Emotional pressure relieving and monitoring method combined with virtual reality technology

Abstract

The invention discloses an emotional pressure relieving and monitoring method combined with a virtual reality technology, and the method comprises the following steps: S1, providing a virtual environment for a user through a virtual reality device, and enabling the user to interact with the environment through wearing a virtual reality head-mounted device and using a controller; s2, collecting physiological data of the user in real time through a biosensor, wherein the physiological data comprises heart rate, brain wave, electrodermal response and oxyhemoglobin saturation; s3, inputting the physiological data in the S2 into a long-short-term memory network model for processing, and predicting emotional changes; s4, adjusting the virtual reality environment in real time to adapt to the current emotional state of the user based on the emotional state analysis and prediction result of the long-short-term memory network model; and S5, repeating the steps S2 to S4 to monitor the emotion of the user. According to the method, solid technical support is provided for emotion pressure management and monitoring based on the virtual reality technology, and a traditional static emotion regulation method can be effectively replaced.

Description

Technical Field

[0001] The present invention relates to the technical field of virtual reality devices, and in particular to an emotional stress relief and monitoring method combined with virtual reality technology. Background Art

[0002] With the rapid development of modern society, emotional stress has become one of the important factors affecting people's quality of life and health. Traditional methods of emotion regulation mainly rely on psychological counseling, drug therapy or static meditation. However, these methods often have problems such as insufficient personalization, insufficient real-time feedback and delayed effects, and cannot fully meet users' emotional management needs in different situations.

[0003] In recent years, the development of virtual reality (VR) technology has provided new possibilities for emotion regulation. Through an immersive virtual environment, users can temporarily escape from the real world and enter a virtual space that integrates multiple senses such as vision, hearing, and touch, thereby effectively relaxing and relieving stress. However, the existing virtual reality emotion regulation system still has the following technical bottlenecks: (1) There is a lack of real-time monitoring and analysis of the user's emotional state, and the virtual environment cannot be dynamically adjusted according to the user's emotional changes; (2) The design of the virtual environment is relatively fixed and cannot be adjusted according to the user's personalized needs; (3) Existing emotion regulation systems mostly rely on simple user self-reports or external observations, and cannot accurately predict and intervene in emotional changes.

[0004] Therefore, how to develop an emotional stress relief and monitoring system that can combine virtual reality technology, real-time emotion monitoring and dynamic feedback has become an urgent problem to be solved by those skilled in the art. The system should be able to accurately identify and predict the user's emotional state through real-time collection and analysis of physiological data, and dynamically adjust the virtual reality environment according to the prediction results, thereby achieving personalized, real-time emotion management and stress relief. Summary of the invention

[0005] The present invention provides a solid technical support for emotional stress management and monitoring based on virtual reality technology, has significant advantages in personalization, real-time and effectiveness, and can effectively replace traditional static emotion regulation methods.

[0006] According to an embodiment of the present invention, a method for alleviating and monitoring emotional stress in combination with virtual reality technology comprises the following steps:

[0007] S1. Provide users with a virtual environment through virtual reality devices. The virtual environment is built on the Unity platform. Users interact with the environment by wearing virtual reality headsets and using controllers.

[0008] S2, collect the user's physiological data in real time through biosensors, including heart rate, brain waves, skin galvanic response and blood oxygen saturation;

[0009] S3, inputting the physiological data of S2 into the long short-term memory network model for processing, analyzing the user's emotional state, and predicting his emotional changes;

[0010] S4. Based on the emotional state analysis and prediction results of the long short-term memory network model, the visual, sound and interactive elements of the virtual reality environment are adjusted in real time, including adjusting the ambient light, sound effects and the interactivity of virtual objects to adapt to the user's current emotional state;

[0011] S5. Repeat steps S2 to S4 to achieve dynamic monitoring and relief of the user's emotional stress.

[0012] Optionally, S1 includes the following steps:

[0013] S11. Use the Unity platform to build a virtual reality environment to meet users' emotional regulation needs by adjusting visual elements, sound effects, interactive elements and scene changes;

[0014] S12. Through virtual reality headsets, high-definition display, wide viewing angle and low-latency visual effects are provided, and audio output devices are provided to achieve immersive sound experience;

[0015] S13. Use a handheld controller to interact, and support users to navigate, manipulate objects, and switch scenes in a virtual environment through an interactive device, and have a tactile feedback function to enhance the user experience;

[0016] S14, when the system is initialized, the virtual scene preset by the user is loaded, and the visual elements of the scene including color, brightness and sound effects including background music and natural sound effects are adjusted in real time according to the emotion adjustment mode selected by the user;

[0017] S15. Define virtual scene parameters P s is a set of control variables, including light intensity L, hue C, volume V, and interactive element response time T r , the system dynamically adjusts P according to the user's emotional state and real-time feedback s =(L, C, V, T r ) value to optimize the user experience.

[0018] Optionally, the S2 specifically includes:

[0019] S21. The user wears a biosensor device, including a heart rate monitor, which measures the user's heart rate data HR and heart rate variability HRV in real time through photoplethysmography, an EEG headband, which collects the user's brain wave signals through electrodes and records the brain's electrical activity data EEG(t), where t represents the collection time, a galvanic skin response sensor, which detects changes in the conductivity of the user's skin and generates galvanic skin response data GSR(t) to evaluate the user's stress level, and a pulse oximeter, which measures the user's blood oxygen saturation SpO2 and respiratory rate RR through an optical sensor;

[0020] S22, the biosensor device transmits the physiological data D(t) collected in real time to the data processing unit through the wireless communication module, where D(t) = {HR, HRV, EEG(t), GSR(t), SpO2, RR};

[0021] S23, the data processing unit pre-processes the physiological data, including signal denoising and normalization, and synchronizes the multi-channel data to generate a unified time series data D s (t);

[0022] S24, the preprocessed time series data D s (t) is passed to the long short-term memory network model as input data for emotional state analysis and prediction.

[0023] Optionally, the S3 specifically includes:

[0024] S31, the preprocessed time series data D s (t) input into the long short-term memory network model to capture the time-dependent patterns related to emotional changes in the user's physiological data;

[0025] S32. The input of the long short-term memory network model is the physiological data vector D at the current time step t. s (t), physiological data vector D s (t) contains the user's heart rate, brain waves, and skin electrical response physiological data at time t. The model processes the relationship between the data of the current time step and the data of the previous time step through the input gate, forget gate, and output gate of the memory unit to generate a hidden state vector h t :

[0026] h t =σ(W i ·[D s (t), h t-1 ]+b h );

[0027] Among them, h trepresents the hidden state vector generated by the LSTM network model at time step t, and represents the user emotion information currently memorized by the LSTM network. s (t) represents the set of user physiological data collected at time t, h t-1 represents the hidden state vector of the LSTM network model at the previous time step t-1, which carries the memory of the LSTM network about the previous emotional state. σ is the activation function, which is used to convert the calculation result into a value suitable for the next step of processing.

[0028] S33, through the multi-layer physiological data vector D s (t) Iterative calculation of the network to generate the final emotional state prediction vector Y t , which represents the system's prediction of the user's emotional state at time t:

[0029] Y t =softmax(W y ·h t +b y );

[0030] Among them, Y t It represents the predicted value of the emotional state at time t, which is the system's estimation result of the user's current emotion. Softmax is a conversion function used to convert the output of the long short-term memory network into the probability distribution of different emotional states and determine the user's main emotional category;

[0031] S34. Predicting vector Y based on emotional state t , identifies the user’s emotional state category and dynamically adjusts elements of the virtual reality environment based on the state category.

[0032] Optionally, the S34 specifically includes:

[0033] The emotional state prediction vector Y t Sort the probability values ​​of each emotion category in the , and determine the user's current main emotional state category C t :

[0034] C t = argmax(Y t );

[0035] Among them, C t It represents the main emotional state category of the user predicted by the system at time t, including anxiety, relaxation, and stress;

[0036] S342: According to the user's main emotional state category C t , select the corresponding regulation strategy S(C t), where each adjustment strategy contains a set of adjustment rules for virtual reality environment parameters:

[0037] S(C t )={P s :L(C t ), C(C t ), V(C t ), T r (C t )};

[0038] Among them, L(C t ) indicates emotional state C t The matching ambient light intensity, C(C t ) indicates emotional state C t Matching scene color, V(C t ) indicates emotional state C t Matching sound effects, T r (C t ) indicates emotional state C t Matching virtual object interaction response time;

[0039] S343, real-time monitoring of adjusted user physiological data D s (t+1), input it into the long short-term memory network model to generate a new emotional state prediction vector Y t+1 ;

[0040] S344, according to Y t+1 The results are used to determine the changing trend of the user's emotional state;

[0041] S345, when Y t+1 If it is shown that the user's emotions have stabilized, the system will gradually restore the virtual environment to the initial settings, ending the current emotion regulation process.

[0042] Optionally, the S4 specifically includes:

[0043] S41, emotional state prediction vector Y generated based on long short-term memory network model t , determine the user's current main emotional state category C t ,and guide the dynamic adjustment of the VR environment through categories;

[0044] S42, according to emotional state category C t , adjust the light intensity L of the virtual reality environment:

[0045] L=L0×f(C t );

[0046] Among them, L0 is the initial light intensity of the virtual environment, f(C t) is based on the emotional state category C t The lighting adjustment function is used to increase or decrease the lighting intensity;

[0047] S43. Adjust the color tone of the virtual reality environment C:

[0048] C=C0+ΔC(C t );

[0049] Among them, C0 is the initial color parameter of the virtual environment, ΔC(C t ) is based on the emotional state category C t The color adjustment value, ΔC(C t ) determines the warmth or lightness of the hue;

[0050] S44. Adjust the sound effects of the virtual reality environment V:

[0051] V=V0×g(C t );

[0052] Among them, V0 is the initial sound volume of the virtual environment, g(C t ) is based on the emotional state category C t The sound adjustment function is used to increase or decrease the volume of background music or environmental sound effects;

[0053] S45, adjusting the interactive response time T of the virtual object in the virtual reality environment r :

[0054] T r =T r0 ×h(C t );

[0055] Among them, T r0 is the initial response time of the virtual object in the virtual environment, h(C t ) is based on the emotional state category C t A response time adjustment function is used to speed up or delay the interactive response speed between the user and the virtual object;

[0056] S46. Real-time collection of the user's physiological data in the adjusted virtual environment D s (t+1), and input it into the long short-term memory network model, perform emotional state analysis and prediction again, and obtain a new emotional state prediction vector Y t+1 .

[0057] Optionally, the S42 specifically includes:

[0058] S42, based on emotional state category C t Adjusting the light intensity L of the virtual reality environment specifically includes the following steps:

[0059] S421. Predicting vector Y based on emotional state t The probability distribution of different emotional states in the equation determines the main emotional state category C t , and calculate the lighting adjustment weight w t :

[0060]

[0061] Among them, p i Indicates emotional state category C i The probability value represents the possibility that the user is currently in this emotional state, k i Representation and emotional state category C i The associated illumination adjustment coefficient is used to characterize the adjustment direction and amplitude of illumination intensity under the emotional state, where n is the total number of all possible emotional states;

[0062] S422, adjust weight w according to illumination t , calculate the light intensity adjustment function f(C t ) value, which is used to determine the specific intensity adjustment of the light in the virtual reality environment:

[0063] f(C t )=1+α·w t ;

[0064] Among them, α is the sensitivity parameter of light intensity adjustment, which determines the responsiveness of light intensity to changes in emotional state;

[0065] w t To adjust the light weight calculated based on the user's current emotional state, affecting the increase or decrease of light intensity;

[0066] S424, real-time collection and analysis of the user's physiological data under the adjusted lighting environment D s (t+1), input it into the long short-term memory network model to generate a new emotional state prediction vector Y t+1 ;

[0067] S425, predicting vector Y based on the new emotional state t+1 , calculate the lighting adjustment weight w again t+1 And the light intensity adjustment function f(C t+1 ), and the light intensity L t+1 Dynamically adjust to ensure the VR environment always matches the user's emotional state.

[0068] Optionally, the S5 specifically includes:

[0069] S51. While the user is experiencing the virtual reality environment, real-time collection of physiological data Ds (t), including heart rate HR, EEG(t), galvanic skin response GSR(t) and blood oxygen saturation SpO2;

[0070] S52, the collected physiological data D s (t) Input into the long short-term memory network model for emotional state analysis and prediction, generating the emotional state prediction vector Y t and adjusting various parameters of the virtual reality environment in real time through step S4;

[0071] S53, after each time step t, the system compares the adjusted virtual reality environment parameters with the user's new physiological data D s (t+1) and input it into the long short-term memory network model for updating to generate a new emotional state prediction vector Y t+1 ;

[0072] S54. Define the emotion change trend vector ΔY t for:

[0073] ΔY t =Y t+1 -Y t ;

[0074] Among them, ΔY t represents the change in the user’s emotional state from time step t to time step t+1, reflecting the impact of the adjusted virtual reality environment on the user’s emotions. t+1 and Y t Represent the emotional state prediction vectors at time step t+1 and time step t respectively;

[0075] S55, according to the emotion change trend vector ΔY t Based on the results, we can determine the direction of change of the user’s emotional state and continuously optimize the model.

[0076] The beneficial effects of the present invention are:

[0077] The present invention makes full use of virtual reality technology, biosensor technology and long short-term memory network (LSTM) model, and describes in detail a method for relieving and monitoring emotional stress with real-time emotional monitoring and dynamic feedback. By combining the user's physiological data analysis with the dynamic adjustment of the virtual reality environment, the present invention has the following significant advantages:

[0078] 1. Personalized emotion regulation: The system collects the user's heart rate, brain waves, skin electrical response and other physiological data in real time, and combines the LSTM model to accurately analyze and predict the user's emotional state. It can dynamically adjust the virtual reality environment according to the individual differences of each user and provide a highly personalized emotion management solution.

[0079] 2. Real-time feedback and dynamic adjustment: The present invention realizes the real-time linkage between emotional state and virtual reality environment. It can make real-time adjustments to the lighting, color tone, sound effects and interactive elements in the environment according to the instant changes in the user's emotional state, ensuring that the virtual environment always matches the user's current emotional state, thereby achieving better emotion regulation effect.

[0080] 3. Accurate prediction and intervention: Through the time series analysis of physiological data by LSTM model, the present invention can identify the user's emotional fluctuations in advance and take corresponding adjustment measures in time to prevent further aggravation of emotional pressure, providing users with accurate and effective emotional management and intervention means.

[0081] 4. Improved user experience: The present invention significantly improves the user's immersion and participation in the emotion regulation process through a highly interactive and adaptive virtual environment design, increases user satisfaction and the effect of emotion management, and provides a safe, convenient and efficient solution for relieving emotional stress.

[0082] In summary, the present invention provides solid technical support for emotional stress management and monitoring based on virtual reality technology, has significant advantages in personalization, real-time and effectiveness, can effectively replace traditional static emotion regulation methods, and meet the growing demand for emotion management in modern society. BRIEF DESCRIPTION OF THE DRAWINGS

[0083] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0084] Figure 1 This is an overall flow chart of an emotional stress relief and monitoring system based on virtual reality technology proposed by the present invention;

[0085] Figure 2 This is a schematic diagram of real-time data collection and LSTM model processing of the biosensor in the present invention, showing how the biosensor collects the user's heart rate, brain waves, skin electrical response and other physiological data, and analyzes and predicts the emotional state through the LSTM model. DETAILED DESCRIPTION

[0086] The present invention will now be described in further detail with reference to the accompanying drawings. These drawings are simplified schematic diagrams, which only illustrate the basic structure of the present invention in a schematic manner, and therefore only show the components related to the present invention.

[0087] refer to Figure 1 , a method for relieving and monitoring emotional stress combined with virtual reality technology, comprising the following steps:

[0088] S1. Provide users with a virtual environment through virtual reality devices. The virtual environment is built on the Unity platform. Users interact with the environment by wearing virtual reality headsets and using controllers.

[0089] In this implementation, S1 includes the following steps:

[0090] S11. Use the Unity platform to build a virtual reality environment to meet users' emotional regulation needs by adjusting visual elements, sound effects, interactive elements and scene changes;

[0091] S12. Through virtual reality headsets, high-definition display, wide viewing angle and low-latency visual effects are provided, and audio output devices are provided to achieve immersive sound experience;

[0092] S13. Use a handheld controller to interact, and support users to navigate, manipulate objects, and switch scenes in a virtual environment through an interactive device, and have a tactile feedback function to enhance the user experience;

[0093] S14, when the system is initialized, the virtual scene preset by the user is loaded, and the visual elements of the scene including color, brightness and sound effects including background music and natural sound effects are adjusted in real time according to the emotion adjustment mode selected by the user;

[0094] S15. Define virtual scene parameters P s is a set of control variables, including light intensity L, hue C, volume V, and interactive element response time T r , the system dynamically adjusts P according to the user's emotional state and real-time feedback s =(L, C, V, T r ) value to optimize the user experience.

[0095] S2, collect the user's physiological data in real time through biosensors, including heart rate, brain waves, skin galvanic response and blood oxygen saturation;

[0096] In this implementation, S2 specifically includes:

[0097] S21. The user wears a biosensor device, including a heart rate monitor, which measures the user's heart rate data HR and heart rate variability HRV in real time through photoplethysmography, an EEG headband, which collects the user's brain wave signals through electrodes and records the brain's electrical activity data EEG(t), where t represents the collection time, a galvanic skin response sensor, which detects changes in the conductivity of the user's skin and generates galvanic skin response data GSR(t) to evaluate the user's stress level, and a pulse oximeter, which measures the user's blood oxygen saturation SpO2 and respiratory rate RR through an optical sensor;

[0098] S22, the biosensor device transmits the physiological data D(t) collected in real time to the data processing unit through the wireless communication module, where D(t) = {HR, HRV, EEG(t), GSR(t), SpO2, RR};

[0099] S23, the data processing unit pre-processes the physiological data, including signal denoising and normalization, and synchronizes the multi-channel data to generate a unified time series data D s (t);

[0100] S24, the preprocessed time series data D s (t) is passed to the long short-term memory network model as input data for emotional state analysis and prediction.

[0101] S3, inputting the physiological data of S2 into the long short-term memory network model for processing, analyzing the user's emotional state, and predicting his emotional changes;

[0102] In this implementation, S3 specifically includes:

[0103] S31, the preprocessed time series data D s (t) input into the long short-term memory network model to capture the time-dependent patterns related to emotional changes in the user's physiological data;

[0104] S32. The input of the long short-term memory network model is the physiological data vector D at the current time step t. s (t), physiological data vector D s (t) contains the user's heart rate, brain waves, and skin electrical response physiological data at time t. The model processes the relationship between the data of the current time step and the data of the previous time step through the input gate, forget gate, and output gate of the memory unit to generate a hidden state vector h t :

[0105] h t =σ(W i ·[D s (t), h t-1 ]+b h );

[0106] Among them, h t represents the hidden state vector generated by the LSTM network model at time step t, and represents the user emotion information currently memorized by the LSTM network. s (t) represents the set of user physiological data collected at time t, h t-1represents the hidden state vector of the LSTM network model at the previous time step t-1, which carries the memory of the LSTM network about the previous emotional state. σ is the activation function, which is used to convert the calculation result into a value suitable for the next step of processing.

[0107] S33, through the multi-layer physiological data vector D s (t) Iterative calculation of the network to generate the final emotional state prediction vector Y t , which represents the system's prediction of the user's emotional state at time t:

[0108] Y t =softmax(W y ·h t +b y );

[0109] Among them, Y t It represents the predicted value of the emotional state at time t, which is the system's estimation result of the user's current emotion. Softmax is a conversion function used to convert the output of the long short-term memory network into the probability distribution of different emotional states and determine the user's main emotional category;

[0110] S34. Predicting vector Y based on emotional state t , identifies the user’s emotional state category and dynamically adjusts elements of the virtual reality environment based on the state category.

[0111] In this implementation, S34 specifically includes:

[0112] The emotional state prediction vector Y t Sort the probability values ​​of each emotion category in the table to determine the user's current main emotional state category C t :

[0113] C t = argmax(Y t );

[0114] Among them, C t It represents the main emotional state category of the user predicted by the system at time t, including anxiety, relaxation, and stress;

[0115] S342: According to the user's main emotional state category C t , select the corresponding regulation strategy S(C t ), where each adjustment strategy contains a set of adjustment rules for virtual reality environment parameters:

[0116] S(C t )={P s :L(C t ), C(Ct ), V(C t ), T r (C t )};

[0117] Among them, L(C t ) indicates emotional state C t The matching ambient light intensity, C(C t ) indicates emotional state C t Matching scene color, V(C t ) indicates emotional state C t Matching sound effects, T r (C t ) indicates emotional state C t Matching virtual object interaction response time;

[0118] S343, real-time monitoring of adjusted user physiological data D s (t+1), input it into the long short-term memory network model to generate a new emotional state prediction vector Y t+1 ;

[0119] S344, according to Y t+1 The results are used to determine the changing trend of the user's emotional state;

[0120] S345, when Y t+1 If it is shown that the user's emotions have stabilized, the system will gradually restore the virtual environment to the initial settings, ending the current emotion regulation process.

[0121] S4. Based on the emotional state analysis and prediction results of the long short-term memory network model, the visual, sound and interactive elements of the virtual reality environment are adjusted in real time, including adjusting the ambient light, sound effects and the interactivity of virtual objects to adapt to the user's current emotional state;

[0122] In this implementation, S4 specifically includes:

[0123] S41, emotional state prediction vector Y generated based on long short-term memory network model t , determine the user's current main emotional state category C t ,and guide the dynamic adjustment of the VR environment through categories;

[0124] S42, according to emotional state category C t , adjust the light intensity L of the virtual reality environment:

[0125] L=L0×f(C t );

[0126] Among them, L0 is the initial light intensity of the virtual environment, f(C t) is based on the emotional state category C t The lighting adjustment function is used to increase or decrease the lighting intensity;

[0127] S43. Adjust the color tone of the virtual reality environment C:

[0128] C=C0+ΔC(C t );

[0129] Among them, C0 is the initial color parameter of the virtual environment, ΔC(C t ) is based on the emotional state category C t The color adjustment value, ΔC(C t ) determines the warmth or lightness of the hue;

[0130] S44. Adjust the sound effects of the virtual reality environment:

[0131] V=V0×g(C t );

[0132] Among them, V0 is the initial sound volume of the virtual environment, g(C t ) is based on the emotional state category C t The sound adjustment function is used to increase or decrease the volume of background music or environmental sound effects;

[0133] S45, adjusting the interactive response time T of the virtual object in the virtual reality environment r :

[0134] T r =T r0 ×h(C t );

[0135] Among them, T r0 is the initial response time of the virtual object in the virtual environment, h(C t ) is based on the emotional state category C t A response time adjustment function is used to speed up or delay the interactive response speed between the user and the virtual object;

[0136] S46. Real-time collection of the user's physiological data in the adjusted virtual environment D s (t+1), and input it into the long short-term memory network model, perform emotional state analysis and prediction again, and obtain a new emotional state prediction vector Y t+1 .

[0137] S42, according to emotional state category C t , adjust the light intensity L of the virtual reality environment:

[0138] L=L0×f(C t );

[0139] Among them, L0 is the initial light intensity of the virtual environment, f(C t ) is based on the emotional state category C t The lighting adjustment function is used to increase or decrease the lighting intensity;

[0140] In this implementation, S42 specifically includes:

[0141] S42, based on emotional state category C t Adjusting the light intensity L of the virtual reality environment specifically includes the following steps:

[0142] S421, predicting vector Y based on emotional state t The probability distribution of different emotional states in the equation determines the main emotional state category C t , and calculate the lighting adjustment weight w t :

[0143]

[0144] Among them, p i Indicates emotional state category C i The probability value represents the possibility that the user is currently in this emotional state, k i Representation and emotional state category C i The associated illumination adjustment coefficient is used to characterize the adjustment direction and amplitude of illumination intensity under the emotional state, where n is the total number of all possible emotional states;

[0145] S422, adjust weight w according to illumination t , calculate the light intensity adjustment function f(C t ) value, which is used to determine the specific intensity adjustment of the light in the virtual reality environment:

[0146] f(C t )=1+α·w t ;

[0147] Among them, α is the sensitivity parameter of light intensity adjustment, which determines the responsiveness of light intensity to changes in emotional state;

[0148] w t To adjust the light weight calculated based on the user's current emotional state, affecting the increase or decrease of light intensity;

[0149] S424, real-time collection and analysis of the user's physiological data under the adjusted lighting environment D s (t+1), input it into the long short-term memory network model to generate a new emotional state prediction vector Y t+1 ;

[0150] S425, predicting vector Y based on the new emotional state t+1 , calculate the lighting adjustment weight w again t+1 And the light intensity adjustment function f(C t+1 ), and the light intensity L t+1 Dynamically adjust to ensure the VR environment always matches the user's emotional state.

[0151] S5. Repeat steps S2 to S4 to achieve dynamic monitoring and relief of the user's emotional stress.

[0152] In this implementation, S5 specifically includes:

[0153] S51. While the user is experiencing the virtual reality environment, real-time collection of physiological data D s (t), including heart rate HR, EEG(t), galvanic skin response GSR(t) and blood oxygen saturation SpO2;

[0154] S52, the collected physiological data D s (t) Input into the long short-term memory network model for emotional state analysis and prediction, generating the emotional state prediction vector Y t and adjusting various parameters of the virtual reality environment in real time through step S4;

[0155] S53, after each time step t, the system compares the adjusted virtual reality environment parameters with the user's new physiological data D s (t+1) and input it into the long short-term memory network model for updating to generate a new emotional state prediction vector Y t+1 ;

[0156] S54. Define the emotion change trend vector ΔY t for:

[0157] ΔY t =Y t+1 -Y t ;

[0158] Among them, ΔY t represents the change in the user’s emotional state from time step t to time step t+1, reflecting the impact of the adjusted virtual reality environment on the user’s emotions. t+1 and Y t Represent the emotional state prediction vectors at time step t+1 and time step t respectively;

[0159] S55, according to the emotion change trend vector ΔY t Based on the results, we can determine the direction of change of the user’s emotional state and continuously optimize the model.

[0160] Embodiment 1:

[0161] In June 2024, the customer service center of a large telecommunications company was located in Beijing. Employees in this department often had to handle a large number of customer calls and online consultations. Due to the huge volume of business, employees faced high work pressure, especially during daily peak hours (9:00 to 11:00 in the morning and 2:00 to 5:00 in the afternoon). Employees generally felt nervous and anxious, which not only affected their mental health, but also led to a decline in overall work efficiency.

[0162] To improve this situation, the company decided to introduce a virtual reality-based emotional stress relief and monitoring system in the customer service department. The system aims to help employees manage emotional stress in a timely manner at work, thereby improving work efficiency and employee happiness.

[0163] Application in this scenario:

[0164] Since June 1, 2024, 50 employees of the customer service center have been selected as the test group. They have worn a set of biosensor devices consisting of a heart rate monitor, an electroencephalogram (EEG) headband, a galvanic skin response (GSR) sensor, and a pulse oximeter during work. These devices collect employees' physiological data in real time, including heart rate, EEG signals, galvanic skin response, and blood oxygen saturation, and transmit the data to the emotional stress relief and monitoring system via a wireless network.

[0165] In the first two weeks of the test, the system conducted a preliminary analysis of the collected physiological data through the LSTM model and established the baseline emotional state of each employee. These baseline data include the range of physiological parameters of employees in normal, relaxed, anxious and other states. With this data, the system can identify employees' emotional fluctuations in real time in subsequent work.

[0166] In the next month, whenever the system detects that an employee's emotional state deviates from the baseline (for example, increased anxiety, heart rate 15% higher than normal, and significant increase in beta wave activity in EEG), the system will immediately send a notification on the employee's work terminal, suggesting that he or she enter a short virtual reality emotion regulation scene. Employees can choose to do this process during work breaks or lunch breaks, which usually lasts 5 to 10 minutes.

[0167] In actual application, when an employee enters a virtual reality scene, the system first guides him to wear a virtual reality headset and enter a specially designed relaxation scene. This scene is a tranquil lakeside, the surrounding environment is set to soft natural light, and the background music is soft natural sounds such as flowing water and birdsong. According to the real-time physiological feedback of the employee, the system will dynamically adjust the lighting, color tone and sound effects in the environment. For example, if the system detects that the employee's heart rate is still high, the light intensity will be further reduced and the rhythm of the background music will be slowed down to help the employee gradually relax.

[0168] Table 1 Comparison of employees' physiological data

[0169]

[0170] Table 2 Comparison of employee performance

[0171]

[0172]

[0173] Table 3 Employee satisfaction and efficiency improvement

[0174]

[0175] Referring to the above table, during the one-month test period (June 1 to June 30, 2024), the system recorded the physiological data and work performance of 50 participating employees. The test results show that the average heart rate of employees using the system during critical working hours (peak hours) dropped from 85 beats per minute to 72 beats per minute, and the heart rate variability (HRV) increased from 45 milliseconds to 54 milliseconds, indicating that the overall anxiety level of employees was significantly reduced. At the same time, EEG data showed that in the relaxation scene, the activity of alpha waves (associated with a relaxed state) increased by 20%, while the activity of beta waves (associated with tension and stress) decreased by 22%.

[0176] Compared with the data before the test, the work performance of employees after the test also improved significantly. The average response time of each employee was shortened from 30 seconds to 27 seconds, and the number of customer requests processed daily increased from 65 to 72 times, and the work efficiency increased by about 10.7%. At the same time, the satisfaction of employees with their work increased by 15% after the test. Most of the employees reported that the virtual reality adjustment scene significantly improved their emotional state, allowing them to better concentrate in a high-pressure work environment.

[0177] In addition, during the test, the frequency of employees' breaks decreased significantly, with the average number of breaks per employee during peak hours reduced from 1.5 times per hour to 1 time. This change shows that through real-time monitoring and effective regulation of emotional stress, employees' working time has been extended and productivity has been improved.

[0178] This example demonstrates the effectiveness of the emotional stress relief and monitoring system based on virtual reality technology through its application in a real workplace environment. Through real-time data collection of biosensor devices and emotional analysis and prediction of LSTM models, the system can accurately identify employees' emotional fluctuations and achieve personalized emotional regulation by dynamically adjusting the virtual reality environment. Test data proves the significant effect of the system in reducing employee anxiety levels, improving work efficiency, and enhancing job satisfaction, which provides a practical and efficient solution for emotional management in the modern workplace.

[0179] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A method for relieving and monitoring emotional stress in combination with virtual reality technology, characterized in that: The steps include: S1. Provide users with a virtual environment through virtual reality devices. The virtual environment is built on the Unity platform. Users interact with the environment by wearing virtual reality headsets and using controllers. S2, collect the user's physiological data in real time through biosensors, including heart rate, brain waves, skin galvanic response and blood oxygen saturation; S3, inputting the physiological data of S2 into the long short-term memory network model for processing, analyzing the user's emotional state, and predicting his emotional changes; S4. Based on the emotional state analysis and prediction results of the long short-term memory network model, the visual, sound and interactive elements of the virtual reality environment are adjusted in real time, including adjusting the ambient light, sound effects and the interactivity of virtual objects to adapt to the user's current emotional state; S5. Repeat steps S2 to S4 to achieve dynamic monitoring and relief of the user's emotional stress.

2. The method for relieving and monitoring emotional stress in combination with virtual reality technology according to claim 1, characterized in that: The S1 comprises the following steps: S11. Use the Unity platform to build a virtual reality environment to meet users' emotional regulation needs by adjusting visual elements, sound effects, interactive elements and scene changes; S12. Through virtual reality headsets, high-definition display, wide viewing angle and low-latency visual effects are provided, and audio output devices are provided to achieve immersive sound experience; S13. Use a handheld controller to interact, and support users to navigate, manipulate objects, and switch scenes in a virtual environment through an interactive device, and have a tactile feedback function to enhance the user experience; S14, when the system is initialized, the virtual scene preset by the user is loaded, and the visual elements of the scene including color, brightness and sound effects including background music and natural sound effects are adjusted in real time according to the emotion adjustment mode selected by the user; S15. Define virtual scene parameters P s is a set of control variables, including light intensity L, hue C, volume V, and interactive element response time T r , the system dynamically adjusts P according to the user's emotional state and real-time feedback s =(L, C, V, T r ) value to optimize the user experience.

3. The method for relieving and monitoring emotional stress in combination with virtual reality technology according to claim 1, characterized in that: The S2 specifically includes: S21. The user wears a biosensor device, including a heart rate monitor, which measures the user's heart rate data HR and heart rate variability HRV in real time through photoplethysmography, an EEG headband, which collects the user's brain wave signals through electrodes and records the brain's electrical activity data EEG(t), where t represents the collection time, a galvanic skin response sensor, which detects changes in the conductivity of the user's skin and generates galvanic skin response data GSR(t) to evaluate the user's stress level, and a pulse oximeter, which measures the user's blood oxygen saturation SpO2 and respiratory rate RR through an optical sensor; S22, the biosensor device transmits the physiological data D(t) collected in real time to the data processing unit through the wireless communication module, where D(t) = {HR, HRV, EEG(t), GSR(t), SpO2, RR}; S23, the data processing unit pre-processes the physiological data, including signal denoising and normalization, and synchronizes the multi-channel data to generate a unified time series data D s (t); S24, the preprocessed time series data D s (t) is passed to the long short-term memory network model as input data for emotional state analysis and prediction.

4. The emotional stress relief and monitoring method combined with virtual reality technology according to claim 1, characterized in that: The S3 specifically includes: S31, the preprocessed time series data D s (t) input into the long short-term memory network model to capture the time-dependent patterns related to emotional changes in the user's physiological data; S32. The input of the long short-term memory network model is the physiological data vector D at the current time step t. s (t), physiological data vector D s (t) contains the user's heart rate, brain waves, and skin electrical response physiological data at time t. The model processes the relationship between the data of the current time step and the data of the previous time step through the input gate, forget gate, and output gate of the memory unit to generate a hidden state vector h t : h t =σ(W h ·[D s (t),h t-1 ]+b h ); Among them, h t represents the hidden state vector generated by the LSTM network model at time step t, and represents the user emotion information currently memorized by the LSTM network. s (t) represents the set of user physiological data collected at time t, h t-1 represents the hidden state vector of the LSTM network model at the previous time step t-1, which carries the memory of the LSTM network about the previous emotional state. σ is the activation function, which is used to convert the calculation result into a value suitable for the next step of processing. S33, through the multi-layer physiological data vector D s (t) Iterative calculation of the network to generate the final emotional state prediction vector Y t , which represents the system's prediction of the user's emotional state at time t: Y t =softmax(W y ·h t +b y ): Among them, Y t It represents the predicted value of the emotional state at time t, which is the system's estimation result of the user's current emotion. Softmax is a conversion function used to convert the output of the long short-term memory network into the probability distribution of different emotional states and determine the user's main emotional category; S34. Predicting vector Y based on emotional state t , identifies the user’s emotional state category and dynamically adjusts elements of the virtual reality environment based on the state category.

5. The method for relieving and monitoring emotional stress in combination with virtual reality technology according to claim 1, characterized in that: The S34 specifically includes: The emotional state prediction vector Y t Sort the probability values ​​of each emotion category in the , and determine the user's current main emotional state category C t : C t =argmax(Y t ); Among them, C t It represents the main emotional state category of the user predicted by the system at time t, including anxiety, relaxation, and stress; S342: According to the user's main emotional state category C t , select the corresponding regulation strategy S(C t ), where each adjustment strategy contains a set of adjustment rules for virtual reality environment parameters: S(C t )={P s :L(C t ),C(C t ),v(C t ),T r (C t )}; Among them, L(C t ) indicates emotional state C t The matching ambient light intensity, C(C t ) indicates emotional state C t Matching scene color, V(C t ) indicates emotional state C t Matching sound effects, T r (C t ) indicates emotional state C t Matching virtual object interaction response time; S343, real-time monitoring of adjusted user physiological data D s (t+1), input it into the long short-term memory network model to generate a new emotional state prediction vector Y t+1 ; S344, according to Y t+1 The results are used to determine the changing trend of the user's emotional state; S345, when Y t+1 If it is shown that the user's emotions have stabilized, the system will gradually restore the virtual environment to the initial settings, ending the current emotion regulation process.

6. The method for relieving and monitoring emotional stress in combination with virtual reality technology according to claim 1, characterized in that: The S4 specifically includes: S41, emotional state prediction vector Y generated based on long short-term memory network model t , determine the user's current main emotional state category C t ,and guide the dynamic adjustment of the VR environment through categories; S42, according to emotional state category C t , adjust the light intensity L of the virtual reality environment: L=L0×f(C t ); Among them, L0 is the initial light intensity of the virtual environment, f(C t ) is based on the emotional state category C t The lighting adjustment function is used to increase or decrease the lighting intensity; S43. Adjust the color tone of the virtual reality environment C: C=C0+ΔC(C t ); Among them, C0 is the initial color parameter of the virtual environment, ΔC(C t ) is based on the emotional state category C t The color adjustment value, ΔC(C t ) determines the warmth or lightness of the hue; S44. Adjust the sound effects of the virtual reality environment V: V=V0×g(C t ); Among them, V0 is the initial sound volume of the virtual environment, g(C t ) is based on the emotional state category C t The sound adjustment function is used to increase or decrease the volume of background music or environmental sound effects; S45, adjusting the interactive response time T of the virtual object in the virtual reality environment r : T r =T r0 ×h(C t ); Among them, T r0 is the initial response time of the virtual object in the virtual environment, h(C t ) is based on the emotional state category C t The response time adjustment function is used to speed up or delay the interactive response speed between the user and the virtual object; S46. Real-time collection of the user's physiological data in the adjusted virtual environment D s (t+1), and input it into the long short-term memory network model, perform emotional state analysis and prediction again, and obtain a new emotional state prediction vector Y t+1 .

7. The method for relieving and monitoring emotional stress in combination with virtual reality technology according to claim 1, characterized in that: The S42 specifically includes: S42, based on emotional state category C t Adjusting the light intensity L of the virtual reality environment specifically includes the following steps: S421, predicting vector Y based on emotional state t The probability distribution of different emotional states in the equation determines the main emotional state category C t , and calculate the lighting adjustment weight w t : Among them, p i Indicates emotional state category C i The probability value represents the possibility that the user is currently in this emotional state, k i Representation and emotional state category C i The associated illumination adjustment coefficient is used to characterize the adjustment direction and amplitude of illumination intensity under the emotional state, where n is the total number of all possible emotional states; S422, adjust weight w according to illumination t , calculate the light intensity adjustment function f(C t ) value, which is used to determine the specific intensity adjustment of the light in the virtual reality environment: f(C t )=1+α·w t ; Among them, α is the sensitivity parameter of light intensity adjustment, which determines the responsiveness of light intensity to changes in emotional state; w t To adjust the light weight calculated based on the user's current emotional state, affecting the increase or decrease of light intensity; S424, real-time collection and analysis of the user's physiological data under the adjusted lighting environment D s (t+1), input it into the long short-term memory network model to generate a new emotional state prediction vector Y t+1 ; S425, predicting vector Y based on the new emotional state t+1 , calculate the lighting adjustment weight w again t+1 And the light intensity adjustment function f(C t+1 ), and the light intensity L t+1 Dynamically adjust to ensure the VR environment always matches the user's emotional state.

8. The method for relieving and monitoring emotional stress in combination with virtual reality technology according to claim 1, characterized in that: The S5 specifically includes: S51. While the user is experiencing the virtual reality environment, real-time collection of physiological data D s (t), including heart rate HR, EEG(t), galvanic skin response GSR(t) and blood oxygen saturation SpO2; S52, the collected physiological data D s (t) Input into the long short-term memory network model for emotional state analysis and prediction, generating the emotional state prediction vector Y t and adjusting various parameters of the virtual reality environment in real time through step S4; S53, after each time step t, the system compares the adjusted virtual reality environment parameters with the user's new physiological data D s (t+1) and input it into the long short-term memory network model for updating to generate a new emotional state prediction vector Y t+1 ; S54. Define the emotion change trend vector ΔY t for: ΔY t =Y t+1 -Y t ; Among them, ΔY t represents the change in the user’s emotional state from time step t to time step t+1, reflecting the impact of the adjusted virtual reality environment on the user’s emotions. t+1 and Y t Represent the emotional state prediction vectors at time step t+1 and time step t respectively; S55, according to the emotion change trend vector ΔY t Based on the results, we can determine the direction of change of the user’s emotional state and continuously optimize the model.

Images