A sleep optimization system based on brainwave regulation

Abstract

The application discloses a sleep optimization system based on brain wave regulation, and relates to the technical fields of sleep health and electroencephalogram regulation, comprising a hierarchical gap dynamic mapping module, a gap resonance compensation module, a cross-layer interlocking calibration module and a gap feature adaptive memory module, and the four modules form a closed loop collaborative work; in the application, the closed loop collaborative technology of the hierarchical structure of brain waves, three-dimensional gap feature dynamic mapping, targeted resonance compensation suitable for sleep stages, cross-layer interlocking calibration and two-dimensional clustering self-optimization of stage gaps is used, the problems that the existing sleep optimization system does not use the brain wave hierarchical division mechanism, the single frequency band regulation has poor adaptability, the multi-frequency band regulation is easy to produce signal interference and has low efficiency, and the deep sleep is not continuous and the wakefulness is residual are solved, the precise adaptation to sleep stages and individual differences is realized, the frequency band interference is eliminated, the continuity and proportion of deep sleep are improved, and the residual wakefulness and shallow sleep wandering are reduced, so that the safe and efficient personalized sleep optimization effect is achieved.

Description

Technical Field

[0001] This invention relates to the field of sleep health and brainwave regulation technology, specifically to a sleep optimization system based on brainwave regulation. Background Technology

[0002] As people's demand for better sleep quality increases, sleep optimization systems based on brainwave regulation have gained widespread attention due to their non-invasive and precise advantages. The core principle of these systems is to detect brainwave signals during sleep and output targeted regulatory signals to optimize slow-wave oscillations and improve sleep repair efficiency.

[0003] Currently, existing technologies are mainly divided into two categories: one is a single-band modulation scheme, which focuses on specific brainwave frequency bands related to sleep and improves sleep by enhancing the amplitude or frequency stability of that band. However, this type of scheme ignores the multi-dimensional functional needs of brainwaves during sleep, and modulation of only a single frequency band is difficult to adapt to the complex sleep regulation mechanism of the brain, which can easily lead to problems such as discontinuous deep sleep and residual wakefulness. The other type is a multi-band parallel modulation scheme, which adjusts the parameters of multiple slow wave frequency bands at the same time. However, this type of scheme does not consider the intrinsic relationship between the frequency bands and only uses an irregular parameter superposition method, which may not only cause signal interference between frequency bands, but also lead to low modulation efficiency due to lack of targeting.

[0004] In recent years, empirical studies by research institutions such as Beijing Normal University have clearly confirmed that the slow oscillations of brain waves during sleep are not a chaotic collection of signals, but can be divided into six sub-bands from slow-1 to slow-6, further constituting three functional levels: the detection layer (slow-1 / 2 / 3) is responsible for processing somatic sensory signals, the computation layer (slow-4) undertakes information integration, and the regulation layer (slow-5 / 6) dominates sleep state switching and stabilization. This research reveals the hierarchical physiological structure of brain waves during sleep. However, existing sleep optimization systems based on brain wave regulation have not recognized the regulatory value of this hierarchical relationship, remaining at the level of viewing each frequency band in isolation or irregular superposition of multiple frequency bands. They have failed to utilize the brain's inherent hierarchical division of labor mechanism to achieve precise and natural sleep optimization, resulting in a significant gap between the regulatory effect of existing systems and user needs. Therefore, a new regulation scheme based on the physiological hierarchical characteristics of brain waves is urgently needed. In view of this, a sleep optimization system based on brain wave regulation is proposed to overcome the above problems. Summary of the Invention

[0005] The purpose of this invention is to provide a sleep optimization system based on brainwave regulation to solve the problems mentioned in the background art.

[0006] To solve the above-mentioned technical problems, the present invention provides a sleep optimization system based on brainwave regulation, characterized in that it includes a sensor component for collecting brainwave signals and physiological signals and a signal output device, and also includes a hierarchical gap dynamic mapping module, a gap resonance compensation module, a cross-layer interlock calibration module and a gap feature adaptive memory module, the four modules forming a closed loop to work together.

[0007] The hierarchical gap dynamic mapping module is based on a three-level structure consisting of a detection layer, a calculation layer, and a regulation layer composed of brainwave sub-bands from slow1 to slow6, and defines and dynamically captures the three-dimensional gap feature vector.

[0008] The notch resonance compensation module has a built-in dynamic adaptation submodule for the sleep stage, which dynamically adjusts the notch threshold and compensation wave parameters based on the characteristics of the sleep stage for targeted resonance compensation.

[0009] The cross-layer interlock calibration module constructs a three-level ring interlock mechanism, combines the sleep phase to verify the compensation effect and perform abnormal calibration;

[0010] The gap feature adaptive memory module constructs a feature library through two-dimensional clustering of stage gaps, which is used for personalized self-optimization of the system.

[0011] Furthermore, the three-dimensional gap feature vector includes response time difference gap, amplitude complementarity gap, and cooperative matching gap; the response time difference gap is the time difference between the peak value of the detection layer signal and the peak value of the computation layer signal, the amplitude complementarity gap is the ratio of the amplitude of the computation layer signal to the geometric mean of the amplitudes of the detection layer and the control layer signals, and the cooperative matching gap is the harmonic matching degree of the three-level signal frequencies; the hierarchical gap dynamic mapping module separates the three-level signals through a digital filtering algorithm and calculates the three-dimensional gap feature vector once per second using a hierarchical signal peak alignment algorithm.

[0012] Furthermore, the sleep stage dynamic adaptation submodule adopts the brain wave physiological signal fusion recognition method, combining the proportions of slow5 and slow6, slow1 and slow2 in the EEG signal, as well as heart rate variability and respiratory rate to identify sleep stages, including light sleep, deep sleep, REM sleep, and wakefulness. The submodule establishes a sleep stage gap threshold offset table and a stage compensation parameter correction coefficient table, and obtains the staged gap threshold by superimposing the basic gap threshold and the stage offset, and performs effective gap secondary confirmation on the three-dimensional gap feature vector.

[0013] Furthermore, the frequency band of the compensation wave is determined according to the gap type. When the response time difference gap is abnormal, the compensation calculation layer is used; when the amplitude complementarity gap is abnormal, the compensation detection layer and the control layer are used; and when the coordinated matching gap is abnormal, the three-level synchronous compensation is used. The frequency of the compensation wave is calculated by the frequency correction coefficient of the current sleep stage and the target layer signal frequency. The amplitude of the compensation wave is calculated by the amplitude correction coefficient of the current sleep stage and the resonance gain coefficient. The resonance gain coefficient is 1.5 minus the amplitude complementarity gap value.

[0014] Furthermore, the interlock verification logic of the cross-layer interlock calibration module is as follows:

[0015] After the compensation calculation layer, the convergence of the response time difference gap is monitored. After the compensation detection layer or control layer, the convergence of the amplitude complementarity gap is monitored. After the three-level synchronous compensation, the convergence of the coordinated matching gap is monitored. Based on the convergence result, coordinated enhancement or locking operations are triggered. The abnormal calibration mechanism includes the gap reconstruction process and the non-coordinated gap judgment rules. If the gap does not improve after three consecutive reconstructions, it is judged as a non-coordinated gap and compensation is suspended. When switching to the sleep stage, compensation is immediately suspended and the effective gap is reconfirmed.

[0016] Furthermore, the feature library of the gap feature adaptive memory module is stored in an SQLite database, recording the three-dimensional gap feature vector, sleep stage, compensation wave parameters and compensation effect data for each sleep cycle; the module uses the K-means clustering principle to divide the gap into six sub-types, performs staged gap feature updates every 7 sleep cycles, and recalibrates the basic gap threshold range of each stage by statistical median.

[0017] Furthermore, the system workflow includes an initialization phase, a sleep stage identification and gap monitoring phase, an effective gap confirmation and resonance compensation phase, a cross-layer interlock calibration phase, and a self-optimization phase. In the initialization phase, the initial basic gap threshold range is determined by signal acquisition over a complete sleep cycle, and a blank feature library is constructed. In the self-optimization phase, a score is calculated based on the compensation efficiency scoring formula, which is obtained by weighting the inverse value of gap convergence time, the proportion of deep sleep, the inverse value of the number of awakenings, and the inverse value of the stage switching compensation interruption rate.

[0018] Compared with the prior art, the beneficial effects of the present invention are:

[0019] 1. Eliminate incoherence in deep sleep and improve the continuity of deep sleep: Based on the hierarchical gap dynamic mapping module, the module accurately captures the three-level coordination gap of brain waves. Combined with the gap resonance compensation module, the module tightens the basic gap threshold and reduces the compensation wave amplitude during the deep sleep period. The adjustment is only activated when the hierarchical coordination defect is significant, avoiding frequent intervention that disrupts the deep sleep state and greatly improving the continuity of the deep sleep cycle.

[0020] 2. No residual wakefulness, ensuring sleep integrity: By setting rules to disable compensation during the wakefulness period and control the duration of a single compensation to within 3 seconds through the gap resonance compensation module, and with the stage switching interruption mechanism of the cross-layer interlock calibration module, compensation is immediately paused and the effective gap is reconfirmed when the sleep stage switches, avoiding wakefulness caused by energy shock or parameter mismatch, and the incidence of residual wakefulness is extremely low.

[0021] 3. Eliminate frequency band interference and ensure adjustment safety: Relying on the harmonic matching calculation between the compensation wave frequency and the target level signal in the notch resonance compensation module, and the strict control of the phase difference by the phase-locked loop (limited to 3-8 degrees according to the sleep stage), the compensation wave and the target level signal resonate rather than superimpose irregularly, completely eliminating the signal interference problem caused by multi-frequency band adjustment.

[0022] 4. Improve regulation efficiency and increase the proportion of deep sleep: By accurately targeting the three-dimensional gaps of the hierarchical gap dynamic mapping module, combined with the staged parameter correction and targeted compensation logic of the gap resonance compensation module, ineffective regulation is avoided, which significantly improves regulation efficiency and increases the proportion of deep sleep time compared with the existing multi-band solution.

[0023] 5. Smooth transition between sleep stages and reduced light sleep lingering: During the light sleep stage, the compensation wave response speed optimization promotes the collaboration between the detection layer and the computation layer, accelerating the entry into the deep sleep stage; during the deep sleep stage, low amplitude and low frequency compensation maintains the stability of the sleep hierarchy; during the REM stage, the balance adjustment parameters avoid interruption, significantly shortening the transition time between sleep stages and reducing the phenomenon of light sleep lingering.

[0024] 6. Personalized stage adaptation and self-evolution to adapt to individual differences: With the help of the gap feature adaptive memory module's stage gap two-dimensional clustering (K-means clustering into 6 gaps) and parameter iteration every 7 sleep cycles, the system can automatically identify the user's sleep characteristics and continuously optimize the adjustment strategy without manual intervention, with high stage adaptation accuracy.

[0025] 7. Low-power intelligent start-stop, reducing usage costs: By using a cross-layer interlock calibration module to completely shut down the compensation output module when there is no effective gap or during the wake-up period, combined with staged compensation duration control, power consumption is significantly reduced compared to the original system, improving the device's battery life.

[0026] 8. Visualized sleep structure warning with personalized suggestions: Based on the SQLite database storage function of the gap feature adaptive memory module, the system can record and analyze the gap correlation data of each sleep stage. When abnormal gaps occur frequently in the light sleep stage, the risk of difficulty falling asleep is judged; when the gap in the deep sleep stage is extremely low, the risk of excessive deep sleep is judged, and targeted improvement suggestions are output. Attached Figure Description

[0027] Figure 1 This is a schematic diagram of a sleep optimization system based on brainwave regulation according to the present invention. Detailed Implementation

[0028] 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.

[0029] Please see Figure 1 The present invention provides a technical solution:

[0030] See Figure 1 As shown, an embodiment of a sleep optimization system based on brainwave regulation is presented:

[0031] I. System Overall Architecture:

[0032] This system includes a hierarchical gap dynamic mapping module, a gap resonance compensation module, a cross-layer interlock calibration module, and a gap feature adaptive memory module. The gap resonance compensation module has a built-in sleep stage dynamic adaptation sub-module. The four modules form a closed loop of gap identification, stage adaptation, self-compensation adjustment, collaborative locking, and optimization iteration. There are no preset fixed adjustment signals throughout the process, and all adjustment parameters are jointly determined by the gap characteristics of the brain waves themselves and the characteristics of the sleep stages.

[0033] II. Core Modules:

[0034] (I) Hierarchical Gap Dynamic Mapping Module:

[0035] Breaking through the limitations of existing methods that only analyze frequency band signals, this paper defines for the first time the three major functional complementarity gaps between brainwave levels and quantifies and maps them in real time. The specific implementation is as follows:

[0036] 1. Definition of notch parameters:

[0037] Response time gap The time difference between the peak signal of the detection layer and the peak signal of the computing layer, in milliseconds, reflects the collaborative efficiency of body signal processing and information integration.

[0038] Amplitude Complementary Gap The ratio of the amplitude of the computational layer signal to the geometric mean of the amplitudes of the detection layer and the modulation layer signal:

[0039] ;

[0040] Reflects the energy adaptability of information integration;

[0041] in, To calculate the amplitude of the layer signal, The amplitude of the probe layer signal is the arithmetic mean of the amplitudes of the three sub-bands: slow1, slow2, and slow3. The amplitude of the control layer signal is the arithmetic mean of the amplitudes of the two sub-band signals, slow5 and slow6.

[0042] Collaborative matching gap Harmonic matching degree of three-level signal frequencies:

[0043] ;

[0044] in, To calculate the average frequency of the layer signal; The average frequency of the probe layer signal is the arithmetic mean of the frequencies of the three sub-bands: slow1, slow2, and slow3. The average frequency of the control layer signal is the arithmetic mean of the frequencies of the two sub-bands, slow5 and slow6; this parameter reflects the frequency coordination from processing to integration to stability.

[0045] 2. Gap dynamic capture process:

[0046] A non-invasive dry electrode EEG sensor was used, with a sampling frequency of 512 Hz and signal noise controlled within 2 microvolts. The sensor synchronously acquired signals from six sub-bands, from slow1 to slow6. A digital filtering algorithm was used to separate the signals from the detection layer, the computation layer, and the modulation layer. The passband of the digital filtering algorithm was set as follows: 0.5 to 4 Hz for the detection layer, 4 to 8 Hz for the computation layer, and 8 to 12 Hz for the modulation layer.

[0047] A hierarchical signal peak alignment algorithm is adopted, which extracts the peak timestamps of each level of signal using a sliding window method. The window length is set to 100 milliseconds, the step size is set to 10 milliseconds, and the algorithm is calculated once per second. , , Generate a three-dimensional gap feature vector ( , , );

[0048] Set a basic gap threshold range under normal collaborative conditions. The value ranges from 20 milliseconds to 80 milliseconds. The value ranges from 0.8 to 1.2. The value ranges from 0 to 0.1. When the three-dimensional gap feature vector exceeds this range, it is initially determined to be a potential valid gap, and a second confirmation is required in combination with the characteristics of the sleep stage.

[0049] It should be added that: for the first time, the hierarchical collaboration defect is quantified into a captureable gap parameter, which provides a precise target for the subsequent stage of hierarchical dual-target adjustment and solves the core pain point that existing technologies cannot identify the root cause of hierarchical collaboration problems.

[0050] (II) Notch Resonance Compensation Module:

[0051] Abandoning the conventional approach of preset frequency band parameters, this system utilizes a built-in dynamic sleep stage adaptation submodule to achieve dual targeted adjustment of stage adaptation and gap self-compensation. The specific implementation is as follows:

[0052] 1. Sleep Stage Dynamic Adaptation Submodule

[0053] Sleep stage identification method: A brainwave physiological signal fusion identification method is adopted. The proportions of slow5 and slow6 and slow1 and slow2 are collected by EEG sensors. The slow5 and slow6 proportions are the ratio of the sum of the signal energy of the slow5 and slow6 sub-bands to the total brainwave energy, and the slow1 and slow2 proportions are the ratio of the sum of the signal energy of the slow1 and slow2 sub-bands to the total brainwave energy. At the same time, a wristband photoelectric sensor is used to collect heart rate variability and a chest patch piezoelectric sensor is used to collect respiratory rate. All of the above sensors are mass-produced mature components. The sleep stage determination results are output every second through a multi-sensor data fusion algorithm. The algorithm adopts a weighted voting method, with the weight of brainwave signal set to 0.6, the weight of heart rate variability set to 0.2, and the weight of respiratory rate set to 0.2. The determination results include light sleep stage, deep sleep stage, REM sleep stage, and wakefulness stage.

[0054] Stage Gap Adaptation Rules: Establish a sleep stage gap threshold offset table and a stage compensation parameter correction coefficient table, dynamically adjust the gap judgment criteria and compensation wave parameters, and adapt to the hierarchical sensitivity characteristics of different stages. Specific parameters are as follows:

[0055] Table 1: Sleep Stage Gap Threshold Offset Table

[0056] Sleep stage Threshold offset (milliseconds) Threshold offset Threshold offset Light sleep period 10 0.1 0.02 Deep sleep -10 -0.1 -0.02 REM period 5 0 0.01 Awakening period Disable compensation Disable compensation Disable compensation

[0057] Table 2: Stage-Compensation Parameter Correction Coefficient Table

[0058] Sleep stage Frequency correction factor α Amplitude correction factor β Permissible phase difference range (degrees) Light sleep period 1.1 1.2 8 Deep sleep 0.9 0.8 3 REM period 1.0 0.9 5

[0059] The threshold offsets in Table 1 are respectively , , .

[0060] Secondary confirmation of valid gaps: The basic gap threshold is superimposed with the offset of the corresponding sleep stage to obtain the staged gap threshold. Only when the three-dimensional gap feature vector exceeds the staged gap threshold is it determined to be a valid gap and the subsequent compensation process is triggered.

[0061] 2. Logic for generating compensated wave parameters:

[0062] Compensation wave frequency band selection: response time difference gap In case of anomalies, compensation calculation layer; amplitude complementary gap. In case of anomalies, the detection layer and control layer are compensated; gaps are matched collaboratively. In case of anomalies, three-level synchronous compensation is implemented;

[0063] Compensation wave frequency :

[0064] ;

[0065] in The frequency correction coefficient for the current sleep stage (taken from Table 2) ensures that the compensation wave and the target layer signal for stage adaptation form harmonic resonance.

[0066] Compensation amplitude value :

[0067] ;

[0068] in The amplitude correction factor for the current sleep stage (taken from Table 2). The resonant gain coefficient ( =1.5- This helps avoid excess energy during deep sleep or insufficient energy during light sleep.

[0069] 3. Resonance compensation execution method:

[0070] A dual-channel phase-locked output method is adopted. The phase-locked loop realizes the phase synchronization between the compensation wave and the target layer signal. The loop bandwidth of the phase-locked loop is set to 1 Hz to ensure that the phase difference between the compensation wave and the target layer signal is controlled within the allowable phase difference range of the corresponding sleep stage. The output device is a bone conduction audio device or a transcranial microcurrent stimulation module. The frequency response range of the bone conduction audio device is set to 0.5 Hz to 30 Hz, and the current intensity of the transcranial microcurrent stimulation module is controlled within 100 microamps. Precise current control is achieved through a common PCB board.

[0071] Compensation wave duration control: Output is only available when an effective gap exists, and output stops immediately when the gap returns to the staged threshold range. The duration of a single compensation is strictly controlled within 3 seconds. Continuous compensation is allowed during the light sleep period, with an interval of no less than 2 seconds between two triggers. During the deep sleep period, the interval between continuous compensation triggers is no less than 5 seconds to avoid over-adjustment.

[0072] It should be noted that this is the first time that sleep stage characteristics have been linked to hierarchical gaps. By dynamically adjusting thresholds and parameters, the compensation wave has been upgraded from a general self-compensation to a stage-specific self-compensation. This addresses the problem of insufficient adjustment accuracy for different sleep stages from a mechanistic perspective. Furthermore, the sensors and correction logic used are all mature technologies that can be mass-produced in ordinary factories.

[0073] (III) Cross-layer interlock calibration module:

[0074] A circular interlocking mechanism is constructed from the detection layer to the computation layer to the control layer and back to the detection layer. The fusion phase verifies and ensures the compensation effect. The specific implementation is as follows:

[0075] 1. Interlock verification logic:

[0076] After the compensation calculation layer, the response time difference gap is monitored in real time. Change, if If the convergence rate is not less than 10% per second and the current sleep stage has not changed, the compensation is deemed effective. A collaborative enhancement signal is then sent to the detection layer and the control layer. This signal is a low-frequency pulse signal with a frequency of 1 Hz and an amplitude of 10% of the target layer signal amplitude. It is used only to activate the hierarchical response.

[0077] After compensating the detection layer or control layer, monitor the amplitude complementarity gap in real time. Change, if If the signal converges to 1.0 with a convergence rate of not less than 5% per second and the current sleep stage does not switch, the compensation is deemed effective. The compensation is then fed back to the computing layer and its signal amplitude fluctuation threshold is lowered by 10% of the original threshold to improve the stability of information integration.

[0078] After three levels of synchronous compensation, real-time monitoring of the collaborative matching gap is conducted. Changes, light sleep stage and REM sleep stage When it drops to 0.05 or below, deep sleep... When the value drops to 0.03 or below, a cooperative locking mechanism is triggered, maintaining the current compensation wave parameters until the gap completely disappears.

[0079] 2. Abnormal calibration mechanism:

[0080] If the gap does not improve after compensation, i.e. the convergence rate of change is less than 3% per second, then the gap reconstruction process is initiated: the three-dimensional gap feature vector is recalculated, the compensation wave frequency is shifted by 10%, and the amplitude is adjusted by 15% to avoid falling into an ineffective adjustment cycle.

[0081] If the gap does not improve after three consecutive reconstructions, it is determined to be a non-cooperative gap. This type of gap is mostly caused by external interference. At this time, the compensation process is suspended and only signal monitoring is maintained to avoid forced adjustment that may trigger arousal.

[0082] A new stage switching interruption mechanism has been added: the sleep stage determination results are monitored in real time during the compensation process. If a sleep stage switch occurs, the compensation is immediately paused and the effective gap secondary confirmation process is re-executed to avoid interference caused by parameter mismatch after the stage switch.

[0083] It should be noted that combining cross-level interlocking with sleep stage verification ensures that the compensation effect not only meets the needs of hierarchical coordination but also adapts to changes in sleep state, avoiding imbalance at a single level or mismatch at a stage.

[0084] (iv) Adaptive memory module for gap features:

[0085] A two-dimensional clustering approach with stage gaps is adopted to achieve personalized self-optimization, and the specific implementation is as follows:

[0086] 1. Construction of the gap feature library:

[0087] Record the three-dimensional gap feature vector, sleep stage, compensation wave parameters, and compensation effect for each sleep cycle. The compensation effect includes gap convergence time, deep sleep duration ratio, number of awakenings, and compensation interruption rate during stage switching. Generate a user-specific three-dimensional mapping library for stage gap compensation. This mapping library is stored in an SQLite database and supports fast querying and updating.

[0088] A two-dimensional clustering algorithm for stage gaps is adopted. This algorithm is based on the K-means clustering principle and sets the number of clusters to 6. The gaps are divided into six sub-types: light sleep time difference type, light sleep amplitude type, light sleep coordination type, deep sleep time difference type, deep sleep amplitude type, and REM coordination type. The optimal compensation parameters corresponding to each type of gap are stored.

[0089] 2. Self-optimizing iterative logic:

[0090] Compensation efficiency score: Score = Gap convergence time (inverse) × 40% + Deep sleep percentage (inverse) × 30% + Number of awakenings (inverse) × 30% + Compensation interruption rate during phase switching (inverse) × 10%;

[0091] Every 7 sleep cycles, the staged gap characteristics are updated, and the basic gap thresholds for each stage are recalibrated to adapt to changes in physiological state.

[0092] It should be noted that the implementation of phase gaps allows for two-dimensional personalized optimization, enabling the system to more accurately adapt to the hierarchical collaborative characteristics of users at different sleep stages.

[0093] III. System Workflow:

[0094] Initialization phase: When a user uses the system for the first time, the system performs a stage gap baseline calibration for a complete sleep cycle. By collecting brain wave signals and physiological signals during the sleep cycle, the system calculates the statistical values ​​of the three-dimensional gap feature vectors for each sleep stage, determines the initial basic gap threshold range for each sleep stage, and constructs a three-dimensional mapping library for gap compensation in blank stages.

[0095] Sleep stage identification and gap monitoring stage: After the system is started, EEG signals and physiological signals are collected synchronously. The sleep stage determination results are output every second through the brain wave physiological signal fusion identification method. The staged gap threshold is generated by combining the basic gap threshold and the stage offset. The three-dimensional gap feature vector is calculated once per second through the hierarchical signal peak alignment algorithm and continuously monitored.

[0096] Effective gap confirmation and resonance compensation stage: The three-dimensional gap feature vector is compared with the staged gap threshold. After determining that it is an effective gap, the target level is determined according to the gap type. The compensation wave frequency and amplitude are calculated in combination with the correction coefficient of the current sleep stage. The compensation wave is output through the bone conduction audio device or transcranial microcurrent stimulation module.

[0097] Cross-layer interlock calibration phase: While outputting the compensation wave, the system acquires signals from the target layer and related layers in real time. The three-dimensional gap feature vector is updated once per second using a layer signal peak alignment algorithm to determine whether the gap is converging towards the staged threshold range. If the compensation is deemed effective, the current compensation wave parameters are maintained and monitoring continues. If a sleep phase switch is detected, compensation is immediately paused and a second confirmation of the effective gap is performed. If the gap does not improve, gap reconstruction is initiated or compensation is paused according to the abnormal calibration mechanism until the gap converges or a non-cooperative gap is confirmed.

[0098] Self-optimization phase: After the sleep cycle ends, the system automatically extracts the compensation effect data for this sleep, including the convergence time of each effective gap, the ratio of deep sleep duration to total sleep duration (deep sleep duration percentage), the number of awakenings, and the ratio of the number of compensation interruptions caused by stage switching to the total number of compensations (compensation interruption rate during stage switching). These data are then substituted into the compensation efficiency scoring formula to calculate the current score. The current score is compared with the historical average in the three-dimensional mapping library for stage gap compensation, and data updates or parameter calls are completed according to the mapping library update rules. If 7 sleep cycles have been accumulated, the stage gap feature update process is initiated. By statistically analyzing the median of the three-dimensional gap feature vectors of each sleep stage within 7 cycles, the basic gap threshold range for each stage is recalibrated, and personalized parameter iteration is completed.

[0099] Summarize:

[0100] Eliminating deep sleep discontinuity: During deep sleep, by tightening the basal gap threshold and reducing the compensation amplitude, regulation is only initiated when there is a significant hierarchical coordination defect, avoiding frequent interventions that disrupt the deep sleep state, and significantly improving the continuity of the deep sleep cycle;

[0101] No residual arousal: Compensation output is strictly prohibited during the arousal period. Compensation is immediately interrupted when the sleep stage switches during the compensation process, and the duration of each compensation is strictly controlled within a reasonable range to avoid energy shocks that could trigger arousal. The incidence of residual arousal is extremely low.

[0102] Eliminate frequency band interference: The compensation wave frequency is calculated based on the relationship between the target level frequency and harmonics, and the phase difference is strictly controlled within a reasonable range to ensure resonance with the target level signal rather than superposition, thus completely eliminating frequency band interference;

[0103] Improved regulation efficiency: Through dual-targeted compensation with phased adaptation, ineffective regulation is avoided, and the regulation efficiency is significantly improved compared with the existing multi-band solution, resulting in a significant increase in the proportion of deep sleep time.

[0104] and:

[0105] Smoothing of sleep stage transitions: During light sleep, the response speed of compensation waves is increased to promote coordination between the detection layer and the computation layer, accelerating the entry into deep sleep; during deep sleep, low-amplitude and low-frequency compensation is used to maintain hierarchical stability; during REM sleep, the balance parameters are balanced to avoid regulatory interruption, significantly shortening the sleep stage transition time and reducing light sleep lingering.

[0106] Personalized stage adaptation and self-evolution: The system identifies sleep characteristics such as light sleep sensitivity (frequent light sleep gaps) and deep sleep stability (few deep sleep gaps) through two-dimensional clustering of stage gaps. It can continuously optimize adjustment strategies without manual intervention, and the stage adaptation accuracy is high.

[0107] Low-power intelligent start-stop: The compensation output module is completely shut down during the awakening period and when there is no effective gap. Combined with the phased compensation duration control, the power consumption is greatly reduced compared to the original system.

[0108] Sleep structure visualization and early warning: Utilizing stage gap compensation 3D mapping library-stored stage gap correlation data, during light sleep... Frequent abnormalities indicate a risk of difficulty falling asleep; during deep sleep... If the level remains at an extremely low level, it is considered a risk of excessive deep sleep, and the system will output personalized improvement suggestions.

Claims

1. A sleep optimization system based on brainwave regulation, characterized in that, It includes sensor components for acquiring brainwave signals and physiological signals, as well as a signal output device. Its feature is that it also includes a hierarchical gap dynamic mapping module, a gap resonance compensation module, a cross-layer interlock calibration module, and a gap feature adaptive memory module, with the four modules forming a closed loop and working collaboratively. The hierarchical gap dynamic mapping module is based on a three-level structure consisting of a detection layer, a calculation layer, and a regulation layer composed of brainwave sub-bands from slow1 to slow6, and defines and dynamically captures the three-dimensional gap feature vector. The notch resonance compensation module has a built-in dynamic adaptation submodule for the sleep stage, which dynamically adjusts the notch threshold and compensation wave parameters based on the characteristics of the sleep stage for targeted resonance compensation. The cross-layer interlock calibration module constructs a three-level ring interlock mechanism, combines the sleep phase to verify the compensation effect and perform abnormal calibration; The gap feature adaptive memory module constructs a feature library through two-dimensional clustering of stage gaps, which is used for personalized self-optimization of the system.

2. The sleep optimization system based on brainwave regulation as described in claim 1, characterized in that: The three-dimensional gap feature vector includes response time difference gap, amplitude complementarity gap and cooperative matching gap; response time difference gap is the time difference between the peak value of the detection layer signal and the peak value of the computation layer signal, amplitude complementarity gap is the ratio of the amplitude of the computation layer signal to the geometric mean of the amplitudes of the detection layer and the control layer signals, and cooperative matching gap is the harmonic matching degree of the three-level signal frequencies; The hierarchical gap dynamic mapping module separates the three-level signals through a digital filtering algorithm and uses a hierarchical signal peak alignment algorithm to calculate the three-dimensional gap feature vector once per second.

3. The sleep optimization system based on brainwave regulation as described in claim 1, characterized in that: The sleep stage dynamic adaptation submodule uses brainwave physiological signal fusion recognition method, combining the proportions of slow5 and slow6, slow1 and slow2 in EEG signals, as well as heart rate variability and respiratory rate to identify sleep stages, including light sleep, deep sleep, REM sleep, and wakefulness. The submodule establishes a sleep stage gap threshold offset table and a stage compensation parameter correction coefficient table. By superimposing the basic gap threshold and the stage offset, the staged gap threshold is obtained, and the effective gap is reconfirmed for the three-dimensional gap feature vector.

4. The sleep optimization system based on brainwave regulation as described in claim 3, characterized in that: The frequency band of the compensation wave is determined according to the gap type. When the response time difference gap is abnormal, the compensation calculation layer is used; when the amplitude complementarity gap is abnormal, the compensation detection layer and the control layer are used; and when the coordinated matching gap is abnormal, the three-level synchronous compensation is used. The frequency of the compensation wave is calculated by the frequency correction coefficient of the current sleep stage and the target layer signal frequency. The amplitude of the compensation wave is calculated by the amplitude correction coefficient of the current sleep stage and the resonance gain coefficient. The resonance gain coefficient is 1.5 minus the amplitude complementarity gap value.

5. The sleep optimization system based on brainwave regulation as described in claim 1, characterized in that: The interlock verification logic of the cross-layer interlock calibration module is as follows: After the compensation calculation layer, the convergence of the response time difference gap is monitored. After the compensation detection layer or control layer, the convergence of the amplitude complementarity gap is monitored. After the three-level synchronous compensation, the convergence of the coordinated matching gap is monitored. Based on the convergence result, coordinated enhancement or locking operations are triggered. The abnormal calibration mechanism includes the gap reconstruction process and the non-coordinated gap judgment rules. If the gap does not improve after three consecutive reconstructions, it is judged as a non-coordinated gap and compensation is suspended. When switching to the sleep stage, compensation is immediately suspended and the effective gap is reconfirmed.

6. The sleep optimization system based on brainwave regulation as described in claim 1, characterized in that: The feature library of the gap feature adaptive memory module is stored in an SQLite database, recording the three-dimensional gap feature vector, sleep stage, compensation wave parameters and compensation effect data for each sleep cycle. The module uses the K-means clustering principle to divide the gap into six sub-types, and performs staged gap feature updates every 7 sleep cycles. The basic gap threshold range of each stage is recalibrated by statistical median.

7. The sleep optimization system based on brainwave regulation as described in claim 1, characterized in that: The system workflow includes an initialization phase, a sleep stage identification and gap monitoring phase, an effective gap confirmation and resonance compensation phase, a cross-layer interlock calibration phase, and a self-optimization phase. In the initialization phase, the initial basic gap threshold range is determined by signal acquisition over a complete sleep cycle, and a blank feature library is constructed. In the self-optimization phase, a score is calculated based on the compensation efficiency scoring formula. This score is obtained by weighting the inverse value of the gap convergence time, the proportion of deep sleep, the inverse value of the number of awakenings, and the inverse value of the stage switching compensation interruption rate.

Images