pressure values;
[0009] A vital sign data local time series association extraction module, used for performing sequence encoding on the time series of the body temperature value, the time series of the heart rate value and the time series of the blood pressure value to obtain a sequence of implicit feature vectors of local time series association of the body temperature value, a sequence of implicit feature vectors of local time series association of the heart rate value and a sequence of implicit feature vectors of local time series association of the blood pressure value;
[0010] A time series correlation feature aggregation module, used to input the sequence of the local time series correlation implicit feature vectors of the temperature value, the sequence of the local time series correlation implicit feature vectors of the heart rate value and the sequence of the local time series correlation implicit feature vectors of the blood pressure value into a feature distribution reverse mask clustering analysis network to obtain a time series correlation feature aggregation representation vector of the temperature value, a time series correlation feature aggregation representation vector of the heart rate value and a time series correlation feature aggregation representation vector of the blood pressure value;
[0011] A vital sign multimodal fusion module, used for multimodally embedding and fusing the temperature value time series associated feature aggregation representation vector, the heart rate value time series associated feature aggregation representation vector and the blood pressure value time series associated feature aggregation representation vector to obtain a vital sign multimodal embedding fusion representation vector as a vital sign multimodal embedding fusion representation feature;
[0012] A monitoring abnormality judgment module, used to obtain a monitoring result based on the multimodal embedding fusion representation feature of the vital signs, wherein the monitoring result is used to indicate whether there is an abnormality in the vital signs of the monitored human subject;
[0013] The satellite communication and positioning module is used to send location information and distress information based on the monitoring results.
[0014] According to another aspect of the present application, a data communication method with an emergency rescue function is provided, comprising:
[0015] Receiving a time series of vital sign data of a monitored human subject sent by a smart bracelet, wherein the vital sign data includes body temperature, heart rate and blood pressure;
[0016] Regularizing the time series of the vital sign data according to the parameter sample dimension to obtain a time series of body temperature values, a time series of heart rate values, and a time series of blood pressure values;
[0017] Sequence-encoding the time series of the body temperature values, the time series of the heart rate values, and the time series of the blood pressure values to obtain a sequence of implicit feature vectors associated with local time series of the body temperature values, a sequence of implicit feature vectors associated with local time series of the heart rate values, and a sequence of implicit feature vectors associated with local time series of the blood pressure values;
[0018] Inputting the sequence of the local temporal association implicit feature vectors of the temperature value, the sequence of the local temporal association implicit feature vectors of the heart rate value and the sequence of the local temporal association implicit feature vectors of the blood pressure value into the feature distribution reverse mask clustering analysis network respectively to obtain the temporal association feature aggregation representation vector of the temperature value, the temporal association feature aggregation representation vector of the heart rate value and the temporal association feature aggregation representation vector of the blood pressure value;
[0019] Performing multimodal embedding fusion on the temperature value time series associated feature aggregation representation vector, the heart rate value time series associated feature aggregation representation vector, and the blood pressure value time series associated feature aggregation representation vector to obtain a vital sign multimodal embedding fusion representation vector as a vital sign multimodal embedding fusion representation feature;
[0020] Based on the multimodal embedding fusion representation feature of the vital signs, a monitoring result is obtained, wherein the monitoring result is used to indicate whether there is an abnormality in the vital signs of the monitored human subject;
[0021] Based on the monitoring results, location information and distress information are sent.
[0022] The present application has at least the following technical effects: Compared with the prior art, the data communication terminal and method with emergency rescue function provided by the present application connect the data communication terminal to the smart bracelet worn by the user, and the smart bracelet is used to collect vital signs data. The data communication terminal processes and analyzes the collected vital signs data to determine whether there is an abnormality in the physiological state. If there is an abnormality, the location information and distress message are sent through the satellite communication antenna. Specifically, an artificial intelligence-based data analysis and processing algorithm is used to perform time series feature analysis on the vital signs data, and then to determine whether there is an abnormality in the vital signs of the monitored human subject, and send the location information and distress message through the satellite communication and positioning module when an abnormality is detected. The present application can more accurately identify and predict the health risks of the monitored person and can also achieve effective communication and distress in places where traditional communication networks are not covered, providing protection for emergency rescue work for outdoor personnel. BRIEF DESCRIPTION OF THE DRAWINGS
[0023] Figure 1 4 is a block diagram of a data communication terminal with emergency rescue function according to an embodiment of the present application.
[0024] Figure 2 Schematic diagram of the architecture of a data communication terminal with emergency rescue function according to an embodiment of the present application.
[0025] Figure 3 It is a block diagram of a timing correlation feature aggregation module in a data communication terminal with emergency rescue function according to an embodiment of the present application.
[0026] Figure 4 The present invention is a block diagram of a monitoring abnormality judgment module in a data communication terminal with an emergency rescue function according to an embodiment of the present application.
[0027] Figure 5 Flow chart of a data communication method with emergency rescue function according to an embodiment of the present application. DETAILED DESCRIPTION
[0028] 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.
[0029] Activities such as offshore operations and outdoor adventures often take place in remote areas with limited communications. In these situations, the timeliness and effectiveness of emergency rescue are particularly critical. The existing communication technology and communication equipment of the emergency response mechanism can provide rapid medical assistance and rescue support for outdoor personnel in these areas.
[0030] However, existing communication equipment with emergency rescue functions is often limited by specific network coverage. In some remote or offshore areas, due to limited communication networks, outdoor personnel are often unable to send accurate positioning information and distress messages when emergency rescue is needed, resulting in the inability to provide emergency rescue in a timely manner. Secondly, in an emergency, a quick and accurate assessment of the victim's health status is crucial for rescue. However, existing equipment mostly relies on monitoring vital signs such as heart rate and body temperature, and comparing them with preset standards to trigger rescue signals. This method has a limited level of intelligence and is easily affected by environmental changes or individual differences, resulting in false alarms, thereby reducing the efficiency and reliability of rescue services.
[0031] Therefore, in response to the above technical problems, the present application proposes an optimized data communication terminal with emergency rescue function, which connects the data communication terminal to the smart bracelet worn by the user, and the smart bracelet is used to collect vital signs data. The data communication terminal processes and analyzes the collected vital signs data to determine whether there is an abnormality in the physiological state. If there is an abnormality, the location information and the distress message are sent through the satellite communication antenna. Specifically, by receiving the time series of the vital signs data (body temperature value, heart rate value and blood pressure value) of the monitored human object sent by the smart bracelet, and using the data analysis and processing algorithm based on artificial intelligence to encode and correlate the time series of the vital signs data, it is automatically determined whether there is an abnormality in the vital signs of the monitored human object according to the multimodal fusion representation characteristics of the body temperature value, the heart rate value and the blood pressure value in the time series, and sends the location information and the distress message through the satellite communication and positioning module when an abnormal situation is detected. The present application can use artificial intelligence algorithms to analyze and process vital signs data in real time, so as to more accurately identify and predict the health risks of the monitored person by using multiple vital signs time series association semantics. At the same time, by using satellite communication and positioning modules, effective communication and rescue can be achieved even in remote areas or at sea where traditional communication networks cannot reach.
[0032] Figure 1 4 is a block diagram of a data communication terminal with emergency rescue function according to an embodiment of the present application. Figure 2 FIG. 1 is a schematic diagram of the architecture of a data communication terminal with emergency rescue function according to an embodiment of the present application. Figure 1 and Figure 2As shown, in the data communication terminal 100 with emergency rescue function, it includes: a human vital sign data receiving module 110, which is used to receive the time series of vital sign data of the monitored human object sent by the smart bracelet, wherein the vital sign data includes body temperature value, heart rate value and blood pressure value; a vital sign data regularization module 120, which is used to regularize the time series of the vital sign data according to the parameter sample dimension to obtain the time series of body temperature value, the time series of heart rate value and the time series of blood pressure value; a vital sign data local time series association extraction module 130, which is used to sequence encode the time series of body temperature value, the time series of heart rate value and the time series of blood pressure value to obtain a sequence of local time series association implicit feature vectors of body temperature value, a sequence of local time series association implicit feature vectors of heart rate value and a sequence of local time series association implicit feature vectors of blood pressure value; a time series association feature aggregation module 140, which is used to regularize the time series of the vital sign data according to the parameter sample dimension to obtain a time series of body temperature value, a time series association implicit feature vector of heart rate value and a sequence of local time series association implicit feature vector of blood pressure value The sequence of the implicit feature vectors of the local time series association of the temperature value, the sequence of the implicit feature vectors of the local time series association of the heart rate value and the sequence of the implicit feature vectors of the local time series association of the blood pressure value are respectively input into the feature distribution reverse mask clustering analysis network to obtain the temperature value time series association feature aggregation representation vector, the heart rate value time series association feature aggregation representation vector and the blood pressure value time series association feature aggregation representation vector; the vital sign multimodal fusion module 150 is used to perform multimodal embedding fusion on the temperature value time series association feature aggregation representation vector, the heart rate value time series association feature aggregation representation vector and the blood pressure value time series association feature aggregation representation vector to obtain the vital sign multimodal embedding fusion representation vector as the vital sign multimodal embedding fusion representation feature; the monitoring abnormality judgment module 160 is used to obtain the monitoring result based on the multimodal embedding fusion representation feature of the vital sign, and the monitoring result is used to indicate whether there is an abnormality in the vital sign of the monitored human subject;
[0033] The satellite communication and positioning module 170 is used to send location information and distress information based on the monitoring results.
[0034] In an embodiment of the present application, the human vital signs data receiving module 110 is used to receive the time series of vital signs data of the monitored human object sent by the smart bracelet, wherein the vital signs data include body temperature, heart rate and blood pressure. Specifically, the three vital signs indicators of body temperature, heart rate and blood pressure are important indicators reflecting the health status of the human body. The time series data of the vital signs indicators contains rich time series information, which can be used to reveal the dynamic changes in the health status of the monitored human object. By analyzing the time series data of body temperature, heart rate and blood pressure, it is used to determine whether there is an abnormality in the vital signs of the monitored human object, and then decide whether to trigger an emergency rescue signal. In particular, in a specific embodiment of the present application, the time series data of body temperature, heart rate and blood pressure can be obtained by infrared sensors, photoelectric sensors and pulse wave conduction velocity algorithms on smart bracelets, respectively.
[0035] In an embodiment of the present application, the vital signs data regularization module 120 is used to regularize the time series of the vital signs data according to the parameter sample dimension to obtain the time series of body temperature values, the time series of heart rate values, and the time series of blood pressure values. Accordingly, it is considered that each vital sign in the time series of the vital signs data is constantly changing with the passage of time, and each vital sign data has a mutual correlation in the time dimension. Based on this, in the technical solution of the present application, the time series of the vital signs data is regularized according to the parameter sample dimension to obtain the time series of body temperature values, the time series of heart rate values, and the time series of blood pressure values. In this way, the correlation between different vital signs can be observed more clearly, providing a better basis for subsequent time series feature analysis and understanding.
[0036] In an embodiment of the present application, the vital sign data local temporal association extraction module 130 is used to perform sequence encoding on the time series of the body temperature values, the time series of the heart rate values, and the time series of the blood pressure values to obtain a sequence of implicit feature vectors of local temporal association of body temperature values, a sequence of implicit feature vectors of local temporal association of heart rate values, and a sequence of implicit feature vectors of local temporal association of blood pressure values. Specifically, in an embodiment of the present application, the vital sign data local temporal association extraction module is used to: input the time series of the body temperature values, the time series of the heart rate values, and the time series of the blood pressure values into a sequence encoder based on a 1D-CNN model to obtain a sequence of implicit feature vectors of local temporal association of body temperature values, a sequence of implicit feature vectors of local temporal association of heart rate values, and a sequence of implicit feature vectors of local temporal association of blood pressure values. Specifically, the time series of the body temperature values, the time series of the heart rate values, and the time series of the blood pressure values all contain time series feature information on the local time series, such as the short-term fluctuation trend in the local time, and the 1D-CNN model using a one-dimensional convolution kernel is good at extracting local time series features from sequence data, which can provide a rich representation for vital signs data. Therefore, in the technical solution of the present application, the time series of the body temperature values, the time series of the heart rate values, and the time series of the blood pressure values are input into a sequence encoder based on the 1D-CNN model to respectively capture and mine the time series feature information of each vital sign parameter on the local time scale, thereby obtaining a sequence of implicit feature vectors associated with the local time series of the body temperature values, a sequence of implicit feature vectors associated with the local time series of the heart rate values, and a sequence of implicit feature vectors associated with the local time series of the blood pressure values.
[0037] In an embodiment of the present application, the time series correlation feature aggregation module 140 is used to input the sequence of the local time series correlation implicit feature vectors of the body temperature value, the sequence of the local time series correlation implicit feature vectors of the heart rate value, and the sequence of the local time series correlation implicit feature vectors of the blood pressure value into the feature distribution reverse mask clustering analysis network to obtain the body temperature value time series correlation feature aggregation representation vector, the heart rate value time series correlation feature aggregation representation vector, and the blood pressure value time series correlation feature aggregation representation vector. Accordingly, considering that each local time series correlation implicit feature vector in the sequence of the local time series correlation implicit feature vectors of the body temperature value, the sequence of the local time series correlation implicit feature vectors of the heart rate value, and the sequence of the local time series correlation implicit feature vectors of the blood pressure value represents the change trend and fluctuation of the body temperature value, the heart rate value, and the blood pressure value within a local time period. Therefore, in order to strengthen the temporal interdependence between various vital sign elements in each local time period and integrate them into a comprehensive timing model so as to more accurately reflect and depict the overall behavior of the data in the entire time dimension, in the technical solution of the present application, the sequence of the local temporal associated implicit feature vectors of the body temperature value, the sequence of the local temporal associated implicit feature vectors of the heart rate value and the sequence of the local temporal associated implicit feature vectors of the blood pressure value are respectively input into the feature distribution reverse mask clustering analysis network to obtain the body temperature value temporal associated feature aggregation representation vector, the heart rate value temporal associated feature aggregation representation vector and the blood pressure value temporal associated feature aggregation representation vector.
[0038] In particular, the processing method of the sequence of the implicit feature vectors of the local time series association of the heart rate value and the sequence of the implicit feature vectors of the local time series association of the blood pressure value is consistent with the processing method of the sequence of the implicit feature vectors of the local time series association of the body temperature value. In this application, the processing of the sequence of the implicit feature vectors of the local time series association of the body temperature value is taken as an example. Specifically, first, a clustering method is applied to perform cluster analysis on the sequence of the implicit feature vectors of the local time series association of the body temperature value to reveal their inherent potential time series patterns and structures to obtain the implicit cluster center vectors of the local time series association of the body temperature value representing a specific behavior pattern. Then, by determining the correlation coefficient between the implicit feature vectors of the local time series association of each body temperature value and the implicit cluster center vector of the local time series association of the body temperature value, the similarity of each feature vector with the cluster center is measured. Next, the reciprocal of each correlation coefficient is obtained to form an anti-correlation coefficient. Such a conversion can amplify the differences of those feature vectors with low correlation with the cluster center of the body temperature value, making them more prominent in the analysis. Next, the anti-correlation coefficients obtained are normalized using the Sigmoid function so as to map them to a specific numerical range while maintaining the properties of the local temporal association implicit features of the original body temperature values, providing accurate numerical input for the feature weighting link. Then, the normalized anti-correlation weight coefficients are used to weight the local temporal association implicit feature vectors of the body temperature values, so as to reversely filter out irrelevant features, thereby obtaining a series of suppressed local temporal association implicit feature vectors of the body temperature values. Finally, the difference between each local temporal association implicit feature vector of the body temperature values and the suppressed local temporal association implicit feature vector of the body temperature values is calculated to filter out noise or irrelevant features, thereby forming a set of enhanced feature vectors. Finally, by averaging the feature vectors in these sets, an aggregate representation vector of the temporal association features of the body temperature values containing key feature information is obtained.
[0039] Specifically, Figure 3 FIG. 1 is a block diagram of a timing correlation feature aggregation module in a data communication terminal with an emergency rescue function according to an embodiment of the present application. Figure 3As shown, the time series correlation feature aggregation module 140 includes: a body temperature value associated implicit feature clustering calculation unit 141, which is used to cluster the sequence of the body temperature value local time series associated implicit feature vectors to obtain the body temperature value local time series associated implicit clustering center vector; a body temperature value time series correlation coefficient calculation unit 142, which is used to calculate the correlation coefficient between each body temperature value local time series associated implicit feature vector in the sequence of the body temperature value local time series associated implicit feature vector and the body temperature value local time series associated implicit clustering center vector to obtain a sequence of body temperature value local time series correlation coefficients; a body temperature value time series anti-correlation coefficient calculation unit 143, which is used to take the inverse of each body temperature value local time series correlation coefficient in the sequence of the body temperature value local time series correlation coefficient to obtain a sequence of body temperature value local time series anti-correlation coefficients; a body temperature value time series anti-correlation coefficient normalization unit 144, which is used to use a Sigmoid function to normalize the sequence of the body temperature value local time series anti-correlation coefficients to obtain a body temperature value local time series anti-correlation weight. A temperature value feature weighted suppression unit 145 is used to weight each temperature value local time series associated implicit feature vector in the sequence of the temperature value local time series associated implicit feature vectors using each temperature value local time series anti-correlation weight coefficient in the sequence of the temperature value local time series anti-correlation weight coefficients as a weight to obtain a sequence of suppressed temperature value local time series associated implicit feature vectors; a temperature value local time series feature enhancement unit 146 is used to calculate the positional difference between each group of the temperature value local time series associated implicit feature vectors and the suppressed temperature value local time series associated implicit feature vectors in the sequence of the temperature value local time series associated implicit feature vectors and the suppressed temperature value local time series associated implicit feature vectors to obtain a sequence of enhanced temperature value local time series associated implicit feature vectors; a temperature value feature aggregation unit 147 is used to calculate the positional mean vector of the sequence of enhanced temperature value local time series associated implicit feature vectors to obtain the temperature value time series associated feature aggregation representation vector.
[0040] More specifically, in an embodiment of the present application, the body temperature value associated implicit feature clustering calculation unit is used to: calculate the mean vector of the sequence of the body temperature value local time series associated implicit feature vectors to obtain the body temperature value local time series associated implicit clustering center vector.
[0041] More specifically, in an embodiment of the present application, the body temperature value time series correlation coefficient calculation unit is used to: vector multiply the transposed vector of the body temperature value local time series association implicit feature vector with the body temperature value local time series association implicit cluster center vector to obtain the body temperature value local time series association implicit correlation value; respectively calculate the norm of the body temperature value local time series association implicit feature vector and the body temperature value local time series association implicit cluster center vector to obtain the body temperature value local time series association implicit norm value and the body temperature value local time series association implicit cluster norm value; multiply the body temperature value local time series association implicit norm value and the body temperature value local time series association implicit cluster norm value to obtain the body temperature value local time series association interaction value; divide the body temperature value local time series association implicit correlation value by the body temperature value local time series association interaction value to obtain the body temperature value local time series correlation coefficient corresponding to the body temperature value local time series association implicit feature vector.
[0042] More specifically, in an embodiment of the present application, the body temperature value time series anti-correlation coefficient normalization unit is used to: use the negative number of each local time series anti-correlation coefficient of the body temperature value in the sequence of local time series anti-correlation coefficients of the body temperature value as the exponent of the natural constant to calculate the exponential function value with the natural constant as the base according to the position to obtain a sequence of local time series anti-correlation class support coefficients of the body temperature value; calculate the reciprocal of the sum of each local time series anti-correlation class support coefficient of the body temperature value in the sequence of local time series anti-correlation class support coefficients of the body temperature value and the constant one to obtain a sequence of local time series anti-correlation weight coefficients of the body temperature value.
[0043] In the embodiment of the present application, specifically, the time series correlation feature aggregation module is used to: input the sequence of the local time series correlation implicit feature vector of the body temperature value into the feature distribution reverse mask clustering analysis network, and process it with the following clustering formula to obtain the body temperature value time series correlation feature aggregation representation vector; wherein the clustering formula is:
[0044]
[0045]
[0046]
[0047]
[0048]
[0049]
[0050]
[0051]
[0052] in, is the sequence of implicit feature vectors associated with the local temporal sequence of the body temperature value, is the number of feature vectors in the sequence of implicit feature vectors associated with the local temporal series of body temperature values, is the number of the sequence of the implicit feature vectors of the local temporal association of the body temperature value The implicit eigenvector of the local temporal correlation of individual temperature values, is the implicit cluster center vector of the local temporal association of the body temperature value, It is The transposed vector of the implicit eigenvector of the local temporal correlation of individual temperature values, is the one-norm of the eigenvector, is the number of the local time series correlation coefficients of the body temperature values Local time series correlation coefficient of individual temperature values, is the first value in the sequence of the local time series anti-correlation coefficient of the body temperature value The local time series anti-correlation coefficient of individual temperature values, is the first in the sequence of the local temporal anti-correlation weight coefficients of the body temperature value The local time series anti-correlation weight coefficient of individual temperature values, is the sequence of implicit feature vectors of the local temporal association of the suppressed body temperature value, is the first in the sequence of implicit feature vectors for suppressing the local temporal association of body temperature values. implicit feature vectors that suppress the local temporal association of body temperature values, is the sequence of implicit feature vectors of the local temporal association of the enhanced body temperature value, is the first in the sequence of implicit feature vectors of the local temporal association of the enhanced body temperature value The implicit feature vector of the enhanced local temporal association of body temperature values, is the temperature value temporal correlation feature aggregation representation vector. Similarly, the encoding method of the sequence of the heart rate value local temporal correlation implicit feature vector and the sequence of the blood pressure value local temporal correlation implicit feature vector is consistent with the encoding method of the sequence of the temperature value local temporal correlation implicit feature vector.
[0053] In an embodiment of the present application, the vital sign multimodal fusion module 150 is used to perform multimodal embedding fusion on the temperature value temporal correlation feature aggregation representation vector, the heart rate value temporal correlation feature aggregation representation vector, and the blood pressure value temporal correlation feature aggregation representation vector to obtain a vital sign multimodal embedding fusion representation vector as a vital sign multimodal embedding fusion representation feature. Specifically, in an embodiment of the present application, the vital sign multimodal fusion module is used to: input the temperature value temporal correlation feature aggregation representation vector, the heart rate value temporal correlation feature aggregation representation vector, and the blood pressure value temporal correlation feature aggregation representation vector into an embedding layer-based vital sign multimodal embedding encoder to obtain the vital sign multimodal embedding fusion representation vector. Specifically, in order to effectively and comprehensively integrate data from different vital signs, so as to provide a comprehensive feature representation, which is helpful for a comprehensive understanding of the health status and to facilitate a more accurate subsequent assessment of whether the vital signs of the monitored human subject are abnormal, in the technical solution of the present application, the temperature value time series associated feature aggregation representation vector, the heart rate value time series associated feature aggregation representation vector and the blood pressure value time series associated feature aggregation representation vector are input into the vital sign multimodal embedding encoder based on the embedding layer to obtain the vital sign multimodal embedding fusion representation vector. That is, the vital sign multimodal embedding encoder based on the embedding layer can integrate data information of different vital signs, thereby enhancing the model's ability to learn the relationship between vital signs, thereby obtaining a more rich and representational vital sign multimodal embedding fusion representation vector.
[0054] In the embodiment of the present application, the monitoring abnormality judgment module 160 is used to obtain the monitoring result based on the multimodal embedding fusion representation feature of the vital signs, and the monitoring result is used to indicate whether the vital signs of the monitored human object are abnormal, and based on the monitoring result, the location information and the distress message are sent through the satellite communication and positioning module 170. That is, the multimodal embedding fusion representation feature of the vital signs obtained by multimodal embedding encoding using the temperature value time series association feature aggregation representation vector, the heart rate value time series association feature aggregation representation vector and the blood pressure value time series association feature aggregation representation vector is classified and processed, so as to automatically determine whether the vital signs of the monitored human object are abnormal, and send the location information and the distress message through the satellite communication and positioning module 170. The present application can use artificial intelligence algorithms to analyze and process vital sign data in real time, so as to use multiple vital sign time series association semantics to more accurately identify and predict the health risks of the monitored person. At the same time, by using the satellite communication and positioning module 170, effective communication and distress can be achieved even in remote areas or places where traditional communication networks cannot cover, such as at sea.
[0055] Specifically, Figure 4FIG. 1 is a block diagram of a monitoring abnormality judgment module in a data communication terminal with an emergency rescue function according to an embodiment of the present application. Figure 4 As shown, the monitoring abnormality judgment module 160 includes: a monitoring result generation unit 161, which is used to input the multimodal embedding fusion representation vector of the vital signs into the classifier-based vital sign monitor to obtain the monitoring result, and the monitoring result is used to indicate whether there is an abnormality in the vital signs of the monitored human object; an abnormal vital sign distress sending unit 162, which is used to send location information and distress information through the satellite communication and positioning module 170 in response to the monitoring result indicating that there is an abnormality in the vital signs of the monitored human object.
[0056] Specifically, in the technical solution of the present application, the temperature value temporal association feature aggregation representation vector, the heart rate value temporal association feature aggregation representation vector and the blood pressure value temporal association feature aggregation representation vector respectively represent the temporal aggregation representation of the local temporal features of the temperature value, the heart rate value and the blood pressure value after the feature distribution reverse mask clustering. Considering that the time series of the temperature value, the time series of the heart rate value and the time series of the blood pressure value have relatively significant modal differences in the source domain temporal distribution, the multimodal fusion embedding feature representation of the multimodal embedding fusion representation vector of the vital signs obtained by inputting the temporal association feature aggregation representation vector of the temperature value, the temporal association feature aggregation representation vector of the heart rate value and the temporal association feature aggregation representation vector of the blood pressure value into the vital sign multimodal embedding encoder based on the embedding layer will result in insufficient embedding fusion coding coverage due to the difference in the original feature modalities, thereby causing the outlier class regression inference mapping deviation, affecting the accuracy of the monitoring results obtained by inputting the multimodal embedding fusion representation vector of the vital signs into the classifier-based vital sign monitor. Based on this, when the multimodal embedding fusion representation vector of the vital sign is input into a classifier-based vital sign monitor to obtain a monitoring result, the multimodal embedding fusion representation vector of the vital sign is optimized.
[0057] Specifically, in a preferred embodiment of the present application, the multimodal embedding fusion representation vector of the vital signs is input into a classifier-based vital sign monitor to obtain the monitoring result, specifically including:
[0058] Calculating the feature mean of the multimodal embedding fusion representation vector of the vital sign, and dividing the feature mean by the difference between the maximum eigenvalue and the minimum eigenvalue of the multimodal embedding fusion representation vector of the vital sign to obtain a multimodal embedding fusion distribution representation value of the vital sign;
[0059] Subtracting the vital sign multimodal embedding fusion distribution representation value from one and dividing it by the vital sign multimodal embedding fusion distribution representation value to obtain a vital sign multimodal embedding fusion distribution modulation value;
[0060] Activating the vital sign multimodal embedding fusion representation vector through a probabilistic function to obtain a probabilistic vital sign multimodal embedding fusion representation vector;
[0061] After performing point subtraction on the probabilistic vital sign multimodal embedding fusion representation vector and the vital sign multimodal embedding fusion distribution modulation value, taking the absolute value and calculating the negative of the logarithmic value with base 2 to obtain the probabilistic vital sign multimodal embedding fusion distribution modulation information vector;
[0062] After dividing the vital sign multimodal embedding fusion distribution representation value by one minus the difference of each eigenvalue of the probabilistic vital sign multimodal embedding fusion representation vector, summing all eigenvalues of the probabilistic vital sign multimodal embedding fusion representation vector and dividing by the length of the vital sign multimodal embedding fusion representation vector to obtain a probabilistic vital sign multimodal embedding fusion distribution modulation bias value;
[0063] Performing a multiplication of the probabilistic vital sign multimodal embedding fusion distribution modulation information vector and the probabilistic vital sign multimodal embedding fusion distribution modulation bias value and a weight as a hyperparameter to obtain an optimized vital sign multimodal embedding fusion representation vector;
[0064] The optimized vital sign multimodal embedding fusion representation vector is input into a classifier-based vital sign monitor to obtain the monitoring result.
[0065] Among them, the optimization expression of the multimodal embedding fusion representation vector of the vital signs is:
[0066]
[0067] in, is the probabilistic multimodal embedding fusion representation vector of vital signs, The probabilistic multimodal embedding fusion representation vector of the vital signs is represented by eigenvalues, is the multimodal embedding fusion distribution representation value of the vital signs, represents the logarithmic function value with base 2, is the length of the multimodal embedding fusion representation vector of the vital signs, is the weight as a hyperparameter, It means to make a difference by position. It means adding by position. Represents the optimized vital sign multimodal embedding fusion representation vector.
[0068] And among them:
[0069]
[0070] in, is the multimodal embedding fusion distribution representation value of the vital signs, represents the feature mean of the multimodal embedding fusion representation vector of the vital signs, and They respectively represent the maximum eigenvalue and the minimum eigenvalue of the multimodal embedding fusion representation vector of the vital signs.
[0071] That is, in the above preferred example, the probability information distribution planning based on the eigenvalue of the multimodal embedding fusion representation vector of the vital sign is carried out through the Bernoulli probability modulation distribution of the multimodal embedding fusion representation vector of the vital sign relative to the eigenvalue distribution, and the probability reverse mapping of the probability feature of the multimodal embedding fusion representation vector of the vital sign is used as the extended coverage of the set mapping space of the multimodal embedding fusion representation vector of the vital sign, so as to independently understand the interaction path between the intuitive probability information distribution and the abstract probability space mapping of the multimodal embedding fusion representation vector of the vital sign, so as to improve the accuracy of the monitoring results obtained by the optimized multimodal embedding fusion representation vector input to the classifier-based vital sign monitor by avoiding the counterfactual reasoning mapping of the outlier feature distribution of the multimodal embedding fusion representation vector of the vital sign to the class regression probability. The present application can use artificial intelligence algorithms to analyze and process vital sign data in real time, so as to more accurately identify and predict the health risks of the monitored person by using multiple time series association semantics of vital signs.
[0072] The satellite communication and positioning module 170 is used to send location information and distress information based on the monitoring results. Specifically, when the abnormal vital signs distress sending unit 162 determines that the monitoring result is that the vital signs of the monitored human object are abnormal, the monitoring result is transmitted to the satellite communication and positioning module 170, and the satellite communication and positioning module 170 starts working and automatically sends the location information of the monitored person, that is, the current precise location positioning information, and can also send the monitored person's distress information; specifically, the monitored person can also manually send the current location positioning information and distress information, so that effective communication and distress can be achieved even in remote areas or at sea where traditional communication networks cannot cover.
[0073] In other embodiments of the present application, the data communication terminal simultaneously supports Iridium satellite burst short message function, global positioning function, voice broadcast prompt and other functions. The terminal can also be connected to a smart phone via Bluetooth, and users can realize location services, Iridium satellite short message communication, GNSS positioning, automatic location reporting, SOS alarm and other functions through a dedicated mobile phone app.
[0074] Specifically, the terminal integrates Iridium antenna, Beidou / GPS antenna, Iridium communication module, Beidou / GPS dual-mode positioning module, main control MCU, Bluetooth function, etc., which can complete satellite communication and positioning functions. In an emergency, the SOS function can push the user's current location information to the emergency contact's mobile phone and email to provide timely assistance. Users can freely set the location reporting service. When the service is turned on, emergency contacts can use the APP to view the user's movement trajectory record. The terminal is small in size, low in power consumption, simple to connect, and easy to operate. It can be widely used in satellite communications, data transmission, ship navigation and positioning, and emergency rescue.
[0075] In summary, a data communication terminal 100 with an emergency rescue function based on an embodiment of the present application is explained, which connects the data communication terminal to a smart bracelet worn by a user, and the smart bracelet is used to collect vital signs data. The data communication terminal processes and analyzes the collected vital signs data to determine whether there is an abnormality in the physiological state. If there is an abnormality, the location information and distress message are sent through a satellite communication antenna. Specifically, an artificial intelligence-based data analysis and processing algorithm is used to perform a time series feature analysis on the vital signs data, and then to determine whether there is an abnormality in the vital signs of the monitored human subject, and when an abnormality is detected, the location information and distress message are sent through a satellite communication and positioning module. The present application can more accurately identify and predict the health risks of the monitored person and can also achieve effective communication and distress in places where traditional communication networks are not covered.
[0076] As described above, the data communication terminal 100 with emergency rescue function according to the embodiment of the present application can be implemented in various terminal devices, such as a server for data communication with emergency rescue function. In one example, the data communication terminal 100 with emergency rescue function according to the embodiment of the present application can be integrated into the terminal device as a software module and / or a hardware module. For example, the data communication terminal 100 with emergency rescue function can be a software module in the operating system of the terminal device, or can be an application developed for the terminal device; of course, the data communication terminal 100 with emergency rescue function can also be one of the many hardware modules of the terminal device.
[0077] Alternatively, in another example, the data communication terminal 100 with emergency rescue function and the terminal device may also be separate devices, and the data communication terminal 100 with emergency rescue function may be connected to the terminal device via a wired and / or wireless network and transmit interactive information in accordance with an agreed data format.
[0078] Figure 5FIG. 1 is a flow chart of a data communication method with emergency rescue function according to an embodiment of the present application. Figure 5 As shown, in a data communication method with an emergency rescue function, the method includes: S110, receiving a time series of vital sign data of a monitored human object sent by a smart bracelet, wherein the vital sign data includes body temperature, heart rate and blood pressure; S120, regularizing the time series of the vital sign data according to the parameter sample dimension to obtain a time series of body temperature, a time series of heart rate and a time series of blood pressure; S130, performing sequence encoding on the time series of body temperature, the time series of heart rate and the time series of blood pressure to obtain a sequence of implicit feature vectors associated with local time series of body temperature values, a sequence of implicit feature vectors associated with local time series of heart rate values and a sequence of implicit feature vectors associated with local time series of blood pressure values; S140, encoding the sequence of implicit feature vectors associated with local time series of body temperature values, the sequence of implicit feature vectors associated with local time series of heart rate values and the sequence of implicit feature vectors associated with local time series of blood pressure values. The sequence of implicit feature vectors and the sequence of implicit feature vectors of local temporal association of blood pressure values are respectively input into the feature distribution reverse mask clustering analysis network to obtain the temperature value temporal association feature aggregation representation vector, the heart rate value temporal association feature aggregation representation vector and the blood pressure value temporal association feature aggregation representation vector; S150, the temperature value temporal association feature aggregation representation vector, the heart rate value temporal association feature aggregation representation vector and the blood pressure value temporal association feature aggregation representation vector are multimodally embedded and fused to obtain a vital sign multimodal embedding fusion representation vector as a vital sign multimodal embedding fusion representation feature; S160, based on the vital sign multimodal embedding fusion representation feature, a monitoring result is obtained, and the monitoring result is used to indicate whether there is an abnormality in the vital sign of the monitored human subject; S170, based on the monitoring result, location information and distress information are sent.
[0079] Specifically, those skilled in the art can understand that the specific operations of each step in the above-mentioned data communication method with emergency rescue function have been referred to above. Figures 1 to 4 The data communication terminal with emergency rescue function has been described in detail, and therefore, its repeated description will be omitted.
[0080] In summary, a data communication method with emergency rescue function based on the embodiment of the present application is explained, which connects a data communication terminal to a smart bracelet worn by a user, and the smart bracelet is used to collect vital signs data. The data communication terminal processes and analyzes the collected vital signs data to determine whether there is an abnormality in the physiological state. If there is an abnormality, the location information and distress message are sent through a satellite communication antenna. Specifically, an artificial intelligence-based data analysis and processing algorithm is used to perform a time series feature analysis on the vital signs data to determine whether there is an abnormality in the vital signs of the monitored human subject, and when an abnormality is detected, the location information and distress message are sent through a satellite communication and positioning module. The present application can more accurately identify and predict the health risks of the monitored person and can also achieve effective communication and distress in places where traditional communication networks are not covered.
[0081] In the several embodiments provided in this application, it should be understood that the disclosed terminal and method can be implemented in other ways. For example, the terminal embodiment described above is only illustrative, for example, the division of the modules is only a logical function division, and there may be other division methods in actual implementation.
[0082] The modules described as separate components may or may not be physically separated, and the components shown as modules may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.
[0083] In addition, each functional module in each embodiment of the present application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of hardware plus software functional modules.
[0084] It is obvious to those skilled in the art that the present application is not limited to the details of the above exemplary embodiments, and that the present application can be implemented in other specific forms without departing from the spirit or essential features of the present application.
Claims
1. Data communication terminal with emergency rescue function, including: A human vital signs data receiving module, used to receive a time series of vital signs data of a monitored human subject sent by a smart bracelet, wherein the vital signs data include body temperature, heart rate and blood pressure; A vital sign data regularization module, used for regularizing the time series of the vital sign data according to the parameter sample dimension to obtain a time series of body temperature values, a time series of heart rate values, and a time series of blood pressure values; A vital sign data local time series association extraction module, used for performing sequence encoding on the time series of the body temperature value, the time series of the heart rate value and the time series of the blood pressure value to obtain a sequence of implicit feature vectors of local time series association of the body temperature value, a sequence of implicit feature vectors of local time series association of the heart rate value and a sequence of implicit feature vectors of local time series association of the blood pressure value; A time series correlation feature aggregation module, used to input the sequence of the local time series correlation implicit feature vectors of the temperature value, the sequence of the local time series correlation implicit feature vectors of the heart rate value and the sequence of the local time series correlation implicit feature vectors of the blood pressure value into a feature distribution reverse mask clustering analysis network to obtain a time series correlation feature aggregation representation vector of the temperature value, a time series correlation feature aggregation representation vector of the heart rate value and a time series correlation feature aggregation representation vector of the blood pressure value; A vital sign multimodal fusion module, used for multimodally embedding and fusing the temperature value time series associated feature aggregation representation vector, the heart rate value time series associated feature aggregation representation vector and the blood pressure value time series associated feature aggregation representation vector to obtain a vital sign multimodal embedding fusion representation vector as a vital sign multimodal embedding fusion representation feature; A monitoring abnormality judgment module, used to obtain a monitoring result based on the multimodal embedding fusion representation feature of the vital signs, wherein the monitoring result is used to indicate whether there is an abnormality in the vital signs of the monitored human subject; The satellite communication and positioning module is used to send location information and distress information based on the monitoring results.
2. The data communication terminal with emergency rescue function according to claim 1, characterized in that: The vital sign data local time series association extraction module is used to: input the time series of the body temperature values, the time series of the heart rate values and the time series of the blood pressure values into a sequence encoder based on the 1D-CNN model to obtain a sequence of implicit feature vectors of local time series association of the body temperature values, a sequence of implicit feature vectors of local time series association of the heart rate values and a sequence of implicit feature vectors of local time series association of the blood pressure values.
3. The data communication terminal with emergency rescue function according to claim 2, characterized in that: The time series correlation feature aggregation module includes: A body temperature value associated implicit feature clustering calculation unit, used for clustering the sequence of the body temperature value local time series associated implicit feature vectors to obtain the body temperature value local time series associated implicit clustering center vector; A temperature value time series correlation coefficient calculation unit, used for calculating the correlation coefficient between each temperature value local time series associated implicit feature vector in the sequence of temperature value local time series associated implicit feature vectors and the temperature value local time series associated implicit clustering center vector to obtain a sequence of temperature value local time series correlation coefficients; A temperature value time series anti-correlation coefficient calculation unit, used for taking the inverse of each temperature value local time series correlation coefficient in the sequence of temperature value local time series correlation coefficients to obtain a sequence of temperature value local time series anti-correlation coefficients; A body temperature value time series anti-correlation coefficient normalization unit, used for normalizing the sequence of the body temperature value local time series anti-correlation coefficients using a Sigmoid function to obtain a sequence of the body temperature value local time series anti-correlation weight coefficients; A body temperature value feature weighted suppression unit, configured to weight each body temperature value local time series associated implicit feature vector in the sequence of body temperature value local time series associated implicit feature vectors by using each body temperature value local time series anti-correlation weight coefficient in the sequence of body temperature value local time series anti-correlation weight coefficients as a weight, so as to obtain a sequence of suppressed body temperature value local time series associated implicit feature vectors; A body temperature value local time series feature enhancement unit, used for calculating the position difference between the body temperature value local time series associated implicit feature vector and the body temperature value local time series associated implicit feature vector corresponding to each group in the sequence of the body temperature value local time series associated implicit feature vector and the sequence of the body temperature value local time series associated implicit feature vector to obtain a sequence of enhanced body temperature value local time series associated implicit feature vector; The body temperature value feature aggregation unit is used to calculate the position mean vector of the sequence of the enhanced body temperature value local time series associated implicit feature vectors to obtain the body temperature value time series associated feature aggregation representation vector.
4. The data communication terminal with emergency rescue function according to claim 3, characterized in that: The body temperature value associated implicit feature clustering calculation unit is used to calculate the mean vector of the sequence of the body temperature value local time series associated implicit feature vectors to obtain the body temperature value local time series associated implicit clustering center vector.
5. The data communication terminal with emergency rescue function according to claim 4, characterized in that: The body temperature value time series correlation coefficient calculation unit is used to: Performing vector multiplication of the transposed vector of the local temporal association implicit feature vector of the body temperature value with the local temporal association implicit cluster center vector of the body temperature value to obtain the local temporal association implicit association value of the body temperature value; Respectively calculating the implicit characteristic vector of the local time series association of the body temperature value and the norm of the implicit cluster center vector of the local time series association of the body temperature value to obtain the implicit norm value of the local time series association of the body temperature value and the implicit cluster norm value of the local time series association of the body temperature value; Multiplying the body temperature value local time series association implicit norm value and the body temperature value local time series association implicit clustering norm value to obtain the body temperature value local time series association interaction value; The local time series association implicit association value of the body temperature value is divided by the local time series association interaction value of the body temperature value to obtain the local time series correlation coefficient of the body temperature value corresponding to the local time series association implicit feature vector of the body temperature value.
6. The data communication terminal with emergency rescue function according to claim 5, characterized in that: The temperature value time series anti-correlation coefficient normalization unit is used for: Taking the negative number of each local time series anti-correlation coefficient of body temperature values in the sequence of local time series anti-correlation coefficients of body temperature values as the exponent of the natural constant to calculate the exponential function value with the natural constant as the base according to the position to obtain the sequence of local time series anti-correlation class support coefficients of body temperature values; The reciprocal of the sum of each local time series anti-correlation class support coefficient of the body temperature value in the sequence of the local time series anti-correlation class support coefficient of the body temperature value and a constant one is calculated to obtain the sequence of the local time series anti-correlation weight coefficients of the body temperature value.
7. The data communication terminal with emergency rescue function according to claim 6, characterized in that: The vital sign multimodal fusion module is used to: input the temperature value time series associated feature aggregation representation vector, the heart rate value time series associated feature aggregation representation vector and the blood pressure value time series associated feature aggregation representation vector into the vital sign multimodal embedding encoder based on the embedding layer to obtain the vital sign multimodal embedding fusion representation vector.
8. The data communication terminal with emergency rescue function according to claim 7, characterized in that: The monitoring abnormality judgment module includes: A monitoring result generating unit, used for inputting the multimodal embedding fusion representation vector of the vital sign into the classifier-based vital sign monitor to obtain the monitoring result, wherein the monitoring result is used for indicating whether there is an abnormality in the vital sign of the monitored human subject; The abnormal vital sign distress sending unit is used to send position information and distress information through the satellite communication and positioning module in response to the monitoring result that the vital sign of the monitored human object is abnormal.
9. A data communication method with emergency rescue function, characterized in that: include: Receive a time series of vital sign data of a monitored human subject sent by a smart bracelet, wherein the vital sign data includes body temperature, heart rate and blood pressure: Regularizing the time series of the vital sign data according to the parameter sample dimension to obtain a time series of body temperature values, a time series of heart rate values, and a time series of blood pressure values; Sequence-encoding the time series of the body temperature values, the time series of the heart rate values, and the time series of the blood pressure values to obtain a sequence of implicit feature vectors associated with local time series of the body temperature values, a sequence of implicit feature vectors associated with local time series of the heart rate values, and a sequence of implicit feature vectors associated with local time series of the blood pressure values; Inputting the sequence of the local temporal association implicit feature vectors of the temperature value, the sequence of the local temporal association implicit feature vectors of the heart rate value and the sequence of the local temporal association implicit feature vectors of the blood pressure value into the feature distribution reverse mask clustering analysis network respectively to obtain the temporal association feature aggregation representation vector of the temperature value, the temporal association feature aggregation representation vector of the heart rate value and the temporal association feature aggregation representation vector of the blood pressure value; Performing multimodal embedding fusion on the temperature value time series associated feature aggregation representation vector, the heart rate value time series associated feature aggregation representation vector, and the blood pressure value time series associated feature aggregation representation vector to obtain a vital sign multimodal embedding fusion representation vector as a vital sign multimodal embedding fusion representation feature; Based on the multimodal embedding fusion representation feature of the vital signs, a monitoring result is obtained, wherein the monitoring result is used to indicate whether there is an abnormality in the vital signs of the monitored human subject; Based on the monitoring results, location information and distress information are sent.
10. The data communication method with emergency rescue function according to claim 9, characterized in that: The time series of the body temperature values, the time series of the heart rate values and the time series of the blood pressure values are sequence-encoded to obtain a sequence of implicit feature vectors associated with local time series of the body temperature values, a sequence of implicit feature vectors associated with local time series of the heart rate values and a sequence of implicit feature vectors associated with local time series of the blood pressure values, including: inputting the time series of the body temperature values, the time series of the heart rate values and the time series of the blood pressure values into a sequence encoder based on a 1D-CNN model to obtain a sequence of implicit feature vectors associated with local time series of the body temperature values, a sequence of implicit feature vectors associated with local time series of the heart rate values and a sequence of implicit feature vectors associated with local time series of the blood pressure values.
