Intelligent dynamic monitoring and evaluation system for altitude sickness and wearable device thereof

Abstract

The invention relates to the technical field of medical equipment, in particular to an intelligent dynamic monitoring and evaluation system for altitude sickness and wearable equipment thereof, and the system comprises a physiological parameter monitoring module, an environmental parameter monitoring module, a data processing and analysis module and a user interaction and feedback module; high-precision physiological parameter monitoring is carried out by combining multi-lead ECG and PPG technologies, intelligent data processing and real-time health early warning are realized by utilizing a multi-layer deep learning neural network and a fuzzy logic system, precise environmental parameter monitoring is carried out by integrating an MEMS sensor technology, and personalized health management suggestions are provided for users.

Description

Technical Field

[0001] The present invention relates to the technical field of medical equipment, and in particular to an intelligent dynamic monitoring and evaluation system for altitude sickness and a wearable device thereof. Background Art

[0002] In plateau areas, people are prone to altitude sickness due to special natural environmental conditions such as thin air, low oxygen content, low air pressure, and large temperature difference between day and night. Altitude sickness is a comprehensive disease caused by lack of oxygen. Its common symptoms include dizziness, headache, nausea, shortness of breath and rapid heartbeat. If not monitored and treated in time, it may lead to more serious health problems such as acute high-altitude pulmonary edema and high-altitude cerebral edema. Therefore, effective monitoring and analysis of altitude sickness is particularly important.

[0003] In response to these challenges, the scientific and medical communities have developed some health monitoring devices that can be used in plateau environments. However, existing technologies and devices still have some limitations and challenges. First, most health monitoring devices currently on the market usually only monitor a single physiological parameter (such as heart rate, blood oxygen level), and often lack multi-parameter integrated systems. These devices can only provide limited information, making it difficult to comprehensively assess a person's overall health status in a plateau environment. Since the occurrence of altitude sickness involves multiple physiological changes, relying solely on a single data may lead to incomplete health assessments and untimely interventions.

[0004] Secondly, the measurement accuracy and data stability of existing equipment are limited by a variety of factors. Users often engage in outdoor sports or move quickly in plateau environments, which may cause fluctuations in monitoring data. In addition, due to changes in ambient light and differences in user skin color, the readings of many optical measurement devices may be biased. This not only affects the accuracy of the data, but may also lead to incorrect health risk assessments.

[0005] Although some devices have certain data analysis capabilities, most of them remain at the level of simple data recording and historical trend display, lacking complex real-time data analysis and anomaly detection functions. As a result, users need to judge and interpret the data on their own, and cannot get timely health risk warnings and recommendations. Summary of the invention

[0006] The purpose of the present invention is to provide an intelligent dynamic monitoring and evaluation system for altitude sickness and its wearable device, which combines multi-lead ECG and PPG technology for high-precision physiological parameter monitoring, uses multi-layer deep learning neural networks and fuzzy logic systems to achieve intelligent data processing and real-time health warnings, and integrates MEMS sensor technology for accurate environmental parameter monitoring to provide users with personalized health management suggestions.

[0007] In order to achieve the above technical objectives and the above technical effects, the present invention is implemented through the following technical solutions:

[0008] The intelligent dynamic monitoring and evaluation system for altitude sickness includes:

[0009] Physiological parameter monitoring module,

[0010] a. Heart rate and ECG monitoring submodule: Flexible electrodes and multi-lead ECG sensors are combined with photoplethysmography (PPG) technology to achieve non-contact wearing while ensuring data accuracy. By integrating bioelectric signals and optical signals, efficient signal processing algorithms are used for data collection and processing. External motion interference and ambient light changes are filtered in real time to ensure accurate data output. ECG data is trained and analyzed through machine learning models to identify abnormal ECG activities such as arrhythmias in real time, thereby quickly providing health status assessment and risk warning.

[0011] b. Blood oxygen saturation monitoring submodule: Blood oxygen monitoring uses dual-wavelength spectral technology (red light and infrared light) combined with a skin optical property compensation module to improve measurement accuracy. The sensor continuously collects the user's blood oxygen data and analyzes it through an adaptive learning algorithm to predict the risk of hypoxia. The difference in light absorption between different groups of people is adjusted through dynamic light compensation technology to ensure consistency and accuracy. The system is calibrated regularly to ensure the reliability of measurements under various environmental conditions.

[0012] c. Body temperature monitoring submodule: Using a combination of micro-thermocouples and non-contact infrared sensors, body temperature monitoring achieves high accuracy and fast response. The system implements multi-point temperature data fusion to model the user's body temperature changes in real time. With the help of a time series analysis model, the system can identify and warn of abnormal body temperature fluctuations caused by the plateau environment and provide personalized body temperature management recommendations.

[0013] Environmental parameter monitoring module,

[0014] a. Altitude and GPS positioning submodule: Combine high-precision MEMS pressure sensor and differential GPS to obtain accurate altitude positioning information. Through multi-source data fusion technology, Geographic Information System (GIS) data is integrated for error correction to ensure the accuracy of real-time altitude. By analyzing the altitude change rate and the user's movement status, the system can provide real-time environmental adaptability analysis and support the formulation of personalized plateau adaptation strategies.

[0015] b. Environmental temperature and humidity monitoring submodule: Temperature and humidity detection relies on high-performance integrated sensors, real-time data analysis combined with historical climate patterns to predict the potential impact of environmental changes on the user's physiological state. The system uses a dynamic adaptive model to provide individualized health and comfort recommendations when temperature and humidity fluctuate, improving response capabilities.

[0016] Data processing and analysis module,

[0017] a. Intelligent data analysis submodule: Through a multi-layer deep learning neural network architecture, the system analyzes various physiological and environmental data to conduct comprehensive risk assessment. The system characterizes the collected parameters, builds a personalized health status model, and identifies long-term physiological abnormal patterns. Through reinforcement learning algorithms, the system continuously optimizes the accuracy and adaptability of the prediction model to ensure efficient health risk identification and prediction capabilities.

[0018] b. Intelligent early warning submodule: Based on the adaptive fuzzy logic algorithm, the early warning threshold is adjusted in real time, and the user's historical physiological data and current environmental data are combined to perform multi-condition cross-validation to avoid false alarms. The event-driven early warning mechanism is implemented to significantly improve the system's response speed and accuracy. When the physiological indicators are detected to reach a dangerous level, the system will immediately issue an alarm.

[0019] User interaction and feedback module,

[0020] a. User interface and operation submodule: The user interface design adopts the principles based on ergonomics, combined with voice assistant and intuitive graphical interface. The system uses speech recognition and natural language processing technology to have natural dialogue and interaction with users, providing real-time health information and suggestions. Through user behavior analysis, the interface can automatically adjust according to user preferences to optimize the user experience.

[0021] b. Information sharing and emergency rescue submodule: integrated with Bluetooth and WiFi transmission functions, supports rapid sharing of health data with medical staff. When serious physiological abnormalities are detected, the system automatically triggers the emergency rescue protocol and transmits the rescue information through encrypted channels, combined with blockchain technology to ensure the integrity and security of data during transmission.

[0022] On the other hand, the present invention proposes a wearable device based on the above system, including a monitoring suit and multiple groups of sensors, the sensors including an electrocardiogram sensor, a PPG sensor, a blood oxygen sensor, and a body temperature monitoring sensor; the electrocardiogram sensor includes a flexible electrode and a multi-lead ECG sensor, which is integrated in the inner layer of the monitoring suit and located at the chest position of the monitoring suit; the PPG sensor is located below the shoulder strap of the monitoring suit, and combines photoelectric technology to obtain blood flow change information; the blood oxygen sensor is located on the side of the monitoring suit near the armpit, and uses red light and infrared light dual-wavelength technology to measure blood oxygen levels; at the same time, the blood oxygen sensor is provided with a dynamic light compensation module integrated in the inner layer of the monitoring suit to adjust for different skin colors and ambient light changes; the body temperature monitoring sensor includes multiple groups of miniature thermocouples and infrared sensors, which are located at multiple points on the front chest, back and armpit of the monitoring suit to obtain multi-point body temperature data.

[0023] Furthermore, the monitoring suit also includes a MEMS pressure sensor, a GPS module, and a temperature and humidity sensor. The MEMS pressure sensor is embedded in the outside of the shoulder of the monitoring suit to sense changes in air pressure and thus calculate the altitude; the GPS module is integrated in the upper part of the rear side of the collar of the monitoring suit to provide accurate geographic location information; the temperature and humidity sensor is located at the collar or cuff of the monitoring suit to monitor the temperature and humidity of the external environment in real time.

[0024] Beneficial effects of the present invention:

[0025] This system uses a combination of multi-lead ECG sensors and PPG technology to achieve wearable monitoring of ECG and heart rate using flexible electrodes. Multi-lead ECG sensors can capture the electrical activity signals of the heart, including P waves, QRS complexes, and T waves. These signals are filtered through an efficient adaptive filtering algorithm to remove myoelectric and environmental noise, achieving clear signal extraction. The R wave detection algorithm calculates the RR interval and its variability (SDNN and LF / HF ratio) by identifying the peak position of the R wave in the envelope detection technology, thereby achieving real-time monitoring and evaluation of arrhythmias. Combined with the machine learning model, the system can quickly identify abnormal ECG activity and achieve dynamic evaluation and risk warning of heart health. Blood oxygen monitoring improves the measurement accuracy through dual-wavelength spectroscopy (red light 660nm and infrared light 940nm) technology combined with a skin optical property compensation module. The AC and DC components collected by the photoelectric sensor calculate the blood oxygen saturation through the Beer-Lambert law, and use an adaptive learning algorithm to predict the individual's hypoxia risk. Body temperature monitoring uses infrared sensing technology and micro-thermocouples, combined with the radiation heat conduction equation, to provide fast-response multi-point temperature data fusion, greatly improving the accuracy of detection and the ability to track individual temperature changes.

[0026] In terms of environmental parameter monitoring, the system combines MEMS pressure sensors with differential GPS technology to achieve high-precision capture of altitude data. The pressure sensor compares the real-time pressure value with the standard atmospheric pressure, uses the pressure-altitude formula to calculate the altitude, and combines GIS data for error correction to ensure accuracy in altitude change monitoring. By analyzing the altitude change rate and user mobility status, personalized environmental adaptability recommendations are provided. Similarly, the integrated ambient temperature and humidity sensor captures ambient temperature and humidity data in real time through high-performance integrated sensors, and dynamically compares it with historical climate patterns. The linear regression model is used to analyze its impact on the user's physiological state. The system can provide users with health and comfort recommendations in environmental changes and improve their adaptability in plateau environments.

[0027] The data processing and analysis module extracts features from complex physiological and environmental data through multi-layer deep learning neural networks and builds a personalized health status model. Through normalization and principal component analysis (PCA), the system effectively reduces redundant information in multi-dimensional data and improves data processing efficiency. Combined with integrated learning frameworks (such as random forests and gradient boosting trees), the classification and prediction accuracy of the system on diversified samples is enhanced. The intelligent early warning mechanism sets monitoring rules for multiple physiological and environmental parameters through a fuzzy logic system, and achieves accurate health status assessment through real-time adaptive threshold adjustment and fuzzy inference engines, avoiding false alarms, and notifying users and providing suggestions immediately when indicators are triggered. This event-driven early warning mechanism relies on accurate fuzzy logic rules, such as multi-parameter cross-rules such as heart rate and blood oxygen saturation, body temperature and ambient temperature, altitude and oxygen saturation, to ensure efficient response speed and accuracy.

[0028] The user interaction design follows the principles of ergonomics, using voice assistants and intuitive graphical interfaces to achieve natural voice recognition and response to users. The interface is continuously optimized through user behavior analysis, and natural dialogue and health advice are provided by combining voice recognition and natural language processing technologies. In terms of information sharing and emergency rescue functions, the system integrates Bluetooth and WiFi transmission technologies, allowing users to quickly share health data and send out distress signals through encrypted communication protocols in emergency situations. At the same time, blockchain technology is used to record data access logs, ensuring the transparency and immutability of data, and fully protecting user privacy and data security. The system analyzes user feedback data, such as interactive response time and satisfaction scores, and uses reinforcement learning mechanisms to continuously optimize system performance to ensure user participation and satisfaction in health monitoring.

[0029] Of course, any product implementing the present invention does not necessarily need to achieve all of the advantages described above at the same time. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for describing the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0031] Figure 1 This is a schematic diagram of the structure of the monitoring suit of the present invention;

[0032] In the picture: 1. Monitoring suit, 2. ECG sensor, 3. PPG sensor, 4. Blood oxygen sensor, 5. Body temperature monitoring sensor, 6. MEMS air pressure sensor, 7. GPS module, 8. Temperature and humidity sensor. DETAILED DESCRIPTION

[0033] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. 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 ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0034] Example 1

[0035] The intelligent dynamic monitoring and evaluation system for altitude sickness described in this embodiment includes:

[0036] Physiological parameter monitoring module,

[0037] a. Heart rate and ECG monitoring submodule: Flexible electrodes and multi-lead ECG sensors are combined with photoplethysmography (PPG) technology to achieve non-contact wearing while ensuring data accuracy. By integrating bioelectric signals and optical signals, efficient signal processing algorithms are used for data collection and processing. External motion interference and ambient light changes are filtered in real time to ensure accurate data output. ECG data is trained and analyzed through machine learning models to identify abnormal ECG activities such as arrhythmias in real time, thereby quickly providing health status assessment and risk warning.

[0038] b. Blood oxygen saturation monitoring submodule: Blood oxygen monitoring uses dual-wavelength spectral technology (red light and infrared light) combined with a skin optical property compensation module to improve measurement accuracy. The sensor continuously collects the user's blood oxygen data and analyzes it through an adaptive learning algorithm to predict the risk of hypoxia. The difference in light absorption between different groups of people is adjusted through dynamic light compensation technology to ensure consistency and accuracy. The system is calibrated regularly to ensure the reliability of measurements under various environmental conditions.

[0039] c. Body temperature monitoring submodule: Using a combination of micro-thermocouples and non-contact infrared sensors, body temperature monitoring achieves high accuracy and fast response. The system implements multi-point temperature data fusion to model the user's body temperature changes in real time. With the help of a time series analysis model, the system can identify and warn of abnormal body temperature fluctuations caused by the plateau environment and provide personalized body temperature management recommendations.

[0040] Environmental parameter monitoring module,

[0041] a. Altitude and GPS positioning submodule: Combine high-precision MEMS pressure sensor and differential GPS to obtain accurate altitude positioning information. Through multi-source data fusion technology, Geographic Information System (GIS) data is integrated for error correction to ensure the accuracy of real-time altitude. By analyzing the altitude change rate and the user's movement status, the system can provide real-time environmental adaptability analysis and support the formulation of personalized plateau adaptation strategies.

[0042] b. Environmental temperature and humidity monitoring submodule: Temperature and humidity detection relies on high-performance integrated sensors, real-time data analysis combined with historical climate patterns to predict the potential impact of environmental changes on the user's physiological state. The system uses a dynamic adaptive model to provide individualized health and comfort recommendations when temperature and humidity fluctuate, improving response capabilities.

[0043] Data processing and analysis module,

[0044] a. Intelligent data analysis submodule: Through a multi-layer deep learning neural network architecture, the system analyzes various physiological and environmental data to conduct comprehensive risk assessment. The system characterizes the collected parameters, builds a personalized health status model, and identifies long-term physiological abnormal patterns. Through reinforcement learning algorithms, the system continuously optimizes the accuracy and adaptability of the prediction model to ensure efficient health risk identification and prediction capabilities.

[0045] b. Intelligent early warning submodule: Based on the adaptive fuzzy logic algorithm, the early warning threshold is adjusted in real time, and the user's historical physiological data and current environmental data are combined to perform multi-condition cross-validation to avoid false alarms. The event-driven early warning mechanism is implemented to significantly improve the system's response speed and accuracy. When the physiological indicators are detected to reach a dangerous level, the system will immediately issue an alarm.

[0046] User interaction and feedback module,

[0047] a. User interface and operation submodule: The user interface design adopts the principles based on ergonomics, combined with voice assistant and intuitive graphical interface. The system uses speech recognition and natural language processing technology to have natural dialogue and interaction with users, providing real-time health information and suggestions. Through user behavior analysis, the interface can automatically adjust according to user preferences to optimize the user experience.

[0048] b. Information sharing and emergency rescue submodule: integrated with Bluetooth and WiFi transmission functions, supports rapid sharing of health data with medical staff. When serious physiological abnormalities are detected, the system automatically triggers the emergency rescue protocol and transmits the rescue information through encrypted channels, combined with blockchain technology to ensure the integrity and security of data during transmission.

[0049] In this embodiment, the physiological parameter monitoring module specifically includes:

[0050] ECG monitoring:

[0051] Data is collected using a multi-lead electrocardiogram (ECG) sensor and a photoplethysmogram (PPG) sensor. The ECG sensor captures the heart's electrical activity, identifying the P wave, QRS complex, and T wave, which represent different stages of heart activity. The PPG sensor supplements heart rate information by detecting changes in blood volume under the skin.

[0052] The data is processed by an adaptive filter to remove myoelectric and environmental noise and obtain a clear ECG signal. Then, the peak position of the R wave in the ECG is identified by the R wave detection algorithm (envelope detection technology). Let the ECG signal be S(t), and a series of R wave peak positions are identified by envelope detection technology, expressed as {R1, R2, ..., Rn}. The RR interval is the time difference between two consecutive R wave peaks, which can be expressed as:

[0053] RR i =R i+1 -R i

[0054] Wherein i=1, 2, ..., n-1.

[0055] Average heart rate can be estimated by taking the inverse of the RR interval, expressed in minutes:

[0056]

[0057] When the calculated HeartRate < 60, bradycardia may be flagged.

[0058] When HeartRate>100, it may be marked as tachycardia.

[0059] For arrhythmia detection, heart rate variability (HRV) analysis was performed, and the standard deviation of RR interval duration (SDNN) and the ratio of low-frequency power (LF) to high-frequency power (HF) were calculated.

[0060]

[0061] where RR i For each RR interval, mean(RR) is the mean RR interval.

[0062] Perform frequency domain analysis to calculate low frequency power (LF) and high frequency power (HF). Use Fast Fourier Transform (FFT). The ratio of low frequency power to high frequency power, LF / HF, can be used to assess the balance of the autonomic nervous system:

[0063]

[0064] P LF Represents the power of the low-frequency part in the spectrum analysis of heart rate variability, P HF Represents the power of the high frequency part in the spectrum analysis of heart rate variability, P LF , P HF Used to assess the activity of the autonomic nervous system.

[0065] If the SDNN is low, it may indicate reduced heart rate variability and the risk of arrhythmia; if the LF / HF ratio is high, it may indicate an abnormal balance of the autonomic nervous system and may also indicate a potential risk of arrhythmia.

[0066] When abnormal cardiac electrical activity is detected, the system alerts the user via a mobile app and recommends further medical examination.

[0067] Blood oxygen saturation monitoring:

[0068] Using dual wavelength spectroscopy of red light (660nm) and infrared light (940nm), the photoelectric sensor measures changes in reflected light intensity, reflecting the proportion of oxygenated hemoglobin in arterial blood. The sensor collects AC (alternating current) and DC (direct current) light signals.

[0069] Using the Beer-Lambert law, the blood oxygen saturation (SpO2) is calculated using the following formula:

[0070]

[0071] AC R :Blood flow changes measured by red light AC component. DC R :Red Light DC Component is used to correct the background light intensity. IR :Blood flow changes measured by Infrared Light AC Component. DC IR :Infrared Light DC Component is used to correct the background light intensity. Normal blood oxygen levels are generally between 95% and 100%. If SpO2 is lower than 90%, there may be a risk of hypoxia. At this time, the system automatically notifies the user and recommends adjusting breathing or using oxygen supplements.

[0072] Temperature monitoring:

[0073] Body temperature monitoring relies on thermocouples combined with infrared sensors to ensure fast response and accuracy. Thermocouples are based on the Seebeck effect and generate a voltage proportional to temperature. The corrected formula is:

[0074] T tc =V mV ·K tc

[0075] K tc : Calibration Coefficient, used to convert voltage to actual temperature value.

[0076] Combined with infrared sensors to obtain body surface temperature, it is calibrated using the radiation heat conduction equation to provide accurate body temperature readings. The normal range is set at 36.5℃ to 37.5℃. If it exceeds this range, it may indicate fever or hypothermia.

[0077] The system will send out an alarm when an abnormal body temperature is detected, and recommend that users take appropriate measures, such as replenishing fluids or seeking medical advice. Users can view detailed body temperature trend graphs through the app to better understand their temperature changes and health status.

[0078] In this embodiment, the environmental parameter monitoring module specifically includes:

[0079] The altitude and GPS positioning submodule uses a MEMS pressure sensor and differential GPS technology to provide accurate altitude and location data. The pressure sensor measures the air pressure value at the current location (denoted as P1) and compares it with the standard sea level atmospheric pressure (denoted as P0) to calculate the altitude (denoted as H now ). The formula is as follows:

[0080]

[0081] Where T0 represents standard temperature (Kelvin), L t is the atmospheric temperature lapse rate, R p is the ideal gas constant, g0 is the acceleration due to gravity, and M is the molar mass of air. Differential GPS technology corrects GPS signal errors by referencing multiple base stations with known ground locations, thereby improving the measurement accuracy of altitude. In addition, by integrating terrain and meteorological data from the Geographic Information System (GIS), the system can perform multi-source data fusion and error correction to ensure the accuracy of real-time altitude data.

[0082] After obtaining the altitude data, the system analyzes the rate of altitude change (dH / dt) and combines the user's movement status (stationary, walking or fast movement) to evaluate the user's adaptability to environmental changes. If the rate of change exceeds the pre-set adaptability threshold, the system generates an early warning and recommends that the user adjust the activity intensity according to the current status, such as slowing down the climbing speed or taking a rest.

[0083] The ambient temperature and humidity monitoring submodule uses a temperature and humidity sensor to record the ambient temperature (T) and humidity (H) in real time. w ) changes. The data is compared with historical climate models to predict the impact of environmental changes on the user's physiological state. This submodule uses a linear regression model for analysis. Through regression analysis, the system can provide users with personalized health suggestions when temperature and humidity fluctuate abnormally, such as adding clothes to keep warm or maintaining a reasonable water intake to improve the ability to adapt to environmental fluctuations.

[0084] In terms of the feedback mechanism, once it is detected that the environmental parameters have reached the warning level, the system will immediately send a notification reminder through the user's device (such as a smartphone) and provide specific action suggestions. For example, when the altitude increases significantly, it is recommended to reduce the intensity of activity or take deep breathing exercises to adapt to the changing air pressure environment. When the temperature and humidity change significantly, the system may suggest that the user increase water intake or reduce activity intensity to avoid discomfort. In addition, the system can also automatically adjust the user's breathing training plan, combined with the user's real-time physiological data, to enhance plateau tolerance by adjusting the breathing rate and intensity.

[0085] In this embodiment, the data processing and analysis module specifically includes:

[0086] The data acquisition module acquires physiological and environmental data through multiple sensors, and these data need to be preprocessed to ensure quality. The preprocessing stage includes signal denoising, data filtering, and standardization. Taking the ECG signal as an example, a bandpass filter is used to remove low-frequency drift and high-frequency noise:

[0087] y[n]=a0·x[n]+a1·x[n-1]+...+a n ·x[n-α]-

[0088] b1·y[n-1]-...-b m y[n-β]

[0089] Where: y[n] is the filtered signal. x[n] is the input signal. i and b i is the filter coefficient, which needs to be calculated according to the design requirements. α and β are the filter orders.

[0090] In the feature extraction process, the system uses principal component analysis (PCA) to reduce the data dimension and extract key information. For electrocardiogram, the features may include the amplitude and duration of the QRS band. The feature extraction formula can be expressed as

[0091] Z=W(X-μ)

[0092] Where Z is the eigenvector, W is the eigenvector matrix, X is the original data, and μ is the mean vector.

[0093] The feature-extracted data is fed into a deep learning model, which can be a convolutional neural network (CNN) or a recurrent neural network (RNN), depending on the type of data and sequence characteristics. During training, the model adjusts the weights to minimize the loss function L0, using the mean square error (MSE):

[0094]

[0095] where y iis the actual value, is the predicted value, n sample is the sample size.

[0096] In order to ensure the adaptability of the model to various populations and environments, the system adopts an ensemble learning strategy, such as random forest or gradient boosting tree. The prediction accuracy and robustness of the model are improved by combining multiple weak learners. In this architecture, the prediction result y of each weak learner is j are weighted averaged to obtain the final output, which is:

[0097]

[0098] where w j is the weight coefficient.

[0099] The intelligent early warning submodule performs real-time monitoring through a fuzzy logic system. This system sets several fuzzy rules, including:

[0100] Heart rate and blood oxygen saturation rules

[0101] 1: IF heart rate is high AND blood oxygen saturation is low THEN the health risk is high.

[0102] 2: IF heart rate is medium AND blood oxygen saturation is low THEN health risk is medium.

[0103] 3: IF heart rate is low AND blood oxygen saturation is low THEN the health risk is medium.

[0104] Body temperature and ambient temperature rules

[0105] 4: IF body temperature is high AND ambient temperature is high THEN the risk of heat stress is high.

[0106] 5: IF body temperature is low AND ambient temperature is low THEN hypothermia risk is high.

[0107] 6: IF body temperature is normal AND ambient temperature is high THEN heat stress risk is medium.

[0108] Altitude and Oxygen Saturation Rules

[0109] 7: IF the altitude is high AND blood oxygen saturation is low THEN the risk of altitude sickness is high.

[0110] 8: IF altitude is medium AND blood oxygen saturation is low THEN risk of altitude sickness is medium.

[0111] 9: IF the altitude is high AND blood oxygen saturation is normal THEN the adaptability is good.

[0112] Heart Rate Variability and Stress Level Rules

[0113] 10: IF heart rate variability is low AND stress level is high THEN cardiac workload is high.

[0114] 11: IF HRV is MEDIUM AND STRESS LEVEL is MEDIUM THEN CARDIAC LOAD is MEDIUM.

[0115] 12: IF heart rate variability is HIGH AND stress level is LOW THEN heart condition is GOOD.

[0116] These rules convert continuous data into fuzzy sets through the fuzzification process and are processed by the inference engine. The fuzzy rule system can be expressed as:

[0117] IF(θ1is A1 ANDθ2is A2)THENηis B

[0118] Where A1, A2 and B are fuzzy sets.

[0119] The adaptive threshold mechanism is another key feature of the system. It adjusts the warning threshold by analyzing the user's historical data and current environment data in real time. This process uses the weighted moving average method (WMA) to smooth the input data. The formula is:

[0120]

[0121] Where V t is the smoothed value, x t is the current observation value, w i is the weight.

[0122] Once any physiological parameter is predicted or detected to exceed the set threshold, the system immediately triggers an alert and notifies the user through a mobile device or other interface. The alert information not only contains the abnormal parameters, but also provides recommended measures, such as rest, oxygen supplementation, or contacting professional medical help.

[0123] In the feedback mechanism, the system records user responses to optimize the model. By analyzing user feedback on warning information, the system uses reinforcement learning algorithms to update its model parameters to make future predictions more accurate. In this way, the data processing and analysis module ensures the safety and health of users in plateau environments through efficient data collection, complex model analysis, and flexible feedback mechanisms.

[0124] In this embodiment, the user interaction and feedback module specifically includes:

[0125] The user interface and operation submodule provides an interface that enables users to interact with the system effectively. First, the interface design follows the principles of ergonomics to improve user comfort and operating efficiency. In this process, the position and size of the user interface elements are obtained by analyzing user behavior data. For each interface element, such as a button, the system calculates an optimal position and size by analyzing the user's click frequency and position preference.

[0126] Assume that the click frequency of each interface element i is f i , and the relative position is p i , then the goal of interface layout optimization is to maximize user interaction efficiency E interaction , which can be expressed as:

[0127]

[0128] in is the position weight of interface element i. On this basis, the interface layout is continuously adjusted through optimization algorithms (such as genetic algorithm or particle swarm optimization) to improve E interaction value.

[0129] In the implementation of natural language processing (NLP) and voice assistants, the user's voice input is converted into text by the speech recognition module, and then parsed by the NLP module. The accuracy of speech recognition depends on the combination of hidden Markov model (HMM) and deep neural network (DNN), where each speech segment is mapped to a state s, and the probability of state transition is P(s t |s t-1 ), and optimize the model parameter λ by maximizing P(O|λ) to maximize the probability of the output sequence O.

[0130] After processing the user input, the system generates a response using the NLP module, which involves sentence intent recognition and semantic understanding. Assume that the user's input text is represented by a set of word vectors V = [v1, v2, ..., v n ]The system determines its response by calculating the semantic vector S of V.

[0131] In the information sharing and emergency rescue submodule, the system uses Bluetooth and WiFi to quickly share health data. When an emergency occurs, the system identifies abnormalities through the physiological parameter monitoring module. For example, when the heart rate H exceeds the safety threshold H th , and blood oxygen saturation S pO2 Below the set threshold S pO2th When the trigger condition C is true: C=(H>H th )AND(S pO2 pO2th ). ​

[0132] Once C is true, the system immediately starts the emergency rescue protocol. The distress message is encrypted using the formula E=enc(M, key), where M is the distress message and key is the encryption key, and is transmitted to the preset emergency contact through a secure channel. The information transmission is recorded in the blockchain to ensure transparency and integrity, and the blockchain records the hash value H(M) to verify that the information has not been tampered with.

[0133] The feedback mechanism optimizes system performance by analyzing user feedback data, such as operation response time, user satisfaction score and other parameters. For example, for each functional module i, the response time t of user feedback i and satisfaction i , the system minimizes the objective function:

[0134] F satisfaction =Σ(ξ i ·τ i -1 )

[0135] F satisfaction is the objective function value, which represents the negative utility of user experience. i is the response time of functional module i. τ i is the user satisfaction score of functional module i.

[0136] To adjust the system configuration to ensure fast system response and high user satisfaction.

[0137] Example 2

[0138] The intelligent dynamic monitoring and evaluation system for altitude sickness described in this embodiment, based on embodiment 1, further includes:

[0139] Energy management and equipment maintenance module,

[0140] a. Power management submodule: The system integrates high-energy-density lithium polymer batteries and is equipped with energy harvesting devices. Through adaptive energy management algorithms, it dynamically adjusts the device's energy consumption mode. It intelligently switches between device standby and normal operation to ensure optimal battery life and provides fast charging options to extend device usage time.

[0141] b. Hardware maintenance submodule: Built-in intelligent diagnosis system continuously monitors the hardware status of the device and prompts users to perform maintenance through intelligent notifications to ensure long-term stable operation of the device.

[0142] Data security and privacy protection module,

[0143] The AES encryption standard and distributed storage technology are used to achieve secure storage and transmission of data. The system records data access logs through blockchain technology to ensure the transparency and non-tamperability of each data operation, and the user's privacy and data security are multiple protections.

[0144] In this embodiment, the energy management submodule includes:

[0145] Dynamically optimizes the device's energy consumption through intelligent algorithms to extend battery life and ensure continuous system operation.

[0146] The system integrates high energy density lithium polymer battery, whose capacity can be expressed as C b (Unit: mAh). The power consumption of the device in different states is P s (standby power consumption) and P a (Active Power Consumption) in mW.

[0147] The total operating time of the equipment is T operation It can be calculated by the formula:

[0148]

[0149] where d is the proportion of time the device is active.

[0150] Adaptive energy management algorithms are used to dynamically adjust the device's switching between standby and active states to maximize T. This algorithm predicts future energy consumption and adjusts the state switching strategy based on historical usage data and current operating patterns.

[0151] The system is equipped with an energy collection device (such as a solar panel) whose output power is P col (Unit: mW). Through energy harvesting, additional charging current I can be provided to the battery col =P col / V b (V b is the battery voltage, unit: V).

[0152] When sufficient energy is harvested, the system prioritizes using the harvested energy to reduce the burden on the battery, preserving battery power for powering devices when light levels are low.

[0153] The system energy management logic evaluates power consumption and collection efficiency, calculates the impact of energy collection on the total endurance time ΔT, and optimizes the energy consumption mode selection based on ΔT.

[0154] P col ·T col >(P a -P col )·d

[0155] (Tcol is the effective time for collecting energy) to confirm the positive effect of energy harvesting.

[0156] Example 3

[0157] The wearable device based on the system described in Examples 1 and 2 as described in this embodiment includes a monitoring suit 1 and multiple groups of sensors, wherein the sensors include an electrocardiogram sensor 2, a PPG sensor 3, a blood oxygen sensor 4, and a body temperature monitoring sensor 5; the electrocardiogram sensor 2 includes a flexible electrode and a multi-lead ECG sensor, which is integrated in the inner layer of the monitoring suit 1 and located at the chest of the monitoring suit 1; the PPG sensor 3 is located below the shoulder strap of the monitoring suit 1, and obtains blood flow change information in combination with photoelectric technology; the blood oxygen sensor 4 is located on the side of the monitoring suit 1 near the armpit, and uses red light and infrared light dual-wavelength technology to measure blood oxygen levels; at the same time, the blood oxygen sensor 4 is provided with a dynamic light compensation module integrated in the inner layer of the monitoring suit 1 to adjust for different skin colors and ambient light changes; the body temperature monitoring sensor 5 includes multiple groups of miniature thermocouples and infrared sensors, which are respectively located at multiple points on the front chest, back and armpit of the monitoring suit 1 to obtain multi-point body temperature data.

[0158] In this embodiment, the monitoring suit 1 also includes a MEMS pressure sensor 6, a GPS module 7, and a temperature and humidity sensor 8. The MEMS pressure sensor 6 is embedded in the outer side of the shoulder of the monitoring suit 1 to sense the air pressure changes and thus calculate the altitude; the GPS module 7 is integrated in the upper part of the rear side of the collar of the monitoring suit 1 to provide accurate geographic location information; the temperature and humidity sensor 8 is located at the collar or cuff of the monitoring suit 1 to monitor the temperature and humidity of the external environment in real time.

[0159] In this embodiment, a high energy density lithium polymer battery is integrated in the lower part of the monitoring suit 1 to ensure the stability of the center of gravity. The energy collection device (solar panel) is embedded in the shoulder and back of the monitoring suit 1, and the solar panel can protrude outside through the folding bracket structure to maximize energy capture.

[0160] The preferred embodiments of the present invention disclosed above are only used to help illustrate the present invention. The preferred embodiments do not describe all the details in detail, nor do they limit the invention to the specific implementation methods described. Obviously, many modifications and changes can be made according to the content of this specification. This specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the present invention, so that those skilled in the art can understand and use the present invention well. The present invention is limited only by the claims and their full scope and equivalents.

Claims

1. An intelligent dynamic monitoring and evaluation system for altitude sickness, characterized in that: include: Physiological parameter monitoring module, Heart rate and ECG monitoring submodule: It uses flexible electrodes and multi-lead ECG sensors combined with photoplethysmography technology to achieve non-contact wearing while ensuring data accuracy; it uses efficient signal processing algorithms to collect and process data by integrating bioelectric signals and optical signals; it filters external motion interference and ambient light changes in real time to ensure accurate data output; ECG data is trained and analyzed through machine learning models to identify abnormal ECG activities such as arrhythmias in real time, thereby quickly providing health status assessment and risk warning; Blood oxygen saturation monitoring submodule: Blood oxygen monitoring uses dual-wavelength spectroscopy technology and combines the skin optical property compensation module to improve measurement accuracy; The sensor continuously collects the user's blood oxygen data and analyzes it through an adaptive learning algorithm to predict the risk of hypoxia. The light absorption differences between different groups of people are adjusted through dynamic light compensation technology to ensure consistency and accuracy. The system is calibrated regularly to ensure the reliability of measurement under various environmental conditions; Body temperature monitoring submodule: Using a combination of micro-thermocouples and non-contact infrared sensors, body temperature monitoring achieves high accuracy and rapid response; the system implements multi-point temperature data fusion to model the user's body temperature changes in real time; with the help of a time series analysis model, the system can identify and warn of abnormal body temperature fluctuations caused by the plateau environment, and provide personalized body temperature management suggestions; Environmental parameter monitoring module, Altitude and GPS positioning submodule: Combines high-precision MEMS pressure sensor and differential GPS to obtain accurate altitude positioning information; integrates geographic information system data through multi-source data fusion technology for error correction to ensure the accuracy of real-time altitude; by analyzing the altitude change rate and the user's movement status, the system can provide real-time environmental adaptability analysis and support the formulation of personalized plateau adaptation strategies; Environmental temperature and humidity monitoring submodule: Temperature and humidity detection relies on high-performance integrated sensors, and real-time data analysis is combined with historical climate patterns to predict the potential impact of environmental changes on the user's physiological state. The system uses a dynamic adaptive model to provide individualized health and comfort recommendations when temperature and humidity fluctuate, improving coping capabilities. Data processing and analysis module, Intelligent data analysis submodule: Through a multi-layer deep learning neural network architecture, the system analyzes various physiological and environmental data to conduct comprehensive risk assessment; the system characterizes the collected parameters, builds a personalized health status model, and identifies long-term physiological abnormal patterns; Through reinforcement learning algorithms, the system continuously optimizes the accuracy and adaptability of the prediction model to ensure efficient health risk identification and prediction capabilities; Intelligent warning submodule: Based on the adaptive fuzzy logic algorithm, the warning threshold is adjusted in real time, and multi-condition cross-validation is performed based on the user's historical physiological data and current environmental data to avoid false alarms; Implementing an event-driven early warning mechanism significantly improves the system's response speed and accuracy. When physiological indicators are detected to reach a dangerous level, the system will immediately issue an alarm. User interaction and feedback module, User interface and operation submodule: The user interface design adopts the principles based on ergonomics, combined with voice assistant and intuitive graphical interface; the system uses speech recognition and natural language processing technology to conduct natural dialogue and interaction with users, providing real-time health information and suggestions; Through user behavior analysis, the interface can automatically adjust according to user preferences to optimize the user experience; Information sharing and emergency rescue sub-module: integrated Bluetooth and WiFi transmission functions, supporting rapid sharing of health data with medical staff; when serious physiological abnormalities are detected, the system automatically triggers the emergency rescue protocol and transmits the rescue information through encrypted channels, combined with blockchain technology to ensure the integrity and security of data during transmission.

2. The intelligent dynamic monitoring and evaluation system for altitude sickness according to claim 1, characterized in that: The physiological parameter monitoring module specifically includes: ECG monitoring: Data is collected using multi-lead ECG sensors and photoplethysmography sensors; ECG sensors capture the heart's electrical activity, identifying P waves, QRS complexes, and T waves, which represent different stages of heart activity; PPG sensors supplement heart rate information by detecting changes in blood volume under the skin; The data is processed by an adaptive filter to remove myoelectric and environmental noise to obtain a clear ECG signal. Then, the peak position of the R wave in the ECG is identified by envelope detection technology. Let the ECG signal be S(t). A series of R wave peak positions are identified by envelope detection technology, expressed as {R1, R2, ..., Rn}. The RR interval is the time difference between two consecutive R wave peaks, which can be expressed as: RR i =R i+1 -R i Where i = 1, 2, ..., n-1; Average heart rate can be estimated by taking the inverse of the RR interval, expressed in minutes: When the calculated HeartRate < 60, it may be marked as bradycardia; When HeartRate>100, it may be marked as tachycardia; For arrhythmia detection, heart rate variability analysis was used to calculate the standard deviation SDNN of RR intervals and the ratio of low-frequency power LF to high-frequency power HF; where RR i For each RR interval, mean(RR) is the mean RR interval; Perform frequency domain analysis to calculate low frequency power LF and high frequency power HF; use fast Fourier transform; the ratio of low frequency power to high frequency power LF / HF can be used to assess the balance of the autonomic nervous system: P LF Represents the power of the low-frequency part in the spectrum analysis of heart rate variability, P HF Represents the power of the high frequency part in the spectrum analysis of heart rate variability, P LF , P HF Used to assess the activity of the autonomic nervous system; If the SDNN is low, it may indicate reduced heart rate variability and arrhythmia risk; if the LF / HF ratio is high, it may indicate an abnormal balance of the autonomic nervous system and may also indicate a potential arrhythmia risk; When abnormal ECG activity is detected, the system alerts the user via a mobile app and recommends further medical examinations; Blood oxygen saturation monitoring: Using dual-wavelength spectroscopy of red and infrared light, the photoelectric sensor measures changes in reflected light intensity, reflecting the proportion of oxygenated hemoglobin in arterial blood; the sensor collects AC and DC light signals; The Beer-Lambert law is used to calculate the blood oxygen saturation SpO2 using the following formula: AC R : Blood flow changes measured by alternating red light; DC R : DC component red light is used to correct the background light intensity; AC IR : Blood flow changes measured by infrared light with alternating current component; DC IR :DC component infrared light is used to correct the background light intensity; normal blood oxygen level is generally 95%-100%. If SpO2 is lower than 90%, there may be a risk of hypoxia. At this time, the system automatically notifies the user and recommends adjusting breathing or using oxygen supplementation; Temperature monitoring: Body temperature monitoring relies on the combination of thermocouples and infrared sensors to ensure fast response and accuracy; thermocouples are based on the Seebeck effect and generate a voltage proportional to temperature. The corrected formula is: T tc =V mV ·K tc K tc : Correction coefficient, used to convert voltage into actual temperature value; Combined with infrared sensors to obtain body surface temperature, it is calibrated using the radiation heat conduction equation to provide accurate body temperature readings; the normal range is set at 36.5℃ to 37.5℃. If it exceeds this range, it may indicate fever or hypothermia; The system will issue an alarm when an abnormal body temperature is detected and recommend that users take appropriate measures, such as replenishing fluids or seeking medical advice; users can view detailed body temperature trend graphs through the application to better understand their own temperature changes and health status.

3. The intelligent dynamic monitoring and evaluation system for altitude sickness according to claim 1, characterized in that: The environmental parameter monitoring module specifically includes: The altitude and GPS positioning submodule uses a MEMS pressure sensor and differential GPS technology to provide accurate altitude and location data; the pressure sensor measures the current location's air pressure value, recorded as P1, and compares it with the standard sea level atmospheric pressure, recorded as P0, to calculate the altitude, recorded as H. now ; The formula is as follows: Among them, T0 represents the standard temperature, L t is the atmospheric temperature lapse rate, R p is the ideal gas constant, g0 is the gravitational acceleration, and M is the molar mass of air; differential GPS technology corrects GPS signal errors by referencing multiple base stations with known ground locations, thereby improving the measurement accuracy of altitude; in addition, by integrating terrain and meteorological data from the geographic information system, the system can perform multi-source data fusion and error correction to ensure the accuracy of real-time altitude data; After obtaining the altitude data, the system analyzes the rate of altitude change and combines it with the user's movement status to assess the user's adaptability to environmental changes. If the rate of change exceeds the pre-set adaptability threshold, the system generates an alert and recommends that the user adjust the intensity of activity according to the current status, such as slowing down the climbing speed or taking a rest. The ambient temperature and humidity monitoring submodule uses a temperature and humidity sensor to record the ambient temperature T and humidity H in real time. w The data is compared with the historical climate model to predict the impact of environmental changes on the user's physiological state. This submodule uses a linear regression model for analysis. Through regression analysis, the system can provide users with personalized health suggestions when temperature and humidity fluctuate abnormally, such as adding clothes to keep warm or maintaining a reasonable water intake to improve the ability to adapt to environmental fluctuations. In terms of the feedback mechanism, once the system detects that the environmental parameters have reached the warning level, the system will immediately send a notification reminder through the user's device and provide specific action suggestions; when the altitude increases significantly, it is recommended to reduce the intensity of activity or take deep breathing exercises to adapt to the changing air pressure environment; when the temperature and humidity change significantly, the system recommends that users increase water intake or reduce activity intensity to avoid discomfort; in addition, the system can also automatically adjust the user's breathing training plan, combined with the user's real-time physiological data, to enhance plateau tolerance by adjusting the breathing rate and intensity.

4. The intelligent dynamic monitoring and evaluation system for altitude sickness according to claim 1, characterized in that: The data processing and analysis module specifically includes: The data acquisition module obtains physiological and environmental data through multiple sensors. These data need to be preprocessed to ensure quality. The preprocessing stage includes signal denoising, data filtering and standardization. Taking the ECG signal as an example, a bandpass filter is used to remove low-frequency drift and high-frequency noise: y[n]=a0·x[n]+a1·x[n-1]+...+a n ·x[n-α]- b1 y[n-1]-...-b m y[n-β] Where: y[n] is the filtered signal; x[n] is the input signal; a i and b i are the filter coefficients, α and β are the filter orders; In the feature extraction process, the system uses principal component analysis to reduce the data dimension and extract key information; for electrocardiogram, the features may include the amplitude and duration of the QRS band; the feature extraction formula can be expressed as Z=W(X-μ) Where Z is the eigenvector, W is the eigenvector matrix, X is the original data, and μ is the mean vector; The feature-extracted data is input into a deep learning model, which can be a convolutional neural network or a recurrent neural network, depending on the type of data and sequence characteristics; during training, the model adjusts the weights to minimize the loss function L0, using the mean square error: where y i is the actual value, is the predicted value, n sample is the sample size; In order to ensure the adaptability of the model to various populations and environments, the system adopts an ensemble learning strategy, such as random forest or gradient boosting tree; by combining multiple weak learners to improve the prediction accuracy and robustness of the model; in this architecture, the prediction result y of each weak learner is j are weighted averaged to obtain the final output, which is: where w j is the weight coefficient; The intelligent early warning submodule performs real-time monitoring through a fuzzy logic system; this system sets several fuzzy rules, including: Heart rate and blood oxygen saturation rules 1: IF heart rate is high AND blood oxygen saturation is low THEN health risk is high; 2: IF heart rate is medium AND blood oxygen saturation is low THEN health risk is medium; 3: IF heart rate is low AND blood oxygen saturation is low THEN health risk is medium; Body temperature and ambient temperature rules 4: IF body temperature is high AND ambient temperature is high THEN heat stress risk is high; 5: IF body temperature is low AND ambient temperature is low THEN hypothermia risk is high; 6: IF body temperature is normal AND ambient temperature is high THEN heat stress risk is medium; Altitude and Oxygen Saturation Rules 7: IF the altitude is high AND blood oxygen saturation is low THEN the risk of altitude sickness is high; 8: IF altitude is medium AND oxygen saturation is low THEN risk of altitude sickness is medium; 9: IF the altitude is high AND blood oxygen saturation is normal THEN the adaptability is good; Heart Rate Variability and Stress Level Rules 10: IF HRV is LOW AND STRESS LEVEL is HIGH THEN CARDIAC WORK is HIGH; 11: IF HRV is medium AND STRESS LEVEL is medium THEN CARDIAC LOAD is medium; 12: IF heart rate variability is high AND stress level is low THEN heart condition is good; These rules convert continuous data into fuzzy sets through the fuzzification process and process them through the inference engine; the fuzzy rule system can be expressed as: IF(θ1is A1 ANDθ2is A2)THENηis B Among them, A1, A2 and B are fuzzy sets; The adaptive threshold mechanism is another key feature of the system. It adjusts the warning threshold by analyzing the user's historical data and current environment data in real time. This process uses the weighted moving average method to smooth the input data. The formula is: Where V t is the smoothed value, x t is the current observation value, w i is the weight; Once any physiological parameter is predicted or detected to exceed the set threshold, the system immediately triggers an alert and notifies the user through a mobile device or other interface; the alert information not only contains the abnormal parameter, but also provides recommended measures, including rest, oxygen supplementation, or contacting professional medical help; In the feedback mechanism, the system records user responses to optimize the model; by analyzing user feedback on warning information, the system uses reinforcement learning algorithms to update its model parameters to make future predictions more accurate; in this way, the data processing and analysis module ensures the safety and health of users in plateau environments through efficient data collection, complex model analysis and flexible feedback mechanisms.

5. The intelligent dynamic monitoring and evaluation system for altitude sickness according to claim 1, characterized in that: The user interaction and feedback module specifically includes: The user interface and operation submodule provides an interface that enables users to interact with the system effectively. First, the interface design follows the principles of ergonomics to improve user comfort and operating efficiency. In this process, the position and size of the user interface elements are obtained by analyzing user behavior data. For each interface element, such as a button, the system calculates an optimal position and size by analyzing the user's click frequency and position preference. Assume the click frequency of each interface element i is f i , and the relative position is p i , then the goal of interface layout optimization is to maximize user interaction efficiency E interaction , which can be expressed as: in is the position weight of interface element i; on this basis, the interface layout is continuously adjusted through the optimization algorithm to improve E interaction value; In the implementation of natural language processing and voice assistants, the user's voice input is converted into text by the speech recognition module and then parsed by the NLP module; the accuracy of speech recognition depends on the combination of hidden Markov models and deep neural networks, where each speech segment is mapped to a state s, and the probability of state transition is P(s t |s t-1 ), and optimize the model parameter λ by maximizing P(O|λ) to maximize the probability of the output sequence O; After processing the user input, the system generates a response using the NLP module. This process involves sentence intent recognition and semantic understanding. Assume that the user's input text is represented by a set of word vectors V = [v1, v2, ..., v n ]The system determines the system response by calculating the semantic vector S of V; In the information sharing and emergency rescue submodule, the system uses Bluetooth and WiFi to quickly share health data; when an emergency occurs, the system identifies abnormalities through the physiological parameter monitoring module; when the heart rate H exceeds the safety threshold H th , and blood oxygen saturation S pO2 Below the set threshold S pO2th When the trigger condition C is true: C=(H>H th )AND(S pO2 pO2th );​ Once C is true, the system immediately initiates the emergency rescue protocol; the distress message is encrypted using the formula E=enc(M, key), where M is the distress message and key is the encryption key, and is transmitted to the preset emergency contact through a secure channel; the information transmission is recorded in the blockchain to ensure transparency and integrity, and the blockchain records the hash value H(M) to verify that the information has not been tampered with; The feedback mechanism optimizes system performance by analyzing user feedback data, such as operation response time, user satisfaction score and other parameters; for example, for each functional module i, the user feedback response time t i and satisfaction i , the system minimizes the objective function: F satisfaction is the objective function value, which represents the negative utility of user experience; i is the response time of functional module i; τ i is the user satisfaction score of functional module i; To adjust the system configuration to ensure fast system response and high user satisfaction.

6. A wearable device according to any one of claims 1 to 5, characterized in that: include: The invention comprises a monitoring suit (1) and a plurality of sensor groups, wherein the sensors comprise an electrocardiogram sensor (2), a PPG sensor (3), a blood oxygen sensor (4), and a body temperature monitoring sensor (5); the electrocardiogram sensor (2) comprises a flexible electrode and a multi-lead ECG sensor, which is integrated into the inner layer of the monitoring suit (1) and located at the chest of the monitoring suit (1); the PPG sensor (3) is located below the shoulder strap of the monitoring suit (1) and combines photoelectric technology to obtain blood flow change information; the blood oxygen sensor (4) is located on the side of the monitoring suit (1) near the armpit and uses red light and infrared light dual wavelength technology to measure the blood oxygen level; at the same time, the blood oxygen sensor (4) is provided with a dynamic light compensation module integrated into the inner layer of the monitoring suit (1) to adjust for different skin colors and changes in ambient light; the body temperature monitoring sensor (5) comprises a plurality of groups of micro thermocouples and infrared sensors, which are respectively located at multiple points on the front chest, back and armpit of the monitoring suit (1) to obtain multi-point body temperature data.

7. The wearable device according to claim 6, wherein: The monitoring suit (1) further comprises a MEMS air pressure sensor (6), a GPS module (7), and a temperature and humidity sensor (8); the MEMS air pressure sensor (6) is embedded in the outer side of the shoulder of the monitoring suit (1) to sense air pressure changes and thereby calculate altitude; the GPS module (7) is integrated in the upper part of the rear side of the collar of the monitoring suit (1) to provide accurate geographical location information; the temperature and humidity sensor (8) is located at the collar or cuff of the monitoring suit (1) to monitor the temperature and humidity of the external environment in real time.

Images