Intelligent safety helmet control method and system based on electroencephalogram characteristics

Abstract

The invention discloses a safety helmet intelligent control method and system based on electroencephalogram characteristics, and the method comprises the steps: decomposing a real-time electroencephalogram signal frequency spectrum to generate cognitive load frequency domain characteristics, and building an electroencephalogram control space; an electroencephalogram control space is scanned to identify a frequency band coupling unbalance point, a cooperative intervention opportunity is reversely derived, and a multi-channel adjustment mode is formed; mapping the stimulation frequency range to generate a frequency matching domain, and performing resonance analysis to determine a resonance intervention window; monitoring a frequency band in a resonance window to obtain an energy migration track, constructing a risk propagation chain to identify a cascade failure node and forming a key blocking anchor point; analyzing risk acceleration of the key blocking anchor points to form a state deterioration rate, fusing frequency domain features to generate dynamic delay tolerance, and generating variable speed control parameters in a fast-slow grading manner; the variable speed control parameter is compared with a threshold value to generate a control deviation signal, reverse phase modulation is carried out to generate a hedging stimulation waveform, a phase compensation execution instruction is formed, and accurate monitoring and safe intervention of the cognitive state are achieved.

Description

Technical Field

[0001] This invention relates to the field of brain-computer interface control technology, and in particular to a smart control method and system for safety helmets based on electroencephalogram (EEG) characteristics. Background Technology

[0002] In high-risk industrial settings such as coal mines, construction, and chemical plants, workers' cognitive state directly impacts operational safety. Fatigue, inattention, or cognitive overload can all lead to operational errors and accidents. Traditional safety management relies on work hour limits and regular rest periods, but these methods cannot reflect workers' true cognitive state in real time and are ill-suited to addressing individual differences and dynamically changing workloads.

[0003] Electroencephalogram (EEG) signals, as direct physiological indicators reflecting cognitive activity, provide a new technical approach for real-time monitoring of workers' cognitive states. However, existing methods still face many challenges: how to extract characteristic representations of cognitive load from multi-band EEG data and establish a correspondence between different fatigue levels and EEG patterns; how to determine the optimal moment for alert intervention and select appropriate stimulation methods and intensities when signs of fatigue or decreased attention are detected in workers; and how to promptly detect and correct deviations in stimulation output parameters due to various factors to avoid failed alerts or overstimulation. Summary of the Invention

[0004] This invention provides an intelligent control method and system for safety helmets based on electroencephalogram (EEG) characteristics. It establishes an EEG control space through spectrum decomposition to achieve hierarchical judgment of cognitive state, realizes forward-looking early warning by identifying frequency band imbalance and inversely deducing the timing of intervention, identifies key intervention nodes by constructing resonance analysis and risk propagation chain, generates dynamic response strategies by fusing state deterioration rates, and achieves adaptive compensation of parameter deviations through anti-phase modulation, thus providing intelligent safety assurance for high-risk industrial operations.

[0005] The first aspect of this invention proposes a smart control method for safety helmets based on electroencephalogram (EEG) characteristics, comprising the following steps: Real-time EEG signals of operators are collected, and the real-time EEG signals are spectrally decomposed to generate cognitive load frequency domain features. Based on the cognitive load frequency domain features, an EEG control space is established. A multi-band energy distribution matrix is ​​formed by scanning the EEG-controlled space. Frequency band coupling imbalance points are identified along the multi-band energy distribution matrix. The timing of collaborative intervention is deduced in reverse from the frequency band coupling imbalance points. A multi-channel regulation mode is formed based on the timing of collaborative intervention. The frequency range of the safety helmet stimulation and the target EEG frequency band are obtained. The stimulation frequency range is mapped to the EEG control space to generate a frequency matching domain. Based on the target EEG frequency band, the resonance intervention window is established by resonant analysis of the frequency matching domain. The multi-channel adjustment mode is used to perform frequency band monitoring within the resonance intervention window to obtain the energy migration trajectory. The risk propagation chain is constructed according to the energy migration trajectory to identify cascade failure nodes. The blocking priority assessment is performed on the cascade failure nodes to form key blocking anchor points. Risk acceleration analysis is performed on the key blocking anchor points to form a state deterioration rate. The state deterioration rate is then fused with the cognitive load frequency domain characteristics to determine the response time and generate a dynamic delay tolerance. Based on the dynamic delay tolerance, fast and slow gears are used to generate variable speed control parameters. The speed control parameters are compared with a preset safety threshold to generate a control deviation signal. The control deviation signal is then inversely modulated to generate a counter-stimulation waveform. A phase compensation execution command is then generated based on the counter-stimulation waveform.

[0006] A second aspect of this invention provides an intelligent control system for a safety helmet based on electroencephalogram (EEG) characteristics, comprising: The signal acquisition module is used to acquire real-time EEG signals from operators, perform spectral decomposition on the real-time EEG signals to generate cognitive load frequency domain features, and establish an EEG control space based on the cognitive load frequency domain features. The frequency band analysis module is used to form a multi-frequency band energy distribution matrix for the EEG control spatial scan, identify frequency band coupling imbalance points along the multi-frequency band energy distribution matrix, deduce the timing of collaborative intervention through the frequency band coupling imbalance points, and form a multi-channel regulation mode based on the collaborative intervention timing. The resonance matching module is used to acquire the stimulation frequency range of the safety helmet and the target EEG frequency band, map the stimulation frequency range to the EEG control space to generate a frequency matching domain, and establish a resonance intervention window based on the resonance analysis of the frequency matching domain on the target EEG frequency band. The propagation tracking module is used to perform frequency band monitoring within the resonance intervention window using the multi-channel adjustment mode to obtain the energy migration trajectory, construct a risk propagation chain according to the energy migration trajectory to identify cascade failure nodes, and perform blocking priority assessment on the cascade failure nodes to form key blocking anchor points. The response scheduling module is used to perform risk acceleration analysis on the key blocking anchor points to form a state deterioration rate, use the state deterioration rate to fuse the cognitive load frequency domain characteristics to determine the response time and generate a dynamic delay tolerance, and perform fast and slow grading to generate variable speed control parameters according to the dynamic delay tolerance. The phase control module is used to compare the speed control parameters with a preset safety threshold to generate a control deviation signal, perform anti-phase modulation on the control deviation signal to generate a counter-stimulation waveform, and form a phase compensation execution command based on the counter-stimulation waveform.

[0007] The beneficial effects of this invention are reflected in the following points: First, by extracting the frequency domain features of cognitive load through spectral decomposition and establishing an EEG control space including a normal working layer, an early warning transition layer, and a loss-of-control intervention layer, and then identifying frequency band coupling imbalance points by scanning the EEG control space and inversely deriving the timing of collaborative intervention, a complete closed loop from static cognitive state grading to dynamic imbalance early warning is achieved. This not only accurately determines the current cognitive load level of the operator, but also proactively predicts the trend of cognitive state deterioration through the imbalance characteristics of inter-frequency band coupling relationships, identifying the optimal intervention window before the operator fully enters a dangerous state. Second, by mapping the stimulation frequency range to generate a frequency matching domain and performing phase locking to extract synchronous frequency components for coupling degree screening, peak resonance points are obtained and resonance intervention windows are determined. Then, energy migration trajectories are obtained within the window and risk propagation chains are constructed to identify cascading failure nodes, achieving precise matching between the intervention frequency and the inherent frequency of EEG and accurate positioning of risk nodes. Through the resonance effect, a large arousal effect is obtained with a small stimulation intensity, while targeted intervention is implemented at the key links most prone to chain collapse. Finally, by performing risk acceleration analysis on key blocking anchor points to form the state deterioration rate, and integrating the frequency domain characteristics of cognitive load to generate dynamic delay tolerance, and performing fast and slow grading to generate variable speed control parameters, the variable speed control parameters are then compared with the safety threshold to generate control deviation signals, and inverse phase modulation is performed to generate counter-stimulation waveforms to form phase compensation execution commands. This achieves dynamic grading of intervention response speed and adaptive compensation for stimulus parameter deviations. It can flexibly adjust the intervention response time according to the urgency of cognitive state deterioration, and can detect and correct the deviation of stimulus parameters in real time through inverse phase modulation technology, ensuring that the stimulus is always kept within a safe and effective range.

[0008] 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

[0009] 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.

[0010] 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.

[0011] Figure 1 This is a flowchart illustrating an intelligent control method for safety helmets based on electroencephalogram (EEG) characteristics according to the present invention.

[0012] Figure 2 This is a structural block diagram of an intelligent control system for a safety helmet based on electroencephalogram (EEG) characteristics according to the present invention. Detailed Implementation

[0013] 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.

[0014] 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.

[0015] 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.

[0016] As used in this application specification and the appended claims, the term "if" may be interpreted, depending on the context, as "when," "once," "in response to determination," or "in response to detection." Similarly, the phrase "if determined" or "if detected [the described condition or event]" may be interpreted, depending on the context, as meaning "once determined," "in response to determination," "once detected [the described condition or event]," or "in response to detection [the described condition or event]."

[0017] Furthermore, in the description of this application and the appended claims, the terms "first," "second," "third," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0018] 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.

[0019] The technical solutions of the embodiments of this application will be described below.

[0020] like Figure 1 As shown, this embodiment of the invention provides a smart control method for safety helmets based on electroencephalogram (EEG) features, including the following steps S110-S160: Step S110: Collect real-time EEG signals from the operator, perform spectral decomposition on the real-time EEG signals to generate cognitive load frequency domain features, and establish an EEG control space based on the cognitive load frequency domain features.

[0021] Specifically, the system collects real-time EEG signals from workers. EEG electrodes, made of flexible conductive material, are deployed on the forehead inside the safety helmet, making contact with the scalp. The EEG acquisition system records the workers' brain activity in real time during work, including brainwave signals in five frequency bands: delta, theta, alpha, beta, and gamma, with a sampling frequency set at 256 Hz. The raw signals are preprocessed: a notch filter is used to eliminate power frequency interference from industrial electrical equipment and lighting systems; a high-pass filter is used to remove low-frequency drift caused by head movements and muscle activity; and a low-pass filter is used to remove high-frequency noise. During normal work, beta wave activity in the prefrontal cortex remains at a moderate level, indicating focused attention and appropriate cognitive load. When workers begin to fatigue after prolonged continuous work, the energy of low-frequency theta and alpha waves gradually increases, while beta wave activity decreases. These changes in brainwave patterns reflect the transition from a state of alertness to a state of fatigue and drowsiness. When miners operate large mining equipment or handle sudden emergencies, both beta and gamma waves are active simultaneously, indicating that the brain is engaged in high-intensity information processing and decision-making activities. Workers operating at heights also exhibited a synergistic enhancement of beta and gamma waves when hoisting heavy components, reflecting an increased cognitive load under high-risk operations.

[0022] Spectral decomposition of real-time EEG signals generates frequency domain features of cognitive load. Fast Fourier Transform (FFT) is performed on five frequency bands of the acquired signal to convert the time-domain signal into a frequency-domain representation. The power spectral density (PSD) of each frequency band within each time window is calculated; PSD reflects the energy distribution of the signal across different frequency components. The temporal variation characteristics of the PSD are analyzed to identify frequency bands where energy increases, decreases, or remains stable. During relatively relaxed work conditions such as inspections or monitoring, the signal shows a high and stable alpha wave PSD, indicating low cognitive load. When instructions are received to troubleshoot equipment malfunctions or handle abnormal situations, the signal shows a rapid increase in the prefrontal cortex beta wave PSD, indicating that the cognitive system has been activated and entered a working state. Based on the signal's PSD, energy ratios between different frequency bands are calculated, including the θ / α ratio and the β / α ratio. These energy ratios reflect the level of cognitive load and alertness. When monitoring complex production processes or operating precision instruments, a significant increase in the β / α ratio is observed in the signal analysis, indicating increased cognitive load and high concentration. As drowsiness begins to set in during the later stages of a night shift, the signal shows a gradual increase in the θ / α ratio, indicating a shift in the cognitive system from a state of wakefulness to drowsiness. The power spectral density, energy ratio, and temporal variation characteristics of the signal are integrated into the frequency domain characteristics of cognitive load.

[0023] In some embodiments, establishing an EEG control space based on the frequency domain features of the cognitive load includes: performing energy spectrum analysis on the frequency domain features of the cognitive load to extract energy distribution features; calibrating boundary transition thresholds based on the energy distribution features; monitoring the fluctuation amplitude of the frequency domain features of the cognitive load according to the boundary transition thresholds to establish a hierarchical mapping rule, the hierarchical mapping rule including a normal working layer, an early warning transition layer, and a loss-of-control intervention layer; and constructing an EEG control space using the hierarchical mapping rule.

[0024] Energy distribution characteristics were extracted from the frequency domain features of cognitive load using energy spectrum analysis. Based on these features, the proportion of energy in each frequency band within the total energy was calculated, reflecting the contribution of different frequency bands to overall EEG activity. During normal work, a relatively balanced proportion of alpha and beta waves was observed in the frequency domain features, indicating that the cognitive system was in a moderately activated state. When workers began to feel fatigued, the frequency domain features showed a gradual increase in the proportion of theta waves and a decrease in the proportion of beta waves, indicating a shift towards drowsiness. The temporal trend of energy proportions in the frequency domain features was analyzed to identify frequency bands where proportions increased, decreased, or remained stable. When workers received emergency tasks, the frequency domain features showed a rapid increase in the proportion of beta waves, indicating a rapid increase in cognitive load. Energy ratio data, including the theta / alpha ratio and the beta / alpha ratio, were extracted from the frequency domain features and integrated with the energy proportions and temporal trends of each frequency band to form energy distribution characteristics. These characteristics include the energy proportions of each frequency band, the energy ratios between frequency bands, and the temporal trends of these indicators, describing the energy distribution of cognitive load as reflected by the frequency domain features.

[0025] Boundary transition thresholds are determined based on energy distribution characteristics. The transition characteristics of energy distribution between different working states are analyzed to identify abrupt changes during state transitions. When a worker transitions from a resting state to a working state, the distribution characteristics show a rapid increase in the proportion of β-wave energy and a corresponding decrease in the proportion of α-wave energy, while the β / α ratio increases significantly. This rapid change in distribution characteristics indicates a state transition. When transitioning from a normal working state to a fatigued state, the distribution characteristics show a continuous increase in the proportion of θ-wave energy, crossing a critical level, which is the boundary transition threshold. A large amount of distribution characteristic data of workers during different state transitions is statistically analyzed to identify statistically significant transition thresholds. For the boundary between normal working state and fatigued state, the proportion of θ-wave energy is extracted from the distribution characteristics. When the proportion of θ-wave energy exceeds 25%, workers begin to show signs of fatigue; therefore, 25% is set as the boundary transition threshold for the fatigued state. For the boundary between normal working state and high-load state, the proportion of β-wave energy is extracted from the distribution characteristics. When the proportion of β-wave energy exceeds 45%, complex tasks begin to be handled; therefore, 45% is set as the boundary transition threshold for the high-load state. When miners handle abnormal gas concentration alarms, the distribution characteristics show that the proportion of β-wave energy rapidly exceeds the 45% threshold, entering a high-load state. By analyzing the transition interfaces of the distribution characteristics, key boundary transition thresholds are determined, with each threshold corresponding to a state transition interface.

[0026] A hierarchical mapping rule is established based on the fluctuation amplitude of the frequency domain characteristics of cognitive load monitored by boundary transition thresholds. The real-time acquired frequency domain characteristics are compared with the boundary transition thresholds. The energy proportion and energy ratio of each frequency band are extracted from the frequency domain characteristics, and the degree of deviation of these indicators from the thresholds is calculated; this degree of deviation is the fluctuation amplitude. When the frequency domain characteristics show that the theta wave energy proportion is 20%, the fluctuation amplitude deviating from the theta wave threshold (25%) is -20%. When the frequency domain characteristics show that the beta wave energy proportion is 50%, the fluctuation amplitude deviating from the beta wave threshold (45%) is +11%. Based on the magnitude of the fluctuation amplitude, a hierarchical mapping rule is established. The mapping rule divides the EEG control space into three levels: the normal working level corresponds to a frequency domain characteristic fluctuation amplitude within ±10%, where the energy distribution of each frequency band is balanced, the β / α ratio is within a reasonable range, and the theta wave energy proportion is low, indicating that the cognitive load of the worker is moderate, attention is focused, and there are no obvious signs of fatigue. The warning transition level corresponds to a frequency domain characteristic fluctuation amplitude within ±10%-30%, where the energy distribution begins to approach the threshold but has not yet fully crossed it. After workers have been working continuously for several hours, the frequency domain characteristics show that the theta wave energy accounts for 22%, with a fluctuation range of -12%. According to the mapping rules, this indicates a transitional warning layer, meaning the worker is beginning to feel fatigued but can still continue working. The out-of-control intervention layer corresponds to a frequency domain characteristic fluctuation range exceeding ±30%, at which point the energy distribution has crossed the threshold. In cases of extreme fatigue, the frequency domain characteristics show that the theta wave energy accounts for more than 32%, with a fluctuation range of +28%. According to the mapping rules, this indicates entry into the out-of-control intervention layer, requiring immediate issuance of a warning or activation of protective measures. The above three-level division and judgment criteria constitute a complete hierarchical mapping rule.

[0027] A hierarchical mapping rule is used to construct an EEG control space. The multidimensional space of cognitive load frequency domain characteristics is divided into three hierarchical regions according to the hierarchical mapping rule. In the EEG control space, the normal working state layer occupies the central region, corresponding to a fluctuation range of ±10% as defined by the mapping rule, representing the frequency domain characteristic range of the operator in an ideal working state. The warning transition layer is located between the central region and the boundary region, corresponding to a fluctuation range of ±10%-30% as defined by the mapping rule, representing the frequency domain characteristic range of the state starting to deviate but not yet reaching a dangerous level. The out-of-control intervention layer is located in the boundary region, corresponding to a fluctuation range exceeding ±30% as defined by the mapping rule, representing the frequency domain characteristic range of the state that has deviated from the normal range and requires timely intervention. Clear boundary conditions are set for each layer, based on a comprehensive judgment of the energy ratio and energy distribution of each frequency band. Buffer bands are set between layers to smooth the state transition process and avoid frequent layer jumps. During a complete shift, the position of the frequency domain features in the EEG control space will move with changes in cognitive state. At the beginning of the shift, when the worker is mentally alert, the frequency domain features are located in the center of the normal working layer. As the work time increases, the frequency domain features gradually drift towards the warning transition layer. If the worker does not get timely rest, the frequency domain features will eventually enter the out-of-control intervention layer. At this time, according to the determination of the mapping rules, a mandatory rest instruction should be issued or the work arrangement should be adjusted.

[0028] Step S120: A multi-band energy distribution matrix is ​​formed by scanning the EEG control space. The frequency band coupling imbalance point is identified along the multi-band energy distribution matrix. The timing of collaborative intervention is deduced in reverse through the frequency band coupling imbalance point. A multi-channel regulation mode is formed based on the timing of collaborative intervention.

[0029] Specifically, a multi-band energy distribution matrix was generated by scanning the EEG control space. Real-time scanning was performed on various regions within the EEG control space, extracting energy values ​​for each frequency band at different time points. The energy values ​​of each frequency band were arranged according to time sequence and frequency band dimension to construct a two-dimensional energy distribution matrix. Rows in the matrix correspond to different time sampling points, columns correspond to the five EEG frequency bands, and matrix element values ​​represent the energy intensity at the corresponding time point and frequency band. The energy distribution matrix was normalized, scaling the energy values ​​of each frequency band to the 0-1 range to eliminate differences in energy magnitude between different frequency bands and make the energy of each frequency band comparable. The characteristics of energy changes in each frequency band over time were analyzed. At the beginning of the shift, the energy distribution matrix showed that the energy of β waves and α waves was relatively balanced, and the energy distribution among the frequency bands was relatively stable. As the continuous working time increased, the values ​​in the column corresponding to θ waves gradually increased, while the values ​​in the column corresponding to β waves gradually decreased, reflecting the evolution of cognitive state from wakefulness to fatigue. When workers suddenly encounter an emergency requiring rapid decision-making, the values ​​of the corresponding columns for β-waves and γ-waves in the matrix rise rapidly and synchronously, indicating a sudden surge in cognitive load.

[0030] Identify frequency band coupling imbalance points along the multi-band energy distribution matrix. Analyze the cooperative relationship between different frequency bands in the energy distribution matrix and calculate the energy correlation between frequency bands. Under normal operating conditions, energy changes in different frequency bands usually exhibit a cooperative pattern; when the energy of some frequency bands increases, the energy of other specific frequency bands decreases accordingly, maintaining overall energy balance. Identify time points in the matrix that deviate from the normal cooperative pattern; these time points correspond to the moments when the energy distribution between frequency bands becomes unbalanced. Calculate the degree of deviation of each frequency band's energy from its baseline level, where the baseline level is the average energy value of that frequency band under normal operating conditions. When the energy of a frequency band deviates from the baseline by more than 30%, it is marked as an imbalanced frequency band. When multiple frequency bands simultaneously exhibit significant deviations, and the direction of deviation does not conform to the normal cooperative pattern, this time point is marked as a frequency band coupling imbalance point. After miners have been working underground for a long time, the energy distribution matrix shows that while the theta wave energy continues to rise, the beta wave energy continues to fall, but the alpha wave energy does not adjust accordingly according to the normal pattern. The cooperative relationship between the three frequency bands is broken, and the frequency band coupling imbalance point is identified. When operators are working on complex equipment, they are suddenly subjected to strong external interference. The energy of beta and gamma waves fluctuates drastically and the wave patterns lose synchronization, causing the coupling relationship between frequency bands to become unbalanced in an instant.

[0031] In some embodiments, the step of inversely deducing the timing of collaborative intervention via the frequency band coupling imbalance point includes: extracting duration data from the frequency band coupling imbalance point; identifying the deterioration acceleration gradient based on the duration data; defining a high-risk time period based on the moment when the deterioration acceleration gradient exceeds a preset threshold; and shifting the start point of the high-risk time period forward to set a buffer duration as the timing of collaborative intervention.

[0032] Duration data was extracted from frequency band coupling imbalance points. The duration from the appearance to the disappearance of each frequency band coupling imbalance point was statistically analyzed; duration reflects the persistence of the imbalance state. The temporal variation pattern of duration was analyzed to identify whether the duration showed an increasing trend. Shorter durations with smaller fluctuations indicate stronger cognitive adjustment capabilities among workers, enabling them to recover quickly from the imbalance state. Longer durations with a continuous increase indicate weakened cognitive adjustment capabilities and a gradually solidified imbalance state. Imbalance points appearing at the beginning of a shift typically had shorter durations and usually recovered after a short rest or task switch. As continuous work time increased, the duration of later imbalance points significantly lengthened, indicating that workers' cognitive recovery capabilities decreased with accumulated fatigue. The duration values ​​of each imbalance point were recorded and arranged chronologically to construct a time series of duration data.

[0033] Identifying the deterioration acceleration gradient based on duration data. Regression analysis is performed on the time series of duration data to fit a trend line T(t) = k × t + b, where T(t) is the predicted duration at time t, k is the slope of the trend line, t is the time variable, and b is the intercept. The slope k of the trend line is the deterioration acceleration gradient; a larger slope indicates a faster deterioration in cognitive state. In the initial stage of work, the duration data is relatively stable, and the deterioration acceleration gradient is close to zero. After entering the fatigue accumulation stage, the duration data begins to rise rapidly, and the deterioration acceleration gradient calculated by regression analysis increases significantly. When the deterioration acceleration gradient suddenly increases, it indicates that the worker's cognitive state is rapidly deteriorating, requiring timely intervention. When miners handle emergencies, multiple imbalance points with rapid increases in duration occur consecutively within a short period, and the deterioration acceleration gradient rises sharply, indicating that cognition is about to enter a state of loss of control.

[0034] High-risk time periods are defined based on when the deterioration acceleration gradient exceeds a preset threshold. A safety threshold for the deterioration acceleration gradient is set, corresponding to the critical point where the cognitive state transitions from controllable to out of control. The safety threshold is dynamically adjusted based on the hazard level of the job type and the individual characteristics of the workers; a stricter threshold is used for high-risk job scenarios, and a more lenient threshold is used for low-risk job scenarios. The numerical changes of the deterioration acceleration gradient are monitored in real time. When the gradient value exceeds the safety threshold, that moment is marked as the starting point of the high-risk time period. Tracking backwards from the starting point until the deterioration acceleration gradient drops below the safety threshold or the worker receives effective intervention, this time interval is the high-risk time period. The high-risk time period corresponds to the period when the worker's cognitive state deteriorates rapidly and the safety risk increases sharply. During continuous work, when the deterioration acceleration gradient first exceeds the safety threshold, that moment and the subsequent period are designated as the high-risk time period, indicating that the worker is about to enter or has already entered a dangerous state. During the high-risk time period, the probability of the worker making operational errors, misjudgments, or accidents increases significantly, requiring immediate protective measures.

[0035] A buffer period is set forward from the start point of the high-risk period as the timing for coordinated intervention. This buffer period is dynamically determined based on the hazard of the task and the rate of cognitive deterioration. For high-risk tasks, a longer buffer period is set to ensure sufficient time for intervention measures to take effect. For situations where cognitive deterioration is rapid, a longer buffer period is set to prevent intervention measures from being implemented too late. Subtracting the buffer period from the start point of the high-risk period gives the timing for coordinated intervention. Implementing preventative interventions at this timing can stop further deterioration of cognitive state before workers enter a high-risk state. When workers are operating hazardous equipment, if the acceleration of deterioration is predicted to exceed the safety threshold, a rest reminder and suggestion to suspend hazardous operations are issued in advance at the coordinated intervention timing, successfully preventing workers from continuing high-risk operations while fatigued. When miners are working underground, multi-channel regulation is activated at the coordinated intervention timing before the arrival of the high-risk period. Through a combination of measures including sound reminders, enhanced lighting, and task adjustments, the deterioration of cognitive state is effectively slowed.

[0036] A multi-channel regulation model is established based on the timing of collaborative interventions. Appropriate combinations of intervention strategies are selected according to the EEG characteristics corresponding to the timing of the collaborative interventions. For imbalances caused by fatigue, a combination of interventions such as rest reminders, task switching, and environmental adjustment is used. For imbalances caused by cognitive overload, a combination of interventions such as task simplification, supportive care, and rhythm regulation is used. A multi-channel regulation model is established, with each channel corresponding to different types of interventions. The audio reminder channel uses the built-in speaker on the safety helmet to emit voice or warning sounds to remind workers to rest or adjust their state. The visual warning channel uses flashing LED indicators on the safety helmet visor to signal abnormal conditions to workers and colleagues. The environmental control channel sends instructions to the environmental control equipment in the work area to adjust lighting brightness, ventilation intensity, or temperature settings to improve the working environment and alleviate fatigue. The task management channel sends suggestions to the work scheduling platform, prompting dispatchers to arrange for workers to take turns resting or adjust their work tasks. When workers show signs of inattention while working at height, safety reminders are played through the audio reminder channel, warning signs are displayed on the visor through the visual warning channel, and a status report is sent to ground monitoring personnel, forming a multi-layered collaborative intervention. When a miner shows signs of fatigue underground, the system not only alerts him to safety through sound, but also sends instructions to the underground lighting equipment to increase the brightness of the local lighting, and suggests to the dispatch center that the miner be brought to the surface early to rest.

[0037] Step S130: Obtain the stimulation frequency range of the safety helmet and the target EEG frequency band, map the stimulation frequency range to the EEG control space to generate a frequency matching domain, and establish the resonance intervention window based on the resonance analysis of the frequency matching domain on the target EEG frequency band.

[0038] Specifically, the stimulation frequency range and target EEG frequency band of the safety helmet are obtained. The types of stimulation output devices equipped on the safety helmet are identified, including sound generators, vibration motors, and LED flashing lights. The operating frequency range of each stimulation device is obtained: sound generators typically cover the low-to-mid frequency range of 200-4000Hz, vibration motors cover the very low frequency range of 5-50Hz, and LED flashing lights cover the low-frequency range of 1-30Hz. The frequency ranges of all stimulation devices together constitute the overall stimulation frequency range of the safety helmet. When the worker is fatigued, the safety helmet plays a specific frequency cue tone through the sound generator. This cue tone is designed to effectively activate beta wave activity in the prefrontal cortex. Based on the worker's current cognitive state and intervention goals, the target EEG frequency band to be adjusted is determined. When signs of fatigue and drowsiness are detected in the worker, the target EEG frequency band is set to beta waves (13-30Hz), and the intervention goal is to enhance beta wave activity to improve alertness. When high cognitive load is detected in workers, the target EEG frequency band is set to alpha waves (8-13Hz), and the intervention goal is to enhance alpha wave activity to alleviate cognitive stress. Miners experience a significant increase in theta wave energy during the later stages of night shifts; therefore, beta waves are set as the target EEG frequency band, and appropriate stimulation frequencies are used to enhance beta wave activity, helping miners regain alertness. When workers are under high stress, alpha waves are set as the target EEG frequency band, and rhythmic stimulation induces alpha wave enhancement to alleviate cognitive overload.

[0039] A frequency matching domain is generated by mapping the stimulation frequency range to the EEG control space. Based on the stimulation frequency range, the correspondence between stimulation frequency and EEG frequency bands is analyzed, showing that external stimuli of different frequencies have differentiated effects on different EEG frequency bands. The frequency ranges of each stimulation device on the safety helmet are projected into the EEG control space, taking into account the stimulus transmission path and brain response characteristics during the projection process. The effective area of ​​the stimulation frequency is marked in the EEG control space, representing the range of EEG states that a specific stimulation frequency can effectively influence. The mid-frequency component of sound stimulation mainly affects the region corresponding to β waves in the EEG control space, effectively activating the alert state. The very low-frequency component of vibration stimulation forms a wide-area influence covering multiple frequency bands in the EEG control space, modulating delta waves, theta waves, and alpha waves. The low-frequency component of LED flashing lights mainly acts on theta and alpha wave regions in the EEG control space, making it suitable for relaxation and soothing interventions. The EEG characteristics of the operator are located at different positions in the control space under different cognitive states, and the most suitable stimulation frequency range is selected according to the current position. When the operator is in the warning transition layer and drifting towards the drowsiness zone, the mid-frequency range of the activated sound stimulation forms a frequency matching domain covering the beta wave region in the EEG control space, guiding the operator's EEG characteristics back to the normal working state. The frequency matching domain represents the area in the EEG control space where the stimulation frequency range has a good matching relationship with the target EEG frequency band; stimulation within this area yields better intervention results. By mapping the frequency ranges of different types of stimulation devices to the EEG control space, a multimodal stimulation synergistic intervention mechanism is established, achieving precise regulation of the operator's cognitive state.

[0040] In some embodiments, establishing a resonance intervention window based on the target EEG frequency band through frequency matching domain resonance analysis includes: performing phase locking on the frequency matching domain to extract frequency components synchronized with the target EEG frequency band; performing coupling degree screening on the frequency components to generate a set of highly coupled frequencies; performing matching degree analysis on the set of highly coupled frequencies and the target EEG frequency band to obtain peak resonance points; and determining the resonance intervention window by extending a set time span to both sides based on the peak resonance points.

[0041] Phase locking was performed on the frequency matching domain to extract frequency components synchronized with the target EEG frequency band. The phase relationship between each stimulation frequency within the frequency matching domain and the target EEG frequency band was analyzed, reflecting the degree of temporal synchronization between the stimulation frequency and the target EEG frequency band. The phase locking degree between the stimulation frequency and the target EEG frequency band was calculated; a high phase locking degree indicates that the stimulation frequency can effectively drive the oscillation of the target EEG frequency band. Stimulation frequencies with a phase locking degree exceeding the synchronization threshold were extracted; these frequencies maintained a phase locking relationship with the target EEG frequency band. The synchronization threshold was set according to the intervention intensity requirements: 0.7 for strong intervention scenarios and 0.5 for mild intervention scenarios. The operator's beta wave exhibited a stable oscillation pattern at a certain moment, with an oscillation frequency of 18Hz. All stimulation frequencies within the frequency matching domain were scanned, and frequency components phase-locked with this beta wave oscillation pattern were identified, including the 1800Hz and 3600Hz frequency components of the sound stimulus. These frequency components can produce a synchronous resonance effect with the beta wave and are candidate frequencies for intervention. The higher the degree of phase locking, the more significant the driving effect of the stimulation frequency on the target EEG frequency band, and the more ideal the intervention effect can be obtained with a smaller stimulation intensity.

[0042] Coupling degree screening was performed on frequency components to generate a high-coupling frequency set. Energy coupling degree analysis was performed on the extracted synchronization frequency components; coupling degree reflects the efficiency of energy transfer from the stimulus frequency to the target EEG frequency band. The coupling degree of each frequency component was calculated as C = ΔE / P_s, where ΔE is the energy change in the target EEG frequency band, and P_s is the stimulus power. Frequency components with high coupling degree can elicit a large energy response in the target EEG frequency band with a smaller stimulus intensity, while frequency components with low coupling degree are unlikely to effectively change the EEG state even with strong stimulation. A coupling degree screening threshold C_threshold = 2.0 was set, and frequency components with coupling degree exceeding the threshold were included in the high-coupling frequency set. When the operator received stimuli of different frequencies, some frequencies could quickly activate β waves and cause their energy to rise continuously. One frequency had a coupling degree C = 4.5, causing a 25% increase in β wave energy, and was identified as a high-coupling frequency. Other frequencies, although also phase-locked with β waves, had a coupling degree of only 0.8, causing only a 5% increase in β wave energy, and were therefore excluded. The high-coupling frequency set contains stimulation frequencies that have a strong effect on the target EEG frequency band, and serves as a frequency resource for implementing precise intervention.

[0043] Peak resonance points are obtained by performing a matching degree analysis between the high-coupling frequency set and the target EEG frequency band. The proximity of each frequency in the high-coupling frequency set to the dominant frequency of the target EEG band is analyzed, with the proximity quantified by the reciprocal of the frequency difference. A small frequency difference indicates a high degree of matching between the stimulation frequency and the inherent frequency of the target EEG band, which can induce a resonance effect. The comprehensive matching degree of each frequency in the high-coupling frequency set is calculated, integrating information from three dimensions: phase locking degree, energy coupling degree, and frequency proximity. The stimulation frequency with the highest comprehensive matching degree is identified as the peak resonance point; stimulation at this frequency point yields the optimal intervention effect. When the dominant frequency of the operator's beta wave is 18Hz, the stimulation frequency that best matches this dominant frequency is searched within the high-coupling frequency set. The 1800Hz frequency component of the sound stimulus corresponds to the 18Hz frequency of the EEG oscillation, showing a frequency ratio of 100:1. This frequency has the highest comprehensive matching degree and is identified as the peak resonance point. By precisely locating the peak resonance point, the optimal match between the intervention stimulus and the operator's EEG characteristics is ensured, maximizing the intervention effect.

[0044] The resonance intervention window is determined by extending the time span to both sides of the peak resonance point. A resonance frequency band is formed by extending the frequency axis to both sides of the peak resonance point as the center frequency, taking into account the frequency fluctuation characteristics of the target EEG band and individual differences. The temporal stability of the target EEG band frequency in the operator is analyzed. A narrower resonance frequency band is set when frequency stability is high, with an extension range of ±5%; a wider resonance frequency band is set when frequency fluctuation is large, with an extension range of ±15%. The time span for continuous stimulation within the resonance frequency band is determined, based on the intervention target and EEG response speed. For rapid arousal interventions, a shorter time span of 10-30 seconds is set to avoid overstimulation. For continuous modulation interventions, a longer time span of 60-180 seconds is set to maintain a stable intervention effect. The determined time span is defined as the resonance intervention window, which represents the time range within which a good resonance effect can be obtained by continuously applying the stimulation frequency corresponding to the peak resonance point. During the intervention process, the operator's EEG state undergoes dynamic changes, and the evolution of the target EEG frequency band is monitored in real time. When the energy of the target EEG frequency band rises to the target level or the operator's EEG characteristics move out of the effective range of the resonance intervention window, the stimulation parameters are adjusted or the stimulation is terminated in a timely manner to avoid weakening the intervention effect or causing negative effects. By reasonably setting the frequency range and time span of the resonance intervention window, safe operation is ensured while avoiding discomfort caused by overstimulation.

[0045] Step S140: Use multi-channel adjustment mode to perform frequency band monitoring within the resonance intervention window to obtain energy migration trajectory, construct risk propagation chain according to energy migration trajectory to identify cascade failure nodes, and perform blocking priority assessment on cascade failure nodes to form key blocking anchor points.

[0046] Specifically, a multi-channel modulation mode was used to perform frequency band monitoring within the resonance intervention window to obtain energy migration trajectories. After the resonance intervention window was activated, sound, vibration, or light stimulation was applied to the workers using the multi-channel modulation mode. Energy changes in five EEG frequency bands (δ, θ, α, β, and γ) were monitored simultaneously, and the temporal data of energy in each frequency band were recorded before, during, and after stimulation. The evolution patterns of energy in each frequency band on the time axis were analyzed to identify characteristics of energy rise, fall, or transfer between different frequency bands. After receiving sound stimulation, the theta wave energy initially decreased, followed by a gradual increase in β wave energy. This energy transfer from low to high frequencies reflected the transition from drowsiness to wakefulness. An energy migration trajectory diagram was plotted, with time on the horizontal axis and energy in each frequency band on the vertical axis. The curves showed the dynamic flow of energy between different frequency bands. In the initial stage of receiving intervention stimulation, the energy migration trajectory showed a rapid decrease in theta wave energy, a brief increase in α wave energy followed by stabilization, and a continuous increase in β wave energy that eventually stabilized at a high level. The entire energy migration process underwent a complete evolution from fatigue recovery to a stable state of wakefulness. The energy migration trajectory obtained through monitoring contains complete information on the changes in energy over time in each frequency band.

[0047] In some embodiments, the step of constructing a risk propagation chain and identifying cascading failure nodes according to the energy migration trajectory includes: dividing the propagation direction window in the energy migration trajectory to generate a set of direction windows; performing propagation path planning based on the set of direction windows to obtain the main propagation channel; performing intensity modulation processing on the main propagation channel to generate a propagation load; and using the propagation load to perform cascading relationship mapping to identify cascading failure nodes.

[0048] A propagation direction window set is generated by dividing the energy migration trajectory into propagation direction windows. The directional characteristics of energy flow in the energy migration trajectory are analyzed to identify the time periods during which energy transfers from the source frequency band to the target frequency band. Based on the energy flow direction, the trajectory is divided into multiple propagation direction windows, each corresponding to a time interval with a clear energy propagation direction. Within a propagation direction window, energy mainly flows out of a specific frequency band and into another specific frequency band, maintaining a relatively stable flow direction. In the initial stage of receiving intervention stimuli, energy mainly transfers from theta waves to alpha waves; this time period constitutes a propagation direction window. Subsequently, energy mainly transfers from alpha waves to beta waves, forming another propagation direction window. The start time, end time, source frequency band, and target frequency band information of each propagation direction window are extracted to construct a data structure for the propagation direction windows. All propagation direction windows are arranged in chronological order to form a direction window set. The direction window set describes the complete temporal process of energy propagation between different frequency bands. For example, a miner's direction window set contains three main windows: theta wave to alpha wave transfer, alpha wave to beta wave transfer, and beta wave remaining stable.

[0049] The propagation path is planned based on a set of directional windows to obtain the main propagation channel. The continuity of energy propagation within the directional window set is analyzed to identify energy transfer links between frequency bands. Multiple windows with energy flowing out of the same source frequency band and multiple windows with energy flowing into the same target frequency band are associated to construct propagation paths between frequency bands. Frequency band links with large energy flow and long propagation duration are identified in the propagation paths; these links constitute the main propagation paths. During the recovery process from fatigue, the main propagation path is typically θ wave → α wave → β wave, with energy flowing sequentially from low frequency bands to high frequency bands. This path corresponds to the normal cognitive recovery mechanism. The temporal relationship within the directional window set is analyzed to calculate the energy transfer volume between frequency bands; the energy transfer volume equals the energy decrease in the source frequency band. The path with the largest energy transfer volume is extracted as the main propagation channel. The main propagation channel carries the main flow during energy migration and is the core channel for the intervention mechanism to function. When a miner receives intervention, θ → α → β constitutes the main propagation channel. When a worker at height experiences cognitive stress, the main propagation channel is β → α, with energy flowing back from high frequency bands to low frequency bands, achieving cognitive relaxation.

[0050] Intensity modulation is applied to the main propagation channel to generate propagation load. The energy flow rate of each propagation stage in the main propagation channel is acquired; energy flow rate equals the change in energy flowing from the source frequency band to the target frequency band per unit time. The temporal variation characteristics of the energy flow rate are analyzed to identify time periods of stable flow, increased flow, and decreased flow. Stable flow indicates smooth energy propagation, increased flow indicates accelerated energy propagation, and decreased flow indicates obstructed energy propagation. The propagation resistance of each propagation stage is assessed; propagation resistance reflects the ease of energy transfer between frequency bands, with high resistance indicating difficult energy transfer and low resistance indicating smooth energy transfer. Propagation resistance is estimated by the ratio of energy flow rate to the energy change rate of the source frequency band; a small ratio indicates high resistance. The propagation load of each propagation stage is acquired, calculated as L = F × T / R, where F is the energy flow rate, T is the propagation duration, and R is the propagation resistance coefficient. A high propagation load indicates that the stage is undertaking a large energy transfer task; excessively high propagation load may lead to propagation channel blockage or collapse. When personnel receive strong stimuli, the load of the propagation stage from theta wave to alpha wave increases rapidly. If the load exceeds the carrying capacity threshold of a particular stage, energy propagation will be hindered, leading to theta wave energy accumulation and a weakened intervention effect. By monitoring the propagation load of each stage in real time, propagation bottlenecks can be identified promptly, and intervention strategies can be adjusted accordingly.

[0051] Cascading failure nodes are identified by mapping cascading relationships using propagation load. The dependencies between propagation links in the main propagation channel are analyzed to identify the impact of upstream link failures on downstream links. When the propagation load of an upstream link is too high, causing propagation obstruction, the energy supply to the downstream link is insufficient, triggering a chain reaction of failures. Links with propagation loads exceeding a carrying threshold are marked as potential failure links; this carrying threshold is dynamically set according to the operation type. The impact of potential failure links on downstream links is analyzed; changes in upstream link load lead to corresponding changes in downstream link load, and the degree of impact is quantified by the ratio of the change in downstream load to the change in upstream load. Links with a high degree of impact constitute strong dependencies. Links located at critical positions in the main propagation channel and with high propagation loads are cascading failure nodes. Once these nodes fail, they will trigger a chain reaction in multiple downstream links, leading to a complete collapse of the cognitive regulation mechanism. When miners receive intervention stimuli in a state of extreme fatigue, the propagation link from theta wave to alpha wave becomes a cascading failure node because its load exceeds the carrying threshold. After identifying this node, the stimulation strategy is adjusted by reducing the stimulation intensity and extending the stimulation time, alleviating the propagation pressure of this link and preventing cascading failures.

[0052] Prioritization assessments are performed on cascading failure nodes to identify key intervention anchors. The position and impact range of each cascading failure node in the risk propagation chain are analyzed. Nodes upstream in the propagation chain affect multiple subsequent links, while nodes downstream have a limited impact. The severity of the consequences of each node's failure is assessed; some node failures lead to localized cognitive decline, while others result in complete cognitive collapse. The intervention priority for each node is calculated as P_i = w1 × U_i + w2 × S_i, where P_i is the intervention priority of node i, U_i is the upstream degree of the node in the propagation chain, S_i is the severity of the node failure, and w1 and w2 are weighting coefficients. The upstream degree U_i is quantified by the number of links from the node to the end of the propagation chain, and the severity S_i is quantified by the percentage decline in cognitive function caused by the node failure. Nodes with high intervention priorities correspond to key intervention anchors; targeted interventions at these nodes can achieve the greatest risk control effect with minimal cost. After continuous night shifts, the theta wave energy of workers continued to rise, becoming the starting point of the risk propagation chain. This point affected multiple downstream frequency bands and exhibited high severity of failure, thus being identified as a critical blocking anchor point. By increasing the combined intensity of sound and vibration stimuli, further increases in theta wave energy were successfully blocked, preventing the cognitive state from progressing to deep fatigue.

[0053] Step S150: Perform risk acceleration analysis on key blocking anchor points to form a state deterioration rate. Use the state deterioration rate to fuse cognitive load frequency domain characteristics to determine the response time and generate dynamic delay tolerance. Perform fast and slow gearing to generate variable speed control parameters according to the dynamic delay tolerance.

[0054] Specifically, risk acceleration analysis is performed on key blocking anchor points to determine the rate of cognitive deterioration. Time-series EEG data corresponding to the key blocking anchor points are extracted, and the rate of energy change in each frequency band at that node is analyzed. The time derivative of the energy change rate is obtained; the derivative reflects the acceleration of energy change. A positive acceleration indicates an increasing rate of energy change and a worsening trend in cognitive deterioration. A negative acceleration indicates a slowing rate of energy change and a slower trend in cognitive deterioration. In the early stages of fatigue, theta wave energy rises slowly, with a relatively small rate of deterioration acceleration. As fatigue deepens, the rate of increase in theta wave energy accelerates significantly, and the rate of deterioration increases substantially. The risk acceleration values ​​at key blocking anchor points are statistically analyzed; a larger risk acceleration indicates a faster rate of cognitive deterioration and a shorter time window for intervention. During continuous underground work, a sudden increase in risk acceleration at a key blocking anchor point indicates a rapid deterioration of the miner's cognitive state, requiring immediate emergency intervention. The risk acceleration is converted into a rate of cognitive deterioration, which quantifies the time required for the cognitive state to deteriorate from the current level to a dangerous level.

[0055] In some embodiments, the step of using the state deterioration rate to fuse the cognitive load frequency domain features to determine the response time and generate dynamic delay tolerance includes: performing trend analysis on the state deterioration rate to identify acceleration mutation points; performing cross-correlation analysis on the acceleration mutation points and the cognitive load frequency domain features to generate an urgency index; classifying tolerance time levels according to the urgency index; and forming dynamic delay tolerance based on the tolerance time levels.

[0056] Trend analysis of the rate of cognitive deterioration is used to identify acceleration points. The time-series curve of the rate of cognitive deterioration is analyzed to identify the point in time when the rate of deterioration suddenly spikes. Before the acceleration point, the rate of cognitive deterioration is relatively stable or changes slowly, exhibiting linear or sublinear characteristics. At the acceleration point, the rate of cognitive deterioration increases rapidly within a short period, and the deterioration process transforms into superlinear or exponential growth. By calculating the magnitude of the rate of deterioration change, the point in time when the magnitude of change suddenly increases is identified as the acceleration point. The acceleration point marks the turning point from slow to rapid cognitive deterioration and is a key early warning signal for intervention measures. During continuous work, workers initially show little fatigue and have a low rate of cognitive deterioration. After reaching a certain critical moment, fatigue rapidly intensifies, and the rate of cognitive deterioration rises sharply; this critical moment is the acceleration point. Miners working underground often experience acceleration points when the working time approaches their personal tolerance limit, after which cognitive function declines rapidly, and reaction speed and judgment accuracy decrease significantly. Identifying acceleration points allows for timely capture of key moments of cognitive transition, providing early warning for rapid response.

[0057] For example, the step of performing cross-correlation analysis between the accelerated mutation point and the cognitive load frequency domain features to generate an urgency index includes: extracting time features from the accelerated mutation point to obtain mutation moment features; synchronously sampling the cognitive load frequency domain features based on the mutation moment features to extract synchronization features; evaluating the correlation between the mutation moment features and the synchronization features to generate a coupling response coefficient; and forming an urgency index based on the coupling response coefficient.

[0058] Temporal features of accelerated mutation points are extracted to obtain mutation moment characteristics. Feature parameters such as time location, mutation amplitude, and mutation rate of accelerated mutation points are extracted. Time location represents the moment the mutation occurs, mutation amplitude represents the change in the rate of deterioration before and after the mutation point, and mutation rate represents the speed of the jump in the rate of deterioration. Analyzing the magnitude and speed of mutation, a large mutation amplitude and fast speed indicate a drastic jump in cognitive state, while a small mutation amplitude and slow speed indicate a more gradual deterioration. When workers encounter sudden stress events, the rate of deterioration may rise sharply in a short period of time, with both mutation amplitude and speed being large, forming significant mutation moment characteristics. In contrast, during normal fatigue accumulation, the rate of deterioration changes relatively gradually, and mutation characteristics are not obvious. A vector representation of mutation moment characteristics is constructed, containing information from multiple dimensions such as time location, mutation amplitude, and mutation rate, providing a quantitative basis for subsequent cross-correlation analysis.

[0059] Synchronous features are extracted from the frequency domain features of cognitive load based on the characteristics of the abrupt change. The time position corresponding to the accelerated abrupt change point is located, and cognitive load frequency domain feature data is extracted at this time position. Synchronous sampling ensures that the abrupt change moment features and the cognitive load frequency domain features are precisely aligned on the time axis, avoiding correlation analysis errors caused by time deviations. The frequency domain features obtained from synchronous sampling are analyzed, including the energy values, energy ratios, and synergistic relationships between frequency bands. Key frequency domain indicators reflecting the cognitive load state are extracted, such as the θ / α ratio, β / α ratio, and energy entropy. The cognitive load frequency domain features of workers at the accelerated abrupt change point reflect the cognitive background conditions that triggered the rapid deterioration. If the θ wave energy proportion is already high, it indicates that the workers were already in a state of deep fatigue before the abrupt change occurred. If the energy distribution of each frequency band is relatively balanced, it indicates that the abrupt change may be caused by an external sudden event rather than internal fatigue accumulation. A vector representation of the synchronous features is constructed, containing the values ​​of multiple key frequency domain indicators, providing cognitive load state information for cross-correlation analysis.

[0060] The coupling response coefficient is generated by assessing the correlation between abrupt change features and synchronous features. The correlation between the abrupt change feature vector and the synchronous feature vector is analyzed, and the correlation strength is quantified using methods such as Pearson correlation coefficient or mutual information. A high correlation indicates that the accelerated state deterioration is closely related to the imbalance of cognitive load, forming a mutually reinforcing deterioration cycle. A low correlation indicates that the accelerated state deterioration is mainly driven by external factors and has a weaker relationship with the current cognitive load state. An evaluation model for the coupling response coefficient is established, which comprehensively considers the strength of the abrupt change feature, the severity of the synchronous feature, and the degree of correlation between the two. The coupling response coefficient is calculated as C = r × (T_mag + F_load) / 2, where r is the correlation coefficient between the abrupt change feature and the synchronous feature, T_mag is the normalized value of the abrupt change amplitude, and F_load is the normalized value of the degree of cognitive load imbalance. A higher coupling response coefficient indicates that cognition is in a dangerous positive feedback loop, with state deterioration and load imbalance mutually aggravating each other. If the cognitive load frequency domain features show a severe imbalance and a high correlation between the two, and the coupling response coefficient is high, it indicates that cognition is rapidly collapsing, even as the worker's state deteriorates rapidly. When miners encounter an alarm for abnormal gas concentration underground, if their cognitive state is already on the verge of fatigue, the sudden stress event will cause their state to deteriorate rapidly. At this time, the abrupt change characteristics are highly coupled with the load imbalance, generating a high coupling response coefficient.

[0061] An urgency index is formed based on the coupling response coefficient. The coupling response coefficient comprehensively reflects the degree of synergy between the rate of state deterioration and the imbalance of cognitive load, and is the core indicator for assessing urgency. The urgency index is adjusted by combining other auxiliary factors such as the hazard of the task, the severity of environmental conditions, and the historical state records of the workers. For high-risk work scenarios, even with a moderate coupling response coefficient, the urgency index needs to be increased to ensure a safety margin. For low-risk work scenarios, the sensitivity of the urgency index can be appropriately reduced to avoid overreaction affecting normal operations. A comprehensive assessment formula for the urgency index is established: U = w1 × C + w2 × R_task + w3 × E_env, where U is the urgency index, C is the coupling response coefficient, R_task is the task hazard coefficient, E_env is the environmental severity coefficient, and w1, w2, and w3 are weighting coefficients satisfying w1 + w2 + w3 = 1. The weighting coefficients are dynamically adjusted according to the type of work and the actual situation. When miners are working underground, if the coupling response coefficient indicates a rapid deterioration in the state and a severe imbalance in cognitive load, a high urgency index is generated, triggering a rapid response mechanism, given the high-risk nature of underground operations and the enclosed nature of the underground environment. Conversely, when workers are conducting routine equipment inspections in open areas on the surface, even if a slight coupling response occurs, the generated urgency index is relatively low due to the low risk of the task and favorable environmental conditions, allowing for milder intervention methods.

[0062] The urgency index is used to classify tolerance time levels. Thresholds are set for the urgency index, dividing the urgency index space into three levels: high urgency, medium urgency, and low urgency. High urgency corresponds to an urgency index exceeding the high threshold, indicating a rapid deterioration in cognitive state and a dangerous workload, requiring immediate intervention. Medium urgency corresponds to an urgency index within the moderate range, indicating a deteriorating cognitive state but with a short preparation time for intervention. Low urgency corresponds to a relatively low urgency index, indicating a slow deterioration in cognitive state and a manageable workload, allowing ample time for intervention. Corresponding tolerance time levels are set for each urgency level: high urgency corresponds to an extremely short tolerance time, medium urgency to a short tolerance time, and low urgency to a standard tolerance time. When workers suddenly exhibit severe inattention while working at height, the urgency index enters the high urgency level, and the tolerance time is set to extremely short, requiring intervention measures to take effect within a very short time. This tiered mechanism enables differentiated responses to different levels of urgency, ensuring the rational allocation of resources.

[0063] Dynamic delay tolerance is established based on tolerance time levels. Tolerance time levels are converted into specific time values ​​to form dynamic delay tolerance. Dynamic delay tolerance corresponding to extremely short tolerance times requires intervention responses to be initiated within a very short time. Dynamic delay tolerance corresponding to short tolerance times allows for brief preparation time, while dynamic delay tolerance corresponding to standard tolerance times allows for ample planning and execution time. Dynamic delay tolerance adjusts in real time according to changes in the worker's cognitive state and environmental conditions. When the rate of deterioration increases or cognitive load rises, the dynamic delay tolerance shortens to accelerate the intervention response. When the state tends to stabilize or cognitive load decreases, the dynamic delay tolerance lengthens to avoid excessively frequent interventions. When miners are working underground, safety helmets continuously monitor changes in their cognitive state, dynamically adjusting the delay tolerance based on fluctuations in the urgency index to ensure that intervention measures are both timely and effective while avoiding excessive frequency, minimizing interference with normal operations while ensuring safety. This dynamic adjustment mechanism achieves adaptive management of the worker's cognitive state.

[0064] Variable-speed control parameters are generated based on the dynamic delay tolerance, categorized into fast, standard, and slow response levels. The fast response level is for situations with extremely short dynamic delay tolerance, requiring immediate activation of a pre-set emergency intervention plan using a high-intensity multi-channel stimulus combination. The standard response level is for situations with moderate dynamic delay tolerance, allowing time to assess the worker's condition and select the most appropriate intervention strategy. The slow response level is for situations with longer dynamic delay tolerance, allowing for a gentle, gradual intervention to avoid excessive stimulation for the worker. Corresponding control parameters are set for each response level, including stimulus intensity, duration, frequency, and multi-channel combination. When a miner encounters an emergency underground and exhibits cognitive overload symptoms, detecting an extremely short dynamic delay tolerance, the system immediately switches to the fast response level, activating the highest-intensity combination of audible alarms and vibration alerts to rapidly increase the miner's alertness. Workers begin to feel drowsy towards the end of their night shift, but their condition deteriorates slowly. A slow response mode is used, employing gradually increasing ambient lighting and intermittent, gentle audio prompts to help workers gradually regain alertness. This tiered control allows for precise intervention and response to varying levels of urgency.

[0065] Step S160: The speed control parameters are compared with the preset safety threshold to generate a control deviation signal. The control deviation signal is then inversely modulated to generate a counter-stimulation waveform. A phase compensation execution command is then generated based on the counter-stimulation waveform.

[0066] Specifically, a control deviation signal is generated by comparing the variable speed control parameters with preset safety thresholds. The numerical values ​​of various indicators in the variable speed control parameters are extracted, including stimulus intensity, stimulus duration, and stimulus frequency. The upper and lower limits of the preset safety thresholds are obtained; these thresholds are set based on the hazard level of the work type and the individual characteristics of the workers. Each control parameter is compared with its corresponding safety threshold to identify parameters exceeding the safe range. When a control parameter exceeds the upper safety limit or falls below the lower safety limit, that parameter is in a deviation state. For example, when workers are working at heights, extreme fatigue may cause the stimulus intensity generated by the rapid response mode to exceed the worker's individual tolerance limit, resulting in a deviation of the stimulus intensity parameter from the safety threshold. Similarly, when miners are working underground, if a rapid deterioration in their cognitive state is detected, the generated stimulus frequency may deviate from the optimal resonant frequency range, failing to effectively activate the target EEG band. The deviation direction and magnitude of each deviation parameter are recorded; the deviation direction is categorized as positive or negative, and the deviation magnitude reflects the severity of the deviation. A control deviation signal is constructed, containing information such as the type, direction, magnitude, and duration of the deviation parameter.

[0067] In some embodiments, generating a counter-stimulation waveform by inverse phase modulation of the control deviation signal includes: expanding the control deviation signal in the phase domain to obtain the dominant phase component; performing inversion detection on the dominant phase component to identify the counter-stimulation point; constructing a reverse compensation sequence based on the counter-stimulation point; and using the reverse compensation sequence to perform phase compensation modulation on the control deviation signal to generate a counter-stimulation waveform.

[0068] The control deviation signal is expanded in the phase domain to obtain the dominant phase component. Phase analysis is performed on the control deviation signal to extract its main phase components. By analyzing the phase energy distribution, the phase component with the largest energy contribution is identified as the dominant phase component. The dominant phase component represents the core phase characteristics of the control deviation and reflects the main oscillation mode of the deviation signal on the time axis. When workers receive stimuli, if the stimulus parameters periodically deviate from the safe range, the deviation signal exhibits a specific periodic phase pattern. The dominant phase component corresponds to the starting phase of the deviation period, which is a key reference point for implementing counter-intervention. Analyzing the temporal characteristics of the dominant phase component provides a benchmark for subsequent reversal detection and compensation design. When miners are working underground, if the stimulus frequency continuously deviates from the target frequency band, the dominant phase component shows obvious phase drift characteristics. By tracking this characteristic, the deviation trend can be accurately grasped.

[0069] For example, the step of detecting and identifying the counter-excitation point by reversing the dominant phase component includes: using the dominant phase component as a detection reference to perform phase deviation detection on adjacent phase intervals to obtain deviation distribution data; performing reversal amplitude analysis on the deviation distribution data to generate an amplitude distribution; continuously locating the reversal position using the amplitude distribution to obtain positioning accuracy; and determining the position as the counter-excitation point when the positioning accuracy meets a preset phase requirement.

[0070] The dominant phase component is used as the detection reference to detect phase deviations in adjacent phase intervals, obtaining deviation distribution data. The phase value of the dominant phase component at each moment is extracted and used as the detection reference phase. Adjacent phase intervals of the dominant phase component are defined, and the deviations of each phase component within these intervals from the detection reference phase are analyzed. The phase deviation equals the phase value of each phase component minus the detection reference phase value; a positive deviation indicates that the component's phase leads the dominant phase, while a negative deviation indicates that it lags behind. The phase deviation values ​​at each moment are statistically analyzed to form deviation distribution data. This deviation distribution data describes the phase dispersion and distribution pattern of the control deviation signal around the dominant phase. During stimulus reception, if the stimulus parameters deviate, causing multiple frequency bands to lose phase coordination, the deviation distribution data shows highly dispersed phase components and a large deviation value distribution range. When the stimulus parameters remain within a safe range, the phase components are concentrated near the dominant phase, and the deviation value distribution range is small.

[0071] Amplitude distribution is generated by performing inversion amplitude analysis on the deviation distribution data. The signal amplitude corresponding to each deviation value in the deviation distribution data is analyzed; phase components with larger deviation values ​​and larger signal amplitudes have a more significant impact on deviation behavior. Locations where the deviation sign changes are extracted from the deviation distribution data; a change in deviation sign indicates that the phase component has changed from a leading dominant phase to a lagging dominant phase, or vice versa. The changes in signal amplitude at these locations are analyzed, and locations with significant amplitude jumps are marked as candidate points for inversion amplitude. These candidate points correspond to moments when both phase and amplitude change significantly simultaneously; these moments are turning points in the deviation signal behavior pattern. The amplitude values ​​of each candidate point are statistically analyzed to form an amplitude distribution. The amplitude distribution describes the distribution pattern of the inversion phenomenon at different amplitude levels. When miners experience abnormal stimulus responses due to fatigue underground, and both the phase and amplitude of the deviation signal show inversion characteristics simultaneously, the amplitude distribution exhibits a distinct bimodal or multimodal structure, with the peak positions corresponding to key inversion moments.

[0072] Positioning accuracy is obtained by continuously locating the reversal point using amplitude distribution. Peak amplitude positions are identified within the amplitude distribution, corresponding to the moments when the reversal phenomenon is most pronounced. Each peak position is time-localized to obtain its precise time point. Continuous tracking and positioning of the peak positions are performed, repeatedly detecting them over multiple consecutive sampling periods to evaluate their temporal stability. Stable peak positions across multiple sampling periods indicate high positioning accuracy, while large fluctuations indicate low accuracy. When the temporal fluctuation range of the peak position is less than a preset threshold and the peak amplitude meets minimum requirements, the position is identified as a reliable counter-current trigger point. During high-risk operations, personnel continuously monitor deviations in stimulus parameters. When the amplitude distribution displays a clear peak and continuous positioning accuracy meets requirements, the location of the counter-current trigger point is precisely identified, ensuring the compensation mechanism is activated at the optimal time. The counter-current trigger point marks the critical moment for implementing reverse compensation; injecting an anti-phase signal at these moments yields the best counter-current effect.

[0073] A reverse compensation sequence is constructed based on the counter-excitation points. The time position and corresponding dominant phase value of each counter-excitation point are extracted. The dominant phase value represents the phase state of the deviation signal at the excitation point. The dominant phase value is inverted, with the inverted phase differing from the original phase by π. The inverted phase corresponds to the phase of the counter-excitation waveform. Reverse compensation pulses are generated at each counter-excitation point. The phase of the compensation pulse is the inverted phase, the amplitude is determined according to the deviation amplitude and compensation requirements, and the duration is set according to the deviation duration. The time intervals between each counter-excitation point are analyzed; short intervals indicate frequent deviations, while long intervals indicate occasional deviations. For frequent deviations, the amplitude of the compensation pulse is increased to enhance the counter-effect. For occasional deviations, a standard amplitude compensation pulse is used to avoid over-intervention. The reverse compensation pulses at each counter-excitation point are arranged in chronological order to form a reverse compensation sequence. The reverse compensation sequence contains the compensation signal parameters at key moments throughout the deviation process. During long-term operations, the stimulation parameters may deviate from the safe range multiple times. Compensation pulses are generated at each deviation counter-excitation point, and these pulses constitute a continuous reverse compensation sequence to continuously counteract the deviation trend.

[0074] A phase-compensated modulation of the control deviation signal using a reverse compensation sequence generates a counter-stimulation waveform. Each compensation pulse in the reverse compensation sequence is superimposed on the control deviation signal at the corresponding time, considering phase relationship and amplitude weighting during the superposition process. Since the phase of the compensation pulse is opposite to that of the deviation signal, a phase cancellation effect occurs after superposition, weakening or eliminating the amplitude of the deviation signal. The amplitude weights of the compensation pulses are adjusted to bring the amplitude of the superimposed signal back to a safe range. The superimposed signal is then smoothed to eliminate any spikes or abrupt changes that may occur during the compensation process. The smoothed signal is the counter-stimulation waveform, which retains the effective components of the original stimulus while suppressing abnormal components that deviate from the safe range. When workers are operating at heights and the stimulus intensity parameter is detected to be persistently high, the counter-stimulation waveform generated by the reverse compensation sequence is superimposed on the original stimulus to maintain the actual stimulus intensity applied to the workers at a safe and effective level. When miners are working underground, if the stimulus frequency deviates from the optimal resonant frequency, the counter-stimulation waveform corrects the frequency back to the target range through inverse phase modulation, ensuring the intervention effect.

[0075] Phase compensation execution commands are generated based on the counter-stimulation waveform. The amplitude and phase information of the counter-stimulation waveform are analyzed to determine the type and intensity of phase compensation. Phase compensation is categorized into positive and negative compensation. Positive compensation enhances insufficient stimulation parameters, while negative compensation weakens excessive stimulation parameters. The phase compensation intensity is dynamically adjusted based on the severity of control deviation and the safety requirements of the work scenario; strong compensation is used for severe deviations, and weak compensation for minor deviations. The phase matching degree between the compensation waveform and the original deviation signal is evaluated; a high matching degree indicates good compensation effect, while a low matching degree indicates the need for further optimization of compensation parameters. A phase compensation execution command is generated, containing parameters such as the frequency, phase, amplitude, and duration of the compensation waveform. These parameters collectively determine the accuracy and effectiveness of the compensation. The phase compensation execution command is sent to the stimulation output device on the safety helmet. The device modulates the original stimulation signal in real time according to the command parameters, achieving deviation correction through phase superposition. During high-risk operations, if a deviation in stimulation parameters is detected, a phase compensation execution command is immediately generated, and the deviation is quickly corrected using the counter-stimulation waveform, ensuring that the intervention is timely and effective without causing discomfort or interference due to overstimulation. When miners encounter emergencies underground, a phase compensation mechanism adjusts stimulation parameters in real time to ensure effective arousal while preventing stimulation intensity from exceeding safe limits and causing secondary injury. The entire closed-loop control process begins with comparing the variable speed control parameters with safety thresholds, then identifies control deviation signals and generates counter-stimulation waveforms, ultimately achieving dynamic adjustment and safety assurance of stimulation parameters through phase compensation execution commands. This ensures that workers receive safe and effective intervention stimuli under various cognitive states.

[0076] To implement the brainwave-based intelligent control method for safety helmets 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 a helmet intelligent control system 200 based on electroencephalogram (EEG) characteristics according to an embodiment of this application. For ease of explanation, only the parts relevant to this embodiment are shown. The helmet intelligent control system 200 based on EEG characteristics provided in this embodiment includes: The signal acquisition module 201 is used to acquire real-time EEG signals of operators, perform spectral decomposition on the real-time EEG signals to generate cognitive load frequency domain features, and establish an EEG control space based on the cognitive load frequency domain features. Frequency band analysis module 202 is used to form a multi-frequency band energy distribution matrix for the EEG control spatial scan, identify frequency band coupling imbalance points along the multi-frequency band energy distribution matrix, deduce the timing of collaborative intervention through the frequency band coupling imbalance points, and form a multi-channel regulation mode based on the collaborative intervention timing. The resonance matching module 203 is used to acquire the stimulation frequency range of the safety helmet and the target EEG frequency band, map the stimulation frequency range to the EEG control space to generate a frequency matching domain, and establish a resonance intervention window based on the resonance analysis of the frequency matching domain on the target EEG frequency band. The propagation tracking module 204 is used to perform frequency band monitoring within the resonance intervention window using the multi-channel adjustment mode to obtain the energy migration trajectory, construct a risk propagation chain according to the energy migration trajectory to identify cascade failure nodes, and perform blocking priority assessment on the cascade failure nodes to form key blocking anchor points. The response scheduling module 205 is used to perform risk acceleration analysis on the key blocking anchor point to form a state deterioration rate, use the state deterioration rate to fuse the cognitive load frequency domain characteristics to determine the response time and generate a dynamic delay tolerance, and perform fast and slow grading to generate variable speed control parameters according to the dynamic delay tolerance. The phase control module 206 is used to compare the speed control parameters with a preset safety threshold to generate a control deviation signal, perform anti-phase modulation on the control deviation signal to generate a counter-stimulation waveform, and form a phase compensation execution command based on the counter-stimulation waveform.

[0077] The aforementioned intelligent helmet control system 200 based on EEG characteristics can implement the intelligent helmet control method based on EEG characteristics 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.

[0078] 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.

[0079] 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 method for intelligent control of a safety helmet based on electroencephalogram (EEG) characteristics, characterized in that, include: Real-time EEG signals of operators are collected, and the real-time EEG signals are spectrally decomposed to generate cognitive load frequency domain features. Based on the cognitive load frequency domain features, an EEG control space is established. The EEG-controlled spatial scan forms a multi-band energy distribution matrix. Frequency band coupling imbalance points are identified along the multi-band energy distribution matrix. The timing of collaborative intervention is deduced in reverse from the frequency band coupling imbalance points. A multi-channel regulation mode is formed based on the collaborative intervention timing. The frequency range of the safety helmet stimulation and the target EEG frequency band are obtained. The stimulation frequency range is mapped to the EEG control space to generate a frequency matching domain. Based on the target EEG frequency band, the resonance intervention window is established by resonant analysis of the frequency matching domain. The multi-channel adjustment mode is used to perform frequency band monitoring within the resonance intervention window to obtain the energy migration trajectory. The risk propagation chain is constructed according to the energy migration trajectory to identify cascade failure nodes. The blocking priority assessment is performed on the cascade failure nodes to form key blocking anchor points. Risk acceleration analysis is performed on the key blocking anchor points to form a state deterioration rate. The state deterioration rate is then fused with the cognitive load frequency domain characteristics to determine the response time and generate a dynamic delay tolerance. Based on the dynamic delay tolerance, fast and slow gears are used to generate variable speed control parameters. The speed control parameters are compared with a preset safety threshold to generate a control deviation signal. The control deviation signal is then inversely modulated to generate a counter-stimulation waveform. A phase compensation execution command is then generated based on the counter-stimulation waveform.

2. The method according to claim 1, characterized in that, The establishment of the EEG control space based on the frequency domain characteristics of the cognitive load includes: Energy distribution characteristics are extracted by performing energy spectral analysis on the frequency domain characteristics of the cognitive load. The boundary transition threshold is calibrated based on the energy distribution characteristics. Based on the fluctuation amplitude of the frequency domain characteristics of the cognitive load monitored by the boundary transition threshold, a hierarchical mapping rule is established, which includes a normal working layer, an early warning transition layer, and a loss of control intervention layer. The aforementioned hierarchical mapping rules are used to construct the EEG control space.

3. The method according to claim 1, characterized in that, The method of deriving the timing of coordinated intervention in reverse from the frequency band coupling imbalance point includes: Extract duration data from the frequency band coupling imbalance point; Based on the duration data, the deterioration acceleration gradient is identified; High-risk time periods are defined based on the moment when the deterioration acceleration gradient exceeds a preset threshold. The starting point of the high-risk time period is moved forward to set a buffer period as the timing for collaborative intervention.

4. The method according to claim 1, characterized in that, The determination of the resonance intervention window based on the frequency-matched domain resonance analysis of the target EEG frequency band includes: Phase-locking is performed on the frequency matching domain to extract frequency components synchronized with the target EEG band; The frequency components are subjected to coupling degree filtering to generate a set of highly coupled frequencies; The peak resonance point is obtained by performing a matching degree analysis between the high-coupling frequency set and the target EEG frequency band. The resonance intervention window is determined by extending the peak resonance point to both sides over a set time span.

5. The method according to claim 1, characterized in that, The step of constructing a risk propagation chain and identifying cascading failure nodes according to the energy migration trajectory includes: In the energy migration trajectory, a propagation direction window is divided to generate a set of direction windows; Based on the aforementioned directional window set, propagation path planning is performed to obtain the main propagation channel; The main propagation channel is subjected to intensity modulation processing to generate a propagation payload; The propagation load is used to perform cascading relationship mapping to identify cascading failure nodes.

6. The method according to claim 1, characterized in that, The step of using the state deterioration rate to fuse the cognitive load frequency domain features to determine the response time and generate dynamic delay tolerance includes: The rate of deterioration of the state is analyzed to identify accelerating mutation points; The urgency index is generated by cross-correlation analysis between the accelerated mutation point and the frequency domain characteristics of the cognitive load. Tolerance time levels are determined based on the aforementioned urgency index; A dynamic delay tolerance is formed based on the tolerance time level.

7. The method according to claim 1, characterized in that, The step of generating a counter-stimulation waveform by inverse phase modulation of the control deviation signal includes: The control deviation signal is expanded in the phase domain to obtain the dominant phase component; The dominant phase component is inverted and the counter-excitation point is identified. A reverse compensation sequence is constructed based on the hedging trigger point; The control deviation signal is phase-compensated and modulated using the reverse compensation sequence to generate a counter-stimulation waveform.

8. The method according to claim 6, characterized in that, The step of generating an urgency index by performing cross-correlation analysis between the accelerated mutation point and the frequency domain characteristics of cognitive load includes: Temporal feature extraction is performed on the accelerated mutation point to obtain the mutation time feature; Synchronization features are extracted by synchronously sampling the frequency domain features of the cognitive load based on the characteristics of the abrupt change time. The correlation between the mutation time feature and the synchronization feature is evaluated to generate a coupling response coefficient; An urgency index is formed based on the coupling response coefficients.

9. The method according to claim 7, characterized in that, The step of detecting and identifying the counter-excitation point by reversing the dominant phase component includes: The dominant phase component is used as the detection benchmark to perform phase deviation detection in adjacent phase intervals to obtain deviation distribution data. The deviation distribution data is subjected to inversion amplitude analysis to generate an amplitude distribution; The positioning accuracy is obtained by continuously locating the reversal position using the amplitude distribution. When the positioning accuracy meets the preset phase requirement, the position is determined as the counter-excitation point.

10. A smart control system for a safety helmet based on electroencephalogram (EEG) characteristics, characterized in that, include: The signal acquisition module is used to acquire real-time EEG signals from operators, perform spectral decomposition on the real-time EEG signals to generate cognitive load frequency domain features, and establish an EEG control space based on the cognitive load frequency domain features. The frequency band analysis module is used to form a multi-frequency band energy distribution matrix for the EEG control spatial scan, identify frequency band coupling imbalance points along the multi-frequency band energy distribution matrix, deduce the timing of collaborative intervention through the frequency band coupling imbalance points, and form a multi-channel regulation mode based on the collaborative intervention timing. The resonance matching module is used to acquire the stimulation frequency range of the safety helmet and the target EEG frequency band, map the stimulation frequency range to the EEG control space to generate a frequency matching domain, and establish a resonance intervention window based on the resonance analysis of the frequency matching domain on the target EEG frequency band. The propagation tracking module is used to perform frequency band monitoring within the resonance intervention window using the multi-channel adjustment mode to obtain the energy migration trajectory, construct a risk propagation chain according to the energy migration trajectory to identify cascade failure nodes, and perform blocking priority assessment on the cascade failure nodes to form key blocking anchor points. The response scheduling module is used to perform risk acceleration analysis on the key blocking anchor points to form a state deterioration rate, use the state deterioration rate to fuse the cognitive load frequency domain characteristics to determine the response time and generate a dynamic delay tolerance, and perform fast and slow grading to generate variable speed control parameters according to the dynamic delay tolerance. The phase control module is used to compare the speed control parameters with a preset safety threshold to generate a control deviation signal, perform anti-phase modulation on the control deviation signal to generate a counter-stimulation waveform, and form a phase compensation execution command based on the counter-stimulation waveform.

Images