Wearable device alarm method and system

Abstract

This application proposes a wearable device alarm method and system, which includes: receiving user monitoring status information, user motion information, and user environment information; receiving user first-class attributes and user second-class attributes; obtaining a first monitoring attribute concentration interval based on the user first-class attributes; obtaining a monitoring attribute system error based on the user second-class attributes; correcting the first monitoring attribute concentration interval according to the monitoring attribute system error to obtain a second monitoring attribute concentration interval; calculating the deviation coefficient between the user monitoring status information and the second monitoring attribute concentration interval, fusing the deviation coefficient and the user's location to generate an alarm signal and execute the alarm. By fusing the user's general attributes and personalized attributes to dynamically correct the monitoring data, the semantic richness of the alarm information is improved, and the alarm accuracy of the wearable device is improved, thereby achieving optimal scheduling of limited rescue resources.

Description

Technical Field

[0001] The present invention relates to the field of intelligent alarms, and in particular to a wearable device alarm method and system. Background Art

[0002] Currently, wearable device alarm mechanisms are primarily based on preset thresholds. This means that an alarm is triggered when the monitored user status information exceeds a predetermined threshold. For example, if a user's heart rate is too high or too low, blood pressure is abnormal, or a fall is detected, the device will issue an alarm signal based on the preset threshold and, combined with the user's location information, notify relevant personnel or agencies for rescue. However, due to significant differences in physiological characteristics, health conditions, and daily habits among different users, a unified threshold standard is difficult to adapt to individual needs. This threshold judgment mechanism has significant shortcomings, leading to inappropriate allocation of rescue resources. For example, for users who frequently engage in high-intensity exercise, an elevated heart rate may be normal; however, for those with weaker heart function, even a small change in heart rate may indicate a serious problem. Therefore, existing wearable device alarms suffer from the technical problem of relying on preset threshold comparisons, lacking the semantic richness of the alarm information, and thus failing to effectively allocate resources based on the varying needs of different users. Summary of the Invention

[0003] The present invention aims to solve the technical problem in the prior art that wearable device alarms are only based on comparisons of predetermined thresholds, the semantic richness of the alarm information is insufficient, and effective resource scheduling cannot be performed according to the needs of different users. A wearable device alarm method and system are provided to solve the problem.

[0004] The technical solution of the present invention to solve the above technical problems is as follows:

[0005] In a first aspect, the present invention provides an alarm method for a wearable device, comprising: receiving user monitoring status information, user motion information and user environment information; receiving user first-class attributes and user second-class attributes, wherein the user first-class attributes are general attributes predefined by an administrator, and the user second-class attributes are personalized attributes that the user can optionally add; based on the user first-class attributes, performing monitoring attribute sample statistics on the user motion information and the user environment information to obtain a first monitoring attribute concentration interval; based on the user second-class attributes, in combination with the user motion information and the user environment information, traversing the monitoring attribute set to perform monitoring attribute sample statistics to obtain a monitoring attribute system error; according to the monitoring attribute system error, correcting the first monitoring attribute concentration interval to obtain a second monitoring attribute concentration interval; calculating the deviation coefficient between the user monitoring status information and the second monitoring attribute concentration interval, fusing the deviation coefficient and the user position to generate an alarm signal to execute an alarm.

[0006] Optionally, the deviation coefficient between the user monitoring status information and the second monitoring attribute concentration interval is calculated, including: comparing the user monitoring status information and the monitoring status deviation vector matrix of the second monitoring attribute concentration interval; analyzing the abnormal event triggering probability that satisfies the monitoring status deviation vector matrix; when the abnormal event triggering probability is less than or equal to the triggering probability threshold, returning the process to start the execution loop; when the abnormal event triggering probability is greater than the triggering probability threshold, setting the abnormal event triggering probability to the deviation coefficient.

[0007] Optionally, analyzing the probability of abnormal event triggering that satisfies the monitoring state deviation vector matrix includes: configuring a first monitoring state deviation vector record matrix up to the Nth monitoring state deviation vector record matrix, where N is an integer, and N≥50000; taking the user-class attribute as a constraint, collecting several user abnormal event trigger identifiers that satisfy the first monitoring state deviation vector record matrix, the user abnormal event trigger identifier is generated according to the decision result of the rescue end in historical data, if it belongs to an abnormal event, the user abnormal event trigger identifier is equal to 1, otherwise, the user abnormal event trigger identifier is equal to 0; counting the number of user abnormal event trigger identifiers equal to 1, and ratio thereof to the number of the several user abnormal event trigger identifiers, and setting it as the first abnormal event trigger probability true value; until the Nth abnormal event trigger probability true value is obtained; taking the first abnormal event trigger probability true value up to the Nth abnormal event trigger probability true value as supervision, taking the first monitoring state deviation vector record matrix up to the Nth monitoring state deviation vector record matrix as input, configuring multiple groups of data, training the abnormal event trigger probability predictor, and analyzing the abnormal event trigger probability that satisfies the monitoring state deviation vector matrix.

[0008] Optionally, the first abnormal event trigger probability true value up to the Nth abnormal event trigger probability true value is used as supervision, the first monitoring state deviation vector record matrix up to the Nth monitoring state deviation vector record matrix is ​​used as input, multiple groups of data are configured, and the abnormal event trigger probability predictor is trained, including: step one: setting a training process quantity threshold, calling the multiple groups of data, and training the first abnormal event trigger probability predictor; step two: extracting the output error vector set of the first abnormal event trigger probability predictor; step three: when the output error vector variance of the output error vector set is greater than or equal to the variance threshold, returning to step one to execute the loop; step four: when the output of the output error vector set is greater than or equal to the variance threshold, returning to step one to execute the loop; step four: when the output of the output error vector set is greater than or equal to the variance threshold, returning to step one to execute the loop; step five: when the output of the output error vector set is greater than or equal to the variance threshold, returning to step one to execute the loop; step six: when the output of the output error vector set is greater than or equal to the variance threshold, returning to step six The error vector variance is less than the variance threshold, the system error vector is counted, and the system error vector modulus is greater than or equal to the convergence threshold, a first residual block is constructed, and it is fused with the first abnormal event trigger probability predictor to generate a second abnormal event trigger probability predictor architecture, and return to step one to execute the loop, wherein the second abnormal event trigger probability predictor architecture is equal to the sum of the outputs of the first abnormal event trigger probability predictor and the first residual block; Step five: when the output error vector variance of the output error vector set is less than the variance threshold, the system error vector is counted, and the system error vector modulus is less than the convergence threshold, the first abnormal event trigger probability predictor is set as the abnormal event trigger probability predictor.

[0009] Optionally, based on the user-type attributes, monitoring attribute sample statistics are performed on the user motion information and the user environment information to obtain a first monitoring attribute concentration interval, including: based on the user-type attributes, according to the user motion information and the user environment information, combined with a predefined attribute deviation threshold, constructing a first query constraint condition; retrieving a first sample set that meets the first query constraint condition, wherein any sample of the first sample set includes user motion record data and user environment record data; based on the user-type attributes, according to the user motion record data and the user environment record data, combined with a predefined attribute deviation threshold, constructing a second query constraint condition; retrieving a second sample set that meets the second query constraint condition; and counting the monitoring attribute concentration interval of the first sample set and the second sample set, and setting it as the first monitoring attribute concentration interval.

[0010] Optionally, based on the second category of user attributes, combined with the user motion information and the user environment information, the monitoring attribute set is traversed to perform monitoring attribute sample statistics to obtain a monitoring attribute system error, including: based on the second category of user attributes, based on the first category of user attributes, combined with the user motion information and the user environment information, the monitoring attribute set is traversed to perform monitoring attribute sample statistics to obtain a third monitoring attribute concentration interval; calculating the first monitoring attribute intersection-and-union ratio of the third monitoring attribute concentration interval and the first monitoring attribute concentration interval, and configuring the first monitoring attribute system error according to the first monitoring attribute intersection-and-union ratio; until the Mth monitoring attribute intersection-and-union ratio of the third monitoring attribute concentration interval and the first monitoring attribute concentration interval is calculated, and the Mth monitoring attribute system error is configured according to the Mth monitoring attribute intersection-and-union ratio; the first monitoring attribute system error until the Mth monitoring attribute system error is added to the monitoring attribute system error.

[0011] Optionally, configuring the first monitoring attribute system error according to the intersection-and-union ratio of the first monitoring attribute includes: when the intersection-and-union ratio of the first monitoring attribute is less than the intersection-and-union ratio threshold, calculating the interval upper boundary deviation vector set and the interval lower boundary deviation vector set of the first monitoring attribute; counting the majority deviation vector of the interval upper boundary deviation vector set, setting it as the first monitoring attribute upper boundary system error, counting the majority deviation vector of the interval lower boundary deviation vector set, setting it as the first monitoring attribute lower boundary system error; adding the first monitoring attribute upper boundary system error and the first monitoring attribute lower boundary system error to the first monitoring attribute system error; when the intersection-and-union ratio of the first monitoring attribute is greater than or equal to the intersection-and-union ratio threshold, the first monitoring attribute system error is set to 0.

[0012] In a second aspect, the present invention provides a wearable device alarm system, comprising: an information receiving module for receiving user monitoring status information, user motion information and user environment information; a user attribute receiving module for receiving user first-class attributes and user second-class attributes, wherein the user first-class attributes are general attributes predefined by an administrator, and the user second-class attributes are personalized attributes that can be optionally added by the user; a general attribute analysis module for performing monitoring attribute sample statistics on the user motion information and the user environment information based on the user first-class attributes to obtain a first monitoring attribute concentration interval; a personalized attribute analysis module for traversing the monitoring attribute set to perform monitoring attribute sample statistics based on the user second-class attributes, combined with the user motion information and the user environment information, to obtain a monitoring attribute system error; a concentration interval correction module for correcting the first monitoring attribute concentration interval according to the monitoring attribute system error to obtain a second monitoring attribute concentration interval; an alarm execution module for calculating the deviation coefficient between the user monitoring status information and the second monitoring attribute concentration interval, fusing the deviation coefficient and the user position to generate an alarm signal to execute the alarm.

[0013] The beneficial effects of the present invention are:

[0014] By collecting user monitoring status, movement, and environmental information, comprehensive user data collection is achieved, ensuring a sufficient data foundation for subsequent analysis. By collecting user primary and secondary attributes, the two types of attribute information are distinguished, laying the foundation for subsequent personalized analysis. Based on the primary attributes, monitoring attribute sample statistics are performed on the user's movement and environmental information to obtain the first monitoring attribute concentration interval. Preliminary statistical analysis of the movement and environmental information using universal attributes is performed to determine the normal state interval based on universal attributes, providing a basic reference range for subsequent personalized corrections. Based on the secondary attributes, combined with the user's movement and environmental information, monitoring attribute sample statistics are traversed through the monitoring attribute set to obtain the monitoring attribute systematic error. By introducing personalized attributes, the systematic error relative to the universal model is calculated, identifying monitoring deviations caused by individual differences and providing a basis for subsequent precise corrections. Based on the monitoring attribute systematic error, the first monitoring attribute concentration interval is corrected to obtain the second monitoring attribute concentration interval, achieving the transition from universal interval to personalized interval, making the monitoring interval more tailored to the user's individual characteristics and improving the accuracy of anomaly detection. By calculating the deviation coefficient between the user monitoring status information and the concentrated interval of the second monitoring attribute, the deviation coefficient and the user location are integrated to generate an alarm signal to execute the alarm. The introduction of the deviation coefficient makes the alarm information differentiated in degree. At the same time, combined with the location information, it can help the rescue party determine the rescue priority and resource allocation strategy, thereby improving the semantic richness of the alarm information.

[0015] Through the above technical solution, a transition from threshold comparison to a dynamic alarm mechanism integrating individual user characteristics is achieved, which solves the problem of insufficient semantic richness of alarm information in the existing technology and provides support for the optimal scheduling of limited rescue resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 A flowchart of a wearable device alarm method provided by the present invention;

[0017] Figure 2 This is a structural schematic diagram of a wearable device alarm system provided by the present invention.

[0018] In the accompanying drawings, the components represented by the reference numerals are as follows:

[0019] Information receiving module 11, user attribute receiving module 12, general attribute analysis module 13, personalized attribute analysis module 14, centralized interval correction module 15, alarm execution module 16. DETAILED DESCRIPTION

[0020] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without making any creative efforts shall fall within the scope of protection of the present invention.

[0021] In the description of the present invention, the terms "first" and "second" are used for descriptive purposes only and should not be understood to indicate or imply relative importance or implicitly specify the number of the technical features indicated. Therefore, a feature specified as "first" or "second" may explicitly or implicitly include one or more of the specified features. In the description of the present invention, "plurality" means two or more, unless otherwise specifically defined.

[0022] In the description of the present invention, the term "for example" is used to mean "used as an example, illustration or illustration". Any embodiment of the present invention described as "for example" is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is given to enable any person skilled in the art to implement and use the present invention. In the following description, details are listed for the purpose of explanation. It should be understood that a person of ordinary skill in the art can recognize that the present invention can be implemented without using these specific details. In other examples, well-known structures and processes are not elaborated in detail to avoid obscuring the description of the present invention with unnecessary details. Therefore, the present invention is not intended to be limited to the embodiments shown, but is consistent with the widest scope consistent with the principles and features disclosed herein.

[0023] Example 1, as Figure 1 As shown, an embodiment of the present invention provides a wearable device alarm method, which is applied to a wearable device, including:

[0024] S100: Receive user monitoring status information, user movement information, and user environment information.

[0025] Specifically, wearable devices collect and receive three types of user-related information in real time through various built-in sensors, namely user monitoring status information, user movement information and user environment information.

[0026] Among them, user monitoring status information refers to the user's physiological parameter data collected by wearable devices, such as heart rate, blood pressure, body temperature, blood oxygen saturation, respiratory rate and other vital signs indicators. These data can reflect the user's current health status. User motion information refers to the user's motion-related data collected by wearable devices through sensors such as accelerometers and gyroscopes, including but not limited to the user's walking steps, movement posture, movement speed, fall detection, static state recognition, etc. These data can reflect the user's movement status and potential abnormal behavior. User environmental information refers to the relevant parameters of the user's environment collected by wearable devices through temperature and humidity sensors, light sensors, air pressure sensors, etc., including but not limited to ambient temperature, humidity, atmospheric pressure, ambient noise, light intensity, etc. These data can reflect the safety status and potential risks of the user's environment.

[0027] By receiving user monitoring status information, user movement information and user environment information, it provides raw data support for subsequent abnormal status analysis and intelligent alarms.

[0028] S200: receiving user first-category attributes and user second-category attributes, wherein the user first-category attributes are general attributes predefined by an administrator, and the user second-category attributes are personalized attributes that can be optionally added by the user.

[0029] Specifically, wearable devices receive two types of user attribute data, namely user type 1 attributes and user type 2 attributes. These two types of attribute data constitute the user's personal profile, providing a personalized reference benchmark for subsequent monitoring and analysis.

[0030] Among them, user-type 1 attributes refer to standardized basic user information pre-defined by the administrator and fall into the category of general attributes. These attributes typically include, but are not limited to, basic physiological characteristics such as the user's gender, age, height, weight, blood type, and basal metabolic rate. General attributes are objective and universal, and are an important basis for basic judgments. They can be used to divide users into different basic population categories and provide a basic reference standard for subsequent abnormal status judgments. User-type 2 attributes refer to personalized information that is independently selected and input by the user based on their personal circumstances and fall into the category of non-standardized personalized attributes. These attributes typically include, but are not limited to, highly personalized information such as the user's eating habits, past medical history, drug allergy history, recent exercise data, and lifestyle habits. Personalized attributes can reflect the user's special circumstances and individual differences, helping to more accurately identify the user's abnormal status, reduce false alarm rates, and improve the accuracy and pertinence of alarms.

[0031] By receiving user first-class attributes and user second-class attributes, we can build a more accurate user monitoring mechanism based on general standards and combined with the user's personalized characteristics, providing comprehensive user feature support for subsequent abnormal status judgment and intelligent alarm.

[0032] S300: Based on the first type of user attributes, perform monitoring attribute sample statistics on the user movement information and the user environment information to obtain a first monitoring attribute concentration interval.

[0033] Specifically, based on the aforementioned received user attributes (i.e., general attributes predefined by the administrator, such as the user's gender, age, height, weight and other basic information), a statistical analysis of monitoring attribute samples is performed on the user's motion information and user environment information to obtain a first monitoring attribute concentration interval.

[0034] First, using user attributes as the basic screening criteria, search the historical database for user groups with similar general attributes to the current user. For example, for an adult male user aged 30-35, 175-180cm tall, and weighing 70-75kg, retrieve historical data of male users with similar age, height, and weight ranges. Then, statistical analysis is performed on the motion and environmental information of the screened user groups to calculate the normal distribution range of each monitored attribute, including but not limited to the average daily step range, normal activity intensity range, heart rate variation range, blood pressure fluctuation range, suitable ambient temperature range, and safe ambient humidity range.

[0035] By statistically analyzing the monitored attribute samples, we obtain a concentrated interval of monitored attributes based on common attributes, namely the first concentrated interval. This interval represents the reasonable fluctuation range under normal circumstances for a group of users with similar user motion and user environment information, providing a preliminary reference standard for subsequent abnormal state judgment.

[0036] S400: Based on the second type of user attributes, combined with the user movement information and the user environment information, traverse the monitoring attribute set to perform monitoring attribute sample statistics to obtain a monitoring attribute system error.

[0037] Specifically, based on the user's second-category attributes (i.e., personalized attributes that users can optionally add, such as diet recipes, past medical history, recent exercise data, etc.), combined with user exercise information and user environment information, the monitoring attribute set is comprehensively traversed and deeply statistically analyzed to obtain the monitoring attribute system error.

[0038] Specifically, the system first uses the user's second-category attributes as additional constraints and searches the historical database for specific user subgroups that meet both the user's first-category attributes (general attributes) and possess similar second-category attributes (personal attributes). For example, for a user with a history of mild hypertension, who regularly engages in low-intensity aerobic exercise and eats a low-salt, low-fat diet, the system identifies a specific user group with similar health conditions, exercise habits, and dietary habits. Statistical analysis is then performed on the historical monitoring data for this specific user group to determine the concentration interval for the monitoring attributes after accounting for personalized factors, known as the third monitoring attribute concentration interval. By comparing the differences between the third monitoring attribute concentration interval and the first monitoring attribute concentration interval, the systematic error of each monitoring attribute is calculated. The systematic error of a monitoring attribute reflects the extent to which a user's personalized characteristics modify the general monitoring standard. For example, for a user with a history of mild hypertension, the upper limit of their normal blood pressure range may need to be slightly higher than the general standard; for a user who regularly engages in aerobic exercise, their normal resting heart rate may be lower than that of the average person of the same age.

[0039] Through personalized statistical analysis, the systematic error of monitoring attributes based on the user's personalized characteristics is obtained, which will be used for subsequent personalized adjustment of monitoring standards, thereby improving the accuracy and pertinence of alarms and reducing false alarms or missed alarms due to individual differences.

[0040] S500: According to the monitoring attribute system error, correct the first monitoring attribute concentration interval to obtain a second monitoring attribute concentration interval.

[0041] Specifically, the obtained monitoring attribute system error is used to accurately correct the established first monitoring attribute concentration interval, thereby obtaining a more personalized second monitoring attribute concentration interval, which can more accurately reflect the normal state range of a specific user.

[0042] Specifically, the corresponding monitoring attribute system error is applied to the upper and lower boundaries of the first monitoring attribute concentration interval. For each monitoring attribute, its upper boundary is added to the corresponding upper boundary system error value, and its lower boundary is added to the corresponding lower boundary system error value, thereby generating a new boundary value that has been individually adjusted. For example, if a user's resting heart rate is usually 5-10 beats / minute lower than that of the same age group due to long-term physical exercise, the lower limit of the user's heart rate normal range will be adjusted down by 5-10 beats / minute accordingly; if a user's normal body temperature is slightly higher than that of ordinary people due to a specific physique, the upper limit of the user's normal body temperature range will be appropriately increased.

[0043] Through a personalized correction process, the first monitoring attribute concentration interval (a group statistical interval based on common attributes) is transformed into a second monitoring attribute concentration interval (a customized interval that takes individual factors into account). This enables personalized and precise monitoring based on group statistics. This avoids the potential misjudgments caused by relying solely on universal standards while overcoming the data shortage issues that can arise with purely individualized monitoring. The second monitoring attribute concentration interval provides a more accurate reference standard for subsequent abnormal state judgments, enhances the semantic richness of alarm information, and effectively improves the accuracy and reliability of wearable device alarm judgments.

[0044] S600: Calculate the deviation coefficient between the user monitoring status information and the second monitoring attribute concentration interval, fuse the deviation coefficient and the user location to generate an alarm signal and execute an alarm.

[0045] Specifically, the user monitoring status information collected in real time is compared and analyzed with the established second monitoring attribute concentration interval, the deviation coefficient is calculated, and combined with the user's location information, an intelligent alarm signal is generated and a precise alarm is executed.

[0046] First, the user's monitored status information (including physiological indicators such as heart rate, blood pressure, body temperature, blood oxygen, and posture, movement status, etc.) is compared item by item with the second monitoring attribute concentration interval. The deviation degree of each monitored attribute is quantified and a comprehensive deviation coefficient is calculated. This deviation coefficient quantifies the degree and urgency of the user's abnormal status. The larger the deviation coefficient, the further the user's status deviates from the normal range, the higher the potential risk, and the higher the priority for rescue. The calculated deviation coefficient is then integrated with the user's real-time location. The user's location includes the user's geographic coordinates, the type of environment (e.g., indoors, outdoors, in water, at high altitude), and the presence of surrounding facilities. By integrating the deviation coefficient and location information, an alarm signal is generated that contains multi-dimensional information such as the degree of abnormality, location accuracy, and environmental risk. An alarm is then executed based on this integrated alarm signal. For example, for minor anomalies, only the user is notified; for moderate anomalies, the user's emergency contacts are notified simultaneously; and for severe anomalies, an emergency call is automatically sent to the rescue center, including detailed user status and precise location.

[0047] Through the intelligent alarm mechanism based on the fusion of deviation coefficient and location information, the problem of insufficient information semantic richness in traditional alarm systems is effectively solved, so that rescue resources can be prioritized according to the actual degree of urgency, and the alarm accuracy and efficiency of wearable devices can be improved.

[0048] Furthermore, calculating the deviation coefficient between the user monitoring state information and the second monitoring attribute concentration interval includes:

[0049] S610: Compare the user monitoring state information and the monitoring state deviation vector matrix of the second monitoring attribute concentration interval;

[0050] S620: Analyze the trigger probability of abnormal events that meet the monitoring state deviation vector matrix;

[0051] S630: When the abnormal event trigger probability is less than or equal to the trigger probability threshold, the process returns to start the execution loop;

[0052] S640: When the abnormal event trigger probability is greater than the trigger probability threshold, the abnormal event trigger probability is set as the deviation coefficient.

[0053] In a preferred embodiment, the user's monitoring state information is first compared with the second monitoring attribute concentration interval to generate a monitoring state deviation vector matrix. Specifically, the user's current physiological indicators (such as heart rate, blood pressure, body temperature, and blood oxygen saturation), motion state parameters (such as posture, acceleration, angular velocity), and environmental parameters (such as ambient temperature and humidity) in the user monitoring state information are compared with the upper and lower limits of the corresponding second monitoring attribute concentration interval. The deviation value and deviation direction of each indicator are calculated. This multi-dimensional deviation data is then organized into a structured deviation vector matrix to obtain a monitoring state deviation vector matrix. This matrix not only contains the deviation degree of each indicator but also reflects the correlation and combination patterns between indicators. Then, based on the obtained monitoring state deviation vector matrix, the triggering probability of various abnormal events that the user may encounter is analyzed. For example, the current monitoring state deviation vector matrix is ​​matched with deviation vector matrices corresponding to known abnormal event patterns in historical data and similarity is calculated to estimate the probability of various abnormal events (such as falls, drowning, heart attacks, hypoxia, hypothermia, etc.) that may be caused by the user's current state. Through probability-based risk assessment, predictions can be made at the early stages of abnormal conditions, without having to wait until indicators deviate seriously before triggering alarms.

[0054] The calculated abnormal event trigger probability is then compared with the preset trigger probability threshold. When the abnormal event trigger probability is less than or equal to the trigger probability threshold, it indicates that although the user's current state has deviated, the risk level is controllable. The system will return to the starting point of the monitoring process and continue the monitoring cycle to maintain continuous observation of the user's state. When the abnormal event trigger probability is greater than the trigger probability threshold, it indicates that the user's current state has deviated to the risk warning line. The abnormal event trigger probability value is directly set as the deviation coefficient, which serves as an important parameter for the subsequent alarm process. This deviation coefficient not only quantifies the degree of abnormality in the user's state, but also reflects the type and urgency of potential risk events, providing a basis for subsequent intelligent alarms.

[0055] By converting multi-dimensional monitoring data into deviation coefficients, the alarm signal can accurately reflect the nature and urgency of the user's abnormal status, thereby supporting the efficient scheduling and precise intervention of rescue resources.

[0056] Furthermore, the probability of triggering an abnormal event that satisfies the monitoring state deviation vector matrix is ​​analyzed, including:

[0057] S621: configuring the first monitoring state deviation vector record matrix up to the Nth monitoring state deviation vector record matrix, where N is an integer and N≥50000;

[0058] S622: Using the user attribute as a constraint, collect a number of user abnormal event trigger flags that satisfy the first monitoring state deviation vector record matrix. The user abnormal event trigger flag is generated based on the decision result of the rescue end in historical data. If it is an abnormal event, the user abnormal event trigger flag is equal to 1; otherwise, the user abnormal event trigger flag is equal to 0.

[0059] S623: Count the number of user abnormal event triggering flags equal to 1, and compare the number of the plurality of user abnormal event triggering flags to set the result as a true value of the first abnormal event triggering probability;

[0060] S624: until the true value of the trigger probability of the Nth abnormal event is obtained;

[0061] S625: Using the first abnormal event trigger probability true value up to the Nth abnormal event trigger probability true value as supervision, using the first monitoring state deviation vector record matrix up to the Nth monitoring state deviation vector record matrix as input, configuring multiple groups of data, training the abnormal event trigger probability predictor, and analyzing the abnormal event trigger probability that satisfies the monitoring state deviation vector matrix.

[0062] In a preferred embodiment, first, a large-scale historical monitoring state deviation vector record matrix data set is configured, including the first monitoring state deviation vector record matrix to the Nth monitoring state deviation vector record matrix, where N is an integer and not less than 50,000. These record matrices contain a large number of historical monitoring data deviation patterns of users in various states, forming a basic data set for abnormal event probability analysis. Each monitoring state deviation vector record matrix is ​​a multidimensional data structure, containing the deviation values ​​and their combination relationships of various physiological indicators, motion parameters and environmental factors. Setting N to be not less than 50,000 ensures the adequacy of the sample and helps to improve the accuracy of the abnormal event trigger probability. Then, with the user's first-class attributes (i.e., general attributes such as gender, age, height, weight, etc.) as constraints, for the first monitoring state deviation vector record matrix, a number of user abnormal event trigger identifiers that match it are collected from the historical database. These trigger flags are binary labels generated based on the rescue end's actual decision-making results in historical data: if the corresponding state is ultimately confirmed as an abnormal event (such as an actual emergency such as a fall, drowning, or heart attack), the user abnormal event trigger flag is set to 1; if the corresponding state is confirmed to be a false alarm or a normal fluctuation, the user abnormal event trigger flag is set to 0. This labeling method based on actual rescue results ensures the authenticity and reliability of the data. Subsequently, a statistical analysis is performed on the collected user abnormal event trigger flags. The ratio of the number of samples with a flag value equal to 1 (i.e., the number of cases where abnormal events actually occurred) to the total number of collected samples is calculated, and this ratio is set as the true value of the first abnormal event trigger probability. The true value of the first abnormal event trigger probability reflects the actual historical frequency of abnormal events under a specific monitoring state deviation pattern.

[0063] Repeat the process of steps S622 and S623, and calculate and obtain the second abnormal event trigger probability true value up to the Nth abnormal event trigger probability true value for the second monitoring state deviation vector record matrix up to the Nth monitoring state deviation vector record matrix in sequence. Subsequently, the obtained first abnormal event trigger probability true value up to the Nth abnormal event trigger probability true value is used as a supervision label, and the corresponding first monitoring state deviation vector record matrix up to the Nth monitoring state deviation vector record matrix is ​​used as an input feature, and multiple sets of training data are configured. A machine learning or deep learning algorithm is applied to train an abnormal event trigger probability predictor. After sufficient training, the predictor can accurately predict the probability of a new monitoring state deviation vector matrix causing an abnormal event, thereby providing a basis for accurate alarming.

[0064] By extracting empirical rules from historical data and constructing an abnormal event trigger probability predictor, the risk assessment of abnormal events is based on a large amount of empirical data, thereby improving the accuracy and reliability of alarms.

[0065] Furthermore, taking the first abnormal event trigger probability true value to the Nth abnormal event trigger probability true value as supervision, taking the first monitoring state deviation vector record matrix to the Nth monitoring state deviation vector record matrix as input, configuring multiple sets of data, and training the abnormal event trigger probability predictor, including:

[0066] Step 1: setting a training process quantity threshold, retrieving the multiple sets of data, and training a first abnormal event trigger probability predictor;

[0067] Step 2: extracting the output error vector set of the first abnormal event trigger probability predictor;

[0068] Step 3: When the output error vector variance of the output error vector set is greater than or equal to the variance threshold, return to step 1 and execute a loop;

[0069] Step 4: When the output error vector variance of the output error vector set is less than the variance threshold, the system error vector is counted, and the system error vector modulus is greater than or equal to the convergence threshold, a first residual block is constructed, and the first abnormal event trigger probability predictor is integrated to generate a second abnormal event trigger probability predictor architecture, and the second abnormal event trigger probability predictor architecture is returned to step 1 to execute the loop, wherein the second abnormal event trigger probability predictor architecture is equal to the sum of the outputs of the first abnormal event trigger probability predictor and the first residual block;

[0070] Step 5: When the output error vector variance of the output error vector set is less than the variance threshold, the system error vector is counted, and the system error vector modulus is less than the convergence threshold, the first abnormal event trigger probability predictor is set as the abnormal event trigger probability predictor.

[0071] In a preferred embodiment, to train an abnormal event trigger probability predictor, a training threshold is first set. This threshold controls the number of training iterations, ensuring a thorough and efficient training process. Subsequently, multiple sets of training data are retrieved and a machine learning algorithm is applied to train a first abnormal event trigger probability predictor. This first abnormal event trigger probability predictor is a model that accepts a monitoring state deviation vector matrix as input and outputs abnormal event trigger probabilities. The training process utilizes a supervised learning approach, continuously adjusting model parameters to minimize the error between the predicted output and the true probability value. Subsequently, the training results of the first abnormal event trigger probability predictor are evaluated, and a set of output error vectors is extracted. These error vectors reflect the deviation between the predicted values ​​of the first abnormal event trigger probability predictor and the true values, serving as indicators for evaluating model performance and guiding further optimization. The variance of the output error vector set is then calculated and compared to a preset variance threshold. If the variance of the output error vectors is greater than or equal to the variance threshold, it indicates that the prediction results of the first abnormal event trigger probability predictor are unstable, subject to significant random fluctuations, and require further optimization. At this point, return to step 1 and re-execute the training cycle to improve the stability and consistency of the first abnormal event trigger probability predictor.

[0072] When the variance of the output error vector set is less than the variance threshold, the prediction results of the first abnormal event trigger probability predictor have achieved good stability. However, further evaluation of the systematic components of the prediction deviation of the first abnormal event trigger probability predictor is required. The average of the error vectors in the output error vector set is calculated to obtain a systematic error vector, and its modulus is calculated and compared with the preset convergence threshold. When the modulus of the systematic error vector is greater than or equal to the convergence threshold, it indicates that the first abnormal event trigger probability predictor has significant systematic deviations and requires structural adjustments. At this point, a dedicated first residual block is constructed based on the current systematic error vector and fused with the first abnormal event trigger probability predictor to generate an improved second abnormal event trigger probability predictor architecture. The output of this architecture is equal to the sum of the output of the first predictor and the output of the first residual block. Through this residual learning mechanism, systematic deviations are effectively compensated for and prediction accuracy is improved. After obtaining the second abnormal event trigger probability predictor architecture, return to step 1 and re-execute the training process using the second abnormal event trigger probability predictor architecture. When the variance of the output error vector set is less than the variance threshold and the system error vector modulus is less than the convergence threshold, the first abnormal event trigger probability predictor has achieved both stability and accuracy, and the training process can be terminated. At this point, the first abnormal event trigger probability predictor is determined as the final abnormal event trigger probability predictor and is used for subsequent abnormal event trigger probability assessments.

[0073] Through adaptive training and optimization, we can build an abnormal event trigger probability predictor with both high precision and high stability, provide reliable decision support for the intelligent alarm of wearable devices, and improve the accuracy and timeliness of abnormal state identification.

[0074] Furthermore, based on the user attribute type, monitoring attribute sample statistics are performed on the user motion information and the user environment information to obtain a first monitoring attribute concentration interval, including:

[0075] S310: Constructing a first query constraint based on the user attribute, the user movement information, and the user environment information, in combination with a predefined attribute deviation threshold;

[0076] S320: Retrieve a first sample set that meets the first query constraint, wherein any sample in the first sample set includes user motion record data and user environment record data;

[0077] S330: Constructing a second query constraint based on the user attribute, the user motion record data and the user environment record data, and a predefined attribute deviation threshold;

[0078] S340: Retrieve a second sample set that meets the second query constraint condition;

[0079] S350: Counting the monitoring attribute concentration intervals of the first sample set and the second sample set, and setting the intervals as the first monitoring attribute concentration intervals.

[0080] In one feasible implementation, a first query constraint is first constructed based on user attributes (i.e., administrator-defined general attributes such as gender, age, height, and weight), combined with currently collected user motion information and user environment information, and with reference to a predefined attribute deviation threshold. This first query constraint is a set of structured query rules used to locate user groups in the historical database that share similar basic characteristics and environmental conditions as the current user. For example, for a 35-year-old male user with a height of 175 cm and a weight of 70 kg, the following query constraints are constructed: Gender = Male, Age Range = 30-40, Height Range = 170-180 cm, Weight Range = 65-75 kg. The query constraints are further refined by considering the user's current motion state (e.g., stationary, walking, running) and environmental conditions (e.g., indoors, temperature range, humidity range, etc.). Then, using this constructed first query constraint, a search operation is performed in the historical database to obtain a first set of samples that meet the criteria. Each sample in the first set consists of two parts: user motion record data and user environment record data. Among them, user motion record data includes but is not limited to historical user's motion status, posture changes, activity intensity and other parameters; user environment record data includes but is not limited to historical user's ambient temperature, humidity, air pressure, light and other parameters.

[0081] Then, based on the user's first-class attributes, combined with the retrieved user motion record data and user environment record data, and with reference to the predefined attribute deviation threshold, a second query constraint is constructed, which can achieve sample expansion and effectively increase the breadth of data collection, so as to capture a more comprehensive reference sample. Then, using the constructed second query constraint, a second round of retrieval operation is performed in the historical database to obtain a second sample set that meets the new conditions. The second sample set is a supplement to the similar samples of the first sample set. Through this multi-level query strategy, the coverage of the reference sample is expanded, the robustness of the statistical analysis is enhanced, and the data foundation established in the first monitoring attribute concentration interval is ensured to be more sufficient and comprehensive. Afterwards, a comprehensive statistical analysis is performed on the obtained first sample set and the obtained second sample set, and the distribution characteristics of each monitoring attribute are calculated, including but not limited to statistical quantities such as mean, standard deviation, and quantile, so as to determine the normal fluctuation range of each monitoring attribute, set as the first monitoring attribute concentration interval, and serve as the basic reference standard for subsequent personalized adjustments.

[0082] Through refined data retrieval and statistical analysis, we can build a monitoring benchmark interval that highly matches the basic characteristics of the current user based on a large amount of historical data, laying the foundation for subsequent personalized optimization and abnormal status judgment.

[0083] Furthermore, based on the second type of user attributes, combined with the user motion information and the user environment information, the monitoring attribute set is traversed to perform monitoring attribute sample statistics to obtain the monitoring attribute system error, including:

[0084] S410: Based on the second-category user attribute, based on the first-category user attribute, combined with the user movement information and the user environment information, traverse the monitoring attribute set to perform monitoring attribute sample statistics to obtain a third monitoring attribute concentration interval;

[0085] S420: Calculate the first monitoring attribute intersection-and-union ratio of the third monitoring attribute concentration interval and the first monitoring attribute concentration interval, and configure the first monitoring attribute system error according to the first monitoring attribute intersection-and-union ratio;

[0086] S430: until the Mth monitoring attribute intersection-and-union ratio of the third monitoring attribute concentration interval and the first monitoring attribute concentration interval is calculated, and the Mth monitoring attribute system error is configured according to the Mth monitoring attribute intersection-and-union ratio;

[0087] S440: Add the first monitoring attribute system error to the Mth monitoring attribute system error into the monitoring attribute system error.

[0088] In an optional implementation, a comprehensive traversal and in-depth statistical analysis of the monitored attribute set is performed based on the user's secondary attributes (i.e., personalized attributes that the user can optionally add, such as dietary habits, medical history, and recent exercise data) and primary attributes (general attributes), combined with user exercise information and user environment information. Through dual-constraint statistical analysis, a third monitored attribute concentration interval is obtained that takes into account the user's personalized characteristics. This third monitored attribute concentration interval reflects the normal fluctuation range of each user's monitored indicators under specific personalized conditions. Then, the intersection-over-union (IoU) of the first monitored attribute between the third monitored attribute concentration interval and the first monitored attribute concentration interval is calculated. The first monitored attribute is one of the monitored attributes in the monitored attribute set, which includes all attributes in the user's secondary attributes and primary attributes. The IoU is a measure of the similarity between two intervals and is calculated as the ratio of the intersection to the union of the two intervals. A higher IoU value indicates greater overlap between the two intervals and a lower impact of the user's personalized characteristics on the general monitoring criteria. A lower IoU value indicates a greater need for adjustment of the monitoring criteria due to personalized factors. According to the calculated intersection-union ratio of the first monitoring attribute, the first monitoring attribute system error is configured, which reflects the adjustment amount that needs to be made to the concentrated interval of the first monitoring attribute.

[0089] Repeat the intersection-over-union calculation and system error configuration in S420, sequentially processing the second monitoring attribute through the Mth monitoring attribute (M is the total number of attributes in the monitoring attribute set). This yields the first monitoring attribute system errors through the Mth monitoring attribute system errors, comprehensively quantifying the impact of the user's personalized characteristics on each monitoring indicator. Subsequently, all obtained first monitoring attribute system errors through the Mth monitoring attribute system errors are added to the monitoring attribute system errors, which contain personalized adjustment parameters for each monitoring attribute. This will be used to subsequently modify the first monitoring attribute set interval, thereby generating a more personalized second monitoring attribute set interval.

[0090] Through intersection-combination analysis and system error configuration, the transformation from general monitoring standards to personalized monitoring standards is achieved, providing a customized monitoring benchmark for the intelligent alarm of wearable devices and improving the accuracy and pertinence of abnormal state judgment.

[0091] Furthermore, configuring the first monitoring attribute system error according to the first monitoring attribute intersection-to-union ratio includes:

[0092] S421: When the intersection-over-union ratio of the first monitoring attribute is less than the intersection-over-union ratio threshold, calculating an interval upper boundary deviation vector set and an interval lower boundary deviation vector set of the first monitoring attribute;

[0093] S422: Counting the mode deviation vector of the upper limit boundary deviation vector set of the interval, setting it as the upper limit systematic error of the first monitoring attribute; counting the mode deviation vector of the lower limit boundary deviation vector set of the interval, setting it as the lower limit systematic error of the first monitoring attribute;

[0094] S423: adding the first monitoring attribute upper limit systematic error and the first monitoring attribute lower limit systematic error to the first monitoring attribute systematic error;

[0095] S424: When the intersection-over-union ratio of the first monitoring attribute is greater than or equal to the intersection-over-union ratio threshold, the system error of the first monitoring attribute is set to 0.

[0096] In a preferred embodiment, the configuration process of the monitoring attribute system error is described by taking the configuration of the first monitoring attribute system error according to the intersection-and-union ratio of the first monitoring attribute as an example. First, the relationship between the intersection-and-union ratio of the first monitoring attribute and the intersection-and-union ratio threshold is determined. When the intersection-and-union ratio of the first monitoring attribute is less than the preset intersection-and-union ratio threshold, it indicates that there is a significant difference between the obtained third monitoring attribute concentration interval and the obtained first monitoring attribute concentration interval, and a systematic adjustment is required. At this time, the interval upper boundary deviation vector set and the interval lower boundary deviation vector set of the first monitoring attribute are calculated. Specifically, the difference between the upper limit value and the lower limit value of the third monitoring attribute concentration interval and the first monitoring attribute concentration interval on the first monitoring attribute is compared to generate a set of deviation vectors that characterize these differences. These deviation vectors reflect the need for monitoring interval boundary adjustment caused by personalized factors.

[0097] Subsequently, a statistical analysis is performed on the obtained set of upper boundary deviation vectors of the interval to identify the most frequently occurring deviation vector (i.e., the mode deviation vector), which is set as the upper systematic error of the first monitoring attribute. Similarly, a similar statistical analysis is performed on the set of lower boundary deviation vectors of the interval to identify the mode deviation vector, which is set as the lower systematic error of the first monitoring attribute. By using the mode as the systematic error, interference from extreme values ​​is effectively avoided, ensuring statistical stability and representativeness. The determined upper and lower systematic errors of the first monitoring attribute are then added to the first monitoring attribute systematic error.

[0098] When the IoU ratio of the first monitoring attribute is greater than or equal to the IoU threshold, the obtained concentration interval of the third monitoring attribute is highly consistent with the concentration interval of the first monitoring attribute, and the personalized feature has a negligible impact on this monitoring attribute. In this case, the systematic error of the first monitoring attribute is directly set to 0. That is, no adjustment is made to the interval boundaries of this monitoring attribute, and its original value in the universal standard is retained.

[0099] Through system error configuration based on the intersection-over-union ratio, differentiated treatment of different monitoring attributes is achieved: for attributes significantly affected by personalized factors, precise quantitative boundary adjustments are made; for attributes less affected by personalized factors, the general standard remains unchanged. This flexible adjustment mechanism avoids unnecessary complexity while ensuring the accuracy and effectiveness of personalized adjustments, providing intelligent alarm systems for wearable devices with both universal and personalized monitoring standards.

[0100] Example 2, as Figure 2 As shown, based on the same inventive concept as the wearable device alarm method provided in Example 1, an embodiment of the present invention further provides a wearable device alarm system, comprising:

[0101] The information receiving module 11 is used to receive user monitoring status information, user movement information and user environment information;

[0102] A user attribute receiving module 12 is configured to receive user first-class attributes and user second-class attributes, wherein the user first-class attributes are general attributes predefined by an administrator, and the user second-class attributes are personalized attributes that can be optionally added by the user;

[0103] A general attribute analysis module 13 is configured to perform monitoring attribute sample statistics on the user motion information and the user environment information based on the user attribute type, and obtain a first monitoring attribute concentration interval;

[0104] The personalized attribute analysis module 14 is configured to perform monitoring attribute sample statistics on the monitoring attribute set based on the user's second-category attributes, combined with the user's motion information and the user's environmental information, to obtain a monitoring attribute system error;

[0105] A concentration interval correction module 15 is configured to correct the first monitoring attribute concentration interval according to the monitoring attribute system error to obtain a second monitoring attribute concentration interval;

[0106] The alarm execution module 16 is used to calculate the deviation coefficient between the user monitoring status information and the second monitoring attribute concentration interval, and fuse the deviation coefficient and the user position to generate an alarm signal to execute an alarm.

[0107] Furthermore, the alarm execution module 16 includes the following execution steps:

[0108] Comparing the user monitoring state information with the monitoring state deviation vector matrix of the second monitoring attribute concentration interval;

[0109] Analyzing the trigger probability of abnormal events that satisfy the monitoring state deviation vector matrix;

[0110] When the trigger probability of the abnormal event is less than or equal to the trigger probability threshold, the return process starts to execute the loop;

[0111] When the abnormal event trigger probability is greater than the trigger probability threshold, the abnormal event trigger probability is set as the deviation coefficient.

[0112] Furthermore, the alarm execution module 16 further includes the following execution steps:

[0113] Configure the first monitoring state deviation vector record matrix up to the Nth monitoring state deviation vector record matrix, where N is an integer, N≥50000;

[0114] Using the user attribute as a constraint, collect a number of user abnormal event trigger flags that satisfy the first monitoring state deviation vector record matrix. The user abnormal event trigger flag is generated based on the decision result of the rescue end in historical data. If it is an abnormal event, the user abnormal event trigger flag is equal to 1; otherwise, the user abnormal event trigger flag is equal to 0;

[0115] Counting the number of user abnormal event triggering flags equal to 1, and setting the ratio of the number of the plurality of user abnormal event triggering flags to a true value of the first abnormal event triggering probability;

[0116] Until the true value of the probability of triggering the Nth abnormal event is obtained;

[0117] Taking the first abnormal event trigger probability true value to the Nth abnormal event trigger probability true value as supervision, taking the first monitoring state deviation vector record matrix to the Nth monitoring state deviation vector record matrix as input, configuring multiple groups of data, training the abnormal event trigger probability predictor, and analyzing the abnormal event trigger probability that satisfies the monitoring state deviation vector matrix.

[0118] Furthermore, the alarm execution module 16 further includes the following execution steps:

[0119] Step 1: setting a training process quantity threshold, retrieving the plurality of sets of data, and training a first abnormal event trigger probability predictor;

[0120] Step 2: extracting the output error vector set of the first abnormal event trigger probability predictor;

[0121] Step 3: When the output error vector variance of the output error vector set is greater than or equal to the variance threshold, return to step 1 and execute a loop;

[0122] Step 4: When the output error vector variance of the output error vector set is less than the variance threshold, the system error vector is counted, and the system error vector modulus is greater than or equal to the convergence threshold, a first residual block is constructed, and the first abnormal event trigger probability predictor is integrated to generate a second abnormal event trigger probability predictor architecture, and the second abnormal event trigger probability predictor architecture is returned to step 1 to execute the loop, wherein the second abnormal event trigger probability predictor architecture is equal to the sum of the outputs of the first abnormal event trigger probability predictor and the first residual block;

[0123] Step 5: When the output error vector variance of the output error vector set is less than the variance threshold, the system error vector is counted, and the system error vector modulus is less than the convergence threshold, the first abnormal event trigger probability predictor is set as the abnormal event trigger probability predictor.

[0124] Furthermore, the general attribute analysis module 13 includes the following execution steps:

[0125] Based on the user attribute, the user movement information and the user environment information, and a predefined attribute deviation threshold, a first query constraint condition is constructed;

[0126] Retrieving a first sample set that meets the first query constraint, wherein any sample in the first sample set includes user motion record data and user environment record data;

[0127] Based on the user's first attribute, according to the user's motion record data and the user's environment record data, combined with a predefined attribute deviation threshold, construct a second query constraint condition;

[0128] Retrieving a second sample set that meets the second query constraint;

[0129] The monitoring attribute concentration interval of the first sample set and the second sample set is counted and set as the first monitoring attribute concentration interval.

[0130] Furthermore, the personalized attribute analysis module 14 includes the following execution steps:

[0131] Based on the second type of user attributes, based on the first type of user attributes, combined with the user movement information and the user environment information, traverse the monitoring attribute set to perform monitoring attribute sample statistics to obtain a third monitoring attribute concentration interval;

[0132] Calculate a first monitoring attribute intersection-and-union ratio of the third monitoring attribute concentration interval and the first monitoring attribute concentration interval, and configure a first monitoring attribute system error according to the first monitoring attribute intersection-and-union ratio;

[0133] Until the Mth monitoring attribute intersection-and-union ratio of the third monitoring attribute concentration interval and the first monitoring attribute concentration interval is calculated, and the Mth monitoring attribute system error is configured according to the Mth monitoring attribute intersection-and-union ratio;

[0134] The first monitoring attribute system error to the Mth monitoring attribute system error are added to the monitoring attribute system error.

[0135] Furthermore, the personalized attribute analysis module 14 further includes the following execution steps:

[0136] When the intersection-over-union ratio of the first monitoring attribute is less than the intersection-over-union ratio threshold, calculating an interval upper boundary deviation vector set and an interval lower boundary deviation vector set of the first monitoring attribute;

[0137] Counting the mode deviation vector of the upper limit boundary deviation vector set of the interval, setting it as the upper limit systematic error of the first monitoring attribute; counting the mode deviation vector of the lower limit boundary deviation vector set of the interval, setting it as the lower limit systematic error of the first monitoring attribute;

[0138] Adding the first monitoring attribute upper limit systematic error and the first monitoring attribute lower limit systematic error to the first monitoring attribute systematic error;

[0139] When the intersection-over-union ratio of the first monitoring attribute is greater than or equal to the intersection-over-union ratio threshold, the system error of the first monitoring attribute is set to 0.

[0140] It should be noted that, in the above embodiments, the description of each embodiment has its own focus. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant description of other embodiments.

[0141] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Thus, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0142] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowcharts and / or block diagrams, as well as combinations of processes and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded computer, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowcharts and / or block diagrams. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0143] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0144] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0145] Although preferred embodiments of the present invention have been described, additional changes and modifications to these embodiments may occur to those skilled in the art once the basic inventive concepts become known.

[0146] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the present invention and its equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A wearable device alarm method, characterized in that: Applied to wearable devices, including: Receive user monitoring status information, user movement information, and user environment information, wherein the user monitoring status information refers to user physiological parameter data collected by the wearable device; Receiving user first-class attributes and user second-class attributes, wherein the user first-class attributes are general attributes predefined by an administrator, and the user second-class attributes are personalized attributes that can be optionally added by the user; Based on the user attribute type, performing monitoring attribute sample statistics on the user motion information and the user environment information to obtain a first monitoring attribute concentration interval; Based on the two types of user attributes, combined with the user movement information and the user environment information, traverse the monitoring attribute set to perform monitoring attribute sample statistics to obtain the monitoring attribute system error; According to the monitoring attribute system error, correcting the first monitoring attribute concentration interval to obtain a second monitoring attribute concentration interval; Calculating a deviation coefficient between the user monitoring state information and the second monitoring attribute concentration interval, fusing the deviation coefficient and the user location to generate an alarm signal and execute an alarm; The method includes: performing sample statistics of monitoring attributes based on the user's second-category attributes, combining the user's motion information and the user's environment information, and obtaining a monitoring attribute system error by traversing the monitoring attribute set. Based on the second type of user attributes, based on the first type of user attributes, combined with the user movement information and the user environment information, traverse the monitoring attribute set to perform monitoring attribute sample statistics to obtain a third monitoring attribute concentration interval; Calculate a first monitoring attribute intersection-and-union ratio of the third monitoring attribute concentration interval and the first monitoring attribute concentration interval, and configure a first monitoring attribute system error according to the first monitoring attribute intersection-and-union ratio; Until the Mth monitoring attribute intersection-and-union ratio of the third monitoring attribute concentration interval and the first monitoring attribute concentration interval is calculated, and the Mth monitoring attribute system error is configured according to the Mth monitoring attribute intersection-and-union ratio; The first monitoring attribute systematic error up to the Mth monitoring attribute systematic error are added to the monitoring attribute systematic error, where M is the total number of attributes in the monitoring attribute set.

2. The method according to claim 1, wherein Calculating a deviation coefficient between the user monitoring state information and the second monitoring attribute concentration interval includes: Comparing the user monitoring state information with the monitoring state deviation vector matrix of the second monitoring attribute concentration interval; Analyzing the trigger probability of abnormal events that satisfy the monitoring state deviation vector matrix; When the trigger probability of the abnormal event is less than or equal to the trigger probability threshold, the return process starts to execute the loop; When the abnormal event trigger probability is greater than the trigger probability threshold, the abnormal event trigger probability is set as the deviation coefficient.

3. The method according to claim 2, wherein Analyzing the trigger probability of an abnormal event that satisfies the monitoring state deviation vector matrix includes: Configure the first monitoring state deviation vector record matrix up to the Nth monitoring state deviation vector record matrix, where N is an integer, N≥50000; Using the user attribute as a constraint, collect a number of user abnormal event trigger flags that satisfy the first monitoring state deviation vector record matrix. The user abnormal event trigger flag is generated based on the decision result of the rescue end in historical data. If it is an abnormal event, the user abnormal event trigger flag is equal to 1; otherwise, the user abnormal event trigger flag is equal to 0; Counting the number of user abnormal event triggering flags equal to 1, and setting the ratio of the number of the plurality of user abnormal event triggering flags to a true value of the first abnormal event triggering probability; Until the true value of the probability of triggering the Nth abnormal event is obtained; Taking the first abnormal event trigger probability true value to the Nth abnormal event trigger probability true value as supervision, taking the first monitoring state deviation vector record matrix to the Nth monitoring state deviation vector record matrix as input, configuring multiple groups of data, training the abnormal event trigger probability predictor, and analyzing the abnormal event trigger probability that satisfies the monitoring state deviation vector matrix.

4. The method according to claim 3, wherein Taking the first abnormal event trigger probability true value to the Nth abnormal event trigger probability true value as supervision, taking the first monitoring state deviation vector record matrix to the Nth monitoring state deviation vector record matrix as input, configuring multiple groups of data, and training the abnormal event trigger probability predictor, including: Step 1: setting a training process quantity threshold, retrieving the plurality of sets of data, and training a first abnormal event trigger probability predictor; Step 2: extracting the output error vector set of the first abnormal event trigger probability predictor; Step 3: When the output error vector variance of the output error vector set is greater than or equal to the variance threshold, return to step 1 and execute a loop; Step 4: When the output error vector variance of the output error vector set is less than the variance threshold, the system error vector is counted, and the system error vector modulus is greater than or equal to the convergence threshold, a first residual block is constructed, and the first abnormal event trigger probability predictor is integrated to generate a second abnormal event trigger probability predictor architecture, and the second abnormal event trigger probability predictor architecture is returned to step 1 to execute the loop, wherein the second abnormal event trigger probability predictor architecture is equal to the sum of the outputs of the first abnormal event trigger probability predictor and the first residual block; Step 5: When the output error vector variance of the output error vector set is less than the variance threshold, the system error vector is counted, and the system error vector modulus is less than the convergence threshold, the first abnormal event trigger probability predictor is set as the abnormal event trigger probability predictor.

5. The method according to claim 1, wherein Based on the user attribute type, performing monitoring attribute sample statistics on the user motion information and the user environment information to obtain a first monitoring attribute concentration interval includes: Based on the user attribute, the user movement information and the user environment information, and a predefined attribute deviation threshold, a first query constraint condition is constructed; Retrieving a first sample set that meets the first query constraint, wherein any sample in the first sample set includes user motion record data and user environment record data; Based on the user's first attribute, according to the user's motion record data and the user's environment record data, combined with a predefined attribute deviation threshold, construct a second query constraint condition; Retrieving a second sample set that meets the second query constraint; The monitoring attribute concentration interval of the first sample set and the second sample set is counted and set as the first monitoring attribute concentration interval.

6. The method according to claim 1, wherein Configuring a first monitoring attribute system error according to the first monitoring attribute intersection-to-union ratio includes: When the intersection-over-union ratio of the first monitoring attribute is less than the intersection-over-union ratio threshold, calculating an interval upper boundary deviation vector set and an interval lower boundary deviation vector set of the first monitoring attribute; Counting the mode deviation vector of the upper limit boundary deviation vector set of the interval, setting it as the upper limit systematic error of the first monitoring attribute; counting the mode deviation vector of the lower limit boundary deviation vector set of the interval, setting it as the lower limit systematic error of the first monitoring attribute; Adding the first monitoring attribute upper limit systematic error and the first monitoring attribute lower limit systematic error to the first monitoring attribute systematic error; When the intersection-over-union ratio of the first monitoring attribute is greater than or equal to the intersection-over-union ratio threshold, the system error of the first monitoring attribute is set to 0.

7. A wearable device alarm system, characterized in that: A wearable device alarm method for implementing any one of claims 1 to 6, applied to a wearable device, comprising: An information receiving module is used to receive user monitoring status information, user movement information and user environment information; A user attribute receiving module, configured to receive user first-class attributes and user second-class attributes, wherein the user first-class attributes are general attributes predefined by an administrator, and the user second-class attributes are personalized attributes optionally added by the user; a general attribute analysis module, configured to perform monitoring attribute sample statistics on the user motion information and the user environment information based on the user attribute class, and obtain a first monitoring attribute concentration interval; A personalized attribute analysis module is used to traverse the monitoring attribute set to perform monitoring attribute sample statistics based on the user's second-category attributes, combined with the user's motion information and the user's environmental information, to obtain a monitoring attribute system error; A concentration interval correction module, configured to correct the first monitoring attribute concentration interval according to the monitoring attribute system error to obtain a second monitoring attribute concentration interval; The alarm execution module is used to calculate the deviation coefficient between the user monitoring status information and the second monitoring attribute concentration interval, fuse the deviation coefficient and the user position to generate an alarm signal and execute the alarm.

Images