Electroencephalogram head ring interaction control method and device based on concentration level

Abstract

The invention discloses an electroencephalogram head ring interaction control method and device based on concentration level, and the method comprises the steps: collecting multi-band electroencephalogram signals and user interaction parameters, carrying out spectrum decoupling analysis to recognize a concentration feature baseline, exporting an attention stability coefficient, and building a cognitive state mapping relation; tracking a high-response frequency band core to implement frequency band fusion analysis to form a control channel, and collecting concentration fluctuation data to construct an interaction control network; executing multi-task response analysis to identify an intention switching point, extracting cognitive load parameters to divide a low-load area and a high-load area, and extracting a recovery rate to construct an attention toughness index in an identification conversion process; determining an optimal trigger position in a low-load area according to an attention toughness index, constructing a brain wave feature chain, aligning the brain wave feature chain with an attention stability coefficient, determining a low-interference response window, and setting a self-adaptive threshold group; and the coupling identification false triggering area extracts redundant brain wave components to execute false triggering suppression to generate an interaction control instruction, so that accurate matching between teaching intervention and the cognitive state of the student is realized.

Description

Technical Field

[0001] This invention relates to the field of brain-computer interface control technology, and in particular to a brainwave headband interactive control method and device based on attention level. Background Technology

[0002] In smart education scenarios, students' concentration levels and cognitive load directly impact learning outcomes. Traditional teaching systems, with their fixed content delivery rhythm, struggle to adapt to changes in students' real-time cognitive states. While some systems attempt to adjust teaching strategies using behavioral data such as answer accuracy and response time, these superficial indicators are lagging and fail to capture the true cognitive load and concentration fluctuations within the brain.

[0003] Electroencephalogram (EEG) signals, as physiological indicators reflecting brain neural activity, provide a technical approach for real-time monitoring of cognitive states. However, existing methods face multiple challenges: how to accurately identify attentional characteristics from multi-band EEG data and establish a mapping relationship with cognitive states; how to identify the dynamic transition process of cognitive load and intervene at the optimal time; and how to effectively suppress false triggers caused by physiological noise such as blinking and muscle activity. Therefore, a method is urgently needed to solve at least one of the above problems. Summary of the Invention

[0004] This invention discloses an EEG headband interactive control method and device based on attention level. It identifies the attention characteristic baseline through spectrum decoupling analysis, tracks the high response frequency band nucleus to form a control channel, determines the optimal trigger position based on the attention resilience index, determines the low interference response window by aligning the brainwave feature chain with the attention stability coefficient, and suppresses false trigger interference to generate interactive control commands, thereby realizing real-time matching between teaching intervention and students' cognitive state.

[0005] The first aspect of this invention proposes a brainwave headband interactive control method based on attention level, comprising the following steps: Multi-band EEG signals and user interaction state parameters are collected from the EEG headband. Spectral decoupling analysis is performed on the multi-band EEG signals to identify the attention feature baseline. An attention stability coefficient is derived from the user interaction state parameters. The attention feature baseline and the attention stability coefficient are correlated to establish a cognitive state mapping relationship. Based on the cognitive state mapping relationship, the high-response frequency band core is tracked and located. Frequency band fusion analysis is performed on the high-response frequency band core to form a control channel. Attention fluctuation data is collected along the control channel. An interactive control network is constructed based on the attention fluctuation data. Multi-task response analysis is performed on the interactive control network to identify intent switching points. Cognitive load parameters of the intent switching points are extracted and classified into low-load and high-load areas. The transition process from the high-load area to the low-load area is identified and the recovery rate is extracted. An attention resilience index is constructed based on the recovery rate. The optimal triggering position is determined in the low-load area based on the attention resilience index. Multi-band EEG data is extracted based on the optimal triggering position to construct a brainwave feature chain. The brainwave feature chain is aligned with the attention stability coefficient to determine a low-interference response window. An adaptive threshold group is set in the low-interference response window. The adaptive threshold group is coupled with the attention fluctuation data to identify false triggering areas. Redundant brainwave components are extracted from the false triggering areas. Based on the redundant brainwave components, the interactive control network is used to perform false triggering suppression to generate interactive control commands.

[0006] A second aspect of this invention provides an EEG headband interactive control device based on attention level, comprising: The signal acquisition module is used to acquire multi-band EEG signals and user interaction state parameters from the EEG headband, perform spectral decoupling analysis on the multi-band EEG signals to identify attention feature baselines, derive attention stability coefficients from the user interaction state parameters, and establish a cognitive state mapping relationship by associating the attention feature baselines with the attention stability coefficients. The channel construction module is used to track and locate the high-response frequency band core according to the cognitive state mapping relationship, perform frequency band fusion analysis on the high-response frequency band core to form a control channel, collect attention fluctuation data along the control channel, and construct an interactive control network based on the attention fluctuation data; The resilience assessment module is used to perform multi-task response analysis on the interactive control network to identify intention switching points, extract the cognitive load parameters of the intention switching points, classify them into low-load and high-load areas, identify the transition process from the high-load area to the low-load area, extract the recovery rate, and construct an attention resilience index based on the recovery rate. The threshold setting module is used to determine the optimal trigger position in the low load area based on the attention resilience index, extract multi-band EEG data based on the optimal trigger position to construct a brainwave feature chain, align the brainwave feature chain with the attention stability coefficient to determine a low interference response window, and set an adaptive threshold group in the low interference response window. The instruction generation module is used to couple the adaptive threshold group with the attention fluctuation data to identify the false triggering area, extract redundant brainwave components from the false triggering area, and perform false triggering suppression to generate interactive control instructions based on the redundant brainwave components.

[0007] The beneficial effects of this invention are reflected in the following points: First, by performing spectral decoupling analysis on multi-band EEG signals to extract dominant frequency components, and identifying the baseline of attention characteristics based on the convergence region of the frequency stability trajectory, a cognitive state mapping relationship is established by combining the attention stability coefficient derived from the interaction parameters, and then a control channel is formed by tracking the high-response frequency band kernel to perform frequency band fusion analysis. Attention fluctuation data is collected along the control channel to construct an interactive control network, thus realizing a complete closed loop from static cognitive feature recognition to dynamic attention fluctuation tracking, providing neurophysiological technical support for real-time monitoring of the evolution of students' cognitive state.

[0008] Second, by identifying the transition process from high-load to low-load areas, a recovery rate is extracted to construct an attention resilience index. Based on this index, the optimal triggering position is determined within the low-load area. Furthermore, a low-interference response window is determined by aligning the brainwave feature chain with the attention stability coefficient. An adaptive threshold group is set within this window, achieving dynamic matching of the timing and intensity of teaching interventions. This ensures that content delivery or difficulty adjustment is implemented during the window period when students have sufficient cognitive resources and are focused, avoiding ineffective interventions during periods of cognitive fatigue or inattention. Third, by coupling the adaptive threshold group with attention fluctuation data, false triggering areas are identified. Redundant brainwave components are extracted and false triggering suppression is implemented, filtering out physiological noise such as blinking and muscle activity, as well as external electromagnetic interference, from contaminating the EEG signal. This avoids misjudging transient external interference or physiological noise as genuine changes in cognitive state, ensuring that interactive control commands are generated based on reliable EEG data, thus improving the accuracy and stability of decision-making.

[0009] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and do not limit this application. Attached Figure Description

[0010] The accompanying drawings illustrate specific examples of the technical solutions described in this invention and, together with the detailed embodiments, form part of the specification, serving to explain the technical solutions, principles, and effects of this invention.

[0011] Unless otherwise specified or defined, the same reference numerals in different figures represent the same or similar technical features, and different reference numerals may be used to represent the same or similar technical features.

[0012] Figure 1 This is a flowchart illustrating an EEG headband interactive control method based on attention level according to the present invention.

[0013] Figure 2 This is a structural block diagram of an EEG headband interactive control device based on attention level according to the present invention. Detailed Implementation

[0014] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods have been omitted so as not to obscure the description of this application with unnecessary detail.

[0015] It should be understood that, when used in this application specification and the appended claims, the term "comprising" indicates the presence of the described features, integrals, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or a collection thereof.

[0016] It should also be understood that the term “and / or” as used in this application specification and the appended claims means any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0017] References to "one embodiment" or "some embodiments" as described in this specification mean that one or more embodiments of this application include a specific feature, structure, or characteristic described in connection with that embodiment. Therefore, the phrases "in one embodiment," "in some embodiments," "in other embodiments," "in still other embodiments," etc., appearing in different parts of this specification do not necessarily refer to the same embodiment, but rather mean "one or more, but not all, embodiments," unless otherwise specifically emphasized. The terms "comprising," "including," "having," and variations thereof mean "including but not limited to," unless otherwise specifically emphasized.

[0018] The technical solutions of the embodiments of this application will be described below.

[0019] like Figure 1 As shown, this embodiment of the invention provides a brainwave headband interactive control method based on attention level, including the following steps S110-S150: Step S110: Collect multi-band EEG signals and user interaction state parameters from the EEG headband; perform spectral decoupling analysis on the multi-band EEG signals to identify the attention feature baseline; derive the attention stability coefficient from the user interaction state parameters; and establish a cognitive state mapping relationship by associating the attention feature baseline with the attention stability coefficient.

[0020] Specifically, multi-band EEG signals and user interaction parameters were collected using an EEG headband. Students wore the headband, with the electrodes contacting corresponding locations on the student's prefrontal cortex. The raw EEG signals were preprocessed: a notch filter was used to eliminate power frequency interference from classroom lighting; a high-pass filter was used to remove baseline drift caused by blinking and head movements; and a low-pass filter was used to remove high-frequency noise such as electromyography (EMG). EEG signal components in each frequency band were extracted using a band-pass filter, including the delta (δ), theta (θ), alpha (α), beta (β), and gamma (γ) bands. When students were focused on listening to a lecture, beta wave activity in the prefrontal cortex significantly increased; when they were distracted or drowsy, the energy of low-frequency theta and alpha waves increased; and when actively thinking about problem-solving, both beta and gamma waves were active. These energy change patterns in different frequency bands reflected the neural activity characteristics of the student's brain under different cognitive states. User interaction parameters were also collected simultaneously, including response time, accuracy rate, page switching frequency, and interaction pause duration. The response time is recorded as the time interval between when a question is presented and when a student submits their answer. The accuracy rate is calculated as the ratio of the number of correct answers to the total number of questions. The page switching frequency is recorded as the number of times a student jumps through pages in the electronic textbook. The interaction pause duration is recorded as the time interval between two consecutive actions by a student. In the early stages of classroom learning, students are focused, answer questions quickly and accurately, and page switching is regular and orderly. As the course progresses and fatigue accumulates, the response time gradually increases, the accuracy rate decreases, and the page switching frequency increases and becomes irregular.

[0021] In some embodiments, the step of performing spectral decoupling analysis on the multi-band EEG signals to identify the attention feature baseline includes: constructing a frequency band energy spectrum based on the multi-band EEG signals; performing peak localization on the frequency band energy spectrum to extract the dominant frequency components; expanding the dominant frequency components in the time dimension to form a frequency stability trajectory; and determining the attention feature baseline based on the convergence region of the frequency stability trajectory.

[0022] A frequency band energy spectrum was constructed based on multi-band EEG signals. Short-time Fourier transforms were performed on the EEG signals of each frequency band to convert the time-domain signals into a time-frequency domain representation. The spectral energy distribution within each time window was calculated, with the spectral energy obtained by squared amplitude. The spectral energy of each time window was arranged according to the frequency dimension, forming a frequency-energy two-dimensional matrix. The frequency-energy matrix of each frequency band was normalized to eliminate differences in energy magnitude between different frequency bands. The normalized frequency band energy E_norm(f,t) = E(f,t) / E_total(t) was calculated, where E_norm(f,t) is the normalized frequency band energy, E(f,t) is the original energy of frequency f at time t, and E_total(t) is the total energy of all frequency bands at time t. The normalization process maps the energy of each frequency band to a uniform range of 0 to 1, making the energy of different frequency bands comparable. The normalized frequency-energy data for each frequency band are integrated into a complete band energy spectrum, which covers the entire frequency range and reflects the energy distribution characteristics of each band in both time and frequency dimensions. When students are focused on listening, a significant energy peak appears in the β band, with a normalized energy value exceeding 0.6. When they are distracted, the energy peak in the α band is prominent, with the normalized value rising to around 0.5. When they are drowsy, the energy in the low-frequency θ band increases, with a normalized value exceeding 0.4. These energy distribution patterns reflect the spectral characteristics of students' cognitive states.

[0023] Peak location is performed on the frequency band energy spectrum to extract the dominant frequency components. The energy peak positions in the frequency band energy spectrum are identified; each peak corresponds to a local frequency point with the highest energy. The frequency value corresponding to each peak point is the dominant frequency component of that band. The dominant frequency component represents the core frequency of neural oscillations in the student's brain, reflecting the main activity patterns of the neural network. The number of dominant frequency components in each band is counted; a single dominant frequency indicates a simple neural activity pattern, while multiple dominant frequencies indicate a complex neural activity pattern. The energy percentage of the dominant frequency component is calculated; the energy percentage is equal to the ratio of the dominant frequency energy to the total energy of that band. A high energy percentage indicates that the dominant frequency occupies an absolute dominant position in that band. When students are deeply focused in class, the β band usually exhibits a single dominant frequency component with a stable frequency within a narrow range, and the energy is highly concentrated, showing a distinct single peak characteristic on the spectrogram. However, when attention is scattered, the dominant frequency in the β band is dispersed over a wider range, with a flat energy distribution, showing multiple low peaks on the spectrogram.

[0024] The dominant frequency components are unfolded over time to form a frequency stability trajectory. The dominant frequency values ​​are arranged chronologically to form a time series of dominant frequencies. A dominant frequency time series curve is plotted, with time on the horizontal axis and dominant frequency values ​​on the vertical axis; the curve reflects the evolution trend of the dominant frequency over time. The fluctuation characteristics of the dominant frequency time series are analyzed, and the mean and standard deviation of the series are calculated. A small standard deviation indicates stable dominant frequencies. Stable and fluctuating segments of the dominant frequency are identified. In stable segments, the dominant frequency values ​​fluctuate slightly around the mean; in fluctuating segments, the dominant frequency values ​​change rapidly, frequently crossing different frequency ranges. The duration of stable segments is calculated; a longer duration indicates that the student's brain can maintain a stable neural oscillation pattern for a longer period. The dominant frequency time series is smoothed to form a frequency stability trajectory. In the early stages of classroom learning, the frequency stability trajectory exhibits significant fluctuations, with the curve rising and falling frequently, corresponding to the student's brain neural network adapting to new learning content and task requirements. As the course progresses and students gradually become more engaged, the trajectory gradually stabilizes, with the curve remaining stable around a certain frequency level, corresponding to the stabilization of neural activity patterns and the establishment of a learning rhythm.

[0025] The baseline of attention characteristics is determined based on the convergence region of the frequency stability trajectory. The fluctuation characteristics of the frequency stability trajectory are analyzed to identify the convergence region. The convergence region is defined as the time segment in the trajectory where the dominant frequency fluctuation amplitude is the smallest and remains stable. The standard deviation of the sliding window of the frequency stability trajectory is calculated; the trough of the sliding window standard deviation corresponds to the candidate position of the convergence region. A convergence region judgment threshold is set; segments with a sliding window standard deviation less than the threshold are marked as convergence regions. Convergence regions with longer durations are selected, and the region with the longest duration among multiple convergence regions is chosen as the final convergence region. After a student enters a state of deep focus, the dominant frequency in the β band fluctuates slightly within a narrow range and remains stable for a long time. The frequency stability trajectory exhibits a clear plateau characteristic; this time period is a typical convergence region, marking that the student's cognition has entered an efficient and stable working mode. The frequency band energy spectrum data for the corresponding time period is reviewed, and the average energy value of each frequency band within the convergence region is calculated. The average energy value represents the typical energy level of each frequency band in a stable focus state. The average value of the dominant frequency of each frequency band within the convergence region is calculated; the average value represents the characteristic frequency of each frequency band in a stable focus state. Calculate the β / α energy ratio and θ / β energy ratio within the convergence region, and integrate the average energy value, characteristic frequency and energy ratio of each frequency band extracted within the convergence region into a focus characteristic baseline.

[0026] The attention stability coefficient was derived from user interaction state parameters. The temporal variation patterns of each interaction parameter were analyzed, and the standard deviation of the response time was calculated. A smaller standard deviation indicates stable response time and strong attention maintenance. The coefficient of variation (COP) of page switching frequency was calculated; the COP is equal to the ratio of the standard deviation to the mean. A smaller COP indicates regular switching behavior and stable attention concentration. The temporal distribution of answer accuracy was statistically analyzed. A consistently high accuracy rate indicates sustained and stable attention and good comprehension, while a gradually decreasing accuracy rate indicates attention decay over time or weak knowledge mastery. The distribution characteristics of interaction pause duration were analyzed. Pause durations concentrated in a moderate range indicate a stable thinking rhythm, while excessively short or long and scattered pause durations indicate unstable attention. During classroom problem-solving, students with stable attention exhibited a regular, periodic change in their answering rhythm, with relatively consistent thinking and answering times for each question, corresponding to a stable thinking-answering rhythm. In contrast, students with unstable attention showed random fluctuations in response time, sometimes answering hastily and sometimes pausing for extended periods, corresponding to a disordered thinking rhythm or frequent inattentiveness. By combining the stability indices of various interaction parameters, an attention stability coefficient S is constructed as follows: S = w1×(1-CV_response)+w2×ACC+w3×(1-CV_switch)+w4×R_pause, where CV_response is the coefficient of variation of the response time, ACC is the accuracy rate of the answer, CV_switch is the coefficient of variation of the page switching frequency, R_pause is the reasonableness score of the interaction pause duration, and w1 to w4 are weight coefficients. The higher the value of S, the stronger the attention maintenance ability.

[0027] A cognitive state mapping relationship is established by correlating the attention characteristic baseline with the attention stability coefficient. The energy ratios of each frequency band in the attention characteristic baseline, including the β / α energy ratio and the θ / β energy ratio, are analyzed. The correlation between the β / α energy ratio and the attention stability coefficient is analyzed. When both the β / α ratio and the attention stability coefficient are high, it indicates that the student is in a highly focused classroom learning state. When the β / α ratio is low but the attention stability coefficient is high, it indicates that the student is in a relaxed but stable state, possibly engaged in relaxed review or independent reading. A cognitive state mapping relationship is established, dividing the energy ratio space of the attention characteristic baseline into multiple regions, each corresponding to a specific learning state type. Regions with high β / α ratios map to a deep focused learning state, regions with moderate β / α ratios map to a normal learning state, regions with low β / α ratios map to a relaxed or distracted state, and regions with high θ / β ratios map to a drowsy state. Within each energy ratio region, the stability level of the learning state is further subdivided based on the value of the attention stability coefficient. During class, attentive listening corresponds to a higher β / α ratio and a higher stability coefficient, while breaks correspond to a lower β / α ratio and a moderate stability coefficient, and dozing off corresponds to a higher θ / β ratio and a lower stability coefficient.

[0028] Step S120: Based on the cognitive state mapping relationship, track and locate the high response frequency band core, perform frequency band fusion analysis on the high response frequency band core to form a control channel, collect attention fluctuation data along the control channel, and construct an interactive control network based on the attention fluctuation data.

[0029] Specifically, high-response frequency bands are located based on the cognitive state mapping relationship. The energy distribution characteristics of EEG frequency bands corresponding to each learning state in the cognitive state mapping relationship are extracted, and the response patterns of each frequency band energy under different learning states are analyzed. The energy change rate of each frequency band during learning state transitions is obtained; the energy change rate is equal to the ratio of the energy difference before and after the state transition to the time interval. The response intensity of each frequency band is statistically analyzed, and the response intensity is comprehensively evaluated by the absolute value of the energy change rate and the frequency of occurrence. A response intensity threshold is set, equal to 1.5 times the average response intensity of all frequency bands; frequency bands with response intensities exceeding the threshold are marked as high-response frequency bands. In classroom learning scenarios, when transitioning from a "distracted state" to a "deeply focused state," the energy of the β frequency band rapidly increases by more than 50% within 2 to 3 seconds, while the energy of the α frequency band simultaneously decreases by more than 30%. The response intensity of these two frequency bands is significantly higher than that of other frequency bands. The core frequency range of the high-response frequency band is identified; the core frequency range corresponds to the frequency sub-interval within the high-response frequency band where energy changes are most drastic. The core frequency range of the beta band is typically 18 to 25 Hz, while the core frequency range of the alpha band is typically 9 to 11 Hz. The high-response frequency nucleus records the EEG frequency bands and their core frequency ranges that are most sensitive to changes in learning states.

[0030] In some embodiments, performing band fusion analysis on the high-response frequency band core to form a control channel includes: extracting the phase consistency of each frequency band from the high-response frequency band core; performing cross-band coupling strength assessment on the phase consistency to generate a band coordination map; identifying the maximum coordination path in the band coordination map; and determining the control channel along the maximum coordination path.

[0031] Phase consistency of each frequency band is extracted from the high-response frequency band kernel. EEG signals from each frequency band in the high-response frequency band kernel are extracted, and Hilbert transform is performed on each signal to obtain the instantaneous phase. The instantaneous phase describes the oscillation position of the signal on the time axis, with a phase value ranging from 0 to 2π. The phase difference between each pair of frequency bands is obtained; the phase difference is equal to the difference between the instantaneous phases of the two frequency bands. A stable phase difference close to zero indicates high synchronization between the two frequency bands; a stable phase difference close to π indicates anti-phase synchronization; and random fluctuations in the phase difference indicate asynchrony. Phase consistency is quantified using the phase lock value (PLV): PLV = |mean(exp(i×Δφ))|, where Δφ is the phase difference, i is the imaginary unit, and mean(exp(i×Δφ)) is the average of the complex vectors over all time points. A high phase lock value indicates high synchronization between frequency bands, and a low phase lock value indicates asynchrony. The phase lock values ​​of all frequency band pairs within the high-response frequency band kernel are obtained to form a phase consistency matrix. When students are deeply focused in class, the beta and gamma bands work together in a highly coordinated manner, while the beta and theta bands operate relatively independently.

[0032] A cross-band coupling strength assessment is performed to evaluate phase consistency and generate a band coordination map. The coupling strength of each band pair is evaluated using the phase lock value in the phase consistency matrix. The coupling strength C_ij = PLV_ij × √(E_i × E_j) is calculated, where C_ij is the coupling strength between band i and band j, PLV_ij is the phase lock value of the two bands, and E_i and E_j are the energy levels of the two bands, respectively. The coupling strength comprehensively considers phase synchronization and energy level; band pairs with phase synchronization and high energy have a strong coupling relationship. A coupling strength threshold is set, which is equal to 1.2 times the average coupling strength of all band pairs. Band pairs with coupling strength exceeding the threshold are marked as strong coupling pairs. The spatial distribution characteristics of strong coupling pairs are analyzed to identify band combinations that form cooperative working clusters. Bands within a cooperative working cluster have strong coupling relationships with each other. The band coordination map uses frequency bands as nodes and coupling strength as the weight of the connecting edges. In the frequency band coordination map, core nodes correspond to frequency bands that play a key role in cognitive regulation, while peripheral nodes correspond to auxiliary frequency bands. During focused classroom learning, the β frequency band is usually located at the core of the map and has strong connections with the γ and α frequency bands, while the δ frequency band is located at the edge of the map.

[0033] Identify the maximum cooperative path in the frequency band cooperative graph. Analyze the topology of the frequency band cooperative graph to identify the main paths connecting the core nodes. The main paths are formed by a series of strongly coupled edges, with the frequency bands on the paths arranged according to the information transmission order. Obtain the cooperative strength of all possible paths in the graph; the cooperative strength of a path is equal to the product or weighted sum of the weights of all connecting edges on the path. Select the path with the highest cooperative strength as the maximum cooperative path. Analyze the starting and ending frequency bands of the maximum cooperative path. The starting frequency band is usually the information source or the initiator of the control signal, and the ending frequency band is usually the execution node or the frequency band related to the behavior output. In the process of classroom attention regulation, the maximum cooperative path usually starts from the β frequency band in the prefrontal cortex, passes through the γ frequency band in the central region, and finally reaches the α frequency band in the parietal cortex. This path corresponds to the complete regulation link from attention initiation, maintenance to relaxation.

[0034] The control channel is determined along the maximum coordination path. EEG signal features of each node frequency band along the maximum coordination path are extracted, including the center frequency, energy level, and phase information of the frequency band. The information transmission characteristics between adjacent nodes along the path are analyzed to obtain the time delay and transmission efficiency. The time delay is obtained by calculating the phase lag between adjacent nodes, and the transmission efficiency is quantified by coherence; high coherence indicates low information transmission loss. Key control nodes on the path are identified. Key control nodes have the strongest influence on downstream nodes and are least affected by upstream nodes. When students are attentively listening, the prefrontal cortex β band acts as a key control node. Changes in its energy state are transmitted downstream along the maximum coordination path to the γ and α bands. An increase in β band energy leads to an increase in γ band energy, while a decrease in α band energy occurs. The control channel uses the EEG signals of the key control nodes as input and the EEG signals of the path's endpoint as output, recording the information flow, transmission delay, and coupling relationship between frequency bands.

[0035] Attention fluctuation data is collected along the control channel. This data includes four key features: peak and trough positions, maintenance period, fluctuation amplitude, and phase lag time. Temporal data of the dominant frequency band of the EEG signal in the control channel are extracted, and envelope detection is performed on the dominant frequency band signal to obtain the envelope curve. The peak and trough positions of the envelope curve are identified; peaks mark moments of high attention, and troughs mark moments of low attention. These peak and trough positions constitute the first type of feature data. The time interval between adjacent peaks is calculated; this time interval is the attention maintenance period, constituting the second type of feature data. In a 45-minute class, students' attention typically exhibits 3 to 4 fluctuation cycles, each lasting 10 to 15 minutes. The amplitude difference between peaks and troughs is calculated; this amplitude difference is the attention fluctuation amplitude, constituting the third type of feature data. Small fluctuation amplitudes in the early stages indicate stable attention, while gradually increasing fluctuation amplitudes with accumulated fatigue indicate unstable attention. The EEG signal of the response frequency band in the control channel is extracted, the phase difference between the response frequency band and the dominant frequency band is calculated, and the phase difference is converted into time, which is the phase lag time. The phase lag time constitutes the fourth type of feature data.

[0036] An interactive control network is constructed based on attention fluctuation data. This network achieves adaptive control through three stages: state localization, trend prediction, and strategy generation. First, state localization involves extracting peak and trough positions from the attention fluctuation data, comparing the current moment with the most recent peak and trough moments to determine whether the student is currently in a peak, trough, or transition period, and outputting a real-time attention state label. Second, trend prediction involves extracting the maintenance period and fluctuation amplitude from the attention fluctuation data, analyzing the regularity of historical fluctuation cycles, and predicting the arrival time of the next peak or trough based on the duration of the previous cycle, outputting the attention change trend for the next 5 to 10 seconds. Finally, strategy generation combines the current state label and future trends to dynamically adjust the timing and difficulty level of learning tasks. When a student is about to enter a trough, a difficulty reduction instruction or a short rest is triggered 5 seconds in advance, based on the predicted trough arrival time. When a student is in a peak period and the trend is stable, a challenging task push instruction is output. Through the coordinated operation of these three stages, the interactive control network transforms attention fluctuation data into precise intervention and control instructions.

[0037] Step S130: Perform multi-task response analysis on the interactive control network to identify intention switching points, extract cognitive load parameters of intention switching points, classify them into low-load and high-load areas, identify the transition process from high-load to low-load areas and extract recovery rate, and construct attention resilience index based on recovery rate.

[0038] Specifically, multi-task response analysis was performed on the interactive control network to identify intent switching points. The strategy adjustment records in the interactive control network were analyzed, including information on task switching times, changes in task type, and adjustments in difficulty level. The temporal characteristics of these strategy adjustment records were analyzed to identify the time points when task types changed. At these task type change time points, abrupt changes in student interactive behavior were detected, including sudden increases in response time, abnormal increases in page dwell time, or sharp increases in error rates. These abrupt changes in interactive behavior indicate that students encountered cognitive challenges while adapting to new task types, requiring a readjustment of attentional resources and cognitive strategies. The EEG signal characteristics corresponding to the task type change time points were analyzed, including instantaneous changes in beta and theta band energy in the prefrontal cortex. For example, after completing a set of simple vocabulary questions, a student suddenly jumped to a function calculation problem requiring multiple steps, with a significantly prolonged page pause and frequent use of scratch paper for auxiliary calculations. This task type switch was captured, and significant changes in prefrontal cortex EEG activity were detected. The common abrupt changes in multiple interactive behaviors and physiological indicators marked the intent switching point. The time points that meet the characteristics of interactive behavior mutation and EEG signal mutation are marked as intention switching points. Intention switching points record the key moments of students' learning task transitions.

[0039] Cognitive load parameters at intention switching points are extracted and categorized into low-load and high-load zones. EEG signal data are analyzed within 30-second time windows before and after the intention switching point to characterize cognitive load levels. The energy ratio R_θ_α of the theta band to the alpha band in the prefrontal cortex is calculated; R_θ_α greater than 1.5 indicates high cognitive load, and R_θ_α less than 0.8 indicates low cognitive load. The P300 amplitude (denoted as P300) in the parietal lobe is analyzed; a large P300 amplitude indicates high attentional resource input and cognitive load. The N200 amplitude (denoted as N200) in the frontal lobe is also analyzed; a large N200 amplitude indicates strong task conflict and high cognitive load. When a student attempts to solve a complex application problem requiring the integrated application of multiple knowledge points, repeatedly trying different solution paths without success, and frequently consulting notes for key formulas, EEG signal analysis indicates that the student has entered a state of high cognitive load. A comprehensive cognitive load score C = w1 × R_θ_α + w2 × P300 + w3 × N200 was constructed using EEG indicators, where each parameter is a normalized value of the corresponding indicator, and w1, w2, and w3 are weighting coefficients. A cognitive load level classification threshold of 0.6 was set; a comprehensive cognitive load score C greater than 0.6 was marked as a high-load zone, and C less than or equal to 0.6 was marked as a low-load zone.

[0040] In some embodiments, the step of identifying the transition process from a high-load region to a low-load region and extracting the recovery rate includes: extracting the prefrontal cortex activity intensity at the boundary between the high-load region and the low-load region; performing instantaneous rate of change tracking on the prefrontal cortex activity intensity to generate an activity decay trajectory; identifying decay inflection point events from the activity decay trajectory; and calibrating the frequency of occurrence of the decay inflection point events as the recovery rate.

[0041] Prefrontal cortex activity intensity was extracted at the boundary between the high-load and low-load zones. The boundary time window was defined as the period from 5 seconds before the end of the high-load zone to 5 seconds after the beginning of the low-load zone. EEG signals in the prefrontal cortex, including electrode locations such as F3, F4, and Fz, were monitored within this time window. Frequency band decomposition of the prefrontal cortex EEG signals was performed to separate the signal components in the β band (13-30 Hz). The instantaneous energy of the β band signal was obtained using the envelope curve obtained through Hilbert transform. Instantaneous energy reflects the intensity level of prefrontal cortex activity; high instantaneous energy indicates strong prefrontal cortex activity, and low instantaneous energy indicates weak prefrontal cortex activity. After completing a proof problem requiring deep thought, a student puts down their pen, leans back in their chair, and looks out the window. Within this boundary time window of transition from focused problem-solving to relaxation, the EEG activity intensity at each electrode location in the prefrontal cortex was captured as it gradually decreased from a high level. Spatial averaging of the instantaneous energy at each electrode location was performed to obtain the prefrontal cortex activity intensity. The time-series curves of prefrontal cortex activity intensity show the dynamic changes in prefrontal cortex activity during the transition from high load to low load.

[0042] For example, the step of performing instantaneous rate of change tracking on the prefrontal cortex activity intensity to generate an activity decay trajectory includes: extracting asynchronous change segments from the time series of the prefrontal cortex activity intensity; performing energy difference accumulation on the asynchronous change segments to obtain cognitive overload events; generating detuning intensity based on the peak amplitude of the cognitive overload events; and forming an activity decay trajectory based on the detuning intensity.

[0043] Asynchronous variation segments were extracted from the time series of prefrontal cortex activity intensity. The β-band activity intensity time series data at the F3 and F4 electrode positions in the left and right hemispheres of the prefrontal cortex were analyzed to assess the synchronicity of activity intensity at the two electrode positions. The cross-correlation function of the F3 and F4 activity intensities was calculated; the cross-correlation function reflects the similarity and time delay relationship between the two signals. A high cross-correlation number indicates high synchronization between the two hemispheres, while a low cross-correlation number indicates asynchrony or time delay. A synchronicity threshold was set, equal to a critical value of 0.6 for the cross-correlation number; time periods with cross-correlation numbers below the threshold were marked as asynchronous variation segments. Within these asynchronous variation segments, significant differences in prefrontal cortex activity were observed between the left and right hemispheres, with one side showing increased activity intensity while the other side showed decreased activity intensity, or the rates of change in activity intensity between the two sides differing significantly. When students were solving a solid geometry problem that required both spatial imagination and logical reasoning, they first tried to draw the solid figure on scratch paper for spatial analysis, and then switched to algebraic derivation to verify the angular relationship. In this process that required the left and right hemispheres to switch back and forth and work together, a significant asynchronous phenomenon was detected in the activity of the prefrontal cortex of the left and right hemispheres. The activity on the left side increased rapidly while the activity on the right side was relatively slow, indicating that there was a coordination imbalance between the left and right hemispheres in the task processing.

[0044] Cognitive overload events are identified by accumulating energy differences during asynchronous change segments. The instantaneous difference in activity intensity between the left and right prefrontal cortexes within the asynchronous change segment is calculated. This instantaneous difference equals the absolute value of the difference between the activity intensity of electrode F3 and electrode F4. A large instantaneous difference indicates a significant difference in activity intensity between the two hemispheres and a severe imbalance between the hemispheres. The instantaneous difference within the asynchronous change segment is integrated over time; the integral value is the accumulated energy difference. The accumulated energy difference reflects the degree of accumulated imbalance between the hemispheres within the asynchronous change segment; a large accumulated energy difference indicates a long duration or a severe degree of imbalance. A threshold for identifying cognitive overload events is set, equal to the upper quartile of the accumulated energy difference distribution. Asynchronous change segments where the accumulated energy difference exceeds the threshold are marked as cognitive overload events. Cognitive overload events indicate a significant imbalance in the student's cognitive system during that period, difficulty in coordinating between the left and right hemispheres, and uneven allocation of cognitive resources. When students were solving comprehensive problems that required both spatial reasoning and logical analysis, they tended to use only one side of their brain for a long period of time, failing to achieve effective coordination and switching between the left and right hemispheres. Through cumulative analysis, this period was determined to be a cognitive overload event, reflecting that the students experienced significant cognitive imbalance and pressure when working on this type of problem.

[0045] Dissonance intensity is generated based on the peak amplitude of cognitive overload events. The maximum difference in activity intensity between the left and right prefrontal cortexes during the cognitive overload event period is analyzed; this maximum value is the peak amplitude. Peak amplitude reflects the most severe imbalance between hemispheres during the cognitive overload event; a large peak amplitude indicates severe imbalance and high cognitive system stress. The duration of the cognitive overload event is calculated, equal to the difference between the end and start times. A longer duration indicates a prolonged imbalance state and prolonged cognitive system stress. Dissonance intensity is calculated as D = A_peak × T_duration, where D is the dissonance intensity, A_peak is the peak amplitude, and T_duration is the duration. Dissonance intensity comprehensively reflects the severity and duration of the cognitive overload event; a large dissonance intensity indicates that the student experienced strong and persistent cognitive imbalance stress. Student A, while answering comprehensive questions, consistently favored one side of the brain for single-type thinking, failing to achieve effective coordination between the left and right hemispheres for an extended period; therefore, strong and persistent cognitive imbalance stress was detected. When faced with similar questions, Student B also experienced an imbalance between the left and right hemispheres of the brain, but the imbalance was less severe and lasted for a shorter period of time.

[0046] An activity decay trajectory is constructed based on the detuning intensity. The detuning intensities of each cognitive overload event are arranged chronologically to form time-series data of detuning intensity. The changing trends of the time-series data of detuning intensity are analyzed to identify the rising, peak, and falling segments of detuning intensity. The time-series data of detuning intensity is normalized, with the normalized values ​​ranging from 0 to 1. Max-min normalization is used. The normalized detuning intensities are mapped to the decay process of prefrontal cortex activity intensity. Periods with high detuning intensity correspond to stages of slow prefrontal cortex activity intensity decay, while periods with low detuning intensity correspond to stages of rapid prefrontal cortex activity intensity decay. An activity decay trajectory is constructed through the mapping relationship between detuning intensity and prefrontal cortex activity intensity. During the recovery process after completing a complex task, students initially maintained a strong imbalance between the left and right hemispheres of their brain. The intensity of prefrontal cortex activity decreased slowly, and then the imbalance gradually eased. Prefrontal cortex activity began to accelerate and decline, and finally the imbalance was basically eliminated. Prefrontal cortex activity rapidly decayed to the baseline level of the relaxed state. The resulting activity decay trajectory fully recorded this recovery process regulated by cognitive imbalance.

[0047] Identifying decay inflection point events from activity decay trajectories. Analyze the curvature characteristics of the activity decay trajectory; curvature represents the degree of curve bending, and locations with large curvature correspond to moments when the decay rate changes significantly. Calculate the second derivative of the activity decay trajectory; the extreme points of the second derivative correspond to the peak positions of the curvature. When the second derivative jumps from a negative value to a positive value, it indicates that the decay rate has changed from accelerating to decelerating, and this moment is marked as a decay inflection point. During cognitive recovery, the intensity of prefrontal cortex activity initially decays rapidly, followed by a sudden slowdown in the decay rate. This turning point in the rate of change marks the transition of the cognitive system from a rapid release phase to a stable adjustment phase, and this turning point is identified as a decay inflection point event. Set the criteria for determining decay inflection points, requiring the transition amplitude of the second derivative to exceed a set threshold to ensure that the identified inflection points have physiological significance rather than random fluctuations. Collect all decay inflection points that meet the criteria within the boundary time window to form a set of decay inflection point events. Decay inflection point events mark key turning points in the cognitive system recovery process.

[0048] The frequency of decay inflection point events is defined as the recovery rate. The frequency of decay inflection point occurrences is obtained by calculating the ratio of the total number of decay inflection point events identified within the boundary time window to the window duration. The recovery rate is equal to the ratio of the total number of decay inflection point events to the total duration, expressed as events per minute or events per second. A large recovery rate indicates frequent decay inflection point events, with the prefrontal cortex activity decaying through multiple speed adjustments and exhibiting strong rhythmicity. A small recovery rate indicates infrequent decay inflection point events, with the prefrontal cortex activity decaying smoothly and monotonously, lacking rhythmicity in the recovery process. After completing a challenging reasoning problem, Student A took a short break. Their prefrontal cortex activity exhibited a rapid, fluctuating recovery pattern, undergoing multiple rhythmic adjustments of tension-relaxation-re-tension-re-relaxation, demonstrating the cognitive system's flexible self-regulation ability. Under the same conditions, Student B's prefrontal cortex activity showed a single, slow downward trend, with a relatively smooth recovery process but lacking obvious rhythmic regulatory characteristics.

[0049] In some embodiments, constructing an attention resilience index based on recovery rate includes: expanding the recovery rate in the frequency domain to extract recovery spectrum features; performing β-wave decay rate tracking on the recovery spectrum features to generate a memory consolidation index; extracting a duration parameter from the memory consolidation index; and establishing an attention resilience index based on the duration parameter.

[0050] The recovery rate is expanded in the frequency domain to extract its spectral features. A Fast Fourier Transform (FFT) is performed on the recovery rate time-series data to convert the time-domain signal into a frequency-domain representation. The power spectral density (PSD) of the recovery rate is calculated, describing the energy distribution of the recovery rate across different frequency components. The PSD curve is analyzed to identify the dominant frequency components and energy peak positions. The dominant frequency components correspond to the main rhythm of the recovery process; higher frequencies indicate a faster recovery pace, while lower frequencies indicate a slower pace. The peak frequency, peak energy, and bandwidth of the PSD curve are extracted. The peak frequency reflects the core rhythm frequency of the recovery process, the peak energy reflects the intensity of the rhythm, and the bandwidth reflects the stability of the recovery rhythm. A student experienced multiple transitions from challenging problems to short breaks during a lesson. Spectral analysis of their recovery rate showed that the student's cognitive recovery process exhibited stable periodic rhythmic characteristics. The rhythm of each recovery process was relatively consistent, and the energy concentration of the recovery rhythm was high with a narrow fluctuation range, indicating that the student's cognitive recovery mechanism had good stability and consistency. The extracted spectral features include peak frequency, peak energy, and bandwidth.

[0051] A memory consolidation index (MCI) is generated by tracking the beta-wave decay rate on the recovered spectrum characteristics. Time-series data of prefrontal beta energy (13-30 Hz) during recovery are analyzed to track the decay of beta-wave energy from peak to baseline. The slope of the beta-wave energy decay curve is calculated; the slope is equal to the ratio of energy change to time change. A larger absolute slope indicates faster beta-wave energy decay, rapid weakening of prefrontal activity, and rapid release of cognitive resources. An optimal beta-wave decay rate range is defined, typically between 0.8 and 1.2 microvolts per second. The degree of proximity between the actual beta-wave decay rate and the optimal range is calculated, with a proximity value between 0 and 1. The memory consolidation index (MCI) is calculated as: MCI = w1 × P_proximity + w2 × E_peak_norm, where P_proximity is the degree of proximity of the beta-wave decay rate to the optimal range, E_peak_norm is the normalized peak energy value of the recovered spectrum, and w1 and w2 are weighting coefficients satisfying w1 + w2 = 1. After completing a set of exercises, students enter a brief, natural rest period. Prefrontal cortex beta wave activity decays smoothly at a moderate rate, avoiding both premature release of cognitive resources due to excessively rapid decay and excessive tension due to excessively slow decay. This well-balanced recovery rhythm is most conducive to the effective consolidation of newly learned knowledge. The memory consolidation index reflects the degree to which cognitive state supports knowledge consolidation during the recovery process.

[0052] The duration parameter was extracted from the memory consolidation index. The temporal curve morphology of the memory consolidation index was analyzed to identify its rising, high-level maintenance, and falling segments. A high-level threshold of 0.7 was set for the memory consolidation index; an index greater than 0.7 indicates a highly efficient memory consolidation state. Continuous time periods where the memory consolidation index exceeded the threshold were identified, with the start point being the moment the index first exceeded the threshold and the end point being the moment the index fell below the threshold. The duration parameter is equal to the time difference between the end and start points, reflecting the duration for which memory consolidation efficiency can be maintained at a high level. Student A's memory consolidation efficiency quickly entered a high level and was maintained for a long time during the recovery process, indicating that after being relieved of cognitive stress, the student was able to maintain an ideal cognitive state conducive to knowledge consolidation. Although Student B could also enter a highly efficient memory consolidation state, the maintenance time was shorter, and they quickly exited the optimal consolidation window.

[0053] An attention resilience index was established based on the duration parameter. The duration parameter was normalized using a maximum-minimum normalization method, resulting in a value between 0 and 1. A longer duration parameter indicates that the student can maintain a state of efficient memory consolidation for an extended period, corresponding to a higher attention resilience index. A shorter duration parameter indicates that the student has difficulty maintaining an optimal cognitive state for a long time, corresponding to a lower attention resilience index. The normalized value of the duration parameter was mapped to the attention resilience index using a sigmoid function: R = 1 / (1 + e^(-k × (DT_norm - 0.5))), where R is the attention resilience index, DT_norm is the normalized value of the duration parameter, and k is the adjustment coefficient. Student A was able to quickly enter and maintain a state of efficient memory consolidation for a long time after completing a challenging task, indicating excellent cognitive recovery ability and learning resilience, effectively coping with learning pressure and making full use of the recovery period to consolidate knowledge. Although Student B could also recover from cognitive pressure, the duration of the efficient consolidation state was shorter, indicating that there is room for improvement in their cognitive recovery ability and learning resilience.

[0054] Step S140: Determine the optimal triggering position in the low-load area based on the attention resilience index, extract multi-band EEG data based on the optimal triggering position to construct a brainwave feature chain, align the brainwave feature chain with the attention stability coefficient to determine the low-interference response window, and set an adaptive threshold group within the low-interference response window.

[0055] Specifically, the optimal triggering position is determined within the low-load zone based on the attention resilience index. The temporal distribution characteristics of the attention resilience index within the low-load zone are analyzed to identify the peak value of the index. The peak value of the attention resilience index represents the moment when the student's cognitive system has the strongest recovery ability, indicating sufficient cognitive resources and high system stability. The trend of attention resilience index changes before and after the peak value is analyzed to identify the starting point of the high-level plateau period. After completing a set of exercises, students enter a resting state. Monitoring shows that the student's cognitive recovery process exhibits a characteristic of rapid initial increase followed by stabilization. When recovery reaches an ideal state and begins to remain stable, this moment marks the student entering the most suitable cognitive window for receiving new learning content. The interactive behavior characteristics of students within the low-load zone are analyzed, including page browsing patterns, mouse movement trajectories, and pause time distribution. Orderly page browsing, smooth mouse movement trajectories, and moderate pause times indicate that students are relaxed but maintaining attention. The optimal triggering position is determined by combining the peak value of the attention resilience index, the starting point of the high-level plateau period, and the stability of interactive behavior. The optimal triggering position marks the most suitable time for teaching intervention or content delivery within the low-load zone.

[0056] A brainwave feature chain was constructed by extracting multi-band EEG data based on the optimal trigger location. The brainwave feature chain consists of three parts: a preparatory phase, the trigger moment, and the subsequent response phase, recording the student's complete neural activity pattern before and after the optimal moment. Using the optimal trigger location as a reference point, a time window of 20 seconds was traced forward and extended backward to construct a time region encompassing the complete cognitive transition process. Energy time-series data of five frequency bands (δ, θ, α, β, and γ) were simultaneously collected within this time region. The energy data of each frequency band were normalized to eliminate individual differences and equipment fluctuations. When the student's rest state was nearing its end, the moment when their cognitive recovery reached its optimal state was used as the core. EEG characteristics of the cognitive preparation phase were captured forward, and EEG changes of the cognitive reactivation phase were traced backward, completely extracting the entire cognitive transition process from relaxation to preparation for re-engagement in learning. The coordinated change patterns of energy in each frequency band within the time window were analyzed, and the phase relationships and energy transfer characteristics between frequency bands were identified. This multi-dimensional EEG information was integrated into the brainwave feature chain. The energy data of each frequency band were concatenated in chronological order to construct a multi-dimensional brainwave feature vector sequence. The brainwave feature chain, with time as the main axis and five frequency bands as the dimensions, forms a complete neurodynamic map of the student before and after the optimal triggering moment.

[0057] In some embodiments, aligning the brainwave feature chain with the attention stability coefficient to determine a low-interference response window includes: performing gamma-wave burst detection on the brainwave feature chain to construct a candidate event set; evaluating the event credibility of the candidate event set based on the attention stability coefficient; selecting high-credibility events from the event credibility to form an understanding event confirmation set; and identifying the time interval of the understanding event confirmation set as a low-interference response window.

[0058] For example, the step of performing gamma wave burst detection on the brainwave feature chain to construct a candidate event set includes: extracting gamma band energy mutation trajectories from the brainwave feature chain; identifying rising and falling edges in the gamma band energy mutation trajectories to form a burst envelope diagram; performing duration screening on the burst envelope diagram to extract time windows that conform to cognitive integration characteristics; and constructing a candidate event set based on the time windows that conform to cognitive integration characteristics.

[0059] The energy mutation trajectory in the gamma band was extracted from the brainwave feature chain. This trajectory records the time path of rapid changes in gamma wave energy, reflecting the dynamic characteristics of students' cognitive integration process. The 30-100Hz gamma band EEG signal components in the brainwave feature chain were separated, and a Hilbert transform was performed on the gamma band signal to obtain the instantaneous amplitude envelope. The instantaneous amplitude envelope reflects the real-time change level of gamma wave energy; the peak of the envelope curve corresponds to the high-energy moment of gamma wave activity. First-order difference operations were performed on the instantaneous amplitude envelope to obtain the time-series data of the energy change rate. The energy change rate V = ΔE / Δt was calculated, where ΔE is the energy change and Δt is the time interval. A positive energy change rate indicates that gamma wave energy is rising, a negative value indicates that energy is falling, and a large absolute value indicates a rapid energy change. When students encounter key information requiring deep thinking during learning, gamma wave energy rises rapidly, and when processing is completed or attention shifts, the energy drops rapidly. These significant energy jumps constitute the key nodes of the gamma band energy mutation trajectory. Identify time points where the absolute value of the energy change rate exceeds a set threshold; these time points correspond to moments when gamma-wave energy undergoes significant abrupt changes. Connect these energy abrupt change time points in chronological order to form a continuous sequence of abrupt change events.

[0060] In the energy mutation trajectory of the gamma band, rising and falling edges are identified to form a burst envelope diagram. The burst envelope diagram clearly shows the complete cycle of each gamma wave burst by marking the start and end times of the energy jump. The changing trend of the energy mutation trajectory of the gamma band is analyzed to identify the rising segment (energy jumping from low to high levels) and the falling segment (energy falling back from high to low levels). The starting point of the rising segment is marked as the rising edge, corresponding to the start time of the gamma wave burst, and the starting point of the falling segment is marked as the falling edge, corresponding to the end time of the gamma wave burst. The pairing relationship between rising and falling edges is analyzed; a complete gamma wave burst event contains one rising edge and one immediately following falling edge, and the time period between them is the burst duration. In the process of students understanding an abstract concept, the rapid rise of gamma wave energy signifies the start of cognitive integration, and the rapid decline of energy after integration signifies the completion of the processing. This complete energy rise-maintenance-fall pattern presents a clear peak structure in the burst envelope diagram. All identified rising and falling edge positions are marked; each rising-falling edge pair corresponds to a complete cognitive burst event in the burst envelope diagram.

[0061] The burst envelope diagram is subjected to duration screening to extract time windows that conform to the characteristics of cognitive integration. Each gamma-wave burst event is extracted from the burst envelope diagram, and the duration of each gamma-wave burst event is calculated. The duration is equal to the difference between the falling edge time and the rising edge time. The distribution characteristics of the burst event durations are analyzed to identify time ranges that conform to the cognitive integration process. True cognitive integration events usually have a reasonable duration; bursts that are too short may be random noise, while bursts that are too long may be persistent high cognitive load rather than instantaneous integration. A duration screening interval for cognitive integration events is set; the lower limit of the interval excludes excessively short noise events, and the upper limit excludes excessively long non-integration events. When a student suddenly thinks of a solution during problem-solving, a brief but strong gamma-wave burst occurs, and its duration precisely matches the time characteristics of insightful cognitive integration. While gamma waves remain at a high level during prolonged calculations, their excessive duration does not conform to the typical characteristics of instantaneous cognitive integration events. Burst events with durations falling within a reasonable interval are selected; these event time windows are the time windows that conform to the characteristics of cognitive integration.

[0062] A candidate event set is constructed based on time windows that conform to cognitive integration characteristics. The start time, end time, and peak energy information of all time windows conforming to cognitive integration characteristics are integrated and structured into the candidate event set. The data structure includes time dimension information, energy dimension information, and duration information. The candidate event set is deduplicated, merging events with high temporal overlap to avoid duplicate labeling of the same cognitive process. The peak gamma-wave energy of each candidate event in the set is analyzed; peak energy reflects the intensity of information processing during cognitive integration. Students experience multiple key points requiring in-depth processing during the learning process. Candidate locations of these cognitive integration moments are captured using gamma-wave burst characteristics. Each candidate location records the time window where information integration may occur and the intensity of neural activity. All candidate events are arranged chronologically to form an ordered event sequence.

[0063] The credibility of events generated from a candidate event set is evaluated based on the attention stability coefficient. For each candidate event in the set, the attention stability coefficient at the moment of occurrence is analyzed to determine the degree of concentration and stability of the student's attention at that time. A high attention stability coefficient indicates that the student's attention is highly concentrated at that moment, and the gamma wave burst is likely to reflect genuine cognitive processing. A low attention stability coefficient indicates that the student's attention is scattered at that moment, and the gamma wave burst may be caused by random noise or external interference. An event credibility evaluation model is established, where the event credibility is equal to the weighted combination of the gamma wave burst intensity and the attention stability coefficient. The gamma wave burst intensity reflects the intensity of neural activity, and the attention stability coefficient reflects the reliability of the cognitive environment. If a student suddenly experiences a gamma wave burst while focusing on reading learning materials, and their attention remains highly concentrated, this event is judged as a highly credible moment of genuine cognitive integration. Although a gamma wave burst is detected when the student's attention is scattered, the event is judged to have lower credibility due to unstable attention, and may be a interference signal. The generated event credibility quantifies the authenticity level of the candidate events.

[0064] The process involves selecting high-confidence events from a pool of event credibility to form a confirmation set of comprehension events. This confirmation set records genuine moments of deep understanding that have undergone multiple verifications and serves as a key basis for identifying low-interference response windows. A credibility threshold is set to ensure the selected events have high authenticity. Candidate events with credibility exceeding the threshold are identified and included in the initial candidate list of the confirmation set. Further verification is performed on the events in the initial candidate list, analyzing the continuity of EEG signals before and after the high-confidence candidate events to verify whether gamma wave bursts are accompanied by other cognitive processing features such as transient inhibition of alpha waves or co-enhancement of beta waves. Simultaneous analysis of the coordinated activity patterns of the prefrontal and parietal lobes at the corresponding moments of the high-confidence candidate events verifies whether a characteristic pattern of multi-brain-region coordination is present. When a student comprehends a complex concept, not only do they experience gamma wave bursts and highly focused attention, but alpha wave activity is synchronously suppressed while beta wave activity is co-enhanced. The prefrontal and parietal lobes exhibit a highly synchronized co-activation pattern. These multiple EEG features collectively verify that this is a genuine moment of deep understanding, and this moment is formally confirmed and added to the confirmation set of comprehension events.

[0065] The time gaps between comprehension event confirmation sets are identified as low-interference response windows. The temporal distribution of each comprehension event in the confirmation set is analyzed to identify the time intervals between adjacent comprehension events. These time gaps represent the transitional phase after a student completes a cognitive breakthrough but before entering the next stage of deeper understanding. During this transition, the student's cognitive system is relatively stable; they have integrated the previous concept but haven't yet begun processing the next complex information. The cognitive load is moderate, and attention remains focused. After understanding the method for determining the monotonicity of a function, students begin browsing related exercises. At this point, they haven't delved into the problem-solving process and are in the preparation phase from understanding to application. The cognitive system remains alert but not under high-intensity processing pressure. This phase is identified as a suitable supplementary window for providing prompts or guidance. A minimum duration requirement is set for each time gap; gaps that are too short are insufficient to form a stable intervention window. Time gaps that meet the duration requirement are identified and designated as low-interference response windows, marking additional time regions where the student's cognitive system is in an open and stable state.

[0066] An adaptive threshold group is set within the low-interference response window. The adaptive threshold group is dynamically adjusted according to the student's real-time cognitive state and includes two core dimensions, namely, the cognitive load threshold T_load and the concentration fluctuation amplitude threshold T_amplitude. Analyze the student's cognitive load level within the analysis window, and characterize the current cognitive state through electroencephalogram indicators such as the θ / α energy ratio R_θ_α, P300 wave amplitude, and N200 wave amplitude. The cognitive load threshold T_load is dynamically adjusted based on the comprehensive cognitive load score C: when C ≤ 0.4, set T_load to 0.7 to encourage exploration; when 0.4 < C ≤ 0.6, set T_load to 0.6 to maintain challenge; when C > 0.6, set T_load to 0.5 to provide additional support. The concentration fluctuation amplitude threshold T_amplitude is personalized based on the student's historical fluctuation pattern. Extract the mean μ_amplitude and standard deviation σ_amplitude of the fluctuation amplitude. For students with stable fluctuations, set T_amplitude = μ_amplitude + 1.5×σ_amplitude to sensitively detect anomalies; for students with variable fluctuations, set T_amplitude = μ_amplitude + 2.5×σ_amplitude to avoid misjudgment. The adaptive threshold group is adjusted in real time according to the change of the student's cognitive state, realizing the personalization and precision of teaching intervention.

[0067] Step S150: Couple the adaptive threshold group with the concentration fluctuation data to identify the mis-triggering area, extract the redundant brain wave components from the mis-triggering area, and perform mis-triggering suppression on the interaction control network based on the redundant brain wave components to generate an interaction control instruction.

[0068] In some embodiments, the coupling of the adaptive threshold group with the concentration fluctuation data to identify the mis-triggering area includes: setting the adaptive threshold group as a dynamic envelope boundary; performing out-of-bounds detection on the concentration fluctuation data based on the dynamic envelope boundary to extract the out-of-bounds time sequence; identifying short-term out-of-bounds events in the out-of-bounds time sequence; and determining the mis-triggering area based on the short-term out-of-bounds events.

[0069] An adaptive threshold group is set as the dynamic envelope boundary. The dynamic envelope boundary consists of an upper and lower boundary. The upper boundary corresponds to the upper limit of each dimension's threshold, and the lower boundary corresponds to the lower limit of the threshold. Together, they form the permissible range for attention fluctuation data. The threshold parameters of each dimension in the adaptive threshold group are converted into time-varying functions, which describe the dynamic adjustment of the thresholds over time. For the cognitive load threshold, adjustments are made based on the progress stage of the learning task and the student's accumulated fatigue, mapping the adjusted threshold to the vertical range of the dynamic envelope boundary. For the attention fluctuation amplitude threshold, it is set based on the student's individual characteristics and historical fluctuation patterns. Students with stable fluctuation patterns correspond to a narrower boundary range, while students with variable fluctuation patterns correspond to a wider boundary range. These personalized parameters determine the width of the dynamic envelope boundary. In the early stages of classroom learning, when the student's cognitive system is in the adaptation phase, the dynamic envelope boundary is set relatively loosely to accommodate normal fluctuation adjustments. As learning enters a stable period, the boundary gradually tightens to improve the sensitivity of anomaly detection. A dynamic envelope boundary curve is plotted, with time on the horizontal axis and the parameter values ​​of each dimension on the vertical axis. The curve clearly shows the evolution trajectory of the thresholds over time and the limits of permissible fluctuation range.

[0070] Based on the dynamic envelope boundary, boundary crossing detection is performed on attention fluctuation data to extract the boundary crossing time sequence. The boundary crossing time sequence records all time points when attention fluctuations exceed the allowable range, which is the basic data for identifying abnormal states and false triggers. The time series curves of each dimension of attention fluctuation data are superimposed and compared with the upper and lower boundary curves of the dynamic envelope boundary to monitor whether the feature curves cross the boundary in real time. The time point when the feature curve crosses the envelope boundary is identified. The moment when the curve crosses from inside the boundary to outside the boundary is marked as the boundary crossing start point and recorded in the boundary crossing time sequence. The moment when the curve returns from outside the boundary to inside the boundary is marked as the boundary crossing end point and recorded simultaneously. During class, if a student's mind wanders to something outside the class and briefly loses focus, their alpha wave energy suddenly rises and exceeds the upper boundary of the dynamic envelope boundary. The boundary crossing start point is captured and added to the boundary crossing time sequence. A few seconds later, the student realizes their inattentiveness and refocuses their attention, the alpha wave energy falls back within the boundary, and the boundary crossing end point is also recorded in the sequence. The boundary crossing time points of all dimension feature curves are counted, and the boundary crossing start points and end points of each dimension are integrated in chronological order. The out-of-bounds time sequence contains complete out-of-bounds event timestamps, providing a detailed time index for subsequent short-term out-of-bounds event identification and false trigger zone determination.

[0071] Identify short-duration boundary crossing events in the boundary crossing time sequence. Calculate the duration of each boundary crossing event in the sequence, where duration equals the time difference between the boundary crossing end point and the boundary crossing start point. Analyze the distribution characteristics of the duration of boundary crossing events and identify events with durations significantly shorter than the average. Set a threshold for judging short-duration boundary crossing events; events with durations less than this threshold are marked as short-duration boundary crossing events. For example, a student might be briefly distracted by a classmate standing up while working on a problem, causing a momentary fluctuation in brainwaves that crosses the envelope boundary. However, the student's attention quickly returns to the problem, and the entire boundary crossing process lasts less than two seconds. This brief boundary crossing event is more likely a false trigger caused by external interference than a genuine entry into a new cognitive state. Analyze the attention fluctuation pattern before and after the short-duration boundary crossing event to verify whether the state before and after the event remains consistent. If the state before and after the boundary crossing event is highly similar, it indicates that the boundary crossing is a transient fluctuation caused by random perturbation. Filter out boundary crossing events with short durations and consistent states before and after the event; these events are short-duration boundary crossing events.

[0072] False trigger zones are identified based on short-term boundary violations. These zones indicate time segments with low data reliability and need to be excluded from control decisions. The spatial distribution characteristics of short-term boundary violations are analyzed to identify the EEG frequency bands and electrode locations involved. Boundary violations involving only a single frequency band or electrode location indicate a limited interference source, likely a specific type of physiological noise or poor local electrode contact. Boundary violations involving multiple frequency bands and electrode locations but lacking coordination indicate a complex or highly random interference source. For example, when a student wears an educational headband, a sudden increase in contact impedance at one electrode due to hair obstruction causes a short-term abnormal fluctuation in the EEG signal recorded at that electrode location, while signals at other electrode locations remain normal. Spatial distribution analysis identifies this as a false trigger caused by a local electrode problem rather than a change in whole-brain cognitive state; this time segment is marked as part of the false trigger zone. The frequency distribution of short-term boundary violations is analyzed to statistically determine the occurrence frequency per unit time. Time segments with frequent occurrences of short-term boundary violations indicate severe interference and low data reliability, and should be included in the false trigger zone. Define criteria for identifying false trigger zones, requiring that the density of short-term boundary crossing events within the zone exceeds a set threshold or that the abnormality of a single boundary crossing event exceeds a significant level. Time segments that meet these criteria are marked as false trigger zones.

[0073] Redundant brainwave components were extracted from the false trigger zone. Redundant brainwave components refer to abnormal frequency energy components that do not reflect real cognitive activity, including electromyographic interference, electrooculographic interference, and equipment noise. The energy distribution characteristics of multi-frequency brainwaves were analyzed within the time window of the false trigger zone to identify abnormal frequency energy components inconsistent with typical cognitive state patterns. Normal cognitive state transitions are usually accompanied by coordinated changes in specific frequency bands, while false triggers often manifest as isolated mutations in a single frequency band or irregular random fluctuations in multiple frequency bands. When students are attentively listening to a lecture, electrooculographic interference signals generated by blinking suddenly enter the EEG recording. This interference is mainly concentrated in the low frequency band and lasts for a very short time, which is significantly different from the multi-frequency coordinated adjustment patterns accompanying real cognitive state changes. These abnormal low-frequency energy components were identified and classified as redundant brainwave components. The temporal variation characteristics of the energy in each frequency band within the false trigger zone were analyzed, and the energy deviation D = |E_obs - E_base| / E_base was calculated, where E_obs is the observed energy and E_base is the baseline energy. Frequency band components with large energy deviations and isolated variation patterns are labeled as redundant brainwave components. Redundant brainwave components within the false triggering zone are identified and separated from the original brainwave signal. The separated redundant brainwave components record the time-frequency characteristics of various interference signals, providing clear filtering targets for false triggering suppression.

[0074] Interactive control instructions are generated by suppressing false triggering based on redundant brainwave components in the interactive control network. These instructions include parameters such as intervention type, timing, intensity, and duration, guiding the system to implement precise teaching interventions at the optimal time. The time-frequency domain characteristics of redundant brainwave components are analyzed, and targeted signal filtering strategies are designed to separate or suppress different types of interference components using appropriate filtering methods. A false triggering suppression mechanism is established in the interactive control network. This mechanism monitors attention fluctuation data in real time, and temporarily suspends control decisions based on data from that time period when redundant brainwave components are detected. The time period corresponding to the false triggering zone is removed from the effective decision window, ensuring that the interactive control network makes intervention decisions only based on reliable cognitive state data. The generation rules for interactive control instructions are constructed by integrating the time position of the low-interference response window, the parameter settings of the adaptive threshold group, and the filtering results of the false triggering suppression mechanism. The intervention timing of the interactive control instructions is locked within the low-interference response window, ensuring that intervention is implemented when students' attention is focused and their cognitive load is moderate. The intervention intensity is dynamically adjusted according to the adaptive threshold group, providing challenging tasks when cognitive resources are sufficient and supportive assistance when cognitive load is high. After completing a challenging problem, a student enters a brief cognitive recovery period. The system identifies that the student's attention resilience is high and they have entered a low-interference response window, making it the perfect time to push the next learning task. A content push command is generated, presenting the new problem at the student's optimal moment of focused attention and abundant cognitive resources. This maintains the continuity of learning while preventing the accumulation of cognitive fatigue. The generated interactive control commands provide the teaching system with precise intervention decisions, achieving intelligent adaptive control based on EEG characteristics.

[0075] To implement the EEG headband interactive control method based on attention level corresponding to the above method embodiments, and to achieve the corresponding functions and technical effects. See also Figure 2 , Figure 2 This diagram illustrates a structural block diagram of an EEG headband interactive control device 200 based on attention level, according to an embodiment of this application. For ease of explanation, only the parts relevant to this embodiment are shown. The EEG headband interactive control device 200 based on attention level provided in this embodiment includes: The signal acquisition module 201 is used to acquire multi-band EEG signals and user interaction state parameters of the EEG headband, perform spectral decoupling analysis on the multi-band EEG signals to identify the attention feature baseline, derive the attention stability coefficient from the user interaction state parameters, and establish a cognitive state mapping relationship by associating the attention feature baseline with the attention stability coefficient. The channel construction module 202 is used to track and locate the high-response frequency band core according to the cognitive state mapping relationship, perform frequency band fusion analysis on the high-response frequency band core to form a control channel, collect attention fluctuation data along the control channel, and construct an interactive control network based on the attention fluctuation data; The resilience assessment module 203 is used to perform multi-task response analysis on the interactive control network to identify intention switching points, extract the cognitive load parameters of the intention switching points, classify them into low-load and high-load areas, identify the transition process from the high-load area to the low-load area, extract the recovery rate, and construct an attention resilience index based on the recovery rate. The threshold setting module 204 is used to determine the optimal trigger position in the low load area based on the attention resilience index, extract multi-band EEG data based on the optimal trigger position to construct a brainwave feature chain, align the brainwave feature chain with the attention stability coefficient to determine a low interference response window, and set an adaptive threshold group in the low interference response window. The instruction generation module 205 is used to couple the adaptive threshold group with the attention fluctuation data to identify the false triggering area, extract redundant brainwave components from the false triggering area, and perform false triggering suppression to generate interactive control instructions based on the redundant brainwave components.

[0076] The aforementioned EEG headband interactive control device 200 based on attention level can implement the EEG headband interactive control method based on attention level described in the above method embodiments. The options in the above method embodiments are also applicable to this embodiment, and will not be detailed here. The remaining content of this application embodiment can be referred to the content of the above method embodiments, and will not be repeated in this embodiment.

[0077] The purpose of the above embodiments is to reproduce and derive the technical solution of the present invention by way of example, and to fully describe the technical solution, purpose and effect of the present invention. The purpose is to enable the public to have a more thorough and comprehensive understanding of the disclosure of the present invention, and not to limit the scope of protection of the present invention.

[0078] The above embodiments are not an exhaustive list based on the present invention, and there may be many other embodiments not listed. Any substitutions and improvements made without departing from the concept of the present invention are within the protection scope of the present invention.

Claims

1. A brainwave headband interactive control method based on attention level, characterized in that, include: Multi-band EEG signals and user interaction state parameters are collected from the EEG headband. Spectral decoupling analysis is performed on the multi-band EEG signals to identify the attention feature baseline. An attention stability coefficient is derived from the user interaction state parameters. The attention feature baseline and the attention stability coefficient are correlated to establish a cognitive state mapping relationship. Based on the cognitive state mapping relationship, the high-response frequency band core is tracked and located. Frequency band fusion analysis is performed on the high-response frequency band core to form a control channel. Attention fluctuation data is collected along the control channel. An interactive control network is constructed based on the attention fluctuation data. Multi-task response analysis is performed on the interactive control network to identify intent switching points. Cognitive load parameters of the intent switching points are extracted and classified into low-load and high-load areas. The transition process from the high-load area to the low-load area is identified and the recovery rate is extracted. An attention resilience index is constructed based on the recovery rate. The optimal triggering position is determined in the low-load area based on the attention resilience index. Multi-band EEG data is extracted based on the optimal triggering position to construct a brainwave feature chain. The brainwave feature chain is aligned with the attention stability coefficient to determine a low-interference response window. An adaptive threshold group is set in the low-interference response window. The adaptive threshold group is coupled with the attention fluctuation data to identify false triggering areas. Redundant brainwave components are extracted from the false triggering areas. Based on the redundant brainwave components, the interactive control network is used to perform false triggering suppression to generate interactive control commands.

2. The method according to claim 1, characterized in that, The step of performing spectral decoupling analysis on the multi-band EEG signals to identify attention feature baselines includes: A frequency band energy spectrum is constructed based on the multi-band EEG signals; Peak location was performed on the energy spectrum of the frequency band to extract the dominant frequency components; The dominant frequency components are expanded in the time dimension to form a frequency stability trajectory; The baseline of attention characteristics is determined based on the convergence region of the frequency stability trajectory.

3. The method according to claim 1, characterized in that, The step of performing frequency band fusion analysis on the high-response frequency band core to form a control channel includes: Extract the phase consistency of each frequency band from the high-response frequency band core; Perform cross-band coupling strength assessment on the phase consistency to generate a frequency band coordination map; Identify the maximum cooperative path in the frequency band cooperative spectrum; The control channel is determined along the maximum cooperative path.

4. The method according to claim 1, characterized in that, The extraction of the recovery rate during the transition from the high-load region to the low-load region includes: Prefrontal activity intensity is extracted at the boundary between the high-load region and the low-load region; Instantaneous rate of change tracking is performed on the prefrontal cortex activity intensity to generate an activity decay trajectory; Identify decay inflection point events from the activity decay trajectory; The frequency of occurrence of the decay inflection point event is calibrated as the recovery rate.

5. The method according to claim 1, characterized in that, The construction of the attention resilience index based on the recovery rate includes: The recovery rate is expanded in the frequency domain to extract the recovery spectral features; Perform β-wave decay rate tracking on the recovered spectral features to generate a memory consolidation index; Extract the retention duration parameter from the memory consolidation index; An attention resilience index is established based on the aforementioned duration parameter.

6. The method according to claim 1, characterized in that, The step of aligning the brainwave feature chain with the attention stability coefficient to determine the low-interference response window includes: Perform gamma wave burst detection on the brainwave feature chain to construct a candidate event set; The credibility of events generated from the candidate event set is evaluated based on the attention stability coefficient. From the stated event credibility, select high-credibility events to form an event confirmation set for understanding; The time intervals of the understanding event confirmation set are identified as low-interference response windows.

7. The method according to claim 1, characterized in that, The step of coupling the adaptive threshold group with the attention fluctuation data to identify the false trigger zone includes: Set the adaptive threshold group as the dynamic envelope boundary; Based on the dynamic envelope boundary, perform out-of-bounds detection on the attention fluctuation data to extract the out-of-bounds time sequence; Identify short-term boundary crossing events in the boundary crossing time sequence; The false triggering zone is determined based on the short-term out-of-bounds event.

8. The method according to claim 4, characterized in that, The step of performing instantaneous rate of change tracking on the prefrontal cortex activity intensity to generate an activity decay trajectory includes: Extract asynchronous variation segments from the time series of prefrontal cortex activity intensity; The energy difference is accumulated to obtain the cognitive overload event for the asynchronous change segment; Detuning intensity is generated based on the peak amplitude of the cognitive overload event; An activity attenuation trajectory is formed based on the aforementioned detuning intensity.

9. The method according to claim 6, characterized in that, The step of performing gamma wave burst detection on the brainwave feature chain to construct a candidate event set includes: Extract the energy mutation trajectory of the γ band from the brainwave feature chain; The rising and falling edges are identified in the energy mutation trajectory of the γ-band to form a burst envelope diagram; The duration of the burst envelope is filtered to extract time windows that conform to cognitive integration characteristics; A candidate event set is constructed based on the time window that conforms to the cognitive integration characteristics.

10. A brainwave headband interactive control device based on attention level, characterized in that, include: The signal acquisition module is used to acquire multi-band EEG signals and user interaction state parameters from the EEG headband, perform spectral decoupling analysis on the multi-band EEG signals to identify attention feature baselines, derive attention stability coefficients from the user interaction state parameters, and establish a cognitive state mapping relationship by associating the attention feature baselines with the attention stability coefficients. The channel construction module is used to track and locate the high-response frequency band core according to the cognitive state mapping relationship, perform frequency band fusion analysis on the high-response frequency band core to form a control channel, collect attention fluctuation data along the control channel, and construct an interactive control network based on the attention fluctuation data; The resilience assessment module is used to perform multi-task response analysis on the interactive control network to identify intention switching points, extract the cognitive load parameters of the intention switching points, classify them into low-load and high-load areas, identify the transition process from the high-load area to the low-load area, extract the recovery rate, and construct an attention resilience index based on the recovery rate. The threshold setting module is used to determine the optimal trigger position in the low load area based on the attention resilience index, extract multi-band EEG data based on the optimal trigger position to construct a brainwave feature chain, align the brainwave feature chain with the attention stability coefficient to determine a low interference response window, and set an adaptive threshold group in the low interference response window. The instruction generation module is used to couple the adaptive threshold group with the attention fluctuation data to identify the false triggering area, extract redundant brainwave components from the false triggering area, and perform false triggering suppression to generate interactive control instructions based on the redundant brainwave components.

Images