An AI-powered dynamic stress and emotion analysis method and system

Abstract

This application relates to the field of psychophysiological monitoring and data analysis technology, specifically an artificial intelligence-based dynamic stress and emotion analysis method and system. The method includes the following steps: collecting students' physiological data and body movement data while they are performing hazardous tasks; analyzing the physiological data and body movement data based on a preset algorithm to determine if the students have negative emotions, including anxiety, tension, fear, anger, and lethargy; issuing an early warning signal when negative emotions are detected and initiating a response procedure to notify relevant personnel, including users, rescue personnel, and students; generating a health feedback report and sending it to users, including teachers, instructors, and project leaders. This application has the capability to efficiently and in real-time monitor and analyze the psychological stress and emotional state of large groups in complex and high-risk work environments, particularly in educational practice training and emergency situations.

Description

Technical Field

[0001] This application relates to the field of psychophysiological monitoring and data analysis technology, and in particular to an artificial intelligence dynamic stress emotion analysis method and system. Background Technology

[0002] The application of smart wearable technology has permeated many areas of daily life, including medical monitoring, sports training, and health management, and its important role is self-evident. These technologies can monitor users' physiological indicators in real time, such as heart rate and blood pressure, playing a crucial role in assessing users' health status and emotional responses. In most one-to-many monitoring environments, smart wearable technology faces similar challenges, although these challenges are not limited to a specific field. For example, in medical monitoring, doctors may need to monitor the vital signs of multiple patients simultaneously; in sports training, coaches may need to track the physical condition of the entire team. In these situations, while smart wearable technology excels in data collection, its speed and accuracy often fall short when processing large amounts of data, leading to delays or inaccuracies in monitoring results.

[0003] However, these limitations of wearable technology are particularly prominent in vocational education. In professional teaching and practical operations such as rail transit, high-speed rail overhead contact line operation, high-voltage power grid pole climbing, and high-altitude work, students and workers face high-risk working environments, which places higher demands on the application of wearable technology. Firstly, in vocational education scenarios, wearable technology struggles to process data generated simultaneously by a large number of students. This limitation in processing capacity leads to delays in monitoring results, a problem that urgently needs to be addressed in vocational education settings where immediate feedback is crucial to ensuring student safety.

[0004] Secondly, accuracy is also a weakness of existing technologies in vocational education applications. In complex teaching environments, especially in one-to-many monitoring scenarios, algorithms struggle to accurately identify changes in each student's psychological state, and data noise further impacts the accuracy of monitoring results. These limitations are particularly pronounced in high-risk vocational education environments, such as high-altitude operations on high-speed rail overhead contact lines. In these scenarios, students and workers face immense psychological pressure, making real-time and accurate tracking of their psychological states crucial for accident prevention and personnel safety. However, current smart wearable technologies cannot yet meet the demands for efficient and accurate data processing in such complex one-to-many environments. Therefore, improvements are needed. Summary of the Invention

[0005] To address the technical challenge of lacking efficient and real-time monitoring and analysis of psychological stress and emotional states in large-scale populations in complex and high-risk work environments, especially in educational practice and emergency situations, this application provides an artificial intelligence-based dynamic stress and emotion analysis method and system.

[0006] The first objective of this invention is achieved through the following technical solution:

[0007] An AI-powered dynamic stress sentiment analysis method includes the following steps:

[0008] Collect students' physiological and physical movement data while they are performing dangerous tasks;

[0009] Based on a preset algorithm, the physiological data and the body movement data are analyzed to determine whether the student has negative emotions, including anxiety, tension, fear, anger, and lethargy.

[0010] When negative emotions arise, a warning signal is issued and a response procedure is initiated to notify relevant personnel, including users, rescue personnel, and students.

[0011] A health feedback report is generated and sent to the user terminals, including teacher terminals, instructor terminals, and project team leader terminals.

[0012] In a preferred embodiment, the step of collecting students' physiological data and limb movement data while students are performing dangerous tasks includes the following steps:

[0013] Based on GPS, the student's specific location is input into a preset database to match the corresponding hazardous work area;

[0014] Based on the hazardous work area, determine the type of hazardous work;

[0015] Input the type of hazardous work into the preset database, and collect students' physiological data and limb movement data at the corresponding preset frequency;

[0016] The hazardous work types include working at heights, working with hazardous substances, operating machinery, precision operations, and physically demanding work.

[0017] The physiological data includes heart rate data, blood pressure data, body temperature data, and skin conductivity data;

[0018] The limb movement data includes movement frequency data, movement amplitude data, and movement speed data.

[0019] In a preferred embodiment, the step of analyzing the physiological data and the body movement data based on a preset algorithm to determine whether the student has negative emotions, including anxiety, tension, fear, anger, and lethargy, includes the following steps:

[0020] Based on an adaptive filtering method, the physiological data and the limb movement data are preprocessed, and the adaptive filtering method is combined with deep learning technology.

[0021] Based on time series analysis technology, features are extracted from the preprocessed physiological data and limb movement data;

[0022] The forward propagation process based on deep neural networks is formulated as f(x) = σ(W). T ·φ(X)+b), where f(x) is the output feature vector, W is the weight matrix, φ(X) is the input feature vector, b is the bias term, and σ is the activation function, which fuses the extracted features into a multidimensional feature vector;

[0023] Based on machine learning algorithms, the multidimensional feature vectors are optimized to output the optimal combination of features;

[0024] Based on an adaptive weight allocation mechanism and optimal combination, a preset weight is assigned to each element of the multidimensional feature vector, the weighted sum is calculated, and the score is output.

[0025] In a preferred embodiment, the step of extracting features from the preprocessed physiological data and limb movement data based on time series analysis technology includes the following steps:

[0026] Heart rate features are extracted from the preprocessed heart rate data, including mean heart rate, heart rate variability, and peak heart rate.

[0027] Blood pressure features are extracted from the preprocessed blood pressure data, including blood pressure variability, blood pressure fluctuation amplitude, and blood pressure load.

[0028] Body temperature features are extracted from the preprocessed body temperature data, including average body temperature, body temperature variability, and peak body temperature.

[0029] Skin conductivity features are extracted from the preprocessed skin conductivity data. These features include skin conductivity fluctuation amplitude, skin conductivity variability, and skin conductivity peak value.

[0030] Action frequency features are extracted from the preprocessed action frequency data. These features include the average action frequency, action frequency variability, and action frequency peak value.

[0031] Motion amplitude features are extracted from the preprocessed motion amplitude data. These features include motion amplitude variability, motion amplitude fluctuation, and motion amplitude load.

[0032] Action speed features are extracted from the preprocessed action speed data. These features include action speed variability, average action speed, and peak action speed.

[0033] In a preferred embodiment, the step of analyzing the physiological data and the body movement data based on a preset algorithm to determine whether the student has negative emotions, including anxiety, tension, fear, anger, and lethargy, further includes the following steps:

[0034] Input students' information into a preset database and match them with corresponding emotional response groups;

[0035] The student information includes student ID and name;

[0036] Based on the emotion response group, match the corresponding preset threshold in the preset database;

[0037] The score is compared with a preset threshold. When the score is greater than 25% of the preset threshold, negative emotions are indicated.

[0038] In a preferred embodiment, the step of issuing an early warning signal and initiating a response procedure to notify relevant personnel when negative emotions occur, including user terminals, rescue personnel terminals, and student terminals, includes the following steps:

[0039] When experiencing negative emotions, assess the degree of those emotions.

[0040] The degree of negative emotion includes mild negative emotion, moderate negative emotion, and severe negative emotion;

[0041] Based on preset fuzzy logic rules, the input score is fed into a preset fuzzy rule base and mapped to a fuzzy set corresponding to the degree of negative emotion;

[0042] Based on a pre-defined deblurring method, the degree of negative sentiment is output.

[0043] The preset fuzzy logic rules include:

[0044] When the score is between 25% and 50% of the preset threshold, the score is judged to be low, indicating mild negative emotions.

[0045] When the score is between 50% and 75% of the preset threshold, the score is judged to be moderate, indicating moderate negative emotions.

[0046] When the score is greater than 75% of the preset threshold, the score is considered high, indicating severe negative emotions.

[0047] In a preferred embodiment, the step of issuing an early warning signal and initiating a response procedure to notify relevant personnel when negative emotions occur, including user terminals, rescue personnel terminals, and student terminals, further includes the following steps:

[0048] Based on the level of negative emotions and GPS data, a corresponding early warning signal is issued, which is combined with the location information.

[0049] The warning signals include yellow warning signals, orange warning signals, and red warning signals;

[0050] When mild negative emotions are detected, a yellow warning signal is issued based on location information, and the basic response procedure is initiated.

[0051] The basic response procedure includes sending relaxation exercise instructions to the student's device;

[0052] When moderate negative emotions are present, an orange alert signal combined with location information is issued, and the general response procedure is initiated.

[0053] The general response procedure includes notifying the user and sending relaxation exercise instructions to the student.

[0054] When there are severe negative emotions, the system will issue a red alert signal combined with location information and initiate an emergency response procedure.

[0055] The emergency response procedure includes notifying rescue personnel, notifying users, and sending emergency instructions to students.

[0056] In a preferred embodiment, the step of generating a health feedback report and sending it to a user terminal, where the user terminal includes a teacher terminal, an instructor terminal, and a project leader terminal, includes the following steps:

[0057] The health feedback report includes the level of negative emotions and the warning signals;

[0058] The health feedback report is encrypted based on a preset encryption algorithm;

[0059] Based on a pre-set secure communication channel, the encrypted health feedback report is sent to the authorized user terminal.

[0060] The second objective of this invention is achieved through the following technical solution:

[0061] Data collection module: Collects students' physiological data and limb movement data when students are performing dangerous tasks;

[0062] Judgment module: Based on a preset algorithm, analyze the physiological data and the body movement data to determine whether the student has negative emotions, including anxiety, tension, fear, anger, and lethargy;

[0063] Matching module: When negative emotions occur, the matching module issues an early warning signal and initiates a response procedure to notify relevant personnel, including users, rescue personnel, and students.

[0064] Sending module: Generates health feedback reports and sends them to user terminals, including teacher terminals, instructor terminals, and project team leader terminals.

[0065] The above-mentioned objective three of this application is achieved through the following technical solution:

[0066] A computer device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the aforementioned artificial intelligence dynamic stress emotion analysis method.

[0067] The fourth objective of this application is achieved through the following technical solution:

[0068] A computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the aforementioned artificial intelligence dynamic stress emotion analysis method.

[0069] In summary, this application includes at least one of the following beneficial technical effects:

[0070] By monitoring students' physiological and physical movement data in real time during hazardous activities, the system uses a pre-set algorithm to analyze and identify negative emotions. Once a negative emotion is detected, the system issues an alert and initiates a response procedure to notify relevant personnel. Simultaneously, the system generates a health feedback report and sends it to the user's device in encrypted form, ensuring information security and privacy. This process enables real-time monitoring of students' emotional changes, rapid response to prevent psychological problems and safety incidents, allows users to understand students' mental state promptly, provide appropriate support, and safeguard student well-being and a safe learning environment. Attached Figure Description

[0071] Figure 1 This is a flowchart of an implementation of an embodiment of an artificial intelligence dynamic stress and emotion analysis method according to this application;

[0072] Figure 2 This is a flowchart of step S10 in an embodiment of an artificial intelligence dynamic stress emotion analysis method of this application;

[0073] Figure 3This is a flowchart of step S20 in an embodiment of an AI-based dynamic stress and emotion analysis method of this application;

[0074] Figure 4 This is a flowchart of step S202 in an embodiment of an artificial intelligence dynamic stress emotion analysis method of this application;

[0075] Figure 5 This is another implementation flowchart of step S20 in an embodiment of an artificial intelligence dynamic stress emotion analysis method of this application;

[0076] Figure 6 This is a flowchart of step S30 in an embodiment of an artificial intelligence dynamic stress emotion analysis method of this application;

[0077] Figure 7 This is another implementation flowchart of step S30 in an embodiment of an artificial intelligence dynamic stress emotion analysis method of this application;

[0078] Figure 8 This is a flowchart of step S40 in an embodiment of an AI-based dynamic stress and emotion analysis method of this application;

[0079] Figure 9 This is a schematic block diagram of a computer device according to this application. Detailed Implementation

[0080] The following is in conjunction with the appendix Figure 1-9 This application will be described in further detail.

[0081] In one embodiment, such as Figure 1 As shown, this application discloses an artificial intelligence-based dynamic stress emotion analysis method, which specifically includes the following steps:

[0082] S10: Collect students' physiological data and limb movement data when students are performing dangerous tasks;

[0083] S20: Based on a preset algorithm, analyze the physiological data and the limb movement data to determine whether the student has negative emotions, including anxiety, tension, fear, anger, and lethargy.

[0084] S30: When negative emotions occur, a warning signal is issued and a response procedure is initiated to notify relevant personnel, including user terminals, rescue personnel terminals, and student terminals;

[0085] S40: Generate a health feedback report and send it to the user terminal, which includes the teacher terminal, the instructor terminal, and the project team leader terminal.

[0086] In this embodiment, by monitoring students' physiological and physical movement data in real time during hazardous activities, a preset algorithm is used to analyze and identify negative emotions. Once a negative emotion is detected, the system issues an alert and initiates a response procedure to notify relevant personnel. Simultaneously, the system generates a health feedback report and sends it to the user's device in encrypted form, ensuring information security and privacy. This process enables real-time monitoring of students' emotional changes, rapid response to prevent psychological problems and safety incidents, allows users to understand students' mental state in a timely manner, provides appropriate support, and safeguards student well-being and a safe learning environment.

[0087] Figure 2 Step S10 includes the following steps:

[0088] S101: Based on GPS, input the student's specific location into a preset database and match the corresponding hazardous work area;

[0089] S102: Determine the type of hazardous operation based on the hazardous work area;

[0090] S103: Input the type of hazardous operation into the preset database and collect students' physiological data and limb movement data at the corresponding preset frequency;

[0091] S104: The hazardous work types include working at height, work involving contact with hazardous substances, machinery operation, precision operation, and high-intensity physical labor.

[0092] S105: The physiological data includes heart rate data, blood pressure data, body temperature data, and skin conductivity data;

[0093] S106: The limb movement data includes movement frequency data, movement amplitude data, and movement speed data.

[0094] In this embodiment, the system uses GPS to locate students and matches their positions against a database of hazardous work areas to identify potentially dangerous environments. The system sets the data collection frequency based on the type of work, monitoring physiological and limb movement data in real time. By monitoring students' conditions during high-altitude work, exposure to hazardous substances, and other high-risk operations, the system can promptly issue warnings of abnormalities, prevent accidents, ensure student safety, and improve the efficiency of hazardous work safety management.

[0095] Figure 3 Step S20 includes the following steps:

[0096] S201: Based on an adaptive filtering method, the physiological data and the limb movement data are preprocessed, and the adaptive filtering method is combined with deep learning technology;

[0097] S202: Based on time series analysis technology, extract features from the preprocessed physiological data and limb movement data;

[0098] S203: Formula for the forward propagation process based on deep neural networks: f(x)=σ(W) T ·φ(X)+b), where f(x) is the output feature vector, W is the weight matrix, φ(X) is the input feature vector, b is the bias term, and σ is the activation function, which fuses the extracted features into a multidimensional feature vector;

[0099] S204: Based on machine learning algorithms, optimize the multidimensional feature vector and output the optimal combination of features;

[0100] S205: Based on an adaptive weight allocation mechanism and optimal combination, a preset weight is assigned to each element of the multidimensional feature vector, the weighted sum is calculated, and the score is output.

[0101] In this embodiment, before preprocessing the physiological and limb movement data, the data needs to be normalized. S201: An adaptive filtering method combined with deep learning technology is used for data preprocessing. This method can dynamically adjust the filter parameters according to the characteristics of the physiological and limb movement data to effectively remove noise and irrelevant information, improving data quality. The application of deep learning technology enables the filter to learn more complex patterns and features from the data, thereby more accurately capturing signals related to emotional states. S202: Time series analysis technology is used to extract features from the preprocessed data. This includes time series analysis of data such as heart rate, blood pressure, body temperature, skin conductivity, movement frequency, amplitude, and speed to identify the periodicity, trends, and outliers of the data. These features are crucial for understanding emotional states. S203: A deep neural network is used to fuse the extracted features into a multi-dimensional feature vector. This network can learn complex relationships between features and combine them into a high-dimensional representation, which helps improve the accuracy of subsequent emotion recognition. S204: Machine learning algorithms are used to optimize the multi-dimensional feature vector and output the optimal combination of features. By training different machine learning models, the feature combinations that contribute most to emotion prediction can be found, thereby improving model performance. S205: The adaptive weight allocation mechanism assigns a preset weight to each element of the multi-dimensional feature vector based on the optimal feature combination. This mechanism dynamically adjusts the weights according to the importance of different features in predicting emotional states, and outputs a score that reflects the student's emotional state by calculating a weighted sum.

[0102] Figure 4 Step S202 includes the following steps:

[0103] SA1: Extract heart rate features from the preprocessed heart rate data, including mean heart rate, heart rate variability, and peak heart rate.

[0104] SA2: Extract blood pressure features from the preprocessed blood pressure data, including blood pressure variability, blood pressure fluctuation amplitude, and blood pressure load;

[0105] SA3: Extract body temperature features from the preprocessed body temperature data, including average body temperature, body temperature variability, and peak body temperature.

[0106] SA4: Extract skin conductivity features from the preprocessed skin conductivity data. The skin conductivity features include skin conductivity fluctuation amplitude, skin conductivity variability, and skin conductivity peak value.

[0107] SA5: Extract motion frequency features from the preprocessed motion frequency data. The motion frequency features include the average motion frequency, motion frequency variability, and motion frequency peak value.

[0108] SA6: Extract motion amplitude features from the preprocessed motion amplitude data. The motion amplitude features include motion amplitude variability, motion amplitude fluctuation, and motion amplitude load.

[0109] SA7: Extract motion speed features from the preprocessed motion speed data. The motion speed features include motion speed variability, average motion speed, and peak motion speed.

[0110] In this embodiment, steps SA1-SA7 extract key features to quantify emotional state by analyzing preprocessed physiological and limb movement data. Features such as heart rate, blood pressure, body temperature, skin conductivity, and the frequency, amplitude, and speed of movements, including average value, variability, and peak value, can reveal dynamic changes in emotions. These features are respectively correlated with cardiac activity, vascular pressure, metabolism, sweat gland activity, and limb movements, collectively constituting a comprehensive assessment of emotional state. After feature extraction, normalization is performed to facilitate subsequent analysis and early warning, effectively improving the accuracy and response speed of monitoring individual emotional states.

[0111] Figure 5 Step S20 also includes the following steps:

[0112] SB1: Input the student's information into the preset database and match the corresponding emotional response group;

[0113] SB2: The student information includes student ID and name;

[0114] SB3: Based on the emotion response group, match the corresponding preset threshold in the preset database;

[0115] SB4: Compare the score with a preset threshold. When the score is greater than 25% of the preset threshold, there is a negative emotion.

[0116] In this embodiment, potential emotional problems are identified through steps SB1-SB4, combining student personal information and emotional scores. The system matches students' basic information with emotional response groups and extracts corresponding thresholds for score comparison. If the score exceeds 25% of the threshold, it is determined to be a negative emotion. This process effectively identifies students' emotional states, especially in the early detection of negative emotions, helping teachers quickly identify students requiring attention and provide timely support. This mechanism improves the speed of response to students' emotional problems, helps promote mental health, reduces the negative impact of emotional problems on learning and social interaction, and creates a caring and supportive environment.

[0117] Figure 6 Step S30 includes the following steps:

[0118] SB1: When experiencing negative emotions, assess the degree of those emotions;

[0119] SB2: The degree of negative emotion includes mild negative emotion, moderate negative emotion, and severe negative emotion;

[0120] SB3: Based on preset fuzzy logic rules, the input score is fed into a preset fuzzy rule base and mapped to a fuzzy set of corresponding negative emotion levels;

[0121] SB4: Outputs the degree of negative sentiment based on a preset deblurring method;

[0122] SB5: The preset fuzzy logic rules include:

[0123] SB6: When the score is 25%-50% of the preset threshold, the score is judged to be low, indicating mild negative emotion;

[0124] SB7: When the score is 50%-75% of the preset threshold, the score is judged to be moderate, indicating moderate negative emotions.

[0125] SB8: When the score is greater than 75% of the preset threshold, the score is judged to be high, indicating severe negative emotions.

[0126] In this embodiment, the system assesses the degree of negative emotions using a fuzzy logic method, categorizing them into mild, moderate, and severe. First, the system identifies negative emotions and calculates scores. Then, it applies a pre-defined fuzzy logic rule base to map the scores to corresponding fuzzy sets of emotion levels. A defuzzification step transforms the fuzzy sets into specific emotion level classifications, dividing the scores into low, moderate, and high levels to achieve accurate judgment of emotion intensity. This method effectively solves the problem of emotion quantification, improves the accuracy and practicality of emotion recognition, and enables the system to more flexibly handle the uncertainty and ambiguity of emotional states.

[0127] Figure 7 Step S30 also includes the following steps:

[0128] SD1: Based on the level of negative emotion and GPS, a corresponding early warning signal is issued, and the early warning signal is combined with the location information;

[0129] SD2: The warning signals include yellow warning signals, orange warning signals, and red warning signals;

[0130] SD3: When there is mild negative emotion, the system will issue a yellow warning signal combined with location information and initiate the basic response procedure;

[0131] SD4: The basic response procedure includes sending relaxation exercise instructions to the student's device;

[0132] SD5: When there is moderate negative emotion, the system will issue an orange alert signal combined with location information and initiate a general response procedure.

[0133] SD6: The general response procedure includes notifying the user and sending relaxation exercise instructions to the student.

[0134] SD7: When experiencing severe negative emotions, the system will issue a red alert signal based on location information and initiate an emergency response procedure.

[0135] SD8: The emergency response procedure includes notifying rescue personnel, notifying users, and sending emergency instructions to students.

[0136] In this embodiment, in principle, steps S30 (SD1-SD8) is a process that triggers warning signals and response procedures based on the degree of negative emotion. In step SD1, the system automatically matches and issues a warning signal of the corresponding color based on the previously determined degree of negative emotion (mild, moderate, severe). In step SD2, the corresponding warning signals are yellow, orange, and red. SD3, SD5, and SD7 define the warning signal colors and response procedures corresponding to different levels of emotion: a yellow warning signal corresponds to mild negative emotion and initiates the basic response procedure (SD4); an orange warning signal corresponds to moderate negative emotion and initiates the general response procedure (SD6); and a red warning signal corresponds to severe negative emotion and initiates the emergency response procedure (SD8). These response procedures include, but are not limited to, sending relaxation exercise instructions to the student, notifying the user, and contacting rescue personnel. The effect is that this process can respond quickly and specifically to the student's emotional state, ensuring that the student receives timely help when encountering emotional distress. By issuing warning signals in different colors, the system can intuitively convey the severity of emotional problems and simultaneously activate corresponding response procedures, thereby effectively providing students with necessary support. The basic response procedure helps students self-regulate when emotional problems first appear; the general response procedure enhances the user's understanding and intervention of the student's situation; and the emergency response procedure ensures that students can quickly obtain professional assistance and user attention when facing a serious emotional crisis. This system design helps improve the school's efficiency in handling emotional crises and safeguards students' mental health and safety.

[0137] Figure 8 Step S40 includes the following steps:

[0138] S401: The health feedback report includes the level of negative emotions and the warning signals;

[0139] S402: The health feedback report is encrypted based on a preset encryption algorithm;

[0140] S403: Based on a preset secure communication channel, send the encrypted health feedback report to the authorized user terminal.

[0141] In this embodiment, step S40 ensures the secure processing and transmission of students' emotional health data. In S401, the system generates a health feedback report containing the degree of negative emotions and warning signals, providing teachers with an overview of the emotional state. In S402, the report content is protected using encryption algorithms to prevent data leakage. In S403, the encrypted report is sent to authorized users through secure channels. This process not only protects student privacy but also ensures data security, enabling teachers to obtain timely information about students' emotional states while protecting their privacy, allowing them to take appropriate measures to improve service quality and promote students' mental health.

[0142] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0143] In one embodiment, an AI-based dynamic stress and emotion analysis method is provided, and this AI-based dynamic stress and emotion analysis system corresponds to the AI-based dynamic stress and emotion analysis method described in the above embodiment. This AI-based dynamic stress and emotion analysis method includes:

[0144] Data collection module: Collects students' physiological data and limb movement data when students are performing dangerous tasks;

[0145] Judgment module: Based on a preset algorithm, analyze the physiological data and the body movement data to determine whether the student has negative emotions, including anxiety, tension, fear, anger, and lethargy;

[0146] Matching module: When negative emotions occur, the matching module issues an early warning signal and initiates a response procedure to notify relevant personnel, including users, rescue personnel, and students.

[0147] Sending module: Generates health feedback reports and sends them to user terminals, including teacher terminals, instructor terminals, and project team leader terminals.

[0148] Optional, also includes:

[0149] First matching module: Based on GPS, the student's specific location is input into a preset database to match the corresponding hazardous work area;

[0150] First judgment module: Based on the hazardous work area, determine the type of hazardous work;

[0151] The second matching module: inputs the type of hazardous operation into the preset database, and matches the corresponding preset frequency to collect students' physiological data and limb movement data;

[0152] The first module includes: the hazardous work types include working at height, work involving contact with hazardous substances, machinery operation, precision operation, and high-intensity physical labor.

[0153] The second module includes: the physiological data includes heart rate data, blood pressure data, body temperature data, and skin conductivity data;

[0154] The third module includes: the limb movement data includes movement frequency data, movement amplitude data, and movement speed data.

[0155] Optional, also includes:

[0156] Adaptive module: Based on an adaptive filtering method, preprocesses the physiological data and the limb movement data, wherein the adaptive filtering method is combined with deep learning technology;

[0157] Feature extraction module: Based on time series analysis technology, features are extracted from the preprocessed physiological data and limb movement data;

[0158] Fusion Vector Module: Based on the forward propagation process formula of deep neural networks, f(x) = σ(W) T ·φ(X)+b), where f(x) is the output feature vector, W is the weight matrix, φ(X) is the input feature vector, b is the bias term, and σ is the activation function, which fuses the extracted features into a multidimensional feature vector;

[0159] Optimal combination module: Based on machine learning algorithms, it optimizes the multidimensional feature vectors and outputs the optimal combination of features;

[0160] Preset weight module: Based on the adaptive weight allocation mechanism and optimal combination, it assigns a preset weight to each element of the multidimensional feature vector, calculates the weighted sum, and outputs the score.

[0161] Optional, also includes:

[0162] First extraction module: Extracts heart rate features from preprocessed heart rate data, including average heart rate, heart rate variability, and peak heart rate;

[0163] The second extraction module extracts blood pressure features from the preprocessed blood pressure data, including blood pressure variability, blood pressure fluctuation amplitude, and blood pressure load.

[0164] The third extraction module extracts body temperature features from the preprocessed body temperature data, including average body temperature, body temperature variability, and peak body temperature.

[0165] The fourth extraction module extracts skin conductivity features from the preprocessed skin conductivity data. These features include skin conductivity fluctuation amplitude, skin conductivity variability, and skin conductivity peak value.

[0166] The fifth extraction module extracts motion frequency features from the preprocessed motion frequency data. The motion frequency features include the average motion frequency, motion frequency variability, and motion frequency peak value.

[0167] The sixth extraction module extracts motion amplitude features from the preprocessed motion amplitude data. The motion amplitude features include motion amplitude variability, motion amplitude fluctuation, and motion amplitude load.

[0168] The seventh extraction module extracts motion speed features from the preprocessed motion speed data. The motion speed features include motion speed variability, average motion speed, and peak motion speed.

[0169] Optional, also includes:

[0170] The third matching module: Input the student's information into the preset database and match the corresponding emotional response group;

[0171] The fourth module includes: the student information includes student ID and name;

[0172] The fourth matching module: Based on the emotion response group, it matches the corresponding preset threshold in the preset database;

[0173] First judgment module: compare the score with a preset threshold. When the score is greater than 25% of the preset threshold, there is a negative emotion.

[0174] Optional, also includes:

[0175] Second judgment module: When negative emotions exist, judge the degree of negative emotions;

[0176] The fifth module includes: the degree of negative emotion includes mild negative emotion, moderate negative emotion, and severe negative emotion;

[0177] Mapping module: Based on preset fuzzy logic rules, input scores are mapped to a preset fuzzy rule base and then to a fuzzy set of corresponding negative emotion levels.

[0178] First output module: Based on a preset deblurring method, outputs the degree of negative sentiment;

[0179] The sixth module includes: the preset fuzzy logic rules include:

[0180] The third judgment module: When the score is between 25% and 50% of the preset threshold, the score is judged to be low, indicating mild negative emotions;

[0181] The fourth judgment module: When the score is between 50% and 75% of the preset threshold, the score is judged as "medium", which indicates moderate negative emotion.

[0182] The fifth judgment module: When the score is greater than 75% of the preset threshold, the score is judged as high, indicating severe negative emotions.

[0183] Optional, also includes:

[0184] The fifth matching module: Based on the level of negative emotions and GPS, it matches and issues corresponding warning signals, which are combined with location information;

[0185] The seventh module includes: the warning signals include yellow warning signals, orange warning signals, and red warning signals;

[0186] The sixth matching module: When there is mild negative emotion, the matching module issues a yellow warning signal combined with location information and initiates the basic response procedure;

[0187] The eighth module includes: the basic response procedure includes sending relaxation exercise instructions to the student's device;

[0188] The seventh matching module: When there is moderate negative emotion, the matching module issues an orange warning signal combined with location information and initiates the general response procedure;

[0189] The ninth module includes: the general response procedure includes notifying the user terminal and sending relaxation exercise instructions to the student terminal;

[0190] The eighth matching module: When there is severe negative emotion, the matching module issues a red warning signal combined with location information and initiates an emergency response procedure;

[0191] The tenth module includes: the emergency response procedure includes notifying rescue personnel, notifying users, and sending emergency instructions to students.

[0192] Optional, also includes:

[0193] The health feedback report includes the level of negative emotions and the warning signals.

[0194] Encryption processing module: encrypts the health feedback report based on a preset encryption algorithm;

[0195] The first sending module sends the encrypted health feedback report to the authorized user terminal based on a preset secure communication channel.

[0196] For specific limitations regarding an AI-based dynamic stress and emotion analysis system, please refer to the limitations of an AI-based dynamic stress and emotion analysis method described above, which will not be repeated here. Each module in the aforementioned AI-based dynamic stress and emotion analysis system can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the corresponding operations of each module.

[0197] In one embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 9As shown, the computer device includes a processor, memory, network interface, and database connected via a system bus. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The database stores health feedback reports. The network interface communicates with external terminals via a network connection. When the computer program is executed by the processor, it implements an artificial intelligence-based dynamic stress and emotion analysis method.

[0198] In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement an artificial intelligence dynamic stress emotion analysis method.

[0199] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, the computer program being executed by a processor as an artificial intelligence dynamic stress emotion analysis method.

[0200] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAM bus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0201] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is used as an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

Claims

1. An artificial intelligence-based dynamic stress and emotion analysis method, characterized in that, Including the following steps: Collect students' physiological and physical movement data while they are performing dangerous tasks; Based on GPS, the student's specific location is input into a preset database to match the corresponding hazardous work area; Based on the hazardous work area, determine the type of hazardous work; Input the type of hazardous work into the preset database, and collect students' physiological data and limb movement data at the corresponding preset frequency; The hazardous work types include working at heights, working with hazardous substances, operating machinery, precision operations, and physically demanding work. The physiological data includes heart rate data, blood pressure data, body temperature data, and skin conductivity data; The limb movement data includes movement frequency data, movement amplitude data, and movement speed data; Based on a preset algorithm, the physiological data and the body movement data are analyzed to determine whether the student has negative emotions, including anxiety, tension, fear, anger, and lethargy. Based on an adaptive filtering method, the physiological data and the limb movement data are preprocessed, and the adaptive filtering method is combined with deep learning technology. Based on time series analysis technology, features are extracted from the preprocessed physiological data and limb movement data; Formula for forward propagation process based on deep neural network , It outputs feature vectors, It is a weight matrix. It is the input feature vector, It is a bias term. It is an activation function that fuses the extracted features into a multidimensional feature vector; Based on machine learning algorithms, the multidimensional feature vectors are optimized to output the optimal combination of features; Based on an adaptive weight allocation mechanism and optimal combination, a preset weight is assigned to each element of the multidimensional feature vector, the weighted sum is calculated, and the score is output. Input students' information into a preset database and match them with corresponding emotional response groups; The student information includes student ID and name; Based on the emotion response group, match the corresponding preset threshold in the preset database; The score is compared with a preset threshold. When the score is greater than 25% of the preset threshold, negative emotions are indicated. When negative emotions arise, a warning signal is issued and a response procedure is initiated to notify relevant personnel, including users, rescue personnel, and students. When experiencing negative emotions, assess the degree of those emotions. The degree of negative emotion includes mild negative emotion, moderate negative emotion, and severe negative emotion; Based on preset fuzzy logic rules, the input score is fed into a preset fuzzy rule base and mapped to a fuzzy set corresponding to the degree of negative emotion; Based on a pre-defined deblurring method, the degree of negative sentiment is output. The preset fuzzy logic rules include: When the score is between 25% and 50% of the preset threshold, the score is judged to be low, indicating mild negative emotions. When the score is between 50% and 75% of the preset threshold, the score is judged to be moderate, indicating moderate negative emotions. When the score is greater than 75% of the preset threshold, the score is judged to be high, indicating severe negative emotions. A health feedback report is generated and sent to the user terminals, including teacher terminals, instructor terminals, and project team leader terminals.

2. The artificial intelligence dynamic stress emotion analysis method according to claim 1, characterized in that, The step of extracting features from the preprocessed physiological data and limb movement data based on time series analysis technology includes the following steps: Heart rate features are extracted from the preprocessed heart rate data, including mean heart rate, heart rate variability, and peak heart rate. Blood pressure features are extracted from the preprocessed blood pressure data, including blood pressure variability, blood pressure fluctuation amplitude, and blood pressure load. Body temperature features are extracted from the preprocessed body temperature data, including average body temperature, body temperature variability, and peak body temperature. Skin conductivity features are extracted from the preprocessed skin conductivity data. These features include skin conductivity fluctuation amplitude, skin conductivity variability, and skin conductivity peak value. Action frequency features are extracted from the preprocessed action frequency data. These features include the average action frequency, action frequency variability, and action frequency peak value. Motion amplitude features are extracted from the preprocessed motion amplitude data. These features include motion amplitude variability, motion amplitude fluctuation, and motion amplitude load. Action speed features are extracted from the preprocessed action speed data. These features include action speed variability, average action speed, and peak action speed.

3. The artificial intelligence dynamic stress emotion analysis method according to claim 1, characterized in that, The step of matching and issuing an early warning signal when negative emotions occur, and initiating a response procedure to notify relevant personnel, including user terminals, rescue personnel terminals, and student terminals, also includes the following steps: Based on the level of negative emotions and GPS data, a corresponding early warning signal is issued, which is combined with the location information. The warning signals include yellow warning signals, orange warning signals, and red warning signals; When mild negative emotions are detected, a yellow warning signal is issued based on location information, and the basic response procedure is initiated. The basic response procedure includes sending relaxation exercise instructions to the student's device; When moderate negative emotions are present, an orange alert signal combined with location information is issued, and the general response procedure is initiated. The general response procedure includes notifying the user and sending relaxation exercise instructions to the student. When there are severe negative emotions, the system will issue a red alert signal combined with location information and initiate an emergency response procedure. The emergency response procedure includes notifying rescue personnel, notifying users, and sending emergency instructions to students.

4. The artificial intelligence dynamic stress emotion analysis method according to claim 1, characterized in that, The step of generating a health feedback report and sending it to the user terminal, which includes a teacher terminal, an instructor terminal, and a project team leader terminal, includes the following steps: The health feedback report includes the level of negative emotions and the warning signals; The health feedback report is encrypted based on a preset encryption algorithm; Based on a pre-set secure communication channel, the encrypted health feedback report is sent to the authorized user terminal.

5. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the computer program, implements the steps of the artificial intelligence dynamic stress emotion analysis method as described in any one of claims 1 to 4.

6. A computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the artificial intelligence dynamic stress emotion analysis method as described in any one of claims 1 to 4.

Images