Non-contact passenger uncomfortable state early warning system suitable for plateau train

Abstract

The invention relates to the technical field of plateau trains, in particular to a non-contact passenger uncomfortable state early warning system suitable for a plateau train, which comprises a passenger state information real-time monitoring system and a passenger uncomfortable state early warning system, the non-contact passenger uncomfortable state early warning system suitable for the plateau train has the advantages that real-time monitoring of the passenger state is achieved by integrating visual signals and physiological signals of passengers, and higher objectivity and accuracy are achieved. In the early warning mechanism, the system fully considers individual differences of passengers, constructs a personalized early warning standard, introduces a dynamic adjustment mechanism, automatically updates an early warning threshold according to real-time environment working conditions and physiological and visual data, and ensures accurate evaluation and feedback of the state of each passenger.

Description

Technical Field

[0001] The invention relates to the field of plateau trains, and in particular to a contactless passenger discomfort state warning system suitable for plateau trains. Background Art

[0002] At present, plateau trains face a climate environment of "low temperature, low pressure, low oxygen, and strong ultraviolet rays". Due to the terrain characteristics of the plateau, long and large tunnels, dense tunnel groups, and continuous steep slopes frequently appear, and the train service environment is poor. The plateau climate characteristics and terrain characteristics jointly aggravate the deterioration of the interior environment of the car. In addition, most of the passengers on plateau trains are people who rush to the plateau. Entering the plateau in a short period of time can easily cause physiological responses such as a decrease in heart rate and an increase in respiratory rate. In severe cases, it can cause physiological and pathological changes in multiple systems such as circulation, respiration, blood, and neuroendocrine, and induce acute altitude sickness, such as panic, chest tightness, headache, and fatigue. Given the limited medical conditions on plateau trains, timely identification and warning of passengers' discomfort has become an important measure to ensure the safety and comfort of passengers.

[0003] However, there is a relative lack of research on the comfort of this special group of plateau train passengers. In addition, existing monitoring methods mostly rely on wearable devices, such as physiological bracelets. Although they can collect relevant data for early warning, the application of wearable devices is limited in practicality and convenience in plateau train scenarios with high passenger density.

[0004] Therefore, for the special operating scenario of plateau trains, there is an urgent need for a non-contact, non-invasive passenger discomfort warning system to monitor passengers' status changes in real time, identify risks early, and take preventive measures. Summary of the invention

[0005] The purpose of the present invention is to provide a non-contact passenger discomfort state warning system suitable for plateau trains to solve the problems raised in the above-mentioned background technology.

[0006] To achieve the above-mentioned purpose, the present invention provides the following technical solutions: a non-contact passenger discomfort state warning system suitable for plateau trains, the system comprising a passenger state information real-time monitoring system and a passenger discomfort state warning system; The real-time monitoring system for occupant status information includes a data acquisition module, a data processing module and a data analysis module. The data acquisition module collects occupant status data information, the data processing module pre-processes the data and extracts key features, and the data analysis module integrates multi-modal features, removes strongly correlated features, extracts highly sensitive indicators, and finally forms an occupant status data set. The occupant discomfort state warning system issues warnings based on the occupant status data set of the occupant status information real-time monitoring system. The occupant discomfort state warning system includes an occupant heterogeneous information library construction module, a personalized warning standard module, a status level confirmation module, and a visualization page and alarm module.

[0007] Preferably, the data acquisition module includes physiological information acquisition and visual information acquisition; Physiological information collection: millimeter wave radar is used to monitor the heart rate and breathing rate of passengers at a long distance; infrared thermal imagers are used to measure the skin temperature of passengers at a long distance, especially the thermal radiation of the face and forehead area; Visual information collection: High-definition cameras are used to collect passengers’ facial expressions and body movements in real time.

[0008] Preferably, the data processing module includes data preprocessing and feature extraction; data preprocessing includes physiological signal preprocessing and image data enhancement; feature extraction includes physiological feature extraction and visual feature extraction; Physiological signal preprocessing: Multiple filtering processes are performed on physiological signals collected by millimeter-wave radar and infrared thermal imagers to effectively remove environmental noise, equipment interference and invalid signals, improve data purity and reliability, and apply interpolation algorithms to complete missing data in case of signal loss or abnormal interruption to ensure data continuity and integrity; Image data enhancement: Image enhancement technology is used to process facial expression and body movement images collected by high-definition cameras to improve image resolution, contrast, and detail clarity, thereby improving the effect of subsequent visual analysis; Physiological feature extraction: Extract key physiological indicators from physiological signals, including heart rate variability, respiratory rate, and skin temperature physiological indicators. These features effectively reflect the health status of the occupants and identify potential health risks; Visual feature extraction: By analyzing the facial expressions and body movements of the occupants, visual features are extracted, including facial key points and body posture information. This process helps describe the occupants' emotional changes and behavioral characteristics, thereby assisting in monitoring the occupants' status.

[0009] Preferably, the data analysis module integrates physiological and visual features, removes strongly correlated or redundant indicators in the feature set, reduces multicollinearity between features, and ensures data independence; uses a feature selection algorithm to screen out key features that are most sensitive to occupant discomfort and emotional fluctuations, and retains indicators that contribute significantly to the prediction of discomfort; and constructs an occupant status data set after the optimized feature set to provide high-quality data support for the early warning model of discomfort status, ensuring that the model can accurately identify and predict the occupant's potential discomfort status.

[0010] Preferably, the occupant heterogeneity information library construction module includes occupant basic information, basic physiological and visual information, and environmental working condition information; Passenger basic information: Collect the basic information of the passengers, including age, gender, and weight. This information will be sorted and stored in the passenger basic information database to facilitate the establishment of subsequent personalized warning standards; Basic physiological and visual information: Collect the physiological and visual information of passengers at the beginning of the ride as the baseline value to form the physiological and visual index files of passengers; Environmental condition information: By installing sensor equipment at different locations in the car, the air pressure, noise, vibration, altitude, temperature, and air quality data in the passenger compartment are collected to monitor the environmental characteristics of the area where the passengers are located and form an environmental information collection.

[0011] Preferably, the personalized warning standard module includes personalized warning standards and a dynamic adjustment mechanism; Personalized warning standards: Based on the heterogeneous information of passengers, combined with the normal fluctuation range of conventional physiological indicators, and taking into account the age, gender, and basic physiological conditions of passengers, personalized warning standards with different comfort levels are established; Dynamic adjustment mechanism: Based on real-time feedback from passengers and current environmental conditions, the warning standards are continuously updated and adjusted to ensure the adaptability and effectiveness of the warning system and to promptly respond to changes in the passengers' status.

[0012] Preferably, the status level confirmation module, based on the real-time physiological and visual information of the occupants, compares the personalized warning standard thresholds, comprehensively analyzes the deviation degree and duration of various indicators, determines the current comfort level of the occupants, determines whether the occupants are comfortable (high, medium, low) or uncomfortable (high, medium, low), and promptly feeds back changes in the occupants' comfort and triggers a corresponding warning mechanism.

[0013] Preferably, the visualization page and alarm module comprises a visualization page and an alarm mechanism; Visualization page: Through the visualization platform, the status information of passengers is intuitively presented, including the passenger heterogeneity information database, real-time collected physiological and visual data, personalized warning standards, and real-time status level of passengers, helping passengers and train staff to quickly understand the health status of passengers, facilitate operation and decision-making, and achieve efficient management and immediate response; Alarm mechanism: When the system detects that an occupant is in a state of moderate discomfort, the visual page will use a flashing alarm to remind relevant personnel to pay attention to the occupant's condition. When the moderate discomfort state tends to turn into severe discomfort, the system will trigger a voice alarm while the visual page flashes, ensuring that staff are informed in time and take appropriate measures to prevent the occupant's condition from further deteriorating.

[0014] Compared with the prior art, the present invention has the following beneficial effects: The non-contact passenger discomfort state warning system suitable for plateau trains proposed by the present invention realizes real-time monitoring of the passenger status by integrating the visual signals and physiological signals of the passengers, and has higher objectivity and accuracy. In terms of the warning mechanism, the system fully considers the individual differences of the passengers, constructs personalized warning standards, and introduces a dynamic adjustment mechanism to automatically update the warning threshold according to the real-time environmental conditions, physiological and visual data, ensuring accurate evaluation and feedback of the status of each passenger. Compared with traditional contact equipment, the present invention adopts a non-contact design to avoid direct interference with the passengers and improve the comfort of riding. Through millimeter-wave radar and infrared thermal imaging non-invasive sensing technology, the system can continuously and stably collect high-precision physiological and visual information under the premise of normal activities of the passengers, ensuring the continuity, reliability and low interference characteristics of the monitoring process. Therefore, the present invention has extremely strong adaptability and practicality in the plateau train environment, can effectively identify the discomfort state of plateau train passengers, and reduce potential health risks in the plateau environment. BRIEF DESCRIPTION OF THE DRAWINGS Figure 1 This is a block diagram of the real-time monitoring system for passenger status information of the present invention; Figure 2 This is a block diagram of the occupant discomfort state warning system of the present invention. DETAILED DESCRIPTION

[0015] In order to make the purpose and technical solution of the present invention clearly and completely described, and the advantages more clearly understood, the embodiments of the present invention are further described in detail with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are part of the embodiments of the present invention, rather than all of the embodiments, and are only used to explain the embodiments of the present invention, and are not used to limit 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.

[0016] For example, see Figure 1 The present invention provides a technical solution: a non-contact passenger status information real-time monitoring system suitable for plateau trains, through which the passenger status information can be monitored in real time. First, the passenger's heart rate, breathing and other physiological and visual information are detected by non-contact equipment; the data is pre-processed by the data processing module to extract key features; then, the multimodal features are integrated in the data analysis module, the strongly correlated features are removed, and the highly sensitive indicators are extracted, and finally the passenger status data set is formed.

[0017] like Figure 1 As shown, the principle of a non-contact passenger status information real-time monitoring system suitable for plateau trains is as follows: 1) Data acquisition module: a) Millimeter wave radar data acquisition Equipment installation: Install the millimeter-wave radar equipment at an appropriate location in the train compartment to ensure that it covers the chest and abdomen areas of the passengers. Start the equipment and adjust the radar's detection range and angle according to the compartment layout to ensure effective coverage of each passenger.

[0018] Real-time monitoring: The millimeter-wave radar detects tiny displacements of the chest and abdomen of the occupants by emitting millimeter-wave signals and receiving reflected signals. Since heartbeat and breathing cause slight movements of the chest and abdomen, the millimeter-wave radar can calculate the occupants' heart rate and breathing rate based on these displacement signals, while automatically filtering out environmental noise to ensure data accuracy.

[0019] Data transmission: Heart rate and respiratory rate data are transmitted to the data processing module in real time for further analysis and feature extraction.

[0020] b) Infrared thermal imager data acquisition Equipment installation: Install the infrared thermal imager at a suitable viewing angle in the train compartment, and select an area that can effectively cover the passenger's face. Start the equipment and calibrate the temperature measurement range to ensure the accuracy of temperature measurement when the ambient temperature in the compartment changes.

[0021] Real-time monitoring: Infrared thermal imagers measure temperature based on infrared radiation emitted from the surface of the occupants' skin. The device captures the occupants' facial thermal images in real time, focusing mainly on the forehead, face and other areas where body temperature changes significantly. The device calculates the temperature changes of the occupants' skin by analyzing the radiation intensity of different parts of the skin. The system can identify areas with abnormal temperatures, such as changes in skin temperature caused by altitude sickness.

[0022] Data transmission: Each occupant's skin temperature data is transmitted to the data processing module for further analysis.

[0023] c) High-definition camera data acquisition Equipment installation: Install a high-definition camera inside the vehicle, ensuring that the camera's field of view fully covers the passenger's face and the area above the chest. Start the camera system and adjust the angle and focal length to ensure that the passenger's facial expressions and body movements are clearly captured.

[0024] Data collection: High-definition cameras capture the facial images and body postures of passengers in real time. By collecting high-resolution video data, the system records the changes in the passengers' facial expressions and body movement information, providing basic data for subsequent analysis.

[0025] Data transmission: The visual data collected by the high-definition camera will be automatically transmitted to the data processing module.

[0026] 2) Data processing module: a) Data synchronization and time alignment: Match accurate timestamps for physiological and visual information, so that data of different modalities can be aligned on the same time axis to avoid data analysis deviations due to time differences. Combine high-frequency data downsampling and low-frequency data interpolation to achieve time synchronization of data of different frequencies and ensure the accuracy of subsequent analysis.

[0027] b) Physiological signal preprocessing: For the collected physiological signals, use filters to remove high-frequency noise and low-frequency interference in the signals to ensure that the data is clean and reliable; detect and eliminate abnormal data caused by occupant movements or environmental changes, and further optimize signal stability through the Kalman filter algorithm to reduce short-term fluctuations and unstable factors. Standardize the processed physiological signals and convert signals from different sources into a standardized dimension range to ensure that various physiological signals are comparable and consistent in subsequent multi-source data fusion analysis.

[0028] c) Visual signal preprocessing: For image data collected by high-definition cameras, image denoising algorithms are used to reduce the noise of images, eliminate image noise caused by ambient lighting conditions, and improve image quality. Image enhancement methods such as histogram equalization and contrast adjustment are used to improve the visibility of facial and body movements in images, ensuring that the system can accurately capture the facial expressions and posture changes of the occupants under different lighting conditions. In response to camera viewing angle deviations or changes in occupant position, facial and body key points are accurately located and corrected to ensure the accuracy of the data under different occupant postures.

[0029] 3) Feature extraction: a) Extraction of physiological information features: Detect the heartbeat interval (RR interval) from the processed heart rate signal, extract continuous heartbeat data, calculate the time domain indicators of heart rate variability (HRV), including heart rate (HR), RR interval standard deviation (SDNN), root mean square difference between adjacent RR intervals (RMSSD), standard deviation of the difference between adjacent RR intervals (SDSD), etc., apply fast Fourier transform (FFT) to calculate the low frequency (LF) and high frequency (HF) power of heart rate variability; for respiratory signals, use the peak detection algorithm to identify the respiratory cycle and calculate the respiratory frequency (RF); extract the skin temperature information (ST) of the forehead and nose.

[0030]

[0031]

[0032]

[0033]

[0034]

[0035] b) Visual information feature extraction: The facial images of the occupants are collected through high-definition cameras, and the Haar features of the face detection algorithm are used to identify the facial areas of the occupants. The Dlib library is used to identify 68 key points of the face, including points at the eyes, eyebrows, nose, mouth and jaw, and extract facial motion information. Machine vision algorithms or deep learning models, such as OpenPose or PoseNet, are used to capture the occupants' body postures through cameras, extract the occupants' body key points (such as shoulders, elbows, etc.), calculate the relative positions between the key points, and analyze body movements.

[0036] 4) Data analysis module: An automatic encoder is used to learn complex nonlinear feature mappings. By compressing the data into a low-dimensional latent space and then reconstructing it back to the original dimension, the original data is reconstructed while maintaining the main features of the data, removing redundant information, reducing computational costs, reducing the risk of model overfitting, and improving the prediction accuracy of the model, forming a data set containing occupant status information.

[0037] Embodiment 2, based on embodiment 1, refers to the attached Figure 2 As shown in the figure, a non-contact passenger discomfort warning system suitable for plateau trains is proposed. Figure 1 The system uses the passenger status information monitored by the system for early warning. First, a heterogeneous information database is constructed, which includes basic information such as age, gender, and weight of passengers, basic physiological and visual information such as heart rate, skin temperature, and respiration at the beginning of the ride, and environmental information such as air pressure and noise in the area where the passengers are located; then a personalized early warning standard is established, and the personalized early warning standard containing different comfort levels for each passenger is determined according to the basic information and basic physiological and visual information of the passengers, and the early warning standard is dynamically adjusted and updated in real time according to the physiological and visual information and environmental working condition information monitored in real time; based on the passenger status and personalized early warning standard at this moment, the current comfort level is determined by calculating the degree of deviation from the standard level, and displayed on the visualization page, and the alarm module is triggered when an abnormality occurs. Among them, the heterogeneous information database construction module—basic physiological and visual information, personalized early warning module—real-time status of passengers, status level confirmation module—passenger status data, etc. are all from the passenger status data collected by a non-contact passenger status information real-time monitoring system suitable for plateau trains.

[0038] The specific implementation includes: 1) Construction of crew heterogeneity information database: a) Basic passenger information: When passengers enter the train, the system collects basic passenger information, including age, gender, etc., through an automated platform. This information can be automatically entered through the online platform, and the collected data is sorted to form a structured basic information database, ensuring efficient retrieval and accurate matching of personalized data when subsequent data is called.

[0039] b) Basic physiological and visual information of passengers: In the early stage of passengers' riding, the system collects the baseline physiological and visual information of passengers through non-contact sensors, and associates it with their basic information to form a personalized physiological index file. This file will serve as the basis for subsequent dynamic monitoring and early warning evaluation, and provide a reference for accurately judging the status of passengers.

[0040] c) Environmental condition information: Divide the compartment into areas and install noise, air pressure, vibration, altitude, temperature, and air quality sensing equipment at appropriate locations to monitor the environmental information in the passenger compartment in real time and form a real-time data set of the environment in which the passengers are located.

[0041] 2) Personalized warning criteria: a) Personalized warning standards: Based on the data in the occupant heterogeneity information database, combined with medical standards and the impact of the plateau environment on the physiological indicators of different groups of people, personalized warning standards for different comfort levels are generated by limiting the deviation from the baseline data, including high comfort, moderate comfort, low comfort, low discomfort, moderate discomfort, and high discomfort. These standards will take into account the occupant's age, gender, previous health records, and the fluctuation range of personal physiological status. The personalized warning standards are used as thresholds to monitor in real time whether the occupant's physiological data exceeds the level range.

[0042] b) Dynamic adjustment mechanism: The system uses supervised learning to dynamically adjust the occupant's warning standard based on real-time data and previously accumulated historical data, combined with changes in environmental data. For example, if a occupant's heart rate increases after entering a plateau environment but remains within the individual's permitted range, the system will appropriately relax or tighten the threshold based on their physiological tolerance to avoid premature alarms. Ensure that the standard closely matches the occupant's current state and environmental conditions.

[0043] 3) Crew status level confirmation: The Euclidean distance formula is used to compare the distance between the occupant's real-time data and the warning standards of different comfort levels to determine which comfort level the occupant belongs to. Specifically, the real-time monitored physiological data will be compared with the preset standard values ​​of each warning level, and the degree of deviation will be calculated. The occupant's real-time status will be classified into the level with the smallest distance.

[0044] 4) Visualization page and alarm: a) Visualization page: A visualization page is established on the vehicle display screen and mobile terminal to display the basic information of the occupants, real-time physiological and visual data, personalized warning standards, and occupant status levels. The visualization interface uses charts, color coding and simple status descriptions to facilitate quick identification of problems. The page will display a plan view of the occupant's position, and different colors are used to represent the different states of the occupants. Green represents high comfort, blue represents moderate comfort, purple represents low comfort, yellow represents mild discomfort, orange represents moderate discomfort, and red represents severe discomfort.

[0045] b) Alarm mechanism: When the system detects that an occupant is in a state of moderate discomfort, the occupant's position will flash on the visualization page to remind relevant personnel to pay attention to their condition. If the moderate discomfort state is accompanied by a trend of transitioning to high discomfort, the system will continue to flash and trigger a voice alarm at the same time to ensure that the staff is informed in time and take appropriate measures. During this process, the system will automatically send an alarm message to the staff, including the occupant's seat position, current physiological state and recommended emergency measures, so as to quickly deal with it and prevent the occupant's condition from further deteriorating.

[0046] Example 3. On the basis of Example 2, during the implementation of a contactless passenger discomfort warning system suitable for plateau trains, when a plateau train reached an altitude of 4,632 meters, the physiological state of Zhang XX (male, 26 years old, 171 cm, 77 kg) changed compared with the baseline value: the heart rate increased by 22.43% compared with the baseline state, the skin temperature dropped by 1.1 ° C, and the respiratory rate increased by 2.19 times / minute. At the same time, the face showed discomfort, the corners of the mouth drooped, the brows were furrowed, and there were symptoms of night sweats. Based on these physiological and visual information, the system identified his state as moderate discomfort, and the seat 3A in carriage 8 where Zhang XX was located displayed an orange warning. The train staff responded quickly and provided oxygen cylinders, warm water and glucose, and the discomfort symptoms were subsequently relieved. At the same time, Li X's (female, 27 years old, 158 cm, 46 kg) physiological state changed slightly: her heart rate increased by 14.20% compared to the baseline state, her skin temperature dropped by 0.3°C, her respiratory rate increased by 0.76 times / minute, and she did not show any abnormal facial expressions. The system identified her as "mildly comfortable" and no further intervention was required.

[0047] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A non-contact passenger discomfort warning system suitable for plateau trains, characterized by: The system includes a real-time monitoring system for passenger status information and a warning system for passenger discomfort status; The real-time monitoring system for occupant status information includes a data acquisition module, a data processing module and a data analysis module. The data acquisition module collects occupant status data information, the data processing module pre-processes the data and extracts key features, and the data analysis module integrates multi-modal features, removes strongly correlated features, extracts highly sensitive indicators, and finally forms an occupant status data set. The occupant discomfort state warning system issues warnings based on the occupant status data set of the occupant status information real-time monitoring system. The occupant discomfort state warning system includes an occupant heterogeneous information library construction module, a personalized warning standard module, a status level confirmation module, and a visualization page and alarm module.

2. According to claim 1, a non-contact passenger discomfort state warning system suitable for plateau trains is characterized by: The data acquisition module includes physiological information acquisition and visual information acquisition; Physiological information collection: millimeter wave radar is used to monitor the heart rate and breathing rate of passengers at a long distance; infrared thermal imagers are used to measure the skin temperature of passengers at a long distance, especially the thermal radiation of the face and forehead area; Visual information collection: High-definition cameras are used to collect passengers’ facial expressions and body movements in real time.

3. According to claim 2, a non-contact passenger discomfort state warning system suitable for plateau trains is characterized by: The data processing module includes data preprocessing and feature extraction; Data preprocessing includes physiological signal preprocessing and image data enhancement; Feature extraction includes physiological feature extraction and visual feature extraction; Physiological signal preprocessing: Multiple filtering processes are performed on physiological signals collected by millimeter-wave radar and infrared thermal imagers to effectively remove environmental noise, equipment interference and invalid signals, improve data purity and reliability, and apply interpolation algorithms to complete missing data in case of signal loss or abnormal interruption to ensure data continuity and integrity; Image data enhancement: Image enhancement technology is used to process facial expression and body movement images collected by high-definition cameras to improve image resolution, contrast, and detail clarity, thereby improving the effect of subsequent visual analysis; Physiological feature extraction: Extract key physiological indicators from physiological signals, including heart rate variability, respiratory rate, and skin temperature physiological indicators. These features effectively reflect the health status of the occupants and identify potential health risks; Visual feature extraction: By analyzing the facial expressions and body movements of the occupants, visual features are extracted, including facial key points and body posture information. This process helps describe the occupants' emotional changes and behavioral characteristics, thereby assisting in monitoring the occupants' status.

4. A non-contact passenger discomfort state warning system suitable for plateau trains according to claim 3, characterized in that: The data analysis module integrates physiological and visual features, removes strongly correlated or redundant indicators in the feature set, reduces multicollinearity between features, and ensures data independence; A feature selection algorithm is used to screen out the key features that are most sensitive to occupant discomfort and emotional fluctuations, and retain indicators that have a significant contribution to the prediction of discomfort. The optimized feature set is constructed into an occupant status dataset, providing high-quality data support for the early warning model of discomfort, ensuring that the model can accurately identify and predict the potential discomfort of the occupants.

5. According to claim 1, a non-contact passenger discomfort state warning system suitable for plateau trains is characterized by: The occupant heterogeneity information library construction module includes occupant basic information, basic physiological and visual information, and environmental condition information; Passenger basic information: Collect the basic information of the passengers, including age, gender, and weight. This information will be sorted and stored in the passenger basic information database to facilitate the establishment of subsequent personalized warning standards; Basic physiological and visual information: Collect the physiological and visual information of passengers at the beginning of the ride as a baseline value to form a physiological and visual index file of the passengers; Environmental condition information: By installing sensor equipment at different locations in the car, the air pressure, noise, vibration, altitude, temperature, and air quality data in the passenger compartment are collected to monitor the environmental characteristics of the area where the passengers are located and form an environmental information collection.

6. According to claim 1, a non-contact passenger discomfort state warning system suitable for plateau trains is characterized by: The personalized early warning standard module includes personalized early warning standards and dynamic adjustment mechanisms; Personalized warning standards: Based on the heterogeneous information of passengers, combined with the normal fluctuation range of conventional physiological indicators, and taking into account the age, gender, and basic physiological conditions of passengers, personalized warning standards with different comfort levels are established; Dynamic adjustment mechanism: Based on real-time feedback from passengers and current environmental conditions, the warning standards are continuously updated and adjusted to ensure the adaptability and effectiveness of the warning system and to promptly respond to changes in the passengers' status.

7. A non-contact passenger discomfort state warning system suitable for plateau trains according to claim 5, characterized in that: The status level confirmation module, based on the occupant's real-time physiological and visual information, compares the personalized warning standard threshold, comprehensively analyzes the deviation degree and duration of various indicators, determines the occupant's current comfort level, and determines whether the occupant's status is comfortable (high, medium, low) or uncomfortable (high, medium, low), timely feedbacks the occupant's comfort changes and triggers the corresponding warning mechanism.

8. A non-contact passenger discomfort state warning system suitable for plateau trains according to claim 7, characterized in that: The visualization page and alarm module includes a visualization page and an alarm mechanism; Visualization page: Through the visualization platform, the status information of passengers is intuitively presented, including the passenger heterogeneity information database, real-time collected physiological and visual data, personalized warning standards, and real-time status level of passengers, helping passengers and train staff to quickly understand the health status of passengers, facilitate operation and decision-making, and achieve efficient management and immediate response; Alarm mechanism: When the system detects that an occupant is in a state of moderate discomfort, the visual page will use a flashing alarm to remind relevant personnel to pay attention to the occupant's condition. When the moderate discomfort state tends to turn into severe discomfort, the system will trigger a voice alarm while the visual page flashes, ensuring that staff are informed in time and take appropriate measures to prevent the occupant's condition from further deteriorating.

Images