A full-semester portable self-help intelligent psychological counseling and sleep intervention method and system

Abstract

The application relates to the technical field of intelligent psychological intervention, in particular to a full-semester portable self-service intelligent psychological guidance and sleep intervention method and system, which comprises a learning period identifier, and a cognitive load limiting parameter is called; resting physiological data is collected as an initial baseline; an operation scene is obtained; if it is a special adjustment scene, corresponding audio-visual guidance content is output; if it is a sleep improvement scene, voice interaction is shielded and sleep aid audio and video are output; in the output process, real-time physiological data is collected, and a physiological fluctuation difference value is calculated; in the special adjustment scene, if the physiological fluctuation difference value increases in a preset time window, a process record is generated, the audio speed and / or picture dynamic rate of the current output are reduced, and deep breathing guidance is switched to; if the physiological fluctuation difference value decreases and reaches a preset relaxation threshold, the current output content is maintained and positive reinforcement intervention is performed. The application prevents cognitive mismatch and reverse anxiety during self-service intervention through cognitive load pre-restraint and active degradation control.

Description

Technical Field

[0001] This application relates to the technical field of intelligent psychological intervention, and in particular to a portable self-service intelligent psychological counseling and sleep intervention method and system for all school levels. Background Technology

[0002] With the increasing psychological stress across all age groups in modern society, intelligent psychological intervention technologies are being applied more and more widely in daily health management. To meet the public's need for emotional regulation in self-help settings such as during school breaks or at home, where they are not under professional supervision, portable psychological counseling and sleep intervention devices are gradually becoming important auxiliary tools. These devices typically provide specific audiovisual guidance content, combined with physiological state monitoring, attempting to help users autonomously soothe their mental state and improve the quality of their daily rest in a non-therapeutic environment.

[0003] Existing intelligent mental state regulation technologies typically incorporate closed-loop feedback mechanisms to dynamically intervene in the user's physiological state. For example, a typical technical solution is disclosed in Chinese invention patent literature (Publication No.: CN114081490B, Patent Title: A System and Method for Mental State Monitoring and Treatment Based on Closed-Loop Feedback). This system integrates an electroencephalogram (EEG) signal acquisition module and related assessment models, enabling the classification and assessment of the user's mental state before treatment, and then selecting either music therapy or magnetic therapy modes accordingly. During the specific treatment process, the system continuously collects the user's EEG signals in real time, and the processor adjusts the intervention parameters based on the newly acquired real-time data, aiming to achieve the intervention goal through this dynamic acquisition and adjustment method.

[0004] However, in actual self-help psychological counseling, the cognitive abilities of different users often vary greatly due to their age or educational level. Existing devices severely lack a pre-constraint mechanism based on the upper limit of individual cognitive abilities during the intervention initiation phase. This leads to a situation where, in a typical scenario, when a younger user attempts to use the device for relaxation during recess, the system may directly output mindfulness guidance content containing complex, high-order instructions or high-frequency dynamic images. Because this audiovisual content far exceeds the comprehension and cognitive load of younger users, it not only cannot be effectively accepted but also causes cognitive mismatch, triggering significant stress fluctuations in the body. At this point, because existing systems are accustomed to optimizing routine parameters and lack bidirectional dynamic collaborative logic for real-time physiological feedback, they struggle to perceive this worsening of physiological stress caused by instruction overload, and they lack a mechanism to proactively downgrade the guidance content to prevent reverse stress when the user's physiological fluctuations abnormally increase. Finally, this continuous one-way output intervention process, lacking proactive downgrading and fallback control, not only easily leads to the failure of this non-therapeutic self-help psychological counseling but also further exacerbates the user's reverse anxiety. Summary of the Invention

[0005] In order to prevent cognitive mismatch and reverse anxiety during self-help intervention through cognitive load pre-constraint and active degradation control, this application provides a portable self-help intelligent psychological counseling and sleep intervention method and system for all school levels.

[0006] Firstly, this application provides a portable, self-service intelligent psychological counseling and sleep intervention method for all educational levels, employing the following technical solution: A portable, self-service intelligent psychological counseling and sleep intervention method for all educational levels, comprising: Receive the input grade segment identifier, call the cognitive load limitation parameters mapped to the grade segment identifier; collect resting physiological data including initial heart rate, initial skin conductance response, initial respiratory rate and initial muscle tension as the initial baseline; Selected operation scenario: If it is a special adjustment scenario, output the corresponding audio-visual guidance content according to the cognitive load limitation parameters; if it is a sleep improvement scenario, disable voice interaction and output sleep-aiding audio and video. During the output process, real-time physiological data is collected, and the relative rate of change of various indicators, including heart rate, skin conductance, respiratory rate, and muscle tension, relative to the corresponding indicators of the initial baseline is calculated. The physiological fluctuation difference is then obtained by weighting and summing these values ​​with their corresponding calculation weights. In the specific adjustment scenario, if the physiological fluctuation difference increases within the preset time window, a process record is generated, the current output audio speed and / or video dynamic range is reduced, and deep breathing guidance is switched; after switching, the calculation weight corresponding to the breathing frequency is temporarily reset to zero, and the weight is compensated and distributed to the calculation weight corresponding to the heart rate and skin conductance response according to a preset ratio, so as to continuously calculate the physiological fluctuation difference. In the specific adjustment scenario, if the physiological fluctuation difference decreases and reaches the preset relaxation threshold, the current output content is maintained and positive reinforcement intervention is performed.

[0007] Optional, also includes: When preset acquisition conditions are met, the user's terminal physiological data is acquired; The terminal physiological data is compared with the initial baseline to generate evaluation data, and the evaluation data, operation scenario and generated process record are associated and stored as an archive record; If the cumulative trigger frequency of process records in the historical archives reaches a preset frequency threshold, the cognitive load limitation parameter mapped to the corresponding learning segment identifier is updated based on the preset weight, so that it can be invoked when the learning segment identifier is received again.

[0008] Optionally, the step of outputting audiovisual guidance content based on the cognitive load limitation parameter includes: Extract the upper limit of the complexity of the guidance instructions from the cognitive load limiting parameters; If the upper limit of the complexity of the guidance instruction is less than the preset cognitive threshold, the audiovisual guidance content is set as a concrete breathing follow-up animation and a solid color visual scene, wherein the scaling rhythm of the concrete breathing follow-up animation matches the initial breathing frequency. If the upper limit of the complexity of the guidance instruction is not less than the cognitive threshold, the audiovisual guidance content will be set as mindfulness guidance audio and simulated visual scene.

[0009] Optionally, the step of blocking voice interaction and outputting sleep-aiding audio and video includes: Limit screen brightness and communication interaction frequency; Based on the initial baseline, output the corresponding virtual scene and sleep aid audio, wherein the initial amplitude modulation frequency of the sleep aid audio is matched with the initial breathing frequency: if the initial breathing frequency is higher than the upper limit of the preset sleep aid target range, the amplitude modulation frequency is slowly reduced by a preset step size during the output process until it falls into the target range; if it is not higher than the upper limit, the initial amplitude modulation frequency is maintained and output smoothly. During the output process, if the physiological fluctuation difference reaches the preset light sleep threshold, the brightness of the virtual scene is reduced by a preset slope until it is turned off, while only the sleep aid audio is played.

[0010] Optionally, after switching to deep breathing guidance, the method further includes: The physiological fluctuation difference is continuously calculated at a preset period; If the physiological fluctuation difference does not decrease within the preset time window, the current output screen is set to retain the solid color background of the deep breathing guidance visual elements, and the rhythm of the deep breathing guidance is reduced by a preset step size. The above steps are repeated until the physiological fluctuation difference is reduced to within the preset safe difference threshold, and the original calculation weight configuration is restored after exiting the deep breathing guidance.

[0011] Optionally, the cognitive load limitation parameter, which is updated based on a preset weight and mapped to the corresponding learning stage identifier, includes: Extract the maximum physiological fluctuation difference from each process record that constitutes the cumulative trigger frequency; Based on the extracted maximum physiological fluctuation difference, a stress representative value is calculated, and then the stress representative value is matched with a preset interval to obtain the corresponding attenuation coefficient as the preset weight. The cognitive load limitation parameters are attenuated according to the preset weights, and the discrete parameters are rounded down to overwrite the original parameters.

[0012] Optionally, the implementation of positive reinforcement intervention includes: Lock the audio and video parameters of the current output content and extract real-time muscle tension from real-time physiological data; If the real-time muscle tension reaches the preset tension threshold, psychological cues are output at a preset slope and mixed in when the current output audio exists. Dynamic visual features are generated based on the physiological fluctuation difference, and then the dynamic visual features are superimposed and rendered onto the currently output image.

[0013] Optionally, after collecting resting physiological data including initial heart rate, initial skin conductance response, initial respiratory rate, and initial muscle tension as an initial baseline, the method further includes: If the initial heart rate is greater than the upper limit of the preset heart rate range and the initial skin conductance response is greater than the preset skin conductance threshold, an initial pressure state identifier is generated. Before obtaining the operation scenario, the convergence coefficient corresponding to the initial pressure state identifier is called, and the cognitive load limit parameter is attenuated using the convergence coefficient, and the discrete parameters therein are rounded.

[0014] Optionally, the above steps may also include: Record the cumulative number of loops. If the preset number of loops is reached and the physiological fluctuation difference is not reduced to within the safe difference threshold, interrupt the current audio and video output. Play soothing audio on a preset frequency band and temporarily overwrite the cognitive load limit parameter for this intervention with the lower limit of minimalist constraints.

[0015] Secondly, this application provides a portable, self-service intelligent psychological counseling and sleep intervention system for all school levels, employing the following technical solution: A portable, self-service intelligent psychological counseling and sleep intervention system for all school levels, comprising: The baseline establishment module receives the input grade segment identifier, calls the cognitive load limitation parameters mapped to the grade segment identifier, and collects resting physiological data including initial heart rate, initial skin conductance response, initial respiratory rate, and initial muscle tension as the initial baseline. The intervention output module acquires the selected operation scenario: if it is a special adjustment scenario, it outputs the corresponding audio-visual guidance content according to the cognitive load limitation parameters; if it is a sleep improvement scenario, it blocks voice interaction and outputs sleep-aiding audio and video. The dynamic coordination module collects real-time physiological data during the output process, calculates the relative change rate of various indicators, including heart rate, skin conductance response, respiratory rate, and muscle tension, compared to the corresponding indicators of the initial baseline, and sums them with the corresponding calculation weights to obtain the physiological fluctuation difference. In the specific adjustment scenario, if the physiological fluctuation difference increases within a preset time window, a process record is generated, the audio speech rate and / or video dynamic rate of the current output are reduced, and deep breathing guidance is switched. After switching, the calculation weight corresponding to the respiratory rate is temporarily reset to zero, and compensated and distributed to the calculation weights corresponding to the heart rate and skin conductance response according to a preset ratio to continuously calculate the physiological fluctuation difference. In the specific adjustment scenario, if the physiological fluctuation difference decreases and reaches a preset relaxation threshold, the current output content is maintained and positive reinforcement intervention is performed.

[0016] In summary, this application includes the following beneficial technical effects: 1. This application introduces a pre-call mechanism for cognitive load limitation parameters based on school stage identifiers, which sets a safe boundary for users that matches their cognitive ability limit at the intervention initiation stage. This avoids the output of high-level guidance content that exceeds the scope of understanding from the source, effectively solving the core problem that existing technologies are prone to cognitive mismatch and initial stress due to the lack of cognitive pre-constraints. It significantly improves the adaptability and safety of psychological intervention in self-help scenarios across all school stages.

[0017] 2. This application constructs a two-way closed-loop dynamic control logic in specific adjustment scenarios, which can proactively perform downgrade adjustments and switch to deep breathing guidance when the difference in real-time physiological fluctuations increases. In particular, it introduces a dynamic weight compensation mechanism to prevent subjective interference. When the system takes over the user's breathing rhythm, it can intelligently remove false safety drop data caused by the user deliberately following breathing commands, and automatically shift the assessment focus proportionally to heart rate and skin conductance indicators that are not subject to subjective control. This completely breaks the stubborn problem of traditional one-way feedback systems being easily misled by local physiological movements, achieving accurate judgment and dynamic bottom-line relief of the user's true stress deterioration, and greatly mitigating the reverse anxiety dilemma that is easily induced during self-intervention.

[0018] 3. This application establishes a dynamic update mechanism for cognitive load limitation parameters based on the cumulative trigger frequency of process records in historical archives. This enables the initial constraint parameters of the intervention system to evolve individually with the user's long-term stress response data, ensuring that the intervention content continuously matches the user's cognitive ability and stress tolerance. This fundamentally overcomes the shortcomings of existing technologies where intervention parameters cannot be autonomously optimized and iterated. Attached Figure Description

[0019] Figure 1 This is a logical flowchart of the method in the embodiments of this application; Figure 2 This is a timing diagram of bidirectional dynamic coordinated regulation driven by physiological fluctuation difference in an embodiment of this application. Detailed Implementation

[0020] The following combination Figures 1-2 This application will be described in further detail.

[0021] This application relates to the technical field of intelligent psychological intervention, and more particularly to a portable, self-service intelligent psychological counseling and sleep intervention method for all school levels. In the prior art, closed-loop mental state intervention solutions, represented by the patent with authorization publication number CN114081490B, have achieved closed-loop control of physiological signal acquisition, mental state assessment, intervention mode selection, and real-time parameter adjustment. Treatment parameters can be optimized through EEG signal feedback to complete the intervention and regulation of mental state. Similar prior art includes cognitive load management systems based on skin conductance response and music regulation systems with multimodal physiological feedback. These solutions all revolve around the unidirectional positive optimization of physiological signals and intervention parameters, primarily targeting medical or professional guidance scenarios, and cannot adapt to self-service scenarios for all school levels without professional supervision, such as school breaks or home settings.

[0022] Existing technologies suffer from three insurmountable core defects in self-service applications: First, they lack a pre-constraint mechanism based on the user's cognitive capacity limits. Cognitive safety boundaries are not defined before intervention begins, and content can only be dynamically adjusted during intervention. This can easily lead to providing advanced guidance content beyond the comprehension of younger users, causing cognitive mismatch and initial stress. Second, the closed-loop adjustment logic only has a unidirectional positive optimization path, assuming that continuous optimization of intervention parameters is sufficient to improve effectiveness. It cannot identify the worsening of user stress caused by cognitive overload or inappropriate content, and lacks a proactive degradation mechanism to prevent reverse stress, easily leading to intervention failure or even exacerbating user anxiety. Third, they fail to establish a long-term evolutionary correlation between intervention process data and initial intervention parameters, making it impossible to continuously adapt individualized parameters based on the user's historical stress responses. The intervention system cannot autonomously optimize as it is used. This application addresses these defects by constructing a complete technical system through a full-process logical design, encompassing pre-emptive cognitive constraints, precise scenario adaptation, bidirectional closed-loop control, and long-term parameter evolution.

[0023] The hardware carrier used in this embodiment is a lightweight and portable device. The device adopts an integrated structure that supports both handheld and wearable operation, with an overall size not exceeding 15cm × 8cm × 2cm. It has a built-in rechargeable lithium battery that can support continuous intervention for no less than 8 hours when fully charged. The device integrates a sensor array for collecting multidimensional physiological data, a touch-screen interactive display, a voice acquisition and playback unit, a local processing unit, and a local storage unit.

[0024] The sensor suite includes a heart rate sensor, a skin conductance response sensor, a respiratory rate sensor, and a muscle tension sensor. All sensors have a sampling frequency of at least 100Hz to ensure the accuracy and real-time nature of physiological data acquisition. The corresponding detection contacts for each sensor are located in the device's grip area. Specifically, the heart rate and skin conductance response sensor contacts correspond to the grip positions of the middle and ring fingers, allowing users to collect data from resting or short-term states while holding the device. For scenarios requiring long-term continuous monitoring, such as sleep improvement, the device can be worn on the palm or wrist using the included flexible strap, ensuring stable contact with the skin and effective data collection during deep muscle relaxation after falling asleep. The muscle tension sensor connects to the device via adhesive electrode pads. It is recommended to attach the electrode pads to the upper trapezius muscle region on both sides, as this area shows the highest correlation between muscle tension and psychological stress, ensuring the validity of the collected data. The device allows users to create unique personal user identifiers. Each personal user identifier can be bound to a currently effective grade level identifier. Users can modify these identifiers according to their own circumstances. After modification, the device will automatically call the baseline parameters of the corresponding grade level and retain the historical records of the original grade level. All usage data is bound to the personal user identifier, avoiding data confusion in multi-user scenarios.

[0025] Combination Figure 1 As shown in the figure, this diagram intuitively illustrates the entire process logic of the proposed method: starting with matching educational stages and establishing a baseline, the cognitive boundaries are locked through initial cognitive constraint parameter calibration, followed by scenario determination and corresponding intervention output; during the intervention process, the system continuously performs 30-second physiological fluctuation calculations and feedback, thereby driving bidirectional dynamic collaborative regulation, and finally completing the final assessment and parameter historical evolution. This process intuitively demonstrates that the proposed method has a complete closed-loop control system that integrates pre-constraint, bidirectional regulation during the intervention, and long-term evolution after the intervention, effectively preventing the risks of cognitive mismatch and one-way intervention failure during self-help psychological intervention.

[0026] In response to the problem that existing self-help psychological intervention programs often adjust the content based on real-time feedback after the intervention is initiated, without pre-defining a safe boundary that matches the user's cognitive ability, which can easily lead to cognitive mismatch and initial anxiety, this embodiment first completes the locking of cognitive boundaries and the establishment of physiological benchmarks, thus laying a solid safety bottom line for the subsequent intervention process.

[0027] After the device is started, it first displays the school stage selection interface through the touch interactive display screen, and receives the school stage identifier entered by the user. The school stage identifiers here cover the entire school stage from primary school to junior high school, senior high school and university. At the same time, it sets up adaptive identifiers for adults and the elderly for all age groups, for a total of five categories.

[0028] The device's local storage unit contains a pre-set benchmark table of cognitive load limitation parameters, each mapped one-to-one with a specific identifier. Once the user selects the corresponding identifier, the device can directly retrieve the relevant benchmark parameters. If the user has historically updated parameters, the latest benchmark parameters after previous iterations will be prioritized. This set of parameters includes three core constraints: upper limit for the complexity of guidance instructions, upper limit for audio playback speed, and upper limit for the dynamic rate of visual scene elements. Each constraint is precisely limited according to the cognitive development characteristics of different user groups.

[0029] The upper limit of the complexity of guided instructions is quantified by the vocabulary and logical levels of a single instruction. For elementary school students, the upper limit is no more than 10 words and one logical level per instruction. This is based on the cognitive development characteristics of children aged 6-12, who primarily use concrete thinking and have limited working memory, ensuring that younger children can easily keep up with the content. For middle school students, the upper limit is no more than 15 words and two logical levels per instruction, aligning with the initial development of abstract thinking and gradually increasing working memory capacity in teenagers aged 12-15. For high school, university, adult, and senior citizens, the upper limit is no more than 20 words and three logical levels per instruction, consistent with the mature cognitive processing abilities of individuals over 15 years old.

[0030] Regarding the audio playback speed settings, 0.8x speed is provided for elementary school students, perfectly matching their language processing rate of 120-140 words per minute, ensuring they can keep up. 1.0x speed is provided for middle school students, aligning with the standard language processing pace of 150-160 words per minute for teenagers. 1.2x speed is provided for high school students, university students, adults, and seniors, catering to their mature language comprehension and information processing abilities. All speed settings include a minimum allowable speed of 0.5x to prevent issues with insufficient audio processing during subsequent downgrades.

[0031] In the settings for the maximum dynamic range of visual scene elements, for elementary school students, the maximum displacement of screen elements per second should not exceed 5% of the screen width, and flickering or switching effects are not allowed. This is because young children have short attention spans, and high-frequency dynamic images can easily cause visual fatigue and exacerbate emotional fluctuations. For middle school students, the maximum displacement of screen elements per second should not exceed 10% of the screen width, with a screen switching interval of at least 5 seconds, balancing the visual tolerance of teenagers with the smoothness of the visuals. For high school, university, adult, and elderly students, the maximum displacement of screen elements per second should not exceed 15% of the screen width, with a screen switching interval of at least 2 seconds, adapting to the acceptability of the visual content. All dynamic range settings reserve a minimum lower limit of 2% of the screen width to avoid meaningless parameter reductions during subsequent downgrade adjustments.

[0032] After the parameters are retrieved, the device will set this set of cognitive load limit parameters as the highest constraint benchmark for the output of audiovisual content during this intervention process, blocking content that exceeds the user's cognitive ability from the source and avoiding the cognitive overload problem that is prone to occur in existing solutions.

[0033] After retrieving the cognitive load limiting parameters, the device initiates the physiological data acquisition process, collecting physiological data from the user at rest using the integrated sensor array. Before collection begins, the device plays a 3-second rest preparation prompt, guiding the user to maintain a sitting or lying position, refrain from limb movement and verbal communication. Collection officially begins after the prompt ends, with the entire collection period fixed at 60 seconds. This duration fully covers the monitoring cycle of the autonomic nervous system's baseline state, filtering out random errors caused by fluctuations in single physiological signals, ensuring that the collected data is stable and representative.

[0034] During the data acquisition process, all sensors operated continuously at a sampling frequency of no less than 100Hz. After noise reduction and outlier removal, the arithmetic mean of the collected raw data within 60 seconds was taken as the final resting physiological data, which specifically included the initial heart rate, initial skin conductance response, initial respiratory rate, and initial muscle tension, with the corresponding units being breaths / min, μS, breaths / min, and μV, respectively.

[0035] After data processing, the device stores this set of resting physiological data as the initial baseline for this intervention process, serving as the sole reference benchmark for subsequent assessments of changes in the user's physiological state. Using the user's own resting physiological data as the baseline, rather than a general population norm, accurately reflects the user's individual physiological characteristics, avoiding the assessment bias caused by using general standards in existing solutions, and ensuring the accuracy of subsequent physiological state monitoring.

[0036] After establishing the initial baseline, the device extracts the initial heart rate and initial skin conductance response values ​​from the baseline to determine the initial stress state. The device pre-sets normal resting heart rate ranges according to educational level: 70-110 beats / min for primary school students (upper limit 110 beats / min); 60-100 beats / min for junior high school students (upper limit 100 beats / min); and 60-90 beats / min for senior high school, university, adult, and elderly students (upper limit 90 beats / min). Simultaneously, a skin conductance threshold of 5 μS is preset. This threshold corresponds to the skin conductance norm of healthy individuals at rest, indicating that the user's sympathetic nervous system is in a state of sustained excitation, inherently carrying basic psychological stress.

[0037] If the initial heart rate value is detected to be greater than the upper limit of the heart rate zone for the corresponding learning segment, and the initial skin conductance response value is also greater than the skin conductance threshold, the device will generate an initial stress state indicator, determining that the user is already in a state of high stress before this intervention. The device will then apply corresponding convergence coefficients based on the magnitude of the exceedance of these two indicators: a convergence coefficient of 0.8 if the initial heart rate exceeds the upper limit by no more than 10% and the skin conductance exceeds the threshold by no more than 20%; a convergence coefficient of 0.75 if the initial heart rate exceeds the upper limit by more than 10% but not more than 20% and the skin conductance exceeds the threshold by more than 20% but not more than 50%; and a convergence coefficient of 0.7 if the initial heart rate exceeds the upper limit by more than 20% and the skin conductance exceeds the threshold by more than 50%. When different convergence coefficients correspond to different exceedances of the two indicators, the smaller coefficient is used to ensure that the cognitive constraints under high stress are sufficiently strict. This set of coefficients is based on research findings in clinical psychology, which shows that cognitive processing ability declines to varying degrees under high stress. The tiered settings can accurately match the cognitive load tolerance level of users under different stress levels.

[0038] The device then uses this convergence coefficient to perform attenuation calculations on the previously retrieved cognitive load limit parameter baseline values. It multiplies the three core indicators—the upper limit of guidance instruction complexity, the upper limit of audio playback speed, and the upper limit of visual scene element dynamics—by their corresponding convergence coefficients. For the upper limit of guidance instruction complexity, indivisible discrete parameters such as vocabulary size and logical hierarchy are rounded down (if the rounded value is less than 1, it is calculated as 1). This yields the calibrated cognitive load limit parameters, which serve as the final execution baseline for this intervention. This step is equivalent to a secondary tightening of the cognitive boundaries before the intervention begins, based on the user's real-time stress state. This further reduces the risk of cognitive mismatch and compensates for the shortcomings of existing solutions that can only passively adjust parameters during intervention, adapting to the user's real-time state from the very beginning.

[0039] Existing self-help psychological intervention programs often directly output preset content after the user selects a scenario mode, without rigidly binding the scenario requirements with the user's cognitive ability. This can easily lead to a disconnect between content and cognitive ability, increasing the user's cognitive burden and emotional fluctuations. In this embodiment, the content generation process is based on the aforementioned final calibrated cognitive load limit parameters and the established initial baseline. The dual adaptation of scenario requirements and cognitive ability is completed at the source of content output, and no content exceeding the upper limit of the cognitive load limit parameters will be output throughout the process.

[0040] After establishing the initial baseline and calibrating the cognitive load limiting parameters, the device displays an operation scenario selection interface on a touchscreen screen. The selected operation scenarios are divided into two categories: specific adjustment scenarios and sleep improvement scenarios. Specific adjustment scenarios are suitable for short-term self-adjustment needs such as relaxation during breaks, pre-exam emotional calming, and daily anxiety relief. Based on the user's selected operation scenario, the device enters the corresponding logical branch. Throughout the content generation process, the final effective cognitive load limiting parameters serve as the highest constraint benchmark, combined with the established initial baseline, to generate initial intervention content perfectly tailored to the user's individual situation.

[0041] After the user selects a specific adaptation scenario, the device first extracts the upper limit of the complexity of the guidance instructions from the final effective cognitive load limit parameters, and then calls the preset cognitive threshold. The cognitive threshold is set to 12 words per instruction and 2 logical levels. This threshold is determined based on the key nodes of children's and adolescents' cognitive development stages, which can clearly distinguish between users with low cognitive load tolerance and users with high cognitive load tolerance, avoiding the problem of one-size-fits-all content adaptation.

[0042] The judgment rule is as follows: if either of the two indicators of the upper limit of the complexity of the guidance instruction reaches or exceeds the indicator corresponding to the cognitive threshold, it is judged to be not less than the cognitive threshold; if both are less than the indicator corresponding to the cognitive threshold, it is judged to be less than the cognitive threshold.

[0043] If the complexity limit of the guidance instructions is less than the cognitive threshold, the device will directly set the audiovisual guidance content as a concrete breathing-following animation and a solid-color visual scene. The core content of the concrete breathing-following animation is a circular light sphere that scales at a uniform speed. The scaling rhythm perfectly matches the initial breathing frequency value in the initial baseline. The light sphere enlarges uniformly during inhalation and shrinks uniformly during exhalation. The solid-color visual scene uses a low-saturation solid-color background with a color temperature below 4000K, without adding any additional dynamic elements. It strictly adheres to the set upper limit of the dynamic rate of visual scene elements throughout, making it fully suitable for users with low cognitive load tolerance and without causing any comprehension burden to the user.

[0044] If the upper limit of the guided instruction complexity is not less than the cognitive threshold, the device will set the audiovisual guidance content to mindfulness-guided audio and simulated visual scenes. The mindfulness-guided audio retrieves the corresponding audio library stored locally, and the content consists of body scanning and attention-anchoring guidance scripts. The playback speed strictly follows the set upper limit of audio playback speed, and the vocabulary and logical hierarchy of a single guided instruction are also strictly controlled within the upper limit of guided instruction complexity. The simulated visual scenes use simulated natural environments such as forests and streams, and the dynamic range of the visual elements strictly follows the set upper limit of the dynamic range of visual scene elements, adapting to users with high cognitive load tolerance. This forced adaptation logic ensures that the output content is completely in line with the user's cognitive ability, fundamentally avoiding the problem of young users receiving high-level guided content in existing solutions, and also preventing initial stress caused by cognitive overload.

[0045] After the user selects a sleep improvement scenario, the device will first perform an exclusive locking operation before outputting the corresponding sleep-aiding content. The device first disables sound acquisition permissions and voice wake-up function, completely blocking voice interaction to avoid interference from ambient sounds or accidental touches. Then, it limits the screen brightness to a preset dark field threshold of 10 cd / m², a value determined based on the safe brightness range that does not suppress melatonin secretion in nighttime environments, ensuring the normal secretion of sleep-related hormones. Simultaneously, the device limits the frequency of Bluetooth and Wi-Fi communication interactions to a preset communication threshold of once every 30 minutes. Except for necessary system-level prompts such as low battery warnings, all other non-essential communications are suspended, minimizing the impact of communication pop-ups and electromagnetic interference on the user's sleep process. The low battery warning is only triggered when the device battery level is below 5%, and the warning is only indicated by a brief increase in screen brightness and a single vibration, without triggering voice broadcasts or additional communications to avoid interfering with the sleep process.

[0046] After exclusive locking is completed, the device will output the corresponding virtual scene and sleep-aid audio based on the established initial baseline, strictly adhering to the final cognitive load limit parameters throughout the process. The initial brightness of the virtual scene does not exceed 10 cd / m², and it uses low-stimulation images with slow rhythmic dynamics, such as starry skies and ocean waves. The dynamic range of the images strictly follows the set upper limit of the dynamic range of visual scene elements. Next, the sleep-aid audio retrieves locally stored low-frequency soothing audio such as white noise and natural ambient sounds. The initial amplitude modulation frequency of the audio perfectly matches the initial respiratory rate value in the initial baseline. During inhalation, the audio amplitude increases linearly, and during exhalation, it decreases linearly, first guiding the user's breathing rhythm to establish initial synchronization with the audio, conforming to the user's basic physiological rhythm. Subsequently, during the output process, the device will make targeted dynamic adjustments based on the initial respiratory rate value: if the initial respiratory rate value in the initial baseline is higher than the upper limit of the preset sleep-aid target range (e.g., 8 to 12 breaths per minute), the device will slowly reduce the amplitude modulation frequency of the audio by a preset step size until it stabilizes within the target range, thus gently guiding the user's breathing rhythm to decrease synchronously; if the initial respiratory rate value itself is not higher than the upper limit, the device will directly maintain the initial amplitude modulation frequency for stable output. This not only perfectly conforms to the physiological frequency reduction process when the human body falls asleep naturally, but also properly takes into account users whose basic breathing is in the normal range or slightly slow, completely avoiding the vicious cycle of frequency reduction that violates physiological laws, and will not bring any extra burden to the initial stage of falling asleep.

[0047] During the output of sleep-aid audio and video, the device continuously monitors the physiological fluctuation difference. If the physiological fluctuation difference reaches the preset light sleep threshold, the brightness of the virtual scene is reduced by a preset slope until it is turned off, and only the sleep-aid audio is played. The core criterion for the light sleep threshold is a physiological fluctuation difference ≤ -15%, while also meeting the auxiliary verification conditions: the real-time heart rate value decreases by 10% from the initial baseline, and the real-time respiratory rate value falls within the range of 8 to 12 breaths per minute; if the initial baseline respiratory rate value is already below 8 breaths per minute, the respiratory rate auxiliary verification condition is adjusted to the real-time respiratory rate value being stable or decreasing from the initial baseline. This threshold setting not only fully conforms to the core judgment logic of the claims, but also conforms to the real physiological characteristics of a slight decrease in heart rate and a steady decline in indicators such as skin conductance during the human light sleep stage. It can accurately identify the point at which the user enters a light sleep state, and also completely avoids the judgment deadlock problem caused by the conflict between fixed high-frequency guidance and the low-frequency requirements of light sleep. Finally, the preset slope is set to a brightness decrease of 2 cd / m² every 30 seconds. This gradual decrease rate will not cause abrupt stimulation to the user's visual system, thus allowing for a smooth transition to a black screen state, avoiding waking the user, and minimizing the interference of visual stimulation on the sleep process.

[0048] Most existing closed-loop intervention solutions focus solely on unidirectional positive parameter optimization, assuming that simply adjusting parameters to align with physiological indicators will improve intervention effectiveness. They completely ignore self-help scenarios where users lack professional guidance; if stress worsens, continuing to push the same content only exacerbates cognitive burden and emotional fluctuations, potentially triggering reverse anxiety. Furthermore, they lack a corresponding proactive downgrade fallback mechanism. This embodiment's dynamic intervention fully integrates the aforementioned audiovisual guidance content or sleep-aid audio-visual materials. It uses the established initial baseline as the sole reference for physiological state assessment and the final effective cognitive load limit parameters as rigid constraints for content adjustment. Throughout the intervention process, it achieves real-time, bidirectional dynamic control, proactively implementing step-by-step downgrades to alleviate user stress and positively reinforcing and consolidating stress when the user enters a relaxed state, completely breaking the inherent limitations of existing unidirectional closed-loop technologies.

[0049] Combination Figure 2 As shown in the time sequence diagram, the bidirectional control mechanism, driven primarily by the physiological fluctuation difference within a 30-second sliding time window, is clearly presented. When the physiological fluctuation difference increases and enters the stress exacerbation zone, the system actively triggers degradation logic, performing operations such as reducing audio speech rate and / or visual dynamic range, forcibly switching to non-breath-holding deep breathing guidance, and even providing backup for extreme stress. When the physiological fluctuation difference decreases and reaches the preset relaxation threshold (≤-10%), entering the relaxation maintenance zone, the system triggers positive reinforcement, maintaining the current output content and incorporating psychological cues and visual features. This diagram intuitively demonstrates that this system completely breaks through the limitations of the existing one-way closed loop, possessing a substantial bidirectional control capability that accurately identifies stress exacerbation and actively performs degradation to prevent reverse anxiety.

[0050] During audio and video output, the device continuously collects the user's real-time physiological data at a sampling frequency of 100Hz using an integrated sensor array. This real-time physiological data includes real-time heart rate, real-time skin conductance, real-time respiratory rate, and real-time muscle tone, with corresponding units of breaths / min, μS, breaths / min, and μV, respectively. The device calculates the physiological fluctuation difference between the real-time physiological data and the initial baseline at a preset 1-second interval, employing a consistent calculation rule throughout to ensure consistency in the benchmarks of all judgment logic.

[0051] The specific calculation method is as follows: First, calculate the relative change rate of each real-time indicator relative to the corresponding indicator of the initial baseline. The formula for calculating the relative change rate is: (Real-time measurement value - Initial baseline value) / Initial baseline value × 100%. The calculation result retains the positive or negative sign; a positive value represents an increase in the indicator compared to the initial baseline, and a negative value represents a decrease. Then, multiply the four relative change rates by their corresponding weights and sum them to obtain the final physiological fluctuation difference. Among them, the weight of the relative change rate of heart rate is 0.4, the weight of the relative change rate of skin conductance is 0.3, the weight of the relative change rate of respiratory rate is 0.2, and the weight of the relative change rate of muscle tension is 0.1. This weight allocation is determined based on the sensitivity of each physiological indicator to the representation of autonomic nervous system excitability and psychological stress state. Heart rate responds fastest to acute stress, and skin conductance is more stable in representing continuous psychological stress, which is fully consistent with the basic research conclusions of physiological psychology and can accurately reflect the changing trend of the user's real-time stress state.

[0052] The device has a pre-set 30-second sliding time window. This window length is determined based on the latency and duration characteristics of the human body's psychological stress response. The human body typically requires a 20-30 second reaction cycle from exposure to a stressor to the onset of sustained physiological fluctuations. The sliding window effectively filters out occasional fluctuations in single heartbeats and skin conductance signals, accurately capturing the continuous changing trend of the user's stress state. In specific adjustment scenarios, if the difference in physiological fluctuations continues to increase within the 30-second sliding time window, the device determines that the current output content exceeds the user's cognitive load or has not been effectively accepted by the user, posing a risk of triggering reverse anxiety. It immediately generates a process record and performs a downgraded soothing operation.

[0053] The device first reduces the current output audio speed and / or video dynamic range. If the current output includes audio, the audio speed is reduced in increments of 0.1x, with a minimum of 0.5x. The video dynamic range is reduced in increments of 20%, with a minimum of 2% of the screen width. After these parameters are reduced, a set transition duration (e.g., 5 to 10 seconds) is maintained to achieve a smooth transition and degradation of sensory stimulation. Then, the current audiovisual content is interrupted, and deep breathing guidance is switched to the guided deep breathing. The guided deep breathing uses a uniform rhythmic, breath-holding-free abdominal breathing instruction. The initial rhythm calls a safe preset value matched to the learning stage (e.g., 4 seconds of inhalation and 6 seconds of exhalation for adults, and 3 seconds of inhalation and 4 seconds of exhalation for younger children). At this point, the device will adaptively match the modality of the output before downgrading: if the output before downgrading included audio, it will synchronously play guiding voice at a speech rate not exceeding 0.8x; if the output before downgrading was a silent, purely visual scene, the deep breathing guidance will rely solely on a uniformly scaling light sphere against a solid-color background to provide visual guidance, without any additional voice intervention, strictly preventing cross-modal stimulation from further aggravating the user's acute stress. This breathing rhythm is a commonly used relaxation breathing rhythm in clinical psychological intervention. The breath-holding-free design avoids increased chest pressure that could exacerbate sympathetic nerve excitation, effectively activating the parasympathetic nervous system to alleviate acute stress. At the same time, the adaptive, minimalist, and anti-overload design fundamentally avoids the core defect of forcibly imposing additional cognitive burden when the user's stress worsens.

[0054] After the device switches to deep breathing guidance, considering that the user's breathing rhythm has been forcibly taken over by the system, in order to prevent a sudden drop in respiratory indicators caused by the user actively following the instructions from directly interfering with the system's assessment of the true stress state and thus causing the formula to calculate a false safety drop difference, the device will activate an anti-interference mechanism. During this stage, the system will temporarily reset the calculation weight of the relative change rate of respiratory rate to zero, and compensate and distribute its original weight according to the existing ratio to the relative change rates of heart rate and skin conductance response, which are not directly controlled by subjective consciousness, to ensure that the assessment results can truly reflect the excitation level of the sympathetic nervous system.

[0055] After eliminating direct interference from respiratory indicators, the device continues to calculate the physiological fluctuation difference after weighting, with a preset cycle of 1 second. Each verification cycle is fixed at 30 seconds, consistent with the sliding time window of the physiological fluctuation difference. Here, "effective reduction of physiological fluctuation difference" is explicitly defined as the minimum value of the physiological fluctuation difference not decreasing to a threshold sufficient to demonstrate improvement within the 30-second sliding time window (i.e., not lower than 10% of the window's initial value minus its absolute value). If the physiological fluctuation difference does not show an effective reduction within the 30-second sliding time window, the device will set the current output screen to retain a solid-color background with visual elements guiding deep breathing (such as a uniformly scaling light sphere), while simultaneously reducing the rhythm of the deep breathing guidance by a preset step size. The preset step size is an increase of 2 seconds in exhalation time each time, while the inhalation time remains unchanged. The maximum exhalation time is no more than the physiologically safe upper limit of 12 seconds to avoid exceeding the body's normal breathing capacity and causing respiratory depression.

[0056] The device continuously cycles through the above-mentioned step-by-step degradation steps until the physiological fluctuation difference decreases to within the preset safe difference threshold (after the stress state is relieved and the deep breathing guidance is exited, the system will automatically restore the original indicator calculation weight configuration). The safe difference threshold is set when the physiological fluctuation difference falls back to less than 20% of the absolute value of the peak value of this cycle. This value is determined based on relevant research conclusions in clinical psychology. A relative decrease in the stress response can be used to determine that the acute stress has been effectively controlled, and can accurately determine whether the user's stress state has been relieved.

[0057] During the step-by-step, cyclical degradation process, the device simultaneously records the cumulative number of cycle steps. If the cumulative number reaches a preset threshold of 3, this value avoids accidental errors from a single cycle failing to take effect and can quickly trigger a fallback response under extreme stress, preventing the user from being in an uncomfortable intervention process for an extended period. Simultaneously, if the physiological fluctuation difference has not decreased to within the safe difference threshold, the device determines that the user is in an extreme stress state and the current intervention content can no longer produce a soothing effect. It immediately interrupts the current audio and video output and plays soothing audio in a preset frequency band. The soothing audio uses low-frequency sine wave audio in the 80Hz to 120Hz band, with a playback volume not exceeding 40dB and a single playback duration not exceeding 60 seconds. Sound waves in this frequency band can effectively soothe the auditory nerve without causing additional sensory stimulation.

[0058] Simultaneously, the device temporarily overrides the cognitive load limitation parameters that took effect during this intervention with a preset minimalist constraint lower limit. This minimalist constraint lower limit is a built-in safety minimum value (e.g., forcibly limiting the vocabulary of a single instruction to no more than 5 words and the logical level to 0), ensuring that its stringency is absolutely higher than any baseline parameter obtained after stress decay calculation at user startup. This limitation parameter will not be increased during this intervention. This temporary override does not modify the user-specific baseline parameters stored locally, avoiding conflicts with subsequent long-term parameter update logic, and fundamentally eliminating the risk of a crash due to the reverse relaxation of cognitive parameters under extreme conditions.

[0059] The device has a preset relaxation threshold, set at ≤-10% for physiological fluctuation difference. This value is determined based on relevant research on autonomic nervous system function assessment. A decrease of 10 percentage points or more in physiological fluctuation difference indicates a significant reduction in the user's sympathetic nervous system excitability, signifying that they have entered an effective relaxation state. In specific adjustment scenarios, if the physiological fluctuation difference continues to decrease within a 30-second sliding time window and reaches the preset relaxation threshold, the device will maintain the core audio and video parameters of the current output content while adding psychological cues to reinforce the relaxation effect.

[0060] The device first locks onto the core audio and video parameters of the current output content, including audio speed, main volume, basic dynamic range, and basic brightness, to prevent parameter changes from interrupting the user's relaxation state. It then extracts muscle tension indicators from real-time physiological data. If the muscle tension indicator reaches a preset tension threshold (set at 25% decrease from the initial baseline), indicating the user's body has entered a state of deep relaxation, adding psychological cues at this point will not elicit resistance and will be more readily accepted. The device outputs psychological cues at a preset slope. If the current output includes the main audio, the psychological cues are mixed in, with a preset slope of 5% volume increase every 5 seconds, ultimately not exceeding 30% of the main audio volume. If the current output is a silent, purely visual scene, the psychological cues are smoothly faded in with a preset extremely low background noise level to avoid program execution errors due to a lack of main volume reference. The gradual volume increase avoids abrupt auditory stimulation, allowing the psychological cues to naturally integrate into the current intervention context. The psychological suggestion guidance language strictly adheres to the currently effective cognitive load limit parameters. The vocabulary and logical hierarchy of a single instruction do not exceed the corresponding upper limit. The content mainly focuses on positive relaxation and emotional acceptance language, avoiding complex expressions. At the same time, the device generates dynamic visual features based on physiological fluctuation differences. The dynamic visual features are slow linear changes in the background color temperature of the screen. The rate of change perfectly matches the rate of decrease in physiological fluctuation differences. That is, the faster the physiological fluctuation difference decreases, the faster the rate of color temperature change increases, and vice versa. This transforms changes in physiological state into gentle visual feedback, further consolidating the user's relaxed state. Moreover, the superimposed visual features do not change the dynamic rate and basic structure of the core elements of the screen, and will not conflict with the parameter locking requirements.

[0061] Most existing self-help psychological intervention programs stop after a single intervention session, at most providing users with a simple feedback on the result. They do not record the actual effect of the intervention, the stress-triggered process, or deeply link it to the initial cognitive load limiting parameters, nor do they establish a long-term parameter iteration mechanism. The intervention parameters remain fixed and cannot adapt to changes in the user's long-term cognitive abilities and stress tolerance. Even after multiple uses, cognitive mismatch problems still occur. This embodiment utilizes the data from the entire intervention process for final evaluation and parameter updates. It uses the established initial baseline as the evaluation benchmark and the final effective cognitive load limiting parameters as the update basis to complete the final evaluation of the intervention effect. Then, based on historical intervention data, it realizes the dynamic evolution of cognitive load limiting parameters, allowing the intervention system to continuously optimize and adapt as the user uses it.

[0062] When preset acquisition conditions are met, the device acquires the user's terminal physiological data. Specifically, for specific adjustment scenarios, the preset acquisition condition is the complete termination of audio and video output. After the output ends, the device immediately stops all intervention content output and then uses the sensor array integrated into the device to collect physiological data of the user in a 60-second resting state as the terminal physiological data. Before collection, the device will play a 3-second resting preparation prompt, guiding the user to maintain a sitting or lying posture completely consistent with the initial baseline collection, refraining from limb movement and verbal communication to ensure consistency between the collection environment and the initial state, avoiding data deviation caused by environmental differences. For sleep improvement scenarios, after the sleep-aid audio and video playback ends, the device automatically enters a low-power standby state.

[0063] To strictly avoid the severe data pollution caused by strong sympathetic nerve excitation due to the morning cortisol wakefulness response (CAR), the preset acquisition condition is set to extract historical stable data. The device will not collect wakefulness data the next morning, but will automatically extract stable physiological data before and after reaching the light sleep threshold during the sleep stage. If the device fails to reach the light sleep threshold throughout the intervention due to continuous high stress, the device will automatically extract stable physiological data within a preset time window before entering low power standby mode, and use it directly as the final physiological data for unified evaluation in this scenario.

[0064] The sensor array continues to operate at a sampling frequency of no less than 100Hz to ensure that the accuracy of the acquired or extracted data is completely matched with the initial baseline. The indicators of the terminal physiological data are also completely consistent with the indicators of the initial baseline, including terminal heart rate, terminal skin conductance response, terminal respiratory rate, and terminal muscle tension, with corresponding units of breaths / min, μS, breaths / min, and μV, respectively. After denoising and removing outliers, the arithmetic mean of the acquired raw data over 60 seconds is taken as the final terminal physiological data.

[0065] After data acquisition, the device compares the terminal physiological data with the initial baseline item by item, calculates the change in each physiological indicator compared to the initial baseline, and generates assessment data. The assessment data includes three core indicators: the magnitude of stress improvement, the duration of relaxation, and the number of stress triggers. The magnitude of stress improvement is calculated based on the overall rate of change between the terminal physiological data and the initial baseline. The duration of relaxation is the cumulative duration during the intervention when the physiological fluctuation difference is consistently below the relaxation threshold. The number of stress triggers is the number of times the physiological fluctuation difference continuously increases during the intervention, triggering a downgraded soothing action. This data comprehensively and objectively reflects the actual effect of the intervention.

[0066] After generating assessment data, the device associates this assessment data with the user's selected operating scenario. If a process record is triggered due to increased physiological fluctuations during the intervention, it is also linked to generate a complete file record. Each file record is bound to a unique personal user identifier, corresponding educational stage identifier, and usage timestamp, ensuring accurate data traceability and preventing data confusion among multiple users. The device then encrypts the file records using the AES256 encryption algorithm. The encryption key is generated using a SHA-256 hash value based on the user's personal identifier, with each key corresponding to a user identifier. Decryption, viewing, and exporting of the file record are only possible after the user passes preset authentication (such as password or fingerprint verification). The encrypted data is stored in the device's local storage unit. Users can also authorize the process via a touchscreen display to synchronize the encrypted file records to the cloud during non-intervention operation, ensuring data security and enabling data synchronization between multiple devices. The non-intervention operation state refers to the device not being in audio / video output, real-time physiological data collection, or intervention control mode; it is only in standby, parameter setting, or file viewing mode.

[0067] After the device completes the storage of the records, it automatically calculates the cumulative trigger frequency of process records in the historical records corresponding to the same educational stage for that user. The cumulative trigger frequency is specifically defined as the number of independent interventions that have triggered process records under that educational stage. If the cumulative trigger frequency reaches a preset threshold of 3 times, the cognitive load limit parameter update process is initiated. This threshold covers both continuous and non-continuous trigger scenarios, excluding accidental triggers caused by environmental interference or temporary emotional fluctuations of the user in a single intervention. Only when multiple stress responses occur can it be concluded that the current cognitive load limit parameter baseline does not match the user's actual cognitive ability and stress tolerance, requiring systematic parameter adjustments.

[0068] The device then extracts the maximum physiological fluctuation difference from each process record that constitutes the cumulative trigger frequency mentioned above. It then calculates the stress representative value using these values. The stress representative value is the arithmetic mean of multiple maximum physiological fluctuation differences. Using this average smooths out the impact of single extreme fluctuations, more accurately reflecting the user's typical stress response intensity. Next, the device matches the stress representative value with preset intervals, obtaining the corresponding attenuation coefficient as a preset weight. The preset intervals are divided into three levels: a stress representative value less than 30% corresponds to an attenuation coefficient of 0.9; a stress representative value between 30% and 50% (inclusive) corresponds to an attenuation coefficient of 0.8; and a stress representative value greater than 50% corresponds to an attenuation coefficient of 0.7. This interval division follows the severity of the psychological stress response; the higher the stress representative value, the worse the user's tolerance to the current parameters, and the greater the required parameter reduction.

[0069] Finally, the device uses the preset weights to perform attenuation calculations on the baseline values ​​of cognitive load limitation parameters mapped to the user's personal user ID and corresponding educational stage ID. It multiplies the three core indicators—upper limits for guided instruction complexity, audio playback speed, and visual scene element dynamics—by their corresponding attenuation coefficients. For discrete parameters that are indivisible, including vocabulary and logical hierarchy, the calculated values ​​are rounded down (rounded to 1 if the result is less than 1). The new parameters are then used to overwrite the original baseline parameters, completing the dynamic parameter update. After the parameter update, the system automatically marks this part of the calculation process in the user's file as "processed," and it is no longer included in subsequent cumulative trigger frequency statistics, completely avoiding the parameter falling into an infinite one-way attenuation loop. The updated parameters will be directly invoked when the user inputs the corresponding educational stage ID again, allowing the entire intervention system to continuously adapt to the user's individual situation in real time, reducing the probability of cognitive mismatch and reverse anxiety in the long term.

[0070] This embodiment effectively reduces the risk of cognitive mismatch and reverse anxiety in self-help scenarios across all educational stages, and improves the adaptability, safety and effectiveness of psychological intervention. It can be widely applied in scenarios such as school breaks and home where there is no professional guidance for self-help psychological adjustment and sleep improvement.

[0071] This application also discloses a portable self-service intelligent psychological counseling and sleep intervention system for all school levels. This system fully inherits the aforementioned full-process intervention method, builds an integrated software and hardware architecture, and the execution logic of all modules corresponds one-to-one with the previous method steps. Those skilled in the art can rely on this system to fully implement the aforementioned full-process intervention method.

[0072] The hardware of this system adopts a lightweight, integrated structure that supports both handheld and wearable use, with an overall size not exceeding 15cm × 8cm × 2cm. It has a built-in rechargeable lithium battery that can support continuous intervention for no less than 8 hours on a full charge. The device integrates a sensor array for collecting multidimensional physiological data, a touch-screen interactive display, a voice acquisition and playback unit, a local processing unit, and a local storage unit.

[0073] The sensor suite includes heart rate sensors, skin conductance response sensors, respiratory rate sensors, and muscle tension sensors. All sensors have a sampling frequency of at least 100Hz, perfectly matching the sampling requirements of the aforementioned methods, ensuring the accuracy and real-time nature of physiological data acquisition. The system's core functional modules include a baseline establishment module, an intervention output module, a dynamic coordination module, and a visualization and encrypted storage module. All modules are integrated into the local processing unit of the hardware platform, completing data interaction and logic control via an internal wired bus, avoiding delays and interference caused by wireless transmission, and ensuring the real-time nature and stability of the intervention process.

[0074] The baseline establishment module fully corresponds to the aforementioned baseline calibration and cognitive boundary pre-establishment process. The module's input consists of the user's personal identifier and grade level identifier entered by the user via a touchscreen display. The output is the final effective cognitive load limitation parameters for this intervention, as well as the user's initial physiological baseline. After activation, the module first retrieves the historically updated cognitive load limitation parameter baseline value mapped to the input grade level identifier from local storage. Then, it drives the sensor array to collect the user's resting physiological data, establishing the initial baseline. Finally, it determines the initial stress state, pre-calibrates the cognitive load limitation parameters, and outputs the final effective baseline parameters and the initial baseline.

[0075] The intervention output module fully corresponds to the entire execution logic of the aforementioned scenario adaptation and initial intervention content generation. The module's inputs are the cognitive load limit parameters output by the baseline establishment module, the initial baseline, and the operation scenario selected by the user through the touch interactive display screen. The output is audiovisual guidance content or sleep-aiding audio-visual content adapted to the user's individual situation. After receiving the input parameters, the module first enters the corresponding logic branch according to the user's selected operation scenario. In the specific adjustment scenario, it strictly outputs the corresponding audiovisual guidance content according to the cognitive load limit parameters. In the sleep improvement scenario, it first disables sound acquisition permissions and voice wake-up function to complete voice interaction shielding, then limits screen brightness and communication interaction frequency within the corresponding thresholds, and finally outputs the corresponding sleep-aiding audio-visual content according to the initial baseline. Throughout the process, it will not output content that exceeds the cognitive load limit parameters.

[0076] The dynamic coordination module fully corresponds to the entire execution logic of the aforementioned two-way closed-loop dynamic intervention with real-time physiological feedback. The module's inputs are the initial baseline output by the baseline establishment module, the current audio and video output parameters of the intervention output module, and the user's physiological data collected in real time by the sensor array. The output is a dynamic adjustment control command for the intervention content. Throughout the audio and video output process, the module continuously collects real-time physiological data and calculates the physiological fluctuation difference between the real-time physiological data and the initial baseline. In a specific adjustment scenario, if the physiological fluctuation difference continues to increase within a preset sliding time window, a command is issued to the intervention output module to reduce the current audio speech rate and video dynamic rate, and switch to deep breathing guidance, while simultaneously executing subsequent step-by-step cyclical degradation and extreme state fallback control. In a specific adjustment scenario, if the physiological fluctuation difference decreases and reaches a preset relaxation threshold, a command is issued to the intervention output module to maintain the current output content and execute positive reinforcement intervention, achieving dynamic regulation in a two-way closed loop throughout the process.

[0077] The visualization and encrypted storage module fully corresponds to the entire execution logic of the aforementioned terminal assessment and cognitive load limitation parameter evolution and update. The module's inputs include the acquired terminal physiological data, the initial baseline output by the baseline establishment module, the intervention process data recorded by the dynamic coordination module, and the scenario selection records from the intervention output module. The outputs are visualized intervention assessment results, encrypted personal profile records, and updated cognitive load limitation parameters. After acquiring the terminal physiological data, the module first compares the terminal physiological data with the initial baseline to generate assessment data, then visualizes it through a touch-screen interactive display. Simultaneously, it associates the assessment data with the scenario selection records. If process records are generated, they are also associated to create profile records, encrypted, and stored in the local storage unit. Finally, based on the cumulative trigger frequency of process records in the historical profile records corresponding to the learning stage identifier, the cognitive load limitation parameters are dynamically updated. The updated parameters synchronously overwrite the corresponding baseline data in the local storage unit for subsequent use by the baseline establishment module. Cloud data synchronization is only performed with user authorization and when the device is in non-intervention operation mode. Strict communication frequency limits are enforced during the intervention process to avoid conflicts with the exclusive locking logic of the sleep scenario.

[0078] After the system completes the modular architecture construction and single-module logic definition, it sets the linkage and collaboration rules between modules to form a complete execution loop. The output of the baseline establishment module directly serves as the sole constraint benchmark for the intervention output module. The real-time output status of the intervention output module is synchronized to the dynamic collaboration module, allowing it to accurately match the current intervention content and perform dynamic adjustments. Control commands issued by the dynamic collaboration module directly reach the intervention output module, adjusting audio and video output parameters in real time. Simultaneously, the entire intervention record of the dynamic collaboration module is synchronized to the visualization and encrypted storage module in real time, ensuring no data loss. The updated cognitive load limit parameters completed by the visualization and encrypted storage module are synchronously overwritten in the local storage unit. When the user inputs the corresponding grade level identifier again, the baseline establishment module will directly call the updated parameters, forming a complete closed loop from baseline establishment, content output, dynamic adjustment to parameter evolution. This fully adapts to the intervention needs of self-service scenarios across all grade levels and ensures the complete implementation of the aforementioned methods at the system level.

[0079] The above are all preferred embodiments of this application, and are not intended to limit the scope of protection of this application. Therefore, all equivalent changes made in accordance with the structure, shape and principle of this application should be covered within the scope of protection of this application.

Claims

1. A portable, self-service, intelligent psychological counseling and sleep intervention method for all educational stages, characterized in that: include: Receive the input grade level identifier and call the cognitive load limitation parameter mapped to the grade level identifier; Initial baseline resting physiological data including initial heart rate, initial skin conductance response, initial respiratory rate, and initial muscle tone were collected. Selected operation scenario: If it is a special adjustment scenario, output the corresponding audio-visual guidance content according to the cognitive load limitation parameters; if it is a sleep improvement scenario, disable voice interaction and output sleep-aiding audio and video. During the output process, real-time physiological data is collected, and the relative rate of change of various indicators, including heart rate, skin conductance, respiratory rate, and muscle tension, relative to the corresponding indicators of the initial baseline is calculated. The physiological fluctuation difference is then obtained by weighting and summing these values ​​with their corresponding calculation weights. In the specific adjustment scenario, if the physiological fluctuation difference increases within the preset time window, a process record is generated, the current output audio speed and / or video dynamic range is reduced, and deep breathing guidance is switched; after switching, the calculation weight corresponding to the breathing frequency is temporarily reset to zero, and the weight is compensated and distributed to the calculation weight corresponding to the heart rate and skin conductance response according to a preset ratio, so as to continuously calculate the physiological fluctuation difference. In the specific adjustment scenario, if the physiological fluctuation difference decreases and reaches the preset relaxation threshold, the current output content is maintained and positive reinforcement intervention is performed.

2. The portable, self-service intelligent psychological counseling and sleep intervention method for all school levels as described in claim 1, characterized in that, Also includes: When preset acquisition conditions are met, the user's terminal physiological data is acquired; The terminal physiological data is compared with the initial baseline to generate evaluation data, and the evaluation data, operation scenario and generated process record are associated and stored as an archive record; If the cumulative trigger frequency of process records in the historical archives reaches a preset frequency threshold, the cognitive load limitation parameter mapped to the corresponding learning segment identifier is updated based on the preset weight, so that it can be invoked when the learning segment identifier is received again.

3. The portable, self-service intelligent psychological counseling and sleep intervention method for all educational stages as described in claim 1, characterized in that, The step of outputting audiovisual guidance content based on the cognitive load limitation parameters includes: Extract the upper limit of the complexity of the guidance instructions from the cognitive load limiting parameters; If the upper limit of the complexity of the guidance instruction is less than the preset cognitive threshold, the audiovisual guidance content is set as a concrete breathing follow-up animation and a solid color visual scene, wherein the scaling rhythm of the concrete breathing follow-up animation matches the initial breathing frequency. If the upper limit of the complexity of the guidance instruction is not less than the cognitive threshold, the audiovisual guidance content will be set as mindfulness guidance audio and simulated visual scene.

4. The portable, self-service intelligent psychological counseling and sleep intervention method for all educational stages as described in claim 1, characterized in that, The method of blocking voice interaction and outputting sleep-aiding audio and video includes: Limit screen brightness and communication interaction frequency; Based on the initial baseline, output the corresponding virtual scene and sleep aid audio, wherein the initial amplitude modulation frequency of the sleep aid audio is matched with the initial breathing frequency: if the initial breathing frequency is higher than the upper limit of the preset sleep aid target range, the amplitude modulation frequency is slowly reduced by a preset step size during the output process until it falls into the target range; if it is not higher than the upper limit, the initial amplitude modulation frequency is maintained and output smoothly. During the output process, if the physiological fluctuation difference reaches the preset light sleep threshold, the brightness of the virtual scene is reduced by a preset slope until it is turned off, while only the sleep aid audio is played.

5. The portable, self-service intelligent psychological counseling and sleep intervention method for all school levels according to claim 1, characterized in that, After switching to deep breathing guidance, the following is also included: The physiological fluctuation difference is continuously calculated at a preset period; If the physiological fluctuation difference does not decrease within the preset time window, the current output screen is set to retain the solid color background of the deep breathing guidance visual elements, and the rhythm of the deep breathing guidance is reduced by a preset step size. The above steps are repeated until the physiological fluctuation difference is reduced to within the preset safe difference threshold, and the original calculation weight configuration is restored after exiting the deep breathing guidance.

6. The portable, self-service intelligent psychological counseling and sleep intervention method for all school levels according to claim 2, characterized in that, The cognitive load limitation parameters, which are updated based on preset weights and mapped to the corresponding learning stage identifiers, include: Extract the maximum physiological fluctuation difference from each process record that constitutes the cumulative trigger frequency; Based on the extracted maximum physiological fluctuation difference, a stress representative value is calculated, and then the stress representative value is matched with a preset interval to obtain the corresponding attenuation coefficient as the preset weight. The cognitive load limitation parameters are attenuated according to the preset weights, and the discrete parameters are rounded down to overwrite the original parameters.

7. The portable, self-service intelligent psychological counseling and sleep intervention method for all educational stages as described in claim 1, characterized in that, The implementation of positive reinforcement intervention includes: Lock the audio and video parameters of the current output content and extract real-time muscle tension from real-time physiological data; If the real-time muscle tension reaches the preset tension threshold, psychological cues are output at a preset slope and mixed in when the current output audio exists. Dynamic visual features are generated based on the physiological fluctuation difference, and then the dynamic visual features are superimposed and rendered onto the currently output image.

8. The portable, self-service intelligent psychological counseling and sleep intervention method for all school levels as described in claim 1, characterized in that, After collecting resting physiological data including initial heart rate, initial skin conductance response, initial respiratory rate, and initial muscle tone as the initial baseline, the process also includes: If the initial heart rate is greater than the upper limit of the preset heart rate range and the initial skin conductance response is greater than the preset skin conductance threshold, an initial pressure state identifier is generated. Before obtaining the operation scenario, the convergence coefficient corresponding to the initial pressure state identifier is called, and the cognitive load limit parameter is attenuated using the convergence coefficient, and the discrete parameters therein are rounded.

9. A portable, self-service intelligent psychological counseling and sleep intervention method for all school levels as described in claim 5, characterized in that, The process of repeating the above steps also includes: Record the cumulative number of loops. If the preset number of loops is reached and the physiological fluctuation difference is not reduced to within the safe difference threshold, interrupt the current audio and video output. Play soothing audio on a preset frequency band and temporarily overwrite the cognitive load limit parameter for this intervention with the lower limit of minimalist constraints.

10. A portable, self-service intelligent psychological counseling and sleep intervention system for all educational stages, characterized in that: To implement the method according to any one of claims 1-9, comprising: The baseline establishment module receives the input grade segment identifier, calls the cognitive load limitation parameters mapped to the grade segment identifier, and collects resting physiological data including initial heart rate, initial skin conductance response, initial respiratory rate, and initial muscle tension as the initial baseline. The intervention output module acquires the selected operation scenario: if it is a special adjustment scenario, it outputs the corresponding audio-visual guidance content according to the cognitive load limitation parameters; if it is a sleep improvement scenario, it blocks voice interaction and outputs sleep-aiding audio and video. The dynamic coordination module collects real-time physiological data during the output process, calculates the relative change rate of various indicators, including heart rate, skin conductance response, respiratory rate, and muscle tension, compared to the corresponding indicators of the initial baseline, and sums them with the corresponding calculation weights to obtain the physiological fluctuation difference. In the specific adjustment scenario, if the physiological fluctuation difference increases within a preset time window, a process record is generated, the audio speech rate and / or video dynamic rate of the current output are reduced, and deep breathing guidance is switched. After switching, the calculation weight corresponding to the respiratory rate is temporarily reset to zero, and compensated and distributed to the calculation weights corresponding to the heart rate and skin conductance response according to a preset ratio to continuously calculate the physiological fluctuation difference. In the specific adjustment scenario, if the physiological fluctuation difference decreases and reaches a preset relaxation threshold, the current output content is maintained and positive reinforcement intervention is performed.

Images