Emotion regulation method and system based on distributed thermal stimulation

Abstract

The invention provides an emotion regulation method and system based on distributed thermal stimulation, and relates to the technical field of biomedical engineering, and the method comprises the steps: collecting the individual characteristics of a target user, and the resting skin blood perfusion amount and resting nerve electrical activity of each wearing part of distributed thermal stimulation equipment; determining an initial temperature, and outputting an initial pressure intensity and an initial frequency in combination with a pre-trained pressure intensity frequency prediction model; thermal stimulation is carried out on each wearing part according to the initial temperature, the initial pressure intensity and the initial frequency, and the real-time skin blood perfusion amount and the real-time nerve electrical activity of each wearing part are collected; calculating individual sensitivity coefficients; acquiring a group sensitivity coefficient, and correcting the initial temperature, the initial pressure and the initial frequency to obtain a corrected temperature, a corrected pressure and a corrected frequency; and executing thermal stimulation according to the corrected temperature, the corrected pressure intensity and the corrected frequency until a preset emotion regulation target is met. The technical problem that in the prior art, a thermal stimulation emotion adjusting method is insufficient in adjusting accuracy is solved.

Description

Technical Field

[0001] This invention relates to the field of biomedical engineering, and in particular to a method and system for mood regulation based on distributed thermal stimulation. Background Technology

[0002] In the field of biomedical engineering, negative emotions such as anxiety and irritability are important factors affecting people's physical and mental health and quality of life. Thermal stimulation, as a gentle and easily accepted form of physical stimulation, can effectively regulate negative emotions by regulating the activity of the human nervous system and blood circulation.

[0003] However, traditional heat-stimulated mood regulation methods are mostly crude, single-parameter-based, and lack adaptability to individual physiological characteristics and sensitivity differences, resulting in insufficient regulation precision and poor effect stability.

[0004] Therefore, there is an urgent need for an emotion regulation method based on distributed thermal stimulation to overcome the limitations of traditional methods. Summary of the Invention

[0005] This invention addresses the technical problem of insufficient precision in existing heat-stimulated emotion regulation methods by providing a distributed heat-stimulated emotion regulation method and system.

[0006] The technical solution of the present invention to solve the above-mentioned technical problems is as follows: In a first aspect, the present invention provides an emotion regulation method based on distributed thermal stimulation, comprising: Collect individual characteristics of target users and resting skin blood perfusion and resting neural electrical activity at each wearing site of the distributed thermal stimulation device; The initial temperature is determined based on the individual characteristics. The initial temperature and individual characteristics are then input into a pre-trained pressure-frequency prediction model, which outputs the initial pressure and initial frequency. The distributed thermal stimulation device is activated, and thermal stimulation is applied to each wearing site according to the initial temperature, initial pressure, and initial frequency, while real-time skin blood perfusion and real-time nerve electrical activity of each wearing site are collected simultaneously. Based on the dynamic differences between resting and real-time skin blood perfusion, and between resting and real-time neural electrical activity at each wearing site, the individual sensitivity coefficient was calculated. Obtain the group sensitivity coefficient of a group with similar individual characteristics to the target user, and combine the individual sensitivity coefficient to correct the initial temperature, initial pressure, and initial frequency to obtain the corrected temperature, corrected pressure, and corrected frequency; The temperature, pressure, and frequency are adjusted as described above to continuously apply thermal stimulation to each wearing part until the preset emotion regulation target is met.

[0007] Secondly, the present invention provides an emotion regulation system based on distributed thermal stimulation, comprising: The resting vital signs data acquisition module is used to collect the individual characteristics of the target user and the resting skin blood perfusion and resting nerve electrical activity of each wearing site of the distributed thermal stimulation device; The initial stimulus parameter determination module is used to determine the initial temperature based on the individual characteristics, input the initial temperature and individual characteristics into a pre-trained pressure-frequency prediction model, and output the initial pressure and initial frequency. The real-time vital signs data acquisition module is used to activate the distributed thermal stimulation device, perform thermal stimulation on each wearing site according to the initial temperature, initial pressure, and initial frequency, and simultaneously collect the real-time skin blood perfusion and real-time nerve electrical activity of each wearing site. The sensitivity coefficient calculation module is used to calculate the individual sensitivity coefficient based on the dynamic differences between resting skin blood perfusion and real-time skin blood perfusion, and between resting neural electrical activity and real-time neural electrical activity at each wearing site. The stimulus parameter correction module is used to obtain the group sensitivity coefficient of a group with similar individual characteristics to the target user, and to correct the initial temperature, initial pressure, and initial frequency in combination with the individual sensitivity coefficient to obtain the corrected temperature, corrected pressure, and corrected frequency. The output execution module is used to continuously apply thermal stimulation to each wearing part according to the corrected temperature, corrected pressure, and corrected frequency until the preset emotion regulation target is met.

[0008] The beneficial effects of this invention are: Compared to existing technologies, this application first collects the individual characteristics of the target user and the resting skin blood perfusion and resting neural electrical activity of each wearing site of the distributed thermal stimulation device, obtaining accurate individual characteristic data and resting physiological data, providing a precise data foundation for the individualized determination of subsequent stimulation parameters. Secondly, the initial temperature is determined based on individual characteristics, and the initial temperature and individual characteristics are input into a pre-trained pressure-frequency prediction model, outputting initial pressure and initial frequency. This fully considers the deep coupling relationship between temperature, pressure, and frequency during thermal stimulation and accurately outputs the initial control parameter combination adapted to the target user. Thirdly, the distributed thermal stimulation device is activated, and thermal stimulation is performed on each wearing site according to the initial temperature, initial pressure, and initial frequency, while simultaneously collecting real-time skin blood perfusion and real-time neural electrical activity of each wearing site, providing direct real-time feedback support for parameter correction. Next, based on the dynamic differences between resting and real-time skin blood perfusion and resting and real-time neural electrical activity at each wearing site, an individual sensitivity coefficient is calculated, which can comprehensively reflect the overall sensitivity of the target user's wearing site to thermal stimulation. Furthermore, the group sensitivity coefficients of a group with similar individual characteristics to the target user are obtained. These individual sensitivity coefficients are then used to correct the initial temperature, initial pressure, and initial frequency, resulting in corrected temperature, corrected pressure, and corrected frequency. By incorporating the group sensitivity coefficients of similar groups for correction, the adaptability and stability of emotion regulation are effectively improved. Finally, thermal stimulation is continuously applied to each wearing area according to the corrected temperature, corrected pressure, and corrected frequency until the preset emotion regulation target is met, achieving a precise, safe, and long-lasting emotion regulation effect.

[0009] Through the above technical solution, this application determines initial parameters by collecting individual characteristics of the target user and the resting skin blood perfusion and resting neural electrical activity at each wearing site. During the application of thermal stimulation, real-time physiological data is simultaneously captured and individual sensitivity coefficients are calculated. The initial temperature, initial pressure, and initial frequency are then adjusted by combining the group sensitivity coefficients of similar groups before continuous application until the preset emotion regulation goal is achieved, forming a complete closed loop. This effectively overcomes the crudeness of traditional thermal stimulation regulation, achieving precise adaptation of emotion regulation to the individual characteristics and physiological sensitivity differences of the target user, improving the stability and reliability of the regulatory effect, and meeting the needs for personalized and precise emotion regulation. Attached Figure Description

[0010] Figure 1 A flowchart illustrating the emotion regulation method based on distributed thermal stimulation provided by this invention; Figure 2 This is a schematic diagram of the structure of the emotion regulation system based on distributed thermal stimulation provided by the present invention.

[0011] In the attached diagram, the components represented by each number are as follows: The module includes a resting vital signs data acquisition module 11, an initial stimulus parameter determination module 12, a real-time vital signs data acquisition module 13, a sensitivity coefficient calculation module 14, a stimulus parameter correction module 15, and an output execution module 16. Detailed Implementation

[0012] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0013] In the description of this invention, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the stated features. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0014] In the description of this invention, the term "for example" is used to mean "used as an example, illustration, or description." Any embodiment described as "for example" in this invention is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is provided to enable any person skilled in the art to make and use the invention. Details are set forth in the following description for purposes of explanation. It should be understood that those skilled in the art will recognize that the invention can be made without using these specific details. In other instances, well-known structures and processes will not be described in detail to avoid obscuring the description of the invention with unnecessary detail. Therefore, the invention is not intended to be limited to the embodiments shown, but is consistent with the broadest scope of the principles and features disclosed herein.

[0015] Example 1, as Figure 1 As shown, embodiments of the present invention provide an emotion regulation method based on distributed thermal stimulation, including: S10: Collect individual characteristics of the target user and resting skin blood perfusion and resting nerve electrical activity at each wearing site of the distributed thermal stimulation device.

[0016] Traditional heat-stimulated mood regulation methods often use standardized parameters, ignoring individual physiological and physical differences, resulting in insufficient regulatory adaptability.

[0017] Meanwhile, skin blood perfusion and neural electrical activity are core physiological markers of emotional fluctuations, which are respectively associated with vasomotor state and neural excitability. To achieve individualized regulation, we can obtain the individual characteristics of the target user and the skin blood perfusion and neural electrical activity at rest, providing a precise basis for providing personalized thermal stimulation.

[0018] To address the aforementioned issues, this application collects individual characteristics of the target user and resting skin blood perfusion and resting neural electrical activity at each wearing site of the distributed thermal stimulation device.

[0019] Specifically, step S10 in the method includes: Collect individual characteristics of the target users, including age, gender, BMI, and body type; When the target user is in a calm state with a stable heart rate and breathing, the resting skin blood perfusion and resting neural electrical activity of the corresponding wearing area are collected by each wearing unit of the distributed thermal stimulation device.

[0020] In this embodiment, the individual characteristics of the target user are first collected. These characteristics reflect the target user's physiological basis and physical differences. These characteristics include age, gender, BMI, and body type: age directly affects skin thickness, blood circulation speed, and nerve sensitivity; users of different ages show significant differences in their tolerance and response speed to heat stimulation; gender differences lead to different hormone levels and skin structures, thus affecting the conduction and physiological feedback of heat stimulation; BMI, or Body Mass Index, is calculated by dividing weight (kg) by the square of height (m); BMI reflects body fat percentage and metabolic state, and a high or low body fat percentage affects heat penetration efficiency; body type can refer to the classification standards of traditional Chinese medicine, such as balanced constitution, yang deficiency constitution, and yin deficiency constitution, etc., and different body types show significant differences in their tolerance to cold and heat.

[0021] For example, the individual constitution is determined by combining the basic information filled in by the user with the TCM constitution classification and judgment scale, and the individual characteristics of the target user are: 25 years old, female, BMI of 22, and balanced constitution.

[0022] Secondly, when the target user is in a calm state with stable heart rate and respiration, the resting skin blood perfusion and resting neural electrical activity are collected from each wearing unit of the distributed thermal stimulation device at the corresponding wearing site. Each wearing unit of the distributed thermal stimulation device is a component that directly contacts the skin and has a built-in physiological signal acquisition sensor. Each wearing unit corresponds to a different stimulation site, such as the wrist or ankle, and can independently collect physiological data for each wearing site. Resting skin blood perfusion and resting neural electrical activity can reflect the basic physiological state of the target user when they are emotionally stable and relaxed, and can serve as a reference benchmark for emotion regulation. Therefore, the data collection process must eliminate the influence of factors such as emotional fluctuations, exercise, and environmental interference on the physiological data, requiring the target user to be in a calm state with stable heart rate and respiration.

[0023] For example, the resting skin blood perfusion and resting neural electrical activity of each wearing site can be collected by the following method: First, provide the target user with a noise-free, softly lit, and independent environment, requiring them to be emotionally stable and have no recent exercise history, and have them sit quietly in this environment for 5-10 minutes; then, through the physiological monitoring module of the distributed thermal stimulation device, such as a wrist heart rate sensor, confirm that the target user's heart rate is stable at 60-80 beats / minute and respiratory rate is stable at 12-18 breaths / minute, that is, determine that the target user is in a calm state; then, activate the miniature blood flow sensor and flexible neural electrode built into each wearing unit of the distributed thermal stimulation device, and continuously collect data from the corresponding site at a preset acquisition frequency, such as 3 times / second, and calculate the average of multiple sets of data to obtain the resting skin blood perfusion, such as 3.2 mL / (100g・min), and the resting neural electrical activity, such as 25-35 μV.

[0024] In summary, compared to existing technologies, this application collects individual characteristics of the target user and resting skin blood perfusion and resting neural electrical activity at each wearing site of the distributed thermal stimulation device. This provides accurate individual characteristic data and resting physiological data, offering a precise data foundation for the subsequent individualized determination of stimulation parameters.

[0025] S20: Determine the initial temperature based on the individual characteristics, input the initial temperature and individual characteristics into the pre-trained pressure-frequency prediction model, and output the initial pressure and initial frequency.

[0026] In traditional heat-stimulated mood regulation, control parameters often rely on experience-based settings, failing to fully consider individual differences in heat stimulation tolerance based on age, physical condition, and other characteristics, as well as the physiological coupling relationship between temperature, pressure, and frequency.

[0027] Specifically, during thermal stimulation, temperature directly affects the efficiency of heat penetration and the degree of nerve arousal, requiring matching of pressure and frequency: when the temperature rises, the pressure should be increased appropriately to ensure that the heat can effectively penetrate the skin surface and subcutaneous tissue, avoiding local overheating or attenuation of stimulation effect due to insufficient heat conduction despite temperature rise; conversely, when the temperature rises, the frequency should be reduced appropriately to reduce the number of stimulations per unit time, avoiding nerve over-excitation or discomfort caused by high-frequency stimulation due to temperature rise.

[0028] To address the aforementioned issues, this application determines the initial temperature based on the individual characteristics, inputs the initial temperature and individual characteristics into a pre-trained pressure-frequency prediction model, and outputs the initial pressure and initial frequency.

[0029] Specifically, step S20 in the method includes: Based on the individual characteristics of the target users, and combined with the preset individual characteristic-temperature mapping rules, the initial temperature is determined; The pre-trained pressure-frequency prediction model is invoked, wherein the pressure-frequency prediction model is based on a multi-output gradient boosting regression framework and has a built-in temperature-pressure coupling module and a temperature-frequency coupling module. The initial temperature and individual characteristics are input into a pre-trained pressure-frequency prediction model. The model is predicted using the basic framework, and then verified and corrected by the temperature-pressure coupling module and the temperature-frequency coupling module to output the initial pressure and initial frequency.

[0030] In this embodiment, the initial temperature is first determined based on the individual characteristics of the target user and a preset individual characteristic-temperature mapping rule. The preset individual characteristic-temperature mapping rule is established based on a large amount of historical user data, through statistical analysis, to determine the correspondence between different combinations of individual characteristics and safe, effective thermal stimulation temperatures.

[0031] For example, a large amount of historical user data on heat stimulation-induced mood regulation was collected, and statistical analysis was used to screen out the effective heat stimulation temperature ranges for each group without skin discomfort, forming a preset individual characteristic-temperature mapping rule. For example, 36.5-37.5℃ corresponds to the group of women aged 20-30 with a BMI of 20-24 and a balanced constitution, and 38-39.5℃ corresponds to the group of men aged 40-50 with a BMI of 25-28 and a Yang deficiency constitution.

[0032] For example, based on the individual characteristics of the target user, such as being 25 years old, female, having a BMI of 22, and having a balanced temperament, and combined with the aforementioned preset individual characteristic-temperature mapping rules, the initial temperature is determined to be 37℃.

[0033] Secondly, the pre-trained pressure-frequency prediction model is invoked. This model is based on a multi-output gradient boosting regression framework and includes built-in temperature-pressure and temperature-frequency coupling modules. The multi-output gradient boosting regression model has the advantage of simultaneously predicting multiple related output variables, adapting to the need for simultaneous prediction of both pressure and frequency parameters. The temperature-pressure and temperature-frequency coupling modules ensure parameter synergy, guaranteeing that the model's prediction results not only conform to historical data patterns but also follow physiological logic.

[0034] Finally, the initial temperature and individual characteristics are input into the pre-trained pressure-frequency prediction model. The model predicts using the basic framework, and then verifies and corrects the results through the temperature-pressure coupling module and the temperature-frequency coupling module, outputting the initial pressure and initial frequency. The use of both initial temperature and individual characteristics as model inputs simultaneously considers both the parameter requirements of thermal stimulation and the individual differences of users.

[0035] Specifically, the basic framework of the pressure-frequency prediction model is based on the initial input temperature and individual characteristics. Through the mapping relationship learned during training, it predicts the initial pressure and frequency. Subsequently, the temperature-pressure coupling module verifies the initial predicted pressure according to the logic that as temperature increases, pressure increases in a positive direction to improve heat penetration. If the initial predicted pressure and temperature do not meet the preset coupling relationship, it is corrected. At the same time, the temperature-frequency coupling module verifies and corrects the initial predicted frequency according to the logic that as temperature increases, frequency decreases in the opposite direction to avoid over-awakening of nerves. The final output is an initial pressure and initial frequency that conforms to both historical data patterns and physiological logic.

[0036] For example, the initial temperature (e.g., 37°C) and individual characteristics (e.g., 25 years old, female, BMI 22, mild temperament) are input into a pre-trained pressure-frequency prediction model. Through the basic framework prediction, the initial predicted pressure is 2.0 kPa and the initial predicted frequency is 4 Hz. Then, the temperature-pressure coupling module determines that the reasonable pressure range at 37°C is 1.8-2.2 kPa. 2.0 kPa is within this range and does not require correction. At the same time, the temperature-frequency coupling module determines that the reasonable frequency range at 37°C is 3-3.8 Hz. 4 Hz is outside this range, so the boundary value of 3.8 Hz is forcibly taken. Finally, the initial pressure is output as 2.0 kPa and the initial frequency is 3.8 Hz.

[0037] Specifically, the construction process of the "pressure frequency prediction model" includes: A multi-output gradient boosting regression model is adopted, with two prediction channels set up as the basic framework for the pressure frequency prediction model; Historical input features from different users are collected from historical data to form a sample dataset; For the aforementioned sample dataset, the optimal pressure that enables the skin blood perfusion to meet the standard is collected and labeled to form a sample pressure label set. Simultaneously, the optimal frequency that enables the neural electrical activity to meet the standard is collected and labeled to form a sample frequency label set. Two prediction channels are trained using the aforementioned sample dataset and the sample pressure label set and sample frequency label set, respectively, until convergence is verified. After verification of convergence, a pre-built temperature-pressure coupling module and a temperature-frequency coupling module are embedded as a physiological logic verification layer to obtain the pressure-frequency prediction model.

[0038] In this embodiment, a multi-output gradient boosting regression model is first employed, with two prediction channels set up as the basic framework for the pressure-frequency prediction model. The two prediction channels are independent yet share features, and are used for initial pressure prediction and initial frequency prediction, respectively.

[0039] For example, the basic framework of the pressure frequency prediction model constructed using a multi-output gradient boosting regression model mainly consists of an input layer, a feature processing layer, a gradient boosting ensemble layer, and dual output channels, as shown in the following structure: The input layer is responsible for receiving the initial temperature and individual characteristics, and converting them into structured feature vectors.

[0040] The feature processing layer standardizes the input features to eliminate dimensional differences and feature interactions, such as generating age and initial temperature, body type and BMI, thereby enhancing the explanatory power of the features on the output parameters.

[0041] The gradient boosting ensemble layer consists of 100-200 weak learners serially integrated, such as decision trees of depth 3-5 serially integrated. It optimizes parameters by iteratively fitting the prediction residuals of the previous model, while learning the nonlinear mapping relationship between input features and pressure and frequency.

[0042] The dual output channels are connected to the gradient boosting ensemble layer. One output channel focuses on outputting the pressure prediction value, and the other output channel focuses on outputting the frequency prediction value. The two channels share the learning results of the feature processing layer and the gradient boosting ensemble layer, which ensures the independence of parameter prediction while preserving the intrinsic relationship between the two.

[0043] Secondly, historical input features from different users were collected to form a sample dataset. This historical data comes from a large number of users who have completed the emotion regulation experience, ensuring the diversity and representativeness of the data within the sample dataset. The historical input features are consistent with the feature dimensions of the target user input model, including age, gender, BMI, body type, and initial temperature for different users.

[0044] For example, from a large amount of historical data of users who have completed the entire process of emotion regulation, historical input characteristics of different users, including age, gender, BMI, body type and initial temperature, are collected in a targeted manner. Then, through statistical methods such as the 3σ principle and box plot analysis, abnormal samples with missing data or values ​​exceeding the physiologically reasonable range are removed. Finally, a sample dataset covering multiple combinations of individual characteristics and meeting the standards of data integrity and reliability is formed, providing high-quality basic data for training the pressure frequency prediction model.

[0045] Next, for the sample dataset, the optimal pressure that ensures adequate skin blood perfusion is collected and labeled, forming a sample pressure label set. Simultaneously, the optimal frequency that ensures adequate neural electrical activity is collected and labeled, forming a sample frequency label set. Adequate skin blood perfusion refers to skin blood perfusion remaining stable within a reasonable range under resting conditions, such as ±5% of resting skin blood perfusion. Adequate neural electrical activity refers to the frequency and amplitude of neural electrical activity remaining stable within a reasonable range under resting conditions, such as ±1Hz of resting frequency and ±8% of resting amplitude. Optimal pressure refers to the pressure value that ensures adequate skin blood perfusion without causing discomfort due to skin pressure; optimal frequency refers to the frequency value that ensures adequate neural electrical activity, matches the neural physiological rhythm, and avoids over- or under-arousal of nerves.

[0046] The optimal pressure for collecting data corresponding to achieving the target skin blood perfusion, and the optimal frequency for collecting data corresponding to achieving the target neural electrical activity, are chosen because pressure and frequency have clearly defined physiological regulatory targets and are highly compatible with the parameter coupling characteristics of thermal stimulation. The core function of pressure is to change the skin contact tightness and heat penetration efficiency of thermal stimulation, directly affecting the dilation / contraction state of local skin blood vessels. Appropriate pressure allows thermal stimulation to be effectively conducted to the subcutaneous vascular layer, pushing skin blood perfusion closer to the reasonable range of the calm state. Therefore, the optimal pressure selected with the target skin blood perfusion can take into account the optimal value of heat penetration effect and vascular physiological compatibility. The core function of frequency is to regulate the stimulation rhythm of nerves and avoid over-arousal or under-arousal of nerves. The frequency of thermal stimulation is directly related to the fluctuation pattern of neural electrical activity. Appropriate frequency can allow the nervous system to produce a stable response to thermal stimulation, pushing the frequency and amplitude of neural electrical activity to stabilize in the calm state range. Therefore, the frequency selected with the target neural electrical activity can match the optimal value of neural physiological rhythm and avoid emotional fluctuations.

[0047] Furthermore, two prediction channels are trained using the sample dataset and the sample pressure label set and sample frequency label set, respectively, until convergence is verified. For example, the two prediction channels can be trained through the following technical path: 1. Data preparation: The sample dataset is matched one-to-one with the sample pressure label set and the sample dataset with the sample frequency label set, and stratified sampling is used to divide the dataset into training set, validation set and test set in a ratio of 7:1.5:1.5. 2. Model Training: The training process for both prediction channels remains consistent. Taking the pressure prediction channel as an example, the sample data in the training set is used as the input features, and the corresponding sample pressure labels are used as the supervision labels. An Adam optimizer with a learning rate of 0.001 and mean squared error (MSE) is used as the loss function. The model is trained iteratively for 100-200 rounds. After each round of training, the prediction error is evaluated using the validation set. An early stopping strategy is adopted, i.e., if the error on the validation set does not decrease for 5 consecutive rounds, the training is terminated. When the pressure prediction error on the validation set is ≤0.5kPa and the pressure prediction error on the test set is ≤0.6kPa, the model is considered to have converged, and the trained pressure prediction channel is obtained. The frequency prediction channel can be trained using the same optimizer, loss function, number of iterations, and convergence criteria.

[0048] Finally, after verification and convergence, a pre-built temperature-pressure coupling module and a temperature-frequency coupling module are embedded as a physiological logic verification layer to obtain the pressure-frequency prediction model. The pre-built temperature-pressure coupling module and temperature-frequency coupling module are embedded to add physiological logic constraints based on historical data patterns, preventing the model from producing predictions that do not conform to physiological common sense. The physiological logic verification layer, located after the basic framework, performs a secondary correction on the prediction results of the basic framework, resulting in a pressure-frequency prediction model that combines data-driven approaches with physiological logic constraints.

[0049] Furthermore, the construction steps of the "temperature-pressure coupling module and temperature-frequency coupling module" include: Collect temperature parameters within the preset effective temperature range, and collect the corresponding effective pressure data and effective frequency data; The preset effective temperature range is divided into multiple temperature sub-ranges. The temperature parameters and effective pressure data in each temperature sub-range are fitted by linear regression to obtain the temperature-pressure coupling formula for each sub-range, and then integrated to form a temperature-pressure coupling module. Linear regression is used to backfit the temperature parameters and effective frequency data within each temperature sub-interval to obtain the temperature-frequency coupling formula for each interval, and then integrates them to form a temperature-frequency coupling module.

[0050] In this embodiment, temperature parameters within a preset effective temperature range are first collected, along with corresponding effective pressure and effective frequency data. The preset effective temperature range is determined based on the safe tolerance range of human skin and the effect on emotional regulation. Preferably, the preset effective temperature range can be set to 32-45℃, avoiding both insufficient thermal stimulation below 32℃ and the risk of skin burns above 45℃. Effective pressure data refers to the pressure value that ensures adequate skin blood perfusion and meets the safe pressure standards for human skin (e.g., 0-5 kPa); effective frequency data refers to the frequency value that ensures adequate neural electrical activity and meets the safe stimulation standards for nerves (e.g., 0-10 Hz).

[0051] For example, during the data collection process, the preset effective temperature range is determined to be 32-45℃ based on the safe tolerance range of human skin and the effect of emotion regulation. Within the preset effective temperature range, 14 temperature points are set at fixed intervals of 1℃. For each temperature point, at least 10-15 sets of pressure data and frequency data are collected when thermal stimulation is performed. For example, at 32℃, multiple sets of pressure data of 1.2-1.5kPa and multiple sets of frequency data of 8-9Hz are collected; at 38℃, multiple sets of pressure data of 2.3-2.6kPa and multiple sets of frequency data of 4-5Hz are collected. After the data collection for each temperature point is completed, abnormal data that exceeds the safe range or physiological standard conditions are removed, and stable data with an intra-group deviation of ≤5% are retained as the effective pressure data and effective frequency data for that temperature point.

[0052] Secondly, the preset effective temperature range is divided into multiple temperature sub-ranges. Linear regression is used to positively fit the temperature parameters and effective pressure data within each sub-range to obtain the interval-specific temperature-pressure coupling formula, which is then integrated to form a temperature-pressure coupling module. Dividing the preset effective temperature range into multiple temperature sub-ranges is to improve fitting accuracy and avoid excessive fitting errors caused by an overly wide temperature range.

[0053] For example, the preset effective temperature range of 32-45℃ is divided into 14 temperature sub-intervals at 1℃ intervals. A linear regression positive fitting is performed on the temperature parameters and effective pressure data within each sub-interval. This is achieved by solving the linear equation y=ax+b using the least squares method, where y is the effective pressure data, x is the temperature parameter, and a and b are fitting coefficients. The fitting result must satisfy a>0, reflecting the positive coupling relationship between temperature increase and pressure increase. This yields the temperature-pressure coupling formula for each sub-interval. Then, the temperature-pressure coupling formulas for all sub-intervals are integrated to form a complete temperature-pressure coupling module covering the preset effective temperature range, ensuring that the corresponding pressure correction basis can be queried at different temperatures.

[0054] Finally, linear regression is used to backfit the temperature parameters and effective frequency data within each temperature sub-interval to obtain the interval-specific temperature-frequency coupling formula, which is then integrated into a temperature-frequency coupling module. For example, the fitting logic of the temperature-frequency coupling formula is consistent with that of the temperature-pressure coupling formula, but the fitting direction is reversed. Taking each temperature sub-interval as a unit, linear regression is used to backfit the temperature parameters and effective frequency data to solve the linear equation z = cx + d, where z is the effective frequency data, x is the temperature parameter, and c and d are the fitting coefficients. The fitting result must satisfy c < 0, reflecting the coupling relationship that as temperature increases, frequency decreases in the opposite direction. Thus, the temperature-frequency coupling formula for each temperature sub-interval is obtained. Then, the temperature-frequency coupling formulas for all temperature sub-intervals are integrated to form the temperature-frequency coupling module, ensuring that the corresponding frequency correction basis can be found at different temperatures.

[0055] In summary, compared to existing technologies, this application determines the initial temperature based on the individual characteristics, inputs the initial temperature and individual characteristics into a pre-trained pressure-frequency prediction model, and outputs the initial pressure and initial frequency. This fully considers the deep coupling relationship between temperature, pressure, and frequency during thermal stimulation, and accurately outputs a combination of initial control parameters adapted to the target user, avoiding parameter contradictions or insufficient adaptation caused by empirical settings, thus ensuring the safety and effectiveness of thermal stimulation.

[0056] S30: Activate the distributed thermal stimulation device to perform thermal stimulation on each wearing part according to the initial temperature, initial pressure and initial frequency, and simultaneously collect the real-time skin blood perfusion and real-time nerve electrical activity of each wearing part.

[0057] The aforementioned steps obtain the initial temperature, initial pressure, and initial frequency suitable for the target user. Although this can be used to perform thermal stimulation on each wearing part, during the thermal stimulation process, the target user's real-time skin blood perfusion and real-time neural electrical activity will dynamically change with the thermal stimulation and their own emotional state. It is impossible to predict their dynamic response pattern through the initial parameters.

[0058] To address the aforementioned issues, this application employs a distributed thermal stimulation device to apply thermal stimulation to each wearing site according to initial temperature, initial pressure, and initial frequency, while simultaneously collecting real-time skin blood perfusion and real-time neural electrical activity at each wearing site. The real-time skin blood perfusion and real-time neural electrical activity can dynamically reflect the physiological changes of the target user under thermal stimulation, and these physiological changes are directly related to their emotional state.

[0059] Specifically, after the distributed thermal stimulation device is activated, thermal stimulation is applied synchronously to the target area through each wearing unit according to the initial temperature, initial pressure, and initial frequency determined in the aforementioned steps. At the same time, the miniature blood flow sensor and flexible neural electrode built into each wearing unit will be activated synchronously, continuously collecting the real-time skin blood perfusion and real-time neural electrical activity of the corresponding area at a preset collection frequency, such as 3 times / second, so as to achieve zero-delay synchronization between thermal stimulation application and physiological feedback collection.

[0060] In this way, the dynamic changes of physiological indicators under thermal stimulation can be recorded in a true manner, avoiding data distortion caused by the asynchrony between stimulation and collection. The real-time skin blood perfusion and real-time neural electrical activity collected will serve as a reliable basis for subsequent calculation of individual sensitivity coefficients and judgment of emotion regulation effects, providing direct real-time feedback support for parameter correction.

[0061] S40: Calculate the individual sensitivity coefficient based on the dynamic differences between resting skin blood perfusion and real-time skin blood perfusion, and between resting neural electrical activity and real-time neural electrical activity at each wearing site.

[0062] Traditional methods of regulating emotions through heat stimulation lack precise quantitative means for assessing individual real-time dynamic physiological responses. Relying solely on initial control parameters cannot accommodate the personalized sensitivity differences in the skin's blood flow and nervous system during heat stimulation. Furthermore, the dynamic differences between resting and real-time physiological data directly reflect an individual's true response to heat stimulation.

[0063] To address the aforementioned issues, this application calculates an individual sensitivity coefficient based on the dynamic differences between resting and real-time skin blood perfusion at each wearing site, as well as between resting and real-time neural electrical activity.

[0064] Specifically, step S40 in the method includes: The ambient temperature before the thermal stimulus is activated is obtained, and the temperature change is calculated by combining the initial temperature. For each wearing site, the rate of change of skin blood perfusion and the response delay time of blood perfusion were calculated between resting skin blood perfusion and real-time skin blood perfusion. Combined with the temperature change, the temperature-skin blood perfusion sensitivity coefficient was calculated. For each wearing site, the neural electrical activity response amplitude and neural electrical activity fluctuation coefficient of resting neural electrical activity and real-time neural electrical activity are calculated respectively. Combined with the temperature change, the temperature-neural electrical activity sensitivity coefficient is calculated. Based on preset weights, the weighted average of the temperature-skin blood perfusion sensitivity coefficient and the temperature-nerve electrical activity sensitivity coefficient is calculated to obtain the individual sensitivity coefficient for each wearing site.

[0065] In this embodiment, the ambient temperature before the thermal stimulus is initiated is first obtained, and the temperature change is calculated by combining it with the initial temperature. The formula for calculating the temperature change is: Temperature Change = Initial Temperature - Ambient Temperature Before Thermal Stimulation. The temperature change accurately reflects the actual temperature intensity of the thermal stimulus relative to the environment and is a key parameter for quantifying the impact of the thermal stimulus on physiological indicators. For example, if the ambient temperature before the thermal stimulus is initiated is 25°C and the initial temperature is 37°C, then the temperature change is 12°C.

[0066] Secondly, for each wearing site, the rate of change of skin blood perfusion volume and the response delay time of blood perfusion volume were calculated between resting and real-time skin blood perfusion volumes. Combined with temperature changes, a temperature-skin blood perfusion sensitivity coefficient was obtained. Specifically, the rate of change of skin blood perfusion volume reflects the fluctuation range of real-time skin blood perfusion volume relative to resting skin blood perfusion volume, and the response delay time of blood perfusion volume reflects the response speed of real-time skin blood perfusion volume to thermal stimulation. Combined with temperature changes, the calculated temperature-skin blood perfusion sensitivity coefficient can accurately quantify the impact of a unit temperature change on the skin blood flow system: a higher temperature-skin blood perfusion sensitivity coefficient indicates a more significant fluctuation in skin blood perfusion volume caused by a unit temperature change, meaning a stronger sensitivity of the user's skin blood flow system to thermal stimulation; conversely, a lower temperature-skin blood perfusion sensitivity coefficient indicates a smoother fluctuation in skin blood perfusion volume caused by a unit temperature change, meaning a weaker sensitivity of the user's skin blood flow system to thermal stimulation.

[0067] Next, for each wearing site, the neural electrical activity response amplitude and neural electrical activity fluctuation coefficient were calculated for both resting and real-time neural electrical activity. Combined with the temperature change, a temperature-neural electrical activity sensitivity coefficient was calculated. Specifically, the neural electrical activity response amplitude reflects the intensity change of real-time neural electrical activity relative to resting neural electrical activity, and the neural electrical activity fluctuation coefficient reflects the stability of real-time neural electrical activity. The temperature-neural electrical activity sensitivity coefficient, calculated by combining the temperature change, can accurately quantify the impact of a unit temperature change on the nervous system: a higher temperature-neural electrical activity sensitivity coefficient indicates a more pronounced fluctuation in neural electrical activity caused by a unit temperature change, meaning a stronger sensitivity of the user's nervous system to thermal stimulation; conversely, a lower temperature-neural electrical activity sensitivity coefficient indicates a smoother fluctuation in neural electrical activity caused by a unit temperature change, meaning a weaker sensitivity of the user's nervous system to thermal stimulation.

[0068] Finally, based on preset weights, the weighted average of the temperature-skin blood perfusion sensitivity coefficient and the temperature-neural electrical activity sensitivity coefficient is calculated to obtain the individual sensitivity coefficient for each wearing site. The preset weights of the temperature-skin blood perfusion sensitivity coefficient and the temperature-neural electrical activity sensitivity coefficient can be dynamically determined based on the core objectives of emotion regulation and the statistical analysis results of historical regulation effect data. For example, considering the stronger direct correlation between nervous system activity and emotional state, the weight of the temperature-skin blood perfusion sensitivity coefficient can be set to 0.4, and the weight of the temperature-neural electrical activity sensitivity coefficient can be set to 0.6. Those skilled in the art can dynamically adjust these weights according to actual application scenarios and regulation needs.

[0069] For example, if the calculated temperature-skin blood flow perfusion sensitivity coefficient and temperature-neural electrical activity sensitivity coefficient for a certain wearing site are 0.075 and 0.005 respectively, with corresponding preset weights of 0.4 and 0.6, then the individual sensitivity coefficient for that wearing site = 0.075 × 0.4 + 0.005 × 0.6 = 0.033. The individual sensitivity coefficient integrates the personalized sensitivity characteristics of the skin blood flow system and the nervous system to thermal stimulation, breaking through the limitations of a single physiological dimension, and can comprehensively reflect the overall sensitivity of the target user's corresponding area to thermal stimulation.

[0070] Specifically, the phrase "for each wearing site, calculating the rate of change of skin blood perfusion volume between resting and real-time skin blood perfusion volume, the blood perfusion volume response delay time, and combining this with the temperature change to calculate the temperature-skin blood perfusion sensitivity coefficient" includes: For each wearing site, the real-time skin blood perfusion volume is obtained for a continuous preset collection period, and the arithmetic mean is calculated as a real-time statistical value; The deviation between the real-time statistical value and the resting skin blood perfusion is calculated as the rate of change of skin blood perfusion; The time from the moment the thermal stimulation was initiated until the real-time statistical value first exceeded the preset proportion of resting skin blood perfusion was recorded as the blood perfusion response delay time. The temperature-skin blood perfusion sensitivity coefficient is obtained by multiplying the rate of change of skin blood perfusion volume by the response delay time of blood perfusion volume and then dividing by the temperature change.

[0071] In this embodiment, the real-time skin blood perfusion volume is first acquired for each wearing site over a continuously preset collection period, and the arithmetic mean is calculated as the real-time statistical value. The continuously preset collection period is used to avoid random fluctuations in single-collection data and to ensure the reliability and stability of the real-time statistical value.

[0072] For example, a continuous preset acquisition period can be set to 3 seconds to acquire multiple sets of real-time skin blood perfusion data within 3 seconds, such as 3.0 mL / (100g·min), 3.2 mL / (100g·min), 2.9 mL / (100g·min), etc., and then the arithmetic mean is calculated to obtain a real-time statistical value, such as 3.03 mL / (100g·min). The real-time statistical value can reflect the average level of skin blood perfusion within the preset acquisition period.

[0073] Secondly, the deviation between the real-time statistical value and the resting skin blood perfusion volume is calculated as the skin blood perfusion volume change rate. The formula for calculating the skin blood perfusion volume change rate is: Skin blood perfusion volume change rate = (Real-time statistical value - Resting skin blood perfusion volume) / Resting skin blood perfusion volume. If the real-time statistical value is higher than the resting skin blood perfusion volume, the skin blood perfusion volume change rate is positive, indicating that heat stimulation increases skin blood perfusion volume; if the real-time statistical value is lower than the resting skin blood perfusion volume, the skin blood perfusion volume change rate is negative, indicating that heat stimulation decreases skin blood perfusion volume. The larger the absolute value of the skin blood perfusion volume change rate, the more significant the response of blood perfusion volume to heat stimulation.

[0074] Next, the time from the moment the thermal stimulation is initiated to the moment the real-time statistical value first exceeds the preset proportion of resting skin blood perfusion is recorded as the blood perfusion response delay time. The preset proportion defines the effective response threshold of real-time skin blood perfusion to thermal stimulation. Preferably, the preset proportion can be set to 10%, meaning that when the real-time skin blood perfusion first exceeds the resting skin blood perfusion by 10%, it is determined that an effective response to thermal stimulation has been achieved.

[0075] For example, if the resting skin blood perfusion volume is 3.0 mL / (100g·min) and the preset ratio is set to 10%, then the effective response threshold is 3.3 mL / (100g·min). Timing begins from the moment the thermal stimulation is initiated and stops when the real-time statistical value first reaches or exceeds 3.3 mL / (100g·min). This duration is the blood perfusion volume response delay time. The shorter the blood perfusion volume response delay time, the faster and more sensitive the skin blood flow system is to thermal stimulation.

[0076] Finally, the product of the rate of change in skin blood perfusion and the response delay time of blood perfusion is calculated, and then divided by the temperature change to obtain the temperature-skin blood perfusion sensitivity coefficient. The formula for calculating the temperature-skin blood perfusion sensitivity coefficient is: Temperature-skin blood perfusion sensitivity coefficient = Rate of change in skin blood perfusion × Response delay time of blood perfusion / Temperature change.

[0077] For example, if the rate of change in skin blood perfusion is 15%, the response delay time is 6 seconds, and the temperature change is 12°C, then the temperature-skin blood perfusion sensitivity coefficient = (15% × 6) / 12 = 0.075. The temperature-skin blood perfusion sensitivity coefficient combines the fluctuation range and response speed of skin blood perfusion with the intensity of temperature change, accurately reflecting the degree of impact of a unit temperature change on the skin's blood flow system. A higher temperature-skin blood perfusion sensitivity coefficient indicates a stronger sensitivity of the user's skin's blood flow system to thermal stimulation.

[0078] Specifically, the phrase "for each wearing site, calculating the neural electrical activity response amplitude and neural electrical activity fluctuation coefficient of resting and real-time neural electrical activity, and combining this with the temperature change to calculate the temperature-neural electrical activity sensitivity coefficient" includes: For each wearing site, real-time neural electrical activity data are recorded within a continuously preset acquisition cycle after thermal stimulation is performed, forming a real-time neural electrical activity sequence; The peak value of real-time neural electrical activity is extracted from the real-time neural electrical activity sequence, and the difference between the peak value of real-time neural electrical activity and the resting neural electrical activity is calculated as the neural electrical activity response amplitude. Calculate the standard deviation and mean of all data in the real-time neural electrical activity sequence, and then calculate the ratio of the standard deviation to the mean to obtain the neural electrical activity fluctuation coefficient; The temperature-neural electrical activity sensitivity coefficient is obtained by multiplying the amplitude of the neural electrical activity response by the fluctuation coefficient of the neural electrical activity and then dividing by the temperature change.

[0079] In this embodiment, real-time neural electrical activity data is first recorded for each wearing site within a continuously preset acquisition period after thermal stimulation is applied, forming a real-time neural electrical activity sequence. The continuously preset acquisition period is consistent with the period used for the aforementioned real-time skin blood perfusion volume statistics, ensuring data temporal synchronization.

[0080] For example, a continuous preset acquisition period of 3 seconds is set, and multiple sets of real-time neural electrical activity data within this period are acquired simultaneously, such as frequency 8.2Hz, amplitude 0.9μV, frequency 8.5Hz, amplitude 1.0μV, frequency 8.3Hz, amplitude 0.8μV, etc., and arranged in the order of acquisition time to form a real-time neural electrical activity sequence.

[0081] Secondly, the peak value of real-time neural electrical activity is extracted from the real-time neural electrical activity sequence, and the difference between the peak value and the resting neural electrical activity is calculated as the neural electrical activity response amplitude. The peak value is the maximum value in the real-time neural electrical activity sequence, reflecting the strongest level of neural excitation under thermal stimulation. The formula for calculating the neural electrical activity response amplitude is: Neural electrical activity response amplitude = Real-time neural electrical activity peak value - Resting neural electrical activity. If the real-time neural electrical activity peak value is higher than the resting neural electrical activity, the neural electrical activity response amplitude is positive, indicating that the degree of neural excitation is enhanced compared to the resting state; if the real-time neural electrical activity peak value is lower than the resting neural electrical activity, the neural electrical activity response amplitude is negative, indicating that the degree of neural excitation is inhibited.

[0082] For example, if the resting neural electrical activity is 0.7 μV and the peak value in the real-time neural electrical activity sequence is 1.2 μV, then the neural electrical activity response amplitude = 1.2 μV - 0.7 μV = 0.5 μV, which means that the degree of neural excitation is enhanced compared to the resting state.

[0083] Next, the standard deviation and mean of all data in the real-time neural electrical activity sequence are calculated, and then the ratio of the standard deviation to the mean is calculated to obtain the neural electrical activity fluctuation coefficient. The standard deviation reflects the dispersion of the data, the mean reflects the average level of the data, and the ratio of the two can quantify the fluctuation stability of neural electrical activity.

[0084] For example, first calculate the standard deviation and mean of all data in the real-time neural electrical activity sequence, then calculate the ratio of the standard deviation to the mean as the neural electrical activity fluctuation coefficient. For instance, if the standard deviation of all data in the real-time neural electrical activity sequence is calculated to be 0.11 and the mean is 0.9, then the neural electrical activity fluctuation coefficient = 0.11 / 0.9 ≈ 0.12. The higher the value of the neural electrical activity fluctuation coefficient, the greater the dispersion and instability of the neural electrical activity amplitude under thermal stimulation, and the more intense the response to thermal stimulation; conversely, the lower the value of the neural electrical activity fluctuation coefficient, the more concentrated and stable the neural electrical activity amplitude, the smoother the response to thermal stimulation, and the better the physiological adaptability of neural electrical activity.

[0085] Finally, the product of the neural electrical activity response amplitude and the neural electrical activity fluctuation coefficient is calculated, and then divided by the temperature change to obtain the temperature-neural electrical activity sensitivity coefficient. The formula for calculating the temperature-neural electrical activity sensitivity coefficient is: Temperature-neural electrical activity sensitivity coefficient = Neural electrical activity response amplitude × Neural electrical activity fluctuation coefficient / Temperature change.

[0086] For example, if the amplitude of the neural electrical activity response is 0.5 μV, the fluctuation coefficient of the neural electrical activity is 0.12, and the temperature change is 12 °C, then the temperature-neural electrical activity sensitivity coefficient = 0.5 × 0.12 / 12 = 0.005. The temperature-neural electrical activity sensitivity coefficient integrates the intensity of neural excitation and the stability of neural activity, and is quantified by combining the intensity of temperature change. It can accurately reflect the degree of influence of a unit temperature change on the nervous system. The higher the temperature-neural electrical activity sensitivity coefficient, the stronger the sensitivity of the user's nervous system to thermal stimulation; conversely, the lower the value, the weaker the sensitivity.

[0087] In summary, compared to existing technologies, this application calculates an individual sensitivity coefficient based on the dynamic differences between resting and real-time skin blood perfusion, and between resting and real-time neural electrical activity at each wearing site. Thus, the individual sensitivity coefficient integrates the personalized sensitivity characteristics of the skin blood flow system and the nervous system to thermal stimulation, overcoming the limitations of a single physiological dimension and comprehensively reflecting the overall sensitivity of the target user's wearing site to thermal stimulation.

[0088] S50: Obtain the group sensitivity coefficient of a group with similar individual characteristics to the target user, and combine the individual sensitivity coefficient to correct the initial temperature, initial pressure, and initial frequency to obtain the corrected temperature, corrected pressure, and corrected frequency.

[0089] Traditional thermal stimulation parameter adjustments rely solely on individual real-time physiological responses or single-experience settings, lacking common group patterns as a reference. This makes it difficult to accurately balance individual sensitivity differences with universal suitability. Therefore, it is necessary to introduce the group sensitivity coefficient of similar groups of target users and combine it with their own individual sensitivity coefficients to correct the initial parameters. This will enable parameter adjustments to conform to both individual characteristics and group patterns, thereby improving the accuracy of adaptation.

[0090] To address the aforementioned issues, this application obtains the group sensitivity coefficient of a group with individual characteristics similar to the target user, and uses the individual sensitivity coefficient to correct the initial temperature, initial pressure, and initial frequency to obtain corrected temperature, corrected pressure, and corrected frequency.

[0091] Specifically, step S50 in the method includes: Based on the individual characteristics of the target users, user groups with feature similarity greater than or equal to a preset similarity threshold are selected from the group database as similar groups with similar characteristics to the target users; Extract the individual sensitivity coefficients of each wearing part in the similar groups, and calculate the arithmetic mean to obtain the group sensitivity coefficient of each wearing part; For each wearing part, the relative deviation rate between the individual sensitivity coefficient and the group sensitivity coefficient of the target user is calculated and used as a correction coefficient; For each wearing part, based on the correction coefficient, the initial temperature, initial pressure, and initial frequency are dynamically adjusted to obtain the corrected temperature, corrected pressure, and corrected frequency.

[0092] In this embodiment, firstly, based on the individual characteristics of the target user, a user group with a feature similarity greater than or equal to a preset similarity threshold is selected from the group database as a similar group with similar characteristics to the target user. The group database stores a large amount of historical user data, including the individual characteristics of historical users and their corresponding individual sensitivity coefficients, and is regularly updated to supplement new user data, ensuring the timeliness of group patterns. The preset similarity threshold can be dynamically adjusted according to the data size of the group database and the required parameter control precision, flexibly adapting to different application scenarios.

[0093] For example, feature similarity can be calculated based on weighted matching of various indicators of individual characteristics. For instance, the weights of age, gender, BMI, and body type can be set to 0.2, 0.2, 0.3, and 0.3, respectively. The cosine similarity between the target user's individual characteristic indicators and the corresponding indicators of historical users can be calculated, and then summed according to the aforementioned preset weights to obtain the feature similarity between the two. If a preset similarity threshold of 80% is set, historical users with feature similarity ≥80% can be selected from the group database to form a similar group with characteristics similar to the target user. This selection logic ensures both the consistency of features between the similar group and the target user and that the group size meets statistical requirements, thereby accurately reflecting the common patterns of sensitivity to thermal stimuli among users with this type of characteristic.

[0094] Secondly, individual sensitivity coefficients for each wearing site within similar groups were extracted, and their arithmetic mean was calculated to obtain the group sensitivity coefficient for each wearing site. The group sensitivity coefficient is a common sensitivity indicator for similar groups across different wearing sites, reflecting the average response level of similar groups to thermal stimuli.

[0095] For example, for each wearing part, the individual sensitivity coefficients of all users in the similar group are extracted, and then the arithmetic mean of these individual sensitivity coefficients is calculated to obtain the group sensitivity coefficient. By calculating the arithmetic mean, the random differences between individuals can be effectively offset, reflecting the common response characteristics of the group and providing a reference benchmark for individual parameter correction.

[0096] Next, for each wearing part, the relative deviation rate between the individual sensitivity coefficient and the group sensitivity coefficient of the target user is calculated as a correction coefficient. Among them, the correction coefficient is the core indicator for quantifying the difference between individuals and groups. The formula for calculating the correction coefficient is: Correction coefficient = (Individual sensitivity coefficient of the target user - Group sensitivity coefficient) / Group sensitivity coefficient.

[0097] For example, if the individual sensitivity coefficient of the target user is 0.033 and the group sensitivity coefficient is 0.035, then the correction coefficient = (0.033 - 0.035) / 0.035 = -0.002. A positive correction coefficient indicates that the target user's sensitivity to thermal stimulation is higher than the average level of similar groups; a negative correction coefficient indicates that the target user's sensitivity is lower than the average level of similar groups. The larger the absolute value of the correction coefficient, the more significant the difference in sensitivity between the target user and similar groups, and the greater the subsequent adjustment range of the thermal stimulation parameters needs to be to adapt to the personalized needs of the target user.

[0098] Finally, for each wearing location, the initial temperature, initial pressure, and initial frequency are dynamically adjusted based on correction coefficients to obtain the corrected temperature, corrected pressure, and corrected frequency. Specifically, corrected temperature = initial temperature × (1 - correction coefficient), corrected pressure = initial pressure × (1 - correction coefficient), and corrected frequency = initial frequency × (1 - correction coefficient). A positive correction coefficient indicates that the target user's sensitivity to thermal stimulation is higher than the average level of similar groups, requiring reverse correction to lower the parameter values ​​to avoid excessive stimulation. A negative correction coefficient indicates that the target user's sensitivity is lower than the average level of similar groups, requiring positive correction to increase the parameter values ​​to enhance the effectiveness of the stimulation. The larger the absolute value of the correction coefficient, the greater the parameter adjustment range.

[0099] For example, if the correction factor is -0.002, the initial temperature is 37℃, the initial pressure is 2.0kPa, and the initial frequency is 3.8Hz, then the corrected temperature = 37℃ × (1 + 0.002) = 37.074℃, the corrected pressure = 2.0kPa × (1 + 0.002) = 2.004kPa, and the corrected frequency = 3.8Hz × (1 + 0.002) = 3.8076Hz. By dynamically adjusting the correction factor, the corrected temperature, pressure, and frequency are better adapted to individual differences, while relying on group patterns to ensure safety, thereby achieving precise thermal stimulation.

[0100] In summary, compared to existing technologies, this application obtains the group sensitivity coefficient of a group with similar individual characteristics to the target user, and combines the individual sensitivity coefficient to correct the initial temperature, initial pressure, and initial frequency, resulting in corrected temperature, corrected pressure, and corrected frequency. Thus, by fusing the group sensitivity coefficient of similar groups with the individual sensitivity coefficient of the target user to correct the initial thermal stimulation parameters, it takes into account both common group patterns and individual sensitivity differences, achieving personalized and precise optimization of thermal stimulation parameters, and effectively improving the adaptability and stability of emotion regulation.

[0101] S60: Continuously apply thermal stimulation to each wearing part according to the corrected temperature, corrected pressure, and corrected frequency until the preset emotion regulation target is met.

[0102] The aforementioned steps involve calculating the individual sensitivity coefficient in conjunction with the group sensitivity coefficient of similar groups to obtain the corrected temperature, corrected pressure, and corrected frequency that are tailored to the personalized characteristics of the target user. Based on this, thermal stimulation can be continuously applied to each wearing part.

[0103] Based on the aforementioned adjustments to temperature, pressure, and frequency, this application continuously applies thermal stimulation to each wearing area until a preset emotion regulation target is met. For example, the preset emotion regulation target can be set as a dual standard: physiologically, real-time skin blood perfusion should return to a stable range of ±5% of resting skin blood perfusion, and neural electrical activity should remain within the normal fluctuation range corresponding to resting neural electrical activity, without excessive excitation or inhibition; subjectively, the target user can actively report their emotional state through the device's interactive interface, confirming that they have transitioned from negative states such as anxiety and irritability to calmness. Only when both physiological and subjective standards are met simultaneously is the preset emotion regulation target deemed met, effectively avoiding misjudgments of regulation effect caused by a single indicator.

[0104] For example, during the continuous execution of thermal stimulation, the aforementioned preset acquisition frequency of 3 seconds / cycle is maintained to capture the skin blood perfusion and neural electrical activity data of each wearing site in real time, and continuously compare them with the preset emotion regulation target until both standards are met, and then the thermal stimulation is stopped.

[0105] In this way, the heat stimulation can be tailored to the individual sensitivity characteristics of the target users, while avoiding the risks of insufficient or excessive stimulation, ultimately achieving a precise, safe and long-lasting effect on mood regulation.

[0106] In summary, the embodiments of this application have at least the following technical effects: Compared to existing technologies, this application first collects the individual characteristics of the target user and the resting skin blood perfusion and resting neural electrical activity at each wearing site of the distributed thermal stimulation device. This obtains precise individual characteristic data and resting physiological data, providing a precise data foundation for the subsequent individualized determination of stimulation parameters.

[0107] Secondly, this application determines the initial temperature based on the individual characteristics, inputs the initial temperature and individual characteristics into a pre-trained pressure-frequency prediction model, and outputs the initial pressure and initial frequency. In this way, the deep coupling relationship between temperature, pressure, and frequency during thermal stimulation is fully considered, and the initial control parameter combination adapted to the target user is accurately output, avoiding parameter contradictions or insufficient adaptation caused by empirical settings, thus ensuring the safety and effectiveness of thermal stimulation.

[0108] Furthermore, this application activates the distributed thermal stimulation device to apply thermal stimulation to each wearing site according to the initial temperature, initial pressure, and initial frequency, while simultaneously collecting real-time skin blood perfusion and real-time neural electrical activity at each wearing site. This allows for the accurate recording of the dynamic changes in physiological indicators under thermal stimulation, avoiding data distortion caused by asynchrony between stimulation and data collection. The collected real-time skin blood perfusion and real-time neural electrical activity will serve as a reliable basis for subsequent calculations of individual sensitivity coefficients and assessment of emotion regulation effects, providing direct real-time feedback support for parameter correction.

[0109] Furthermore, this application calculates an individual sensitivity coefficient based on the dynamic differences between resting and real-time skin blood perfusion, and between resting and real-time neural electrical activity at each wearing site. Thus, the individual sensitivity coefficient integrates the personalized sensitivity characteristics of the skin blood flow system and the nervous system to thermal stimulation, overcoming the limitations of a single physiological dimension and comprehensively reflecting the overall sensitivity of the target user's wearing site to thermal stimulation.

[0110] Furthermore, this application obtains the group sensitivity coefficient of a group with similar individual characteristics to the target user, and uses the individual sensitivity coefficient to correct the initial temperature, initial pressure, and initial frequency to obtain corrected temperature, corrected pressure, and corrected frequency. Thus, by incorporating the group sensitivity coefficient of similar groups for correction, both common group patterns and individual sensitivity differences are considered, achieving personalized and precise optimization of thermal stimulation parameters, effectively improving the adaptability and stability of emotion regulation.

[0111] Finally, this application continuously applies thermal stimulation to each wearing part according to the modified temperature, modified pressure, and modified frequency until the preset emotion regulation target is met. In this way, it can ensure that the thermal stimulation is both in line with the individual sensitivity characteristics of the target user and avoid the risks of insufficient or excessive stimulation, ultimately achieving a precise, safe, and long-lasting emotion regulation effect.

[0112] Through the above technical solution, this application determines initial parameters by collecting individual characteristics of the target user and the resting skin blood perfusion and resting neural electrical activity at each wearing site. During the application of thermal stimulation, real-time physiological data is simultaneously captured and individual sensitivity coefficients are calculated. The initial temperature, initial pressure, and initial frequency are then adjusted by combining the group sensitivity coefficients of similar groups before continuous application until the preset emotion regulation goal is achieved, forming a complete closed loop. This effectively overcomes the crudeness of traditional thermal stimulation regulation, achieving precise adaptation of emotion regulation to the individual characteristics and physiological sensitivity differences of the target user, improving the stability and reliability of the regulatory effect, and meeting the needs for personalized and precise emotion regulation.

[0113] Example 2, as Figure 2As shown, based on the same inventive concept as the distributed thermal stimulation-based emotion regulation method provided in Embodiment 1, this embodiment of the invention also provides a distributed thermal stimulation-based emotion regulation system, including: The resting vital signs data acquisition module 11 is used to collect the individual characteristics of the target user and the resting skin blood perfusion and resting nerve electrical activity of each wearing site of the distributed thermal stimulation device; The initial stimulus parameter determination module 12 is used to determine the initial temperature based on the individual characteristics, input the initial temperature and individual characteristics into a pre-trained pressure-frequency prediction model, and output the initial pressure and initial frequency. The real-time vital signs data acquisition module 13 is used to start the distributed thermal stimulation device, perform thermal stimulation on each wearing part according to the initial temperature, initial pressure and initial frequency, and simultaneously collect the real-time skin blood perfusion and real-time nerve electrical activity of each wearing part. Sensitivity coefficient calculation module 14 is used to calculate the individual sensitivity coefficient based on the dynamic differences between resting skin blood perfusion and real-time skin blood perfusion, and between resting neural electrical activity and real-time neural electrical activity at each wearing site. The stimulus parameter correction module 15 is used to obtain the group sensitivity coefficient of a group with similar individual characteristics to the target user, and to correct the initial temperature, initial pressure and initial frequency in combination with the individual sensitivity coefficient to obtain the corrected temperature, corrected pressure and corrected frequency. The output execution module 16 is used to continuously apply thermal stimulation to each wearing part according to the corrected temperature, corrected pressure and corrected frequency until the preset emotion regulation target is met.

[0114] The resting vital signs data acquisition module 11 is specifically used for: Collect individual characteristics of the target users, including age, gender, BMI, and body type; When the target user is in a calm state with a stable heart rate and breathing, the resting skin blood perfusion and resting neural electrical activity of the corresponding wearing area are collected by each wearing unit of the distributed thermal stimulation device.

[0115] The initial stimulus parameter determination module 12 is specifically used for: Based on the individual characteristics of the target users, and combined with the preset individual characteristic-temperature mapping rules, the initial temperature is determined; The pre-trained pressure-frequency prediction model is invoked, wherein the pressure-frequency prediction model is based on a multi-output gradient boosting regression framework and has a built-in temperature-pressure coupling module and a temperature-frequency coupling module. The initial temperature and individual characteristics are input into a pre-trained pressure-frequency prediction model. The model is predicted using the basic framework, and then verified and corrected by the temperature-pressure coupling module and the temperature-frequency coupling module to output the initial pressure and initial frequency.

[0116] Specifically, the construction process of the "pressure frequency prediction model" includes: A multi-output gradient boosting regression model is adopted, with two prediction channels set up as the basic framework for the pressure frequency prediction model; Historical input features from different users are collected from historical data to form a sample dataset; For the aforementioned sample dataset, the optimal pressure that enables the skin blood perfusion to meet the standard is collected and labeled to form a sample pressure label set. Simultaneously, the optimal frequency that enables the neural electrical activity to meet the standard is collected and labeled to form a sample frequency label set. Two prediction channels are trained using the aforementioned sample dataset and the sample pressure label set and sample frequency label set, respectively, until convergence is verified. After verification of convergence, a pre-built temperature-pressure coupling module and a temperature-frequency coupling module are embedded as a physiological logic verification layer to obtain the pressure-frequency prediction model.

[0117] Furthermore, the construction steps of the "temperature-pressure coupling module and temperature-frequency coupling module" include: Collect temperature parameters within the preset effective temperature range, and collect the corresponding effective pressure data and effective frequency data; The preset effective temperature range is divided into multiple temperature sub-ranges. The temperature parameters and effective pressure data in each temperature sub-range are fitted by linear regression to obtain the temperature-pressure coupling formula for each sub-range, and then integrated to form a temperature-pressure coupling module. Linear regression is used to backfit the temperature parameters and effective frequency data within each temperature sub-interval to obtain the temperature-frequency coupling formula for each interval, and then integrates them to form a temperature-frequency coupling module.

[0118] The real-time vital sign data acquisition module 13 is specifically used for: The distributed thermal stimulation device is activated, and thermal stimulation is applied to each wearing site according to the initial temperature, initial pressure, and initial frequency. Simultaneously, the real-time skin blood perfusion and real-time nerve electrical activity of each wearing site are collected.

[0119] Specifically, the sensitivity coefficient calculation module 14 is used for: The ambient temperature before the thermal stimulus is activated is obtained, and the temperature change is calculated by combining the initial temperature. For each wearing site, the rate of change of skin blood perfusion and the response delay time of blood perfusion were calculated between resting skin blood perfusion and real-time skin blood perfusion. Combined with the temperature change, the temperature-skin blood perfusion sensitivity coefficient was calculated. For each wearing site, the neural electrical activity response amplitude and neural electrical activity fluctuation coefficient of resting neural electrical activity and real-time neural electrical activity are calculated respectively. Combined with the temperature change, the temperature-neural electrical activity sensitivity coefficient is calculated. Based on preset weights, the weighted average of the temperature-skin blood perfusion sensitivity coefficient and the temperature-nerve electrical activity sensitivity coefficient is calculated to obtain the individual sensitivity coefficient for each wearing site.

[0120] Specifically, the phrase "for each wearing site, calculating the rate of change of skin blood perfusion volume between resting and real-time skin blood perfusion volume, the blood perfusion volume response delay time, and combining this with the temperature change to calculate the temperature-skin blood perfusion sensitivity coefficient" includes: For each wearing site, the real-time skin blood perfusion volume is obtained for a continuous preset collection period, and the arithmetic mean is calculated as a real-time statistical value; The deviation between the real-time statistical value and the resting skin blood perfusion is calculated as the rate of change of skin blood perfusion; The time from the moment the thermal stimulation was initiated until the real-time statistical value first exceeded the preset proportion of resting skin blood perfusion was recorded as the blood perfusion response delay time. The temperature-skin blood perfusion sensitivity coefficient is obtained by multiplying the rate of change of skin blood perfusion volume by the response delay time of blood perfusion volume and then dividing by the temperature change.

[0121] Specifically, the phrase "for each wearing site, calculating the neural electrical activity response amplitude and neural electrical activity fluctuation coefficient of resting and real-time neural electrical activity, and combining this with the temperature change to calculate the temperature-neural electrical activity sensitivity coefficient" includes: For each wearing site, real-time neural electrical activity data are recorded within a continuously preset acquisition cycle after thermal stimulation is performed, forming a real-time neural electrical activity sequence; The peak value of real-time neural electrical activity is extracted from the real-time neural electrical activity sequence, and the difference between the peak value of real-time neural electrical activity and the resting neural electrical activity is calculated as the neural electrical activity response amplitude. Calculate the standard deviation and mean of all data in the real-time neural electrical activity sequence, and then calculate the ratio of the standard deviation to the mean to obtain the neural electrical activity fluctuation coefficient; The temperature-neural electrical activity sensitivity coefficient is obtained by multiplying the amplitude of the neural electrical activity response by the fluctuation coefficient of the neural electrical activity and then dividing by the temperature change.

[0122] The stimulation parameter correction module 15 is specifically used for: Based on the individual characteristics of the target users, user groups with feature similarity greater than or equal to a preset similarity threshold are selected from the group database as similar groups with similar characteristics to the target users; Extract the individual sensitivity coefficients of each wearing part in the similar groups, and calculate the arithmetic mean to obtain the group sensitivity coefficient of each wearing part; For each wearing part, the relative deviation rate between the individual sensitivity coefficient and the group sensitivity coefficient of the target user is calculated and used as a correction coefficient; For each wearing part, based on the correction coefficient, the initial temperature, initial pressure, and initial frequency are dynamically adjusted to obtain the corrected temperature, corrected pressure, and corrected frequency.

[0123] Specifically, the output execution module 16 is used for: The temperature, pressure, and frequency are adjusted as described above to continuously apply thermal stimulation to each wearing part until the preset emotion regulation target is met.

[0124] In summary, the embodiments of this application have at least the following technical effects: Compared to existing technologies, this application firstly acquires individual characteristics of the target user and resting skin blood perfusion and resting neural electrical activity at each wearing site of the distributed thermal stimulation device through a resting vital signs data acquisition module. This obtains accurate individual characteristic data and resting physiological data, providing a precise data foundation for the individualized determination of subsequent stimulation parameters. Secondly, through an initial stimulation parameter determination module, the initial temperature is determined based on individual characteristics. The initial temperature and individual characteristics are input into a pre-trained pressure-frequency prediction model, which outputs initial pressure and initial frequency. This fully considers the deep coupling relationship between temperature, pressure, and frequency during thermal stimulation and accurately outputs a combination of initial control parameters adapted to the target user. Thirdly, through a real-time vital signs data acquisition module, the distributed thermal stimulation device is activated, and thermal stimulation is performed on each wearing site according to the initial temperature, initial pressure, and initial frequency. Simultaneously, real-time skin blood perfusion and real-time neural electrical activity at each wearing site are acquired, providing direct real-time feedback support for parameter correction. Next, the sensitivity coefficient calculation module calculates individual sensitivity coefficients based on the dynamic differences between resting and real-time skin blood perfusion, and between resting and real-time neural electrical activity at each wearing site. This comprehensively reflects the overall sensitivity of the target user's wearing site to thermal stimulation. Furthermore, the stimulation parameter correction module obtains the group sensitivity coefficients of a group with similar characteristics to the target user. These individual sensitivity coefficients are then used to correct the initial temperature, initial pressure, and initial frequency, resulting in corrected temperature, corrected pressure, and corrected frequency. By integrating the group sensitivity coefficients of similar groups, the adaptability and stability of emotion regulation are effectively improved. Finally, the output execution module continuously applies thermal stimulation to each wearing site according to the corrected temperature, corrected pressure, and corrected frequency until the preset emotion regulation target is met, achieving precise, safe, and long-lasting emotion regulation effects.

[0125] In this way, the shortcomings of traditional heat stimulation regulation are effectively compensated for, and the emotion regulation is accurately adapted to the individual characteristics and physiological sensitivity differences of the target users. This improves the stability and reliability of the regulation effect and meets the needs of personalized and precise emotion regulation.

[0126] It should be noted that the descriptions of each embodiment in the above embodiments have different focuses. For parts that are not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0127] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0128] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded computer, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0129] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0130] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0131] Although preferred embodiments of the invention have been described, those skilled in the art, once they have learned the basic inventive concept, can make other changes and modifications to these embodiments.

[0132] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of this invention and its equivalents, this invention also intends to include these modifications and variations.

Claims

1. A method for emotion regulation based on distributed thermal stimulation, characterized in that, The method comprises the steps of: Collecting individual characteristics of a target user and resting skin blood perfusion and resting neural electrical activity of each wearing part of a distributed thermal stimulation device; Determining an initial temperature according to the individual characteristics, inputting the initial temperature and individual characteristics into a pre-trained pressure-frequency prediction model, and outputting an initial pressure and an initial frequency; Starting the distributed thermal stimulation device, performing thermal stimulation on each wearing part at the initial temperature, initial pressure, and initial frequency, and synchronously collecting real-time skin blood perfusion and real-time neural electrical activity of each wearing part; Calculating individual sensitivity coefficients based on the dynamic differences between resting skin blood perfusion and real-time skin blood perfusion and between resting neural electrical activity and real-time neural electrical activity of each wearing part; Obtaining group sensitivity coefficients of a group similar to the individual characteristics of the target user, correcting the initial temperature, initial pressure, and initial frequency in combination with the individual sensitivity coefficients, and obtaining a corrected temperature, a corrected pressure, and a corrected frequency; Continuously performing thermal stimulation on each wearing part at the corrected temperature, corrected pressure, and corrected frequency until a preset emotion regulation target is met.

2. The mood regulation method based on distributed thermal stimulation according to claim 1, characterized in that, The method comprises the steps of: Collecting individual characteristics of a target user and resting skin blood perfusion and resting neural electrical activity of each wearing part of a distributed thermal stimulation device, which comprises: Collecting individual characteristics of the target user, wherein the individual characteristics include age, gender, BMI, and physical type; 3.The distributed thermal stimulation based emotion regulation method of claim 1, wherein, When the target user is in a calm state with stable heart rate and breathing, collecting resting skin blood perfusion and resting neural electrical activity of each wearing part through each wearing unit of the distributed thermal stimulation device. Determining an initial temperature according to the individual characteristics, inputting the initial temperature and individual characteristics into a pre-trained pressure-frequency prediction model, and outputting an initial pressure and an initial frequency, which comprises: Determining an initial temperature based on the individual characteristics of the target user in combination with a preset individual characteristic-temperature mapping rule; Calling a pre-trained pressure-frequency prediction model, wherein the pressure-frequency prediction model is based on a multi-output gradient boosting regression framework and internally embedded with a temperature-pressure coupling module and a temperature-frequency coupling module; 4. The mood regulation method based on distributed thermal stimulation according to claim 3, characterized in that, Inputting the initial temperature and individual characteristics into the pre-trained pressure-frequency prediction model for prediction through the basic framework, and then verifying and correcting through the temperature-pressure coupling module and the temperature-frequency coupling module to output an initial pressure and an initial frequency. The construction process of the pressure-frequency prediction model comprises: Using a multi-output gradient boosting regression model to set two prediction channels as the basic framework of the pressure-frequency prediction model; Collecting historical input characteristics of different users from historical data to form a sample data set; For the sample data set, collecting and labeling the optimal pressure that can make the skin blood perfusion meet the standard to form a sample pressure label set, and synchronously collecting and labeling the optimal frequency that can make the neural electrical activity meet the standard to form a sample frequency label set; Training the two prediction channels using the sample data set, the sample pressure label set, and the sample frequency label set respectively until verification convergence is achieved; After verification convergence is achieved, embedding a pre-constructed temperature-pressure coupling module and a temperature-frequency coupling module as a physiological logic verification layer to obtain the pressure-frequency prediction model.

5. The mood regulation method based on distributed thermal stimulation according to claim 4, characterized in that, The construction steps of the temperature-pressure coupling module and the temperature-frequency coupling module include: Collecting temperature parameters in a preset temperature effective range, and collecting corresponding effective pressure data and effective frequency data; Dividing the preset temperature effective range into multiple temperature sub-ranges, and performing linear regression forward fitting on the temperature parameters and the effective pressure data in each temperature sub-range to obtain a temperature-pressure coupling formula in each sub-range, and integrating to form a temperature-pressure coupling module; Performing linear regression reverse fitting on the temperature parameters and the effective frequency data in each temperature sub-range to obtain a temperature-frequency coupling formula in each sub-range, and integrating to form a temperature-frequency coupling module.

6. The mood regulation method based on distributed thermal stimulation according to claim 1, characterized in that, Based on the dynamic differences between the resting skin blood perfusion and the real-time skin blood perfusion, and the dynamic differences between the resting neural electrical activity and the real-time neural electrical activity of each wearing part, the individual sensitivity coefficient is calculated, including: Obtaining the ambient temperature before the start of the thermal stimulation, and combining the initial temperature to calculate the temperature change amount; For each wearing part, the skin blood perfusion change rate and the blood perfusion response delay time of the resting skin blood perfusion and the real-time skin blood perfusion are calculated respectively, and the temperature-skin blood perfusion sensitivity coefficient is calculated by combining the temperature change amount; For each wearing part, the neural electrical activity response amplitude and the neural electrical activity fluctuation coefficient of the resting neural electrical activity and the real-time neural electrical activity are calculated respectively, and the temperature-neural electrical activity sensitivity coefficient is calculated by combining the temperature change amount; According to a preset weight, the weighted average of the temperature-skin blood perfusion sensitivity coefficient and the temperature-neural electrical activity sensitivity coefficient is calculated to obtain the individual sensitivity coefficient of each wearing part.

7. The mood regulation method based on distributed thermal stimulation according to claim 6, characterized in that, For each wearing part, the skin blood perfusion change rate and the blood perfusion response delay time of the resting skin blood perfusion and the real-time skin blood perfusion are calculated respectively, and the temperature-skin blood perfusion sensitivity coefficient is calculated by combining the temperature change amount, including: For each wearing part, the real-time skin blood perfusion in a continuous preset collection period is obtained, and the arithmetic mean value is calculated as a real-time statistical value; The deviation amplitude of the real-time statistical value and the resting skin blood perfusion is calculated as the skin blood perfusion change rate; The length of time from the start of the thermal stimulation to the first time when the real-time statistical value exceeds the preset proportion of the resting skin blood perfusion is recorded as the blood perfusion response delay time; The product of the skin blood perfusion change rate and the blood perfusion response delay time is calculated, and then divided by the temperature change amount to obtain the temperature-skin blood perfusion sensitivity coefficient.

8. The mood regulation method based on distributed thermal stimulation according to claim 6, characterized in that, For each wearing part, the neural electrical activity response amplitude and the neural electrical activity fluctuation coefficient of the resting neural electrical activity and the real-time neural electrical activity are calculated respectively, and the temperature-neural electrical activity sensitivity coefficient is calculated by combining the temperature change amount, including: For each wearing part, the real-time neural electrical activity data in a continuous preset collection period after the execution of the thermal stimulation is recorded to form a real-time neural electrical activity sequence; The real-time neural electrical activity peak value is extracted from the real-time neural electrical activity sequence, and the difference between the real-time neural electrical activity peak value and the resting neural electrical activity is calculated as the neural electrical activity response amplitude; calculating a standard deviation and a mean of all data in the real-time neural electrical activity sequence, and then calculating a ratio of the standard deviation and the mean to obtain a neural electrical activity fluctuation coefficient; calculating a product of the neural electrical activity response amplitude and the neural electrical activity fluctuation coefficient, and then dividing the product by the temperature change amount to obtain a temperature-neural electrical activity sensitivity coefficient. 9.The distributed thermal stimulation based emotion regulation method of claim 1, wherein, obtaining a group sensitivity coefficient of a group similar to the individual characteristics of the target user, and correcting the initial temperature, the initial pressure, and the initial frequency based on the individual sensitivity coefficient to obtain a corrected temperature, a corrected pressure, and a corrected frequency, including: filtering, based on the individual characteristics of the target user, a user group with a feature similarity greater than or equal to a preset similarity threshold from a group database as a similar group similar to the individual characteristics of the target user; extracting individual sensitivity coefficients of each wearing part in the similar group, and calculating an arithmetic mean to obtain a group sensitivity coefficient of each wearing part; for each wearing part, calculating a relative deviation rate of the individual sensitivity coefficient of the target user and the group sensitivity coefficient as a correction coefficient; for each wearing part, dynamically adjusting the initial temperature, the initial pressure, and the initial frequency based on the correction coefficient to obtain a corrected temperature, a corrected pressure, and a corrected frequency.

10. An emotion regulation system based on distributed thermal stimulation, characterized in that, a method for performing the distributed thermal stimulation-based emotion regulation method of any one of claims 1-9, including: a resting vital sign data acquisition module configured to acquire individual characteristics of a target user and resting skin blood perfusion and resting neural electrical activity of each wearing part of a distributed thermal stimulation device; an initial stimulation parameter determination module configured to determine an initial temperature based on the individual characteristics, input the initial temperature and the individual characteristics into a pre-trained pressure-frequency prediction model, and output an initial pressure and an initial frequency; a real-time vital sign data acquisition module configured to start the distributed thermal stimulation device, perform thermal stimulation on each wearing part according to the initial temperature, the initial pressure, and the initial frequency, and synchronously acquire real-time skin blood perfusion and real-time neural electrical activity of each wearing part; a sensitivity coefficient calculation module configured to calculate an individual sensitivity coefficient based on dynamic differences between resting skin blood perfusion and real-time skin blood perfusion and between resting neural electrical activity and real-time neural electrical activity of each wearing part; a stimulation parameter correction module configured to obtain a group sensitivity coefficient of a group similar to the individual characteristics of the target user, and correct the initial temperature, the initial pressure, and the initial frequency based on the individual sensitivity coefficient to obtain a corrected temperature, a corrected pressure, and a corrected frequency; an output execution module configured to continuously perform thermal stimulation on each wearing part according to the corrected temperature, the corrected pressure, and the corrected frequency until a preset emotion regulation target is met.

Images