Signal quality optimization method, device, equipment and medium for intelligent brain-controlled headset

Abstract

The present application discloses a signal quality optimization method, device, equipment and medium for an intelligent brain-controlled headset. The method includes: obtaining multiple quality levels corresponding to the original signal collected by the intelligent brain-controlled headset and quality data corresponding to each quality level; obtaining interference characteristics and corresponding enhanced signals in the quality data; organizing the enhanced signal in segments, obtaining multiple processing sequences, determining resource allocation information to determine the optimization channel and the corresponding signal inflection point, determining the monitoring process corresponding to the signal tracking, and obtaining the quality monitoring result corresponding to the processing sequence; obtaining the rule chain corresponding to the quality monitoring result, determining the processing scheme and the optimization parameters corresponding to the processing scheme, the corresponding key nodes and the scene mode corresponding to the key nodes, adjusting the optimization parameters based on the scene mode, obtaining the adaptation results corresponding to the optimization parameter adjustment, determining the feature area and converting it into a quality expression, generating an optimization signal according to the quality expression, and completing the quality optimization of the original signal.

Description

Technical Field

[0001] The present application relates to the field of brain-computer interface technology, and in particular to a signal quality optimization method, device, equipment and medium for intelligent brain-controlled headphones. Background Art

[0002] With the development of brain-computer interface technology, smart brain-controlled headphones, as a new type of portable device, are gaining more and more attention for their performance in daily applications. Early signal quality optimization methods mainly used classic digital signal processing techniques, such as fixed-parameter frequency domain filtering and time domain smoothing. These methods lack scene adaptability and are difficult to cope with complex and changing actual usage environments. In recent years, although researchers have proposed optimization techniques including adaptive filtering and independent component analysis, they still face problems such as inaccurate feature extraction and unreasonable parameter configuration when applied to portable brain-controlled headphones.

[0003] In daily use, smart brain-controlled headphones face a variety of complex interference sources. The existing interference feature analysis methods lack systematicity, and it is difficult to establish an accurate spectrum labeling and signal strength evaluation system, resulting in unsatisfactory detection and feature extraction of interference fragments. The signal quality evaluation system is imperfect, lacks multi-dimensional analysis of time domain stability, frequency domain characteristics and spatial distribution, and cannot accurately reflect the actual quality status of the signal. The optimization configuration of the processing channel is too static, and it is difficult to dynamically adjust the processing strategy according to the monitoring results, which affects the real-time performance of signal processing. The parameter configuration in the signal enhancement process lacks scene adaptability, and it is impossible to reasonably adapt the features according to different usage environments and user states, which easily causes distortion of the enhanced signal and loss of key features. In addition, the existing signal quality optimization methods generally have problems such as fragmented processing flow, single quality evaluation, and lagging parameter adjustment, which makes it difficult to meet the continuous optimization needs of portable brain-controlled headphones for signal quality in practical applications.

[0004] Therefore, a method is urgently needed to solve at least one of the above problems. Summary of the invention

[0005] In a first aspect, an embodiment of the present application provides a method for optimizing the signal quality of an intelligent brain-controlled headset, comprising:

[0006] Acquire an original signal collected by the intelligent brain-controlled headset, and acquire multiple quality levels corresponding to the original signal and quality data corresponding to each quality level;

[0007] Obtaining an interference segment from the quality data, generating an interference feature according to the interference segment, obtaining a processing interval corresponding to the interference feature and a processing signal corresponding to the processing interval, performing signal enhancement on the processed signal, and obtaining an enhanced signal;

[0008] Organizing the enhanced signal in sections to obtain multiple processing sequences, determining resource allocation information according to the processing sequences, and determining an optimized channel according to the resource allocation information;

[0009] Obtaining a signal inflection point corresponding to the optimization channel, determining a monitoring process corresponding to the signal tracking according to the signal inflection point, and obtaining a quality monitoring result corresponding to the processing sequence according to the monitoring process;

[0010] Obtaining a rule chain corresponding to the quality monitoring result, and determining a processing scheme and optimization parameters corresponding to the processing scheme according to the rule chain;

[0011] Acquire key nodes corresponding to the optimization parameters and scene modes corresponding to the key nodes, adjust the optimization parameters based on the scene modes, and acquire adaptation results corresponding to the optimization parameter adjustments;

[0012] A characteristic region is determined according to the adaptation result, the characteristic region is converted into a quality expression, an optimization signal is generated according to the quality expression, and quality optimization of the original signal is completed.

[0013] In a second aspect, the present application further provides a signal quality optimization device, comprising:

[0014] A signal acquisition module, used to acquire the original signal collected by the intelligent brain-controlled headset, and acquire multiple quality levels corresponding to the original signal and quality data corresponding to each quality level;

[0015] a fragment acquisition module, configured to acquire interference fragments from the quality data, generate interference features according to the interference fragments, acquire a processing interval corresponding to the interference features and a processing signal corresponding to the processing interval, perform signal enhancement on the processing signal, and acquire an enhanced signal;

[0016] A sequence acquisition module, used to organize the enhanced signal in sections, acquire multiple processing sequences, determine resource allocation information according to the processing sequences, and determine an optimized channel according to the resource allocation information;

[0017] An inflection point acquisition module, used to acquire the signal inflection point corresponding to the optimization channel, determine the monitoring process corresponding to the signal tracking according to the signal inflection point, and acquire the quality monitoring result corresponding to the processing sequence according to the monitoring process;

[0018] A parameter determination module, used to obtain a rule chain corresponding to the quality monitoring result, and determine a processing scheme and an optimization parameter corresponding to the processing scheme according to the rule chain;

[0019] A result acquisition module, used to acquire key nodes corresponding to the optimization parameters and scene modes corresponding to the key nodes, adjust the optimization parameters based on the scene modes, and acquire adaptation results corresponding to the optimization parameter adjustments;

[0020] The optimization completion module is used to determine the characteristic region according to the adaptation result, convert the characteristic region into a quality expression, generate an optimization signal according to the quality expression, and complete the quality optimization of the original signal.

[0021] In a third aspect, the present application also provides a computer device, including a processor and a memory, wherein the memory is used to store a computer program, and when the computer program is executed by the processor, the signal quality optimization method of the intelligent brain-controlled headset as described in the first aspect is implemented.

[0022] In a fourth aspect, the present application also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the signal quality optimization method of the intelligent brain-controlled headset as described in the first aspect.

[0023] Compared with the prior art, this application has at least the following beneficial effects:

[0024] 1The present invention combines time domain, frequency domain and space features to evaluate signal quality by constructing a multi-dimensional quality evaluation system. First, the original signal is collected for spectrum marking and converted into signal strength for quality level evaluation; then interference features are extracted based on quality data, and a feature library is established for type matching; finally, a comprehensive evaluation of signal quality is achieved through signal tracking and quality monitoring of processing channels, which effectively improves the accuracy and reliability of signal quality evaluation.

[0025] 2 The present invention adopts a scenario-aware parameter optimization strategy, extracts rules based on monitoring results and constructs rule chains, determines key nodes through feature adaptation and maps them to scenario modes, and then dynamically adjusts processing parameters based on scenario modes. This adaptive parameter configuration mechanism enables the method to adjust the processing strategy in a timely manner according to the actual use environment and user status, significantly improving the adaptability and real-time performance of signal processing.

[0026] 3 The present invention proposes a signal enhancement method with hierarchical processing, and constructs a multi-level signal channel network by dividing and subdividing the signal processing interval. The segmented organization and channel selection strategy are adopted to realize the progressive processing from signal conditioning to feature enhancement. This hierarchical and progressive processing mechanism effectively improves the accuracy and stability of signal enhancement while ensuring signal continuity.

[0027] It should be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 A schematic diagram of a flow chart of a signal quality optimization method for an intelligent brain-controlled headset according to an embodiment of the present application;

[0029] Figure 2 This is a schematic diagram of the structure of a signal quality optimization device shown in an embodiment of the present application;

[0030] Figure 3 A schematic diagram of the structure of a computer device shown in an embodiment of the present application. DETAILED DESCRIPTION

[0031] In the following description, specific details such as specific system structures, technologies, etc. are provided for the purpose of illustration rather than limitation, so as to provide a thorough understanding of the embodiments of the present application. However, it should be clear to those skilled in the art that the present application may also be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to prevent unnecessary details from obstructing the description of the present application.

[0032] It should be understood that when used in the present specification and the appended claims, the term "comprising" indicates the presence of described features, integers, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or combinations thereof.

[0033] It should also be understood that the term “and / or” used in the specification and appended claims refers to any and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0034] As used in the specification and appended claims of this application, the term "if" can be interpreted as "when" or "uponce" or "in response to determining" or "in response to detecting", depending on the context. Similarly, the phrase "if it is determined" or "if [described condition or event] is detected" can be interpreted as meaning "uponce it is determined" or "in response to determining" or "uponce [described condition or event] is detected" or "in response to detecting [described condition or event]", depending on the context.

[0035] In addition, in the description of the present application specification and the appended claims, the terms "first", "second", "third", etc. are only used to distinguish the descriptions and cannot be understood as indicating or implying relative importance.

[0036] References to "one embodiment" or "some embodiments" etc. described in the specification of this application mean that one or more embodiments of the present application include specific features, structures or characteristics described in conjunction with the embodiment. Therefore, the statements "in one embodiment", "in some embodiments", "in some other embodiments", "in some other embodiments", etc. that appear in different places in this specification do not necessarily refer to the same embodiment, but mean "one or more but not all embodiments", unless otherwise specifically emphasized in other ways. The terms "including", "comprising", "having" and their variations all mean "including but not limited to", unless otherwise specifically emphasized in other ways.

[0037] The technical solution of the embodiment of the present application is introduced below.

[0038] With the development of brain-computer interface technology, smart brain-controlled headphones, as a new type of portable device, are gaining more and more attention for their performance in daily applications. Early signal quality optimization methods mainly used classic digital signal processing techniques, such as fixed-parameter frequency domain filtering and time domain smoothing. These methods lack scene adaptability and are difficult to cope with complex and changing actual usage environments. In recent years, although researchers have proposed optimization techniques including adaptive filtering and independent component analysis, they still face problems such as inaccurate feature extraction and unreasonable parameter configuration when applied to portable brain-controlled headphones.

[0039] In daily use, smart brain-controlled headphones face a variety of complex interference sources. The existing interference feature analysis methods lack methodology, and it is difficult to establish an accurate spectrum labeling and signal strength evaluation system, resulting in unsatisfactory detection and feature extraction of interference fragments. The signal quality evaluation system is imperfect, lacks multi-dimensional analysis of time domain stability, frequency domain characteristics and spatial distribution, and cannot accurately reflect the actual quality status of the signal. The optimization configuration of the processing channel is too static, and it is difficult to dynamically adjust the processing strategy according to the monitoring results, which affects the real-time performance of signal processing. The parameter configuration in the signal enhancement process lacks scene adaptability, and it is impossible to reasonably adapt the features according to different usage environments and user states, which easily causes distortion of the enhanced signal and loss of key features. In addition, the existing signal quality optimization methods generally have problems such as fragmented processing flow, single quality evaluation, and lagging parameter adjustment, which makes it difficult to meet the continuous optimization needs of portable brain-controlled headphones for signal quality in practical applications.

[0040] To solve the above problems, please refer to Figure 1 , Figure 1A flowchart of a method for optimizing the signal quality of a smart brain-controlled headset provided in an embodiment of the present application. The method for optimizing the signal quality of a smart brain-controlled headset in an embodiment of the present application can be applied to computer devices, including but not limited to smart phones, laptops, tablet computers, desktop computers, physical servers, and cloud servers. Figure 1 As shown, the signal quality optimization method of the intelligent brain-controlled headset of this embodiment includes steps S101 to S107, which are described in detail as follows:

[0041] Step S101, obtain the original signal collected by the intelligent brain-controlled headset, obtain multiple quality levels corresponding to the original signal and quality data corresponding to each quality level.

[0042] Specifically, the EEG signals of the brain-controlled headset are collected through a high-precision sampling circuit, with the sampling frequency set to 1024Hz and the sampling accuracy set to 24 bits. Four electrode points (F3, F4, F7, F8) are arranged in the frontal lobe area, and two electrode points (O1, O2) are arranged in the occipital lobe area. A differential amplifier circuit is used to achieve multi-channel synchronous sampling. In daily application scenarios, the method automatically adjusts the sampling parameters according to the use environment. For example, when the user detects significant motion interference during walking, the method reduces the amplifier gain from the standard value of 1000 to the range of 300-500, and increases the cutoff frequency of the digital filter to 45Hz; when the user sits and works indoors, the method increases the amplifier gain to the range of 2000-2500 and reduces the cutoff frequency of the digital filter to 35Hz to fully capture weak EEG signals. The impedance between the electrode and the skin is continuously monitored. When the impedance of a certain electrode point exceeds 10kΩ, the method automatically marks the channel data. The collected original signal is transformed into a spectrum distribution through 256-point fast Fourier transform. Each spectrum marker contains the frequency point location information, the corresponding amplitude and the timestamp record. In practical applications, for example, when the user remains in a sitting state for a long time, the spectrum marker presents a stable low-frequency dominant feature; when the user performs strenuous exercise, the spectrum marker will have an obvious high-frequency component enhancement. The method classifies and labels the spectrum features in these different states to form a spectrum marker dataset containing dimensions such as frequency, amplitude, and time.

[0043] Based on the obtained spectrum marker dataset, the method performs signal strength conversion calculation. For typical EEG rhythms such as the δ band (0.5-4Hz), theta band (4-8Hz), alpha band (8-13Hz) and beta band (13-30Hz), a dynamic calculation window of 2 seconds is set, and the window overlap rate is 50%. The method integrates the spectrum markers of each frequency band to obtain the corresponding energy distribution. In office scene applications, when it is detected that the user is in a focused working state, the spectrum markers of the beta band show continuous high energy characteristics, and the method correspondingly increases the calculation weight of this frequency band to 60%; when the user enters a short rest state, the energy of the spectrum markers of the alpha band is significantly enhanced, and its calculation weight is increased to 50%. For sports scenes, the method can identify typical motion artifact spectrum markers in the range of 0.8-2Hz, and reduce the interference effect by reducing the calculation weight of this frequency band to less than 10%. At the same time, the method also monitors the spectrum markers of the high frequency band (above 35Hz), and automatically adjusts the signal strength calculation ratio of the relevant frequency band when electromyographic signal interference is found. When used in complex environments such as subways, the method analyzes the spectrum marker characteristics of 50Hz power frequency interference and dynamically adjusts the notch filter parameters to ensure the accuracy of signal strength calculation. After multi-dimensional calculation processing, the method finally outputs a data stream containing the real-time signal strength values ​​of each frequency band.

[0044] In some embodiments, obtaining multiple quality levels corresponding to the original signal and quality data corresponding to each quality level includes: obtaining a spectrum mark of the original signal; obtaining a signal strength corresponding to the spectrum mark; calculating the quality level based on the signal strength, and obtaining the quality level and quality data corresponding to each spectrum mark.

[0045] Based on the real-time calculated signal strength data stream, the method established a hierarchical quality assessment mechanism. In the signals collected at the main electrode positions, the excellent quality requires that the signal strength values ​​of the core frequency bands (α waves and β waves) are stably maintained above 85% of the baseline threshold, and the intensity ratio fluctuation range between each frequency band does not exceed 20%. Taking the classroom focused learning scene as an example, when students concentrate, the β band signal strength recorded by the prefrontal electrodes (F3, F4) reaches 90% of the baseline value, the α band is maintained at 87%, and the intensity ratio of the θ wave and the δ wave remains stable. The method determines it as an excellent quality level. For outdoor sports scenes, even if there is the influence of motion artifacts, as long as the high-frequency interference component is controlled within 15% of the total energy, the low-frequency motion artifact does not exceed 25%, and the signal strength of the target EEG rhythm is maintained within the range of 70%-85% of the baseline threshold, it can still be rated as a good level. When users use devices in noisy public places, the method establishes an anti-interference ability scoring index by analyzing the modulation characteristics of power frequency interference. If the modulation depth of the power frequency interference exceeds 30%, even if the signal strength of the core frequency band is still in the range of 60%-70%, it will be downgraded to the general quality level. For dynamic monitoring of signal quality, the method updates the quality assessment results every 500 milliseconds and keeps a continuous 10-second assessment record to analyze the changing trend of signal quality. The final quality data output includes: signal strength time series data of each electrode channel, real-time quality grade assessment results, abnormal frequency band identification information, and quality assessment statistical characteristics under different usage scenarios.

[0046] Step S102, obtaining an interference segment from the quality data, generating an interference feature according to the interference segment, obtaining a processing interval corresponding to the interference feature and a processing signal corresponding to the processing interval, performing signal enhancement on the processed signal, and obtaining an enhanced signal.

[0047] Specifically, the method receives the quality data output in the step. For the data collected by the frontal electrodes (F3, F4), when the beta band signal strength drops from 90% of the baseline value to 45% within 3 consecutive seconds, and the quality level drops from excellent to to be optimized, the method marks the time period as a suspected interference segment. In long-term recording, by analyzing the statistical characteristics of quality assessment, the method identifies a variety of typical interference modes: when the user turns his head or blinks quickly, a transient interference of 200-300ms is generated in the frontal electrode signal, and the quality level drops briefly; when the user grits his teeth or frowns, electromyographic interference lasting 2-3 seconds is introduced, resulting in abnormal high-frequency energy; during walking, all electrodes present a rhythmic quality fluctuation of 0.8-2Hz. For each identified interference interval, the method extracts information such as its start and end time points, the number of the affected electrode, the abnormal frequency band range, and the signal strength change curve. At the same time, based on the scene quality assessment characteristics, the method analyzes the data of 30 seconds before and after the interference occurs to obtain the inducing conditions and attenuation laws of the interference. When used in public places, the method can identify complex situations such as power frequency interference superposition and electromagnetic environment mutation from the quality data. Finally, a complete interference fragment data set is formed, each fragment contains time domain features (duration, amplitude change), frequency domain features (frequency band distribution, energy proportion), space features (electrode distribution, propagation characteristics) and scene features (usage environment, user status).

[0048] In some embodiments, interference fragments are obtained from the quality data, and interference features are generated based on the interference fragments, including: performing feature analysis on the quality data to obtain interference fragments corresponding to the quality data; performing feature matching between the interference fragments and a preset feature library to obtain interference types corresponding to the interference fragments; and obtaining interference features corresponding to the interference types.

[0049] Based on the interference fragment dataset obtained above and the multi-dimensional feature information contained in it, the method constructs and organizes the feature library. The feature library adopts a multi-level classification structure. First, according to the time domain characteristics of the interference fragments, it is divided into transient class (less than 500ms), short-term class (500ms-2s) and continuous class (greater than 2s). Under each time category, it is subdivided into low-frequency interference (0-15Hz), power frequency interference (45-55Hz), broadband interference (0-100Hz) and other subclasses according to the frequency domain characteristics. Finally, based on the spatial characteristics, it is further divided into local type (1-2 electrodes), regional type (3-4 electrodes) and global type (all electrodes). For each type of interference, the method establishes a standard template containing the expected value and fluctuation range of the characteristic parameters. For example, for the transient interference caused by blinking, its standard template defines the amplitude mutation threshold (signal strength reduction of more than 70%), duration range (150-250ms) and frequency band energy distribution characteristics of the prefrontal electrodes. The feature library also includes typical interference combination patterns in different usage environments based on scene characteristics. For example, in the subway scene, the amplitude of the 50Hz power frequency interference will show periodic modulation with the acceleration and deceleration of the train, and low-frequency vibration (2-5Hz) shows significant correlation on all electrodes. Through the analysis of a large amount of actual application data, the feature library continues to enrich its content, forming a hierarchical description structure including basic features, combined features and scene features. Finally, a feature description set is obtained, and each set element contains standardized feature templates, parameter thresholds and scene association information.

[0050] The method uses the feature description set organized in the feature library to perform type matching and feature generation for the newly detected interference fragments. In office scene applications, when the user uses a mobile phone close to the headset, the high-frequency band (35-100Hz) energy of the F7 and F8 electrodes increases, and the signal quality level decreases. The method compares the interference fragment with the feature library template and matches the "local electromagnetic interference" type. Then a feature description containing information such as the spatial position, frequency component, and attenuation characteristics of the interference is generated. In outdoor sports scenes, the method can identify and match multiple superimposed interferences: motion artifacts appear as global low-frequency periodic interference, environmental electromagnetic interference causes power frequency component fluctuations, and myoelectric signals produce irregular high-frequency fluctuations in the forehead area. For each identified interference type, the method generates a combined feature description containing relative strength, primary and secondary relationships, and interactive characteristics. When an unknown interference pattern is detected, the method calculates its similarity with the existing template. When the similarity is lower than the threshold, it is marked as a new interference and a temporary feature template is created. During continuous use, the method continuously updates and optimizes the feature template to improve the interference recognition ability. The generated interference features not only contain type information and feature parameters, but also record time-varying characteristics and scene correlation, providing a comprehensive feature basis for subsequent signal processing. The final output of complete interference feature data includes: basic feature parameters (duration, frequency range, spatial distribution), combined feature description (coupling relationship of multi-source interference, propagation characteristics), scene-related features (usage environment, user behavior correlation) and time-varying feature description (fluctuation law, modulation characteristics).

[0051] The method receives the complete interference feature data output by the step, and first determines the basic attributes of the processing interval according to the basic feature parameters: based on the duration feature, the processing redundancy of 2 times the duration of the interference is set before and after the transient interference, and the processing interval of 3 times the period is set for the continuous interference; based on the spatial distribution feature, the electrode coverage range of the processing interval is determined; based on the frequency range feature, the sampling parameter configuration of the processing interval is adjusted. Secondly, the method adjusts the superposition strategy of the processing interval according to the combined feature description: when there is a coupling relationship between multiple sources of interference, the mutually related interferences are divided into the same processing interval; according to the propagation characteristics of the interference, the propagation delay compensation of the processing interval is set. Then, the method optimizes the processing interval parameters according to the scene-related characteristics: according to the feature change law under different usage environments, the length coefficient of the processing interval is dynamically adjusted; based on the user behavior association characteristics, the trigger conditions of the processing interval are preset. Finally, the method improves the dynamic characteristics of the processing interval in combination with the time-varying feature description: according to the fluctuation law of the interference, the dynamic boundary of the processing interval is set; according to the modulation characteristics, the analysis window of the processing interval is configured. For example, when a combined action of blinking and head turning is detected, the method divides the 800ms comprehensive processing interval based on the transient characteristics of blinking (lasting 200ms) and the gradual characteristics of head turning (lasting 1s), combined with their propagation relationship. All divided processing intervals form a structured data set, each of which contains time range, spatial range, processing parameters, priority information, and correlation attributes between intervals.

[0052] Based on the obtained processing interval data set, the method subdivides the processing segment. Each processing interval adopts different subdivision strategies according to the interference characteristics: the single interference source area is evenly divided with a fixed step size (50ms); the multiple interference superposition area is adaptively divided according to the characteristic parameters of the primary and secondary interference sources, the primary interference determines the basic step size, and the secondary interference affects the adjustment coefficient of the step size. A gradual transition segment is set at the signal mutation point, and the step size is adjusted from short to long to ensure the continuity of processing. In specific implementation, the transient interference area uses a short processing segment (20-50ms), and the continuous interference area uses a long processing segment (100-200ms). For example, when the overlap of blinking and power frequency interference is detected, the blinking area is divided into a 25ms processing segment, and the power frequency area remains a 100ms processing segment. The hierarchical processing segment information finally formed includes the association relationship between segments, processing parameter configuration and intra-segment feature description.

[0053] Using the hierarchical processing segment information, the method constructs a signal channel network. First, the basic processing channel is established, including independent channels and collaborative channels. The independent channel processes the signal of a single electrode, and the channel parameters are optimized according to the electrode position characteristics: the bandwidth of the frontal electrode channel is 0.5-45Hz, and the bandwidth of the occipital electrode channel is 0.5-60Hz. The collaborative channel processes the associated signals of multiple electrodes to achieve spatial filtering and common mode suppression. Secondly, the cascade processing unit is configured, including an adaptive notch filter, a baseline drift corrector, and a feature enhancer. The method dynamically adjusts the working state of each processing unit according to the characteristic parameters of the processing segment. For example, when the user wears headphones to listen to music while working, the method detects the superposition of power frequency and audio equipment interference. At this time, the adaptive notch filter group with adjustable center frequency is enabled to perform directional suppression for interference in different frequency bands; when the user changes from sitting to standing, the baseline drifts due to the change in electrode position. The method automatically activates the drift corrector to achieve rapid stabilization of the signal baseline. The method also establishes a feedback mechanism based on the processing effect of the signal channel to optimize the processing performance by adjusting the channel parameters. The signal channel network finally established forms a complete signal processing system, which includes independent processing channels, series processing channels, parallel processing channels and feedback regulation mechanisms between channels.

[0054] Through the constructed signal channel network, the method performs progressive processing on the processing segment data. First, basic signal processing is completed in independent channels, including interference suppression, noise filtering and signal conditioning. When users use the device in an office environment, independent channels can effectively handle local interference from devices such as monitors and keyboards. Then, multi-level processing is implemented through series channels, and power frequency suppression, baseline correction and feature enhancement are performed in sequence. For example, when users use the device outdoors, the series channel can handle environmental electromagnetic interference, body motion artifacts and baseline drift in sequence. At the same time, spatial filtering is performed in parallel channels to remove common-mode interference components. For example, when users are walking, the parallel channel can effectively separate motion artifacts and target EEG signals by analyzing the signal correlation of multiple electrodes. The method also establishes a collaborative mechanism between channels. When a channel detects an abnormal signal, it can trigger the adaptive adjustment of other channel parameters. The processing results are integrated between channels, and the smoothing algorithm is used to process the boundaries between segments to ensure the continuity of the signal. The final output processed signal retains the key EEG features while improving the stability and anti-interference ability of the signal.

[0055] In some embodiments, the performing signal enhancement on the processed signal to obtain the enhanced signal includes: obtaining an enhancement position corresponding to the processed signal; obtaining an enhancement chain corresponding to the enhancement position; completing signal reconstruction according to the enhancement chain to obtain the enhanced signal.

[0056] The method receives the processed signal output from the step, which has been processed by multiple stages of independent channels, series channels and parallel channels. Based on the single-electrode processed signal output from the independent channel, the method analyzes the characteristic performance of each electrode position: calculates the local energy distribution of the signal after interference suppression, analyzes the spectral structure of the signal after noise filtering, and evaluates the time domain continuity of the signal conditioned data. Based on the processed signal output from the series channel, the method further determines the enhanced position: marks the prominent area of ​​the main frequency component in the signal after removing the power frequency interference, identifies the stable feature interval in the signal segment after baseline correction, and locates the key change point of the signal in the result of feature enhancement. Based on the spatial filtering results output from the parallel channel, the method integrates the spatial correlation of multiple electrodes: marks the electrode combination with significant correlation, determines the directional characteristics of signal propagation, and identifies the coordinated activity pattern of the local area. Through multi-stage signal analysis, the method finally determines the enhanced position. For example, when the energy distribution of the prefrontal beta band (13-30Hz) is detected to exceed 50% of the average value, and the activity shows stable spatial correlation in adjacent electrodes, the method marks it as a position to be enhanced. All determined enhanced positions form a positioning data set, which includes time position, frequency position, spatial position and characteristic description of each position.

[0057] Based on the obtained enhanced positioning data, the method constructs a signal enhancement chain structure. In the time dimension, the method sets forward enhancement units and backward enhancement units around each enhancement position to form a time sequence enhancement chain with a time span of 2 seconds before and after the enhancement position; in the frequency dimension, the cross-band enhancement association is established to construct a spectrum enhancement chain, covering the range of 5Hz above and below the target frequency band; in the spatial dimension, according to the anatomical relationship between electrodes, a spatial enhancement chain is set up, including the association enhancement of adjacent electrodes and symmetrical electrodes. Each enhancement chain node is configured with corresponding enhancement parameters, including gain coefficient (range 0.5-2.0), filter group (adjustable bandwidth) and phase corrector (±45° compensation range). The method establishes a synergistic mechanism between enhancement chains to achieve multi-dimensional linkage enhancement. For example, when the user is concentrating on work, the beta band signal enhancement in the prefrontal region is detected. The method not only activates the enhancement chain in this area, but also drives the synergistic enhancement of adjacent electrodes to form an overall enhancement effect in the local area. The constructed multi-dimensional enhancement chain network contains complete node parameter configuration and inter-chain triggering relationship.

[0058] Using the constructed enhancement chain network, the method carries out signal reconstruction processing. First, local signal reconstruction is performed on each enhancement chain: the temporal enhancement chain realizes the time domain enhancement of the signal through a sliding window with a step size of 50ms; the spectrum enhancement chain uses sub-band energy compensation with a resolution of 1Hz to complete the frequency domain reconstruction; the spatial enhancement chain uses a distance-weighted superposition strategy to reconstruct the spatial signal. Then the method performs multi-level signal synthesis. The first layer completes the signal synthesis in a single dimension, such as the smooth transition of adjacent windows in the time domain, the energy balance of adjacent sub-bands in the frequency domain, and the feature fusion of electrodes close in space; the second layer realizes cross-dimensional feature fusion and integrates the reconstruction results of the time domain, frequency domain and space domain. In practical applications, when the user performs memory tasks in a quiet environment, the method can effectively reconstruct the coordinated activity characteristics of theta waves and beta waves; when the user performs visual attention tasks, the method focuses on strengthening the spatiotemporal correlation patterns of the alpha waves in the occipital lobe and the beta waves in the prefrontal lobe. Through multi-layer synthesis processing, the method finally outputs a comprehensive reconstruction data set containing time domain reconstruction data, frequency domain enhancement features and spatial fusion results.

[0059] According to the above comprehensive reconstruction data set, the method generates the final enhanced signal through feature extraction and signal modulation. First, an enhancement evaluation index system is established: the signal integrity index calculates the retention rate of the characteristic waveform, which is required to be no less than 90%; the feature prominence index evaluates the energy enhancement effect of the target frequency band, and sets an enhancement range of 15-30%; the stability index monitors the degree of signal fluctuation, and limits the instantaneous fluctuation amplitude to within ±20%. Based on the evaluation results, the method adopts adaptive gain control technology to implement differentiated enhancement of signals in different frequency bands and different time periods. A 100ms gradient window is used at the feature mutation point to achieve smooth transition. For example, in educational application scenarios, when students are concentrating, the method highlights the β band activity through time-frequency joint modulation, while maintaining the basic characteristics of the θ band, providing a reliable basis for subsequent attention level evaluation. In clinical monitoring applications, the method can adjust the enhancement strategy according to the characteristic manifestations of different pathological states and highlight the characteristics of abnormal waveforms. The final enhanced signal maintains the physiological characteristics of the original EEG signal and achieves effective enhancement of key features.

[0060] Step S103, organize the enhanced signal in sections, obtain multiple processing sequences, determine resource allocation information according to the processing sequences, and determine the optimized channel according to the resource allocation information.

[0061] Specifically, the enhanced signal is organized in segments and the processing sequence is calibrated. Special processing sequences are configured for different types of time domain segments: for the steady-state segment (2-10s duration), a feature preservation sequence is used, which includes steps such as signal-to-noise ratio evaluation (signal-to-noise ratio threshold>15dB), frequency band energy calculation (using multi-resolution analysis) and feature extraction (wavelet packet decomposition); for the transition segment (feature gradient area), a smooth transition sequence is configured, and an adaptive window (50-200ms) is set for feature tracking; for the mutation segment (change amplitude>50% area), a fast response sequence is used, and a feature detection mechanism with high time resolution (10ms) is enabled. At the same time, based on the frequency domain decomposition results, a corresponding analysis sequence is configured for each frequency band (δ, θ, α, β, γ). In practical applications, such as meditation training scenarios, when the method detects stable θ wave enhancement, the steady-state sequence is started for continuous monitoring; when the user receives external interference, it quickly switches to the mutation sequence to capture the instantaneous changes of EEG features. In addition, based on the spatial grouping information, the method establishes corresponding processing sequence combinations for different brain regions. Each processing sequence contains specific execution steps, processing parameters, execution order, timing constraints and trigger conditions, forming a complete processing sequence configuration data set.

[0062] Based on the execution steps and processing parameters defined in the processing sequence configuration data set, the method performs resource allocation planning. First, the resource requirements of each sequence are evaluated: for feature retention sequences, a large computing load (about 10MB / s) is required for multi-resolution analysis; for smooth transition sequences, a medium storage space (about 2MB) is required for window data cache; for fast response sequences, low latency processing (<50ms) needs to be ensured. Then, the resource allocation strategy is determined according to the priority of the segments: for processing sequences of key feature segments (such as attention-related beta wave analysis sequences), more than 50% of computing resources are allocated; for processing sequences of transition features, about 30% of resources are allocated for feature tracking; and the remaining resources are used for basic signal processing. In a multi-person online conference scenario, the method can dynamically adjust sequence resources according to the status of the speaker: the processing sequences related to the speaker obtain priority resources to ensure the real-time feature extraction; the processing sequences of other participants use lower resource configurations. The method establishes a dynamic resource adjustment mechanism, which automatically triggers resource reallocation by monitoring the backlog of the processing queue (the threshold is set to 100ms), and finally forms a resource planning scheme that includes resource demand evaluation, priority allocation, and dynamic adjustment strategy for processing sequences.

[0063] Based on the resource planning scheme, the method selects channels. Channel selection is based on three dimensions of evaluation: the processing efficiency dimension examines the real-time performance of the channel, requiring feature extraction delay <200ms and throughput >5MB / s; the resource utilization dimension evaluates the resource utilization efficiency of the channel, and the CPU occupancy rate is expected to be maintained in the range of 60-80%; the feature preservation dimension examines the fidelity of the channel to the signal characteristics, and the energy error of the key frequency band should be <5%. According to the evaluation results of each dimension, the method selects channels according to the following standards: priority A channel (all three indicators meet the standard), priority B channel (two indicators meet the standard), and priority C channel (only one indicator meets the standard). In the application of concentration training, the method selects the A channel through evaluation: the prefrontal beta wave channel is equipped with a high-precision feature extractor, and the occipital alpha wave channel uses a fast-response processor. In the sleep monitoring scenario, the method evaluates and selects the B channel combination with multi-band analysis capabilities, which is used to simultaneously process the changing characteristics of delta waves, theta waves, and alpha waves. The final output channel selection result data set includes the performance score, priority classification, and application scenario adaptability of each channel.

[0064] Based on the above channel selection result data set, the method determines the final optimization channel. Each optimization channel is configured with specific working parameters: processing bandwidth range (adjustable from 0.5 to 100 Hz), calculation accuracy (optional 16 / 24 / 32 bits), and data cache size (standard 512KB, expandable to 2MB). The method dynamically allocates resources according to priority: Class A channels get the highest priority and can use 80% of the computing resources; Class B channels get medium priority and can use 50% of the computing resources; Class C channels are used as backup channels and are only enabled when resources are idle. In the actual operation process, the method continuously monitors the performance indicators of each channel. When the performance indicators of a channel decrease (such as the delay exceeds 200ms), the backup channel with the same priority is automatically enabled to replace it. After multi-dimensional evaluation and actual verification, the method finally determines a set of optimization channel combinations with stable performance and efficient resources, including the specific configuration parameters, working status and dynamic adjustment strategy of each priority channel.

[0065] Step S104, obtaining a signal inflection point corresponding to the optimized channel, determining a monitoring process corresponding to the signal tracking according to the signal inflection point, and obtaining a quality monitoring result corresponding to the processing sequence according to the monitoring process.

[0066] Specifically, the method receives the optimized channel combination determined in the step. Based on the working parameters of the optimized channel, including processing bandwidth (0.5-100Hz), calculation accuracy (16 / 24 / 32 bits) and cache size (512KB-2MB), the method deploys a signal tracking strategy: an adaptive window is used to scan the signal in real time, and the window length is adjusted in the range of 50-200ms according to the signal change rate; a multi-band analyzer configured with the optimized channel is used to synchronously track the energy change trend of each band, where the δ band uses a 200ms window, the θ and α bands use a 100ms window, and the β and γ bands use a 50ms window; based on the synergistic relationship between channels, the spatial correlation of multi-electrode signals is monitored, and the spatial analysis window is set to 500ms. In educational scenario applications, when students perform different types of learning tasks in class, the method uses the specific configuration of the optimized channel to track changes in EEG features: the reading comprehension task focuses on tracking the β wave activity of the prefrontal channel, and the problem thinking stage simultaneously tracks the changing characteristics of the θ wave. In the rehabilitation training scenario, the method dynamically adjusts the signal tracking parameters according to the degree of recovery of the patient's cognitive function: a larger tracking window (200ms) is used in the early rehabilitation stage to improve signal stability, and the window is gradually reduced (to 50ms) in the later training to improve response sensitivity. Through continuous tracking, the method identifies three types of key inflection points: amplitude inflection point (signal amplitude change exceeds the threshold of 30%), frequency inflection point (dominant frequency band conversion point) and spatial inflection point (electrode activity pattern change point). Finally, an inflection point data set containing time position, feature type, impact range and degree of change is formed.

[0067] In some embodiments, obtaining the signal inflection point corresponding to the optimized channel and determining the monitoring process corresponding to the signal tracking according to the signal inflection point include: performing signal tracking on the optimized channel to obtain the signal inflection point; the signal inflection point includes at least an amplitude inflection point, a frequency inflection point and a spatial inflection point; generating an inflection point data set based on multiple signal inflection points; and generating the monitoring process based on the inflection point data set.

[0068] Based on the obtained inflection point data set, the method constructs a multi-level monitoring process. According to the different types of inflection points identified, corresponding monitoring strategies are set: baseline drift detection (±5% threshold), mutation analysis (30% change threshold) and trend prediction (10s prediction window) are established for amplitude inflection points; energy migration tracking (energy transfer rate between frequency bands) and spectrum structure analysis (frequency band energy ratio monitoring) are established for frequency inflection points; propagation mode analysis (inter-electrode delay measurement) and regional linkage monitoring (correlation threshold 0.7) are established for spatial inflection points. In the medical monitoring scenario, the method configures a special monitoring process based on the characteristic inflection points under different physiological states: anesthesia depth monitoring focuses on the inflection point changes of δ waves and α waves, and sets a stricter detection threshold (amplitude change threshold is reduced to 20%); intraoperative awakening warning monitors the inflection point of β waves and adopts a shorter analysis window (2s) to improve the timeliness of warning. In the game interaction scenario, the method dynamically adjusts the monitoring parameters according to the complexity of the game task: the stability of α waves is mainly monitored in the simple task stage, and the monitoring frequency of β wave inflection points is strengthened in the difficult task stage. All monitoring processes form a multi-level monitoring system, including trigger mechanisms between processes, data transmission relationships and emergency response strategies.

[0069] Using the established multi-level monitoring system, the method conducts signal quality analysis. First, the signal stability is evaluated based on the amplitude monitoring process: the baseline drift rate (excellent <5%, good 5-10%, general 10-15%, to be optimized >15%), signal-to-noise ratio (excellent >20dB, good 15-20dB, general 10-15dB, to be optimized <10dB) and amplitude stability are calculated. Secondly, the frequency feature reliability is analyzed through the spectrum monitoring process: the uniformity of frequency band energy distribution, spectral peak clarity (spectral peak signal-to-noise ratio >6dB is excellent) and frequency consistency (frequency drift <0.5Hz is excellent) are evaluated. Finally, the spatial feature consistency is tested according to the spatial monitoring process: the inter-electrode correlation matrix, propagation delay diagram and regional activity synchronization index are calculated. In long-term cognitive training, the method evaluates the training effect by analyzing the quality indicators of different time periods: the quality stability of the β band is analyzed in the morning training stage, and the fatigue-related θ wave quality monitoring is strengthened in the afternoon training. In group collaborative tasks, the method can simultaneously evaluate the signal quality of multiple participants and establish a group-level quality benchmark. The quality analysis results include the scores of each dimension (0-100 points) and the comprehensive grade judgment, and also record the analysis of the causes of quality fluctuations.

[0070] According to the scoring and grade judgment of quality analysis, the method generates comprehensive monitoring results. The monitoring results contain three levels of information: the first level is basic state information, which records the real-time signal state of each optimization channel, the location of the key inflection point and the current quality level; the second level is quality diagnosis information, which includes time domain stability score, frequency domain feature score, spatial consistency score and composition analysis of each score; the third level is trend analysis information, which describes the change trend of signal quality, the distribution law of inflection points and the characteristics of state transition. In the distance education scenario, the method generates a learning status report in real time: a brief status report is output every 5 minutes during the course to show the change of attention level; after the course, a complete learning quality analysis is generated, including the concentration curve and key time point annotations. In the application of brain function assessment, the method generates targeted quality reports according to the characteristics of different cognitive tasks: the memory task focuses on reporting the quality changes of theta waves and beta waves, and the executive function task focuses on analyzing the signal quality of the prefrontal region. The final monitoring results include both an accurate description of the current state and a dynamic analysis of the time dimension, providing comprehensive data support for subsequent signal processing and application adjustments.

[0071] Step S105, obtaining a rule chain corresponding to the quality monitoring result, and determining a processing solution and optimization parameters corresponding to the processing solution according to the rule chain.

[0072] Specifically, the monitoring results output by the method receiving step include basic state information, quality diagnosis information and trend analysis information. Based on the basic state information, rule extraction is performed: steady-state rules (standard processing rules are used when the signal-to-noise ratio is greater than 20dB) and fluctuation rules (enhanced processing rules are enabled when the signal fluctuation exceeds 15%) are established according to the signal state; conversion rules are formulated based on the inflection point distribution, such as triggering the smoothing processing rule when the inflection point density exceeds 2 per second; adjustment rules are extracted from the quality level change, and the compensation rule is activated when the level decreases. Rule mapping is performed based on the quality diagnosis information: the time domain score (80-100 points) is mapped to a low-intensity filtering rule, (60-80 points) is mapped to a medium-intensity filtering rule, and (below 60 points) is mapped to a high-intensity filtering rule; the frequency domain score is converted into a frequency band adjustment rule, such as an excellent score corresponding to an adjustment range of ±2Hz, and a good score corresponding to an adjustment range of ±4Hz; the spatial consistency score is mapped to the weight coefficient of the spatial filtering rule. Prediction rules are constructed based on trend analysis information: the quality continuous decline trend triggers the progressive enhancement rule, the frequent occurrence of inflection points triggers the smooth enhancement rule, and the state frequent conversion feature triggers the stable enhancement rule. In the educational scenario, when the student's attention level continuously decreases, the method extracts progressive enhancement processing rules based on the characteristics of β wave quality changes; in rehabilitation training, based on the recovery trend of the patient's EEG characteristics, adaptive adjustment rules are extracted. All extracted rules form a rule data set, which contains rule type, trigger condition, execution action and priority information.

[0073] In some embodiments, the quality monitoring results include basic status information, quality diagnosis information and trend analysis information; obtaining the rule chain corresponding to the quality monitoring results includes: extracting rules based on the basic status information; mapping rules based on the quality diagnosis information; constructing rule predictions based on the trend analysis information; forming a rule data set based on the results corresponding to the rule extraction, rule mapping and rule prediction; the rule data set includes the rule type, trigger condition, execution action and priority information corresponding to each rule; and constructing the rule chain based on the rule data set.

[0074] Based on the formed rule data set, the method constructs a rule chain structure. First, according to the logical order of signal processing, the rules are arranged into a linear chain structure of preprocessing rules, main processing rules and post-processing rules, where the preprocessing rules are responsible for signal conditioning, the main processing rules perform feature enhancement, and the post-processing rules complete the result optimization. Then, an association mechanism between rules is established: a main processing rule can trigger multiple preprocessing rules, such as feature enhancement rules triggering signal conditioning rules and noise suppression rules; multiple post-processing rules can work together on the output of a main processing rule.

[0075] Exemplarily, constructing the rule chain according to the rule data set includes: rearranging each rule in the rule data set according to a preset signal processing logic order to obtain a linear chain structure in which the arrangement order is pre-processing rules, main processing rules and post-processing rules; constructing a time association mechanism and conflict handling strategy corresponding to the linear chain structure to obtain the rule chain.

[0076] The method sets a rule conflict handling strategy: when multiple rules meet the trigger conditions at the same time, they are executed in order of priority; when there is a conflict in the rule execution results, a weighted fusion method is used. For example, in the sleep monitoring application, the method constructs a rule chain including delta wave extraction rules, artifact suppression rules and result optimization rules for real-time evaluation of sleep depth. In the motor imagery task, the rule chain includes μ rhythm detection rules, spatial filtering rules and feature enhancement rules. The final rule chain structure contains a complete rule sequence, rule association relationship and execution mechanism.

[0077] Using the constructed rule chain structure, the method formulates a specific processing plan. Based on the preprocessing rules, the basic processing unit is configured: the center frequency, bandwidth and order of the adaptive filter are set, the gain coefficient of the signal amplifier is adjusted (adjustable from 0.5 to 2.0), and the threshold range of the denoising module is configured. The core processing modules are deployed according to the main processing rules: the time domain processing module configures the smoothing window (adjustable from 20 to 200ms) and the mutation suppression threshold; the frequency domain processing module sets the frequency band energy adjustment range and step size; the spatial processing module determines the electrode weight and phase correction parameters. The optimization unit is configured according to the post-processing rules: the judgment criteria for result verification, the amplitude range of feature enhancement and the limit value of output adjustment are set. In the application of attention detection, when the rule chain determines that the beta wave feature needs to be enhanced, the method configures a bandpass filter with a frequency band of 13-30Hz, a signal gain of 1.5 times and a frequency band adjustment range of ±0.5Hz. The parameter configuration and processing flow of all processing units form a complete processing plan.

[0078] Using the determined processing scheme, the method performs signal optimization processing to generate an optimized parameter set. In the preprocessing stage, the method conditions the original signal according to the filter parameters and gain coefficients to achieve baseline drift correction and noise suppression; in the main processing stage, each optimization module is started: the time domain module performs signal smoothing, the frequency domain module performs energy adjustment, and the spatial module optimizes the electrode signal; in the post-processing stage, the signal is verified and fine-tuned to ensure that the optimization result meets the quality requirements. When the processing effect is found to be unsatisfactory, the method automatically adjusts the processing parameters and re-executes the optimization. In practical applications, such as when insufficient μ rhythm energy is detected in brain-computer interaction tasks, the method gradually increases the gain of the 8-13Hz frequency band until it reaches the target level; in multi-person collaborative scenarios, the method records the respective optimization configurations according to the signal characteristics of different users. After each signal optimization, an optimized parameter set is generated, and the configuration values ​​of core parameters such as filter parameters, gain coefficients, and processing thresholds are recorded.

[0079] Step S106, obtaining key nodes corresponding to the optimization parameters and scene modes corresponding to the key nodes, adjusting the optimization parameters based on the scene modes, and obtaining adaptation results corresponding to the optimization parameter adjustments.

[0080] Specifically, the method receives the optimized parameter set output by the step, which records the optimal configuration values ​​of the filter parameters (center frequency, bandwidth, order), gain coefficient (0.5-2.0) and processing threshold in the optimization process. Based on these optimized parameters, feature adaptation analysis is performed: first, the optimized filter configuration is used to evaluate the filtering effect through multi-resolution analysis, the energy distribution density of the signal in each frequency band is calculated, and the energy distribution is smoothed using the Gaussian kernel function to obtain the frequency band coverage curve; then, the optimized gain coefficient is used to scan and analyze the signal amplitude and phase characteristics in the range of 0.5-2.0 with a step size of 0.1 to evaluate the influence of the gain parameter on the signal quality; finally, according to the optimized processing threshold, sensitivity analysis is performed in its neighborhood range, including baseline drift threshold (±5μV range, 0.5μV step size), mutation detection threshold (10%-50% range, 5% step size) and correlation threshold (0.3-0.9 range, 0.1 step size). For example, when the user uses the headset for daily office work, the method evaluates the parameter adaptation characteristics under different working conditions based on the optimized parameters. Through feature adaptation analysis, the method identifies a group of key nodes: energy turning points of each frequency band, inflection points of gain response, sensitive points of threshold detection and characteristic points of spatial distribution, forming a node feature set including node position, influence range and adjustment limit.

[0081] Based on the obtained node feature set, the method uses the node position and influence range information to establish scene mode mapping. First, the scene feature vector is constructed according to the distribution characteristics of key nodes, including frequency component characteristics (main frequency band, frequency band ratio, spectrum structure), time domain characteristics (waveform morphology, amplitude statistics, stability) and spatial characteristics (channel correlation, spatial distribution, propagation characteristics). Then the feature mapping function is designed, and the deep neural network structure is adopted. The input layer receives the node feature vector, the hidden layer extracts and combines features through the multi-layer perceptron, and the output layer generates the scene category and parameter configuration. According to the combination mode of node features, multiple basic scene modes are defined: working mode (based on β wave node features, 13-30Hz frequency band optimization, gain 1.2-1.8), leisure mode (based on α wave node features, 8-13Hz enhancement, gain 1.0-1.5), and focus mode (based on β and γ band node features). At the same time, composite scene modes are established: audio and video interaction mode (combining β wave and auditory evoked potential nodes), game control mode (combining μ rhythm and β wave nodes), and reading and learning mode (fusion of α and β band nodes). For each scene mode, the method designs the mode activation conditions and exit mechanism based on node features and establishes the state machine for scene switching. The definitions, switching conditions and parameter mappings of all scene modes form a complete scene mode library.

[0082] The method performs parameter adjustment by using the established scene mode library combined with the preset mode definition and switching conditions. A three-level parameter adjustment strategy is configured for each mode in the scene mode library: macro adjustment is responsible for parameter reconstruction during scene switching, and an adjustment period of 100-500ms is set according to the mode switching conditions; meso adjustment processes parameter optimization within the scene, and an update period of 10-50ms is adopted based on the mode characteristics; micro adjustment realizes real-time parameter fine-tuning, and the response period is controlled within 1-5ms according to the mode definition. The method adopts fuzzy control algorithm to design parameter adjuster, calculates deviation based on the current signal state and the preset target parameters of the scene mode, and determines the adjustment step and direction. In the frequency domain parameter adjustment, an adaptive filter group is used to dynamically adjust the center frequency and bandwidth of the filter according to the spectral characteristics of the scene mode. The proportional-integral-differential (PID) control strategy is adopted for the adjustment of gain parameters to achieve smooth transition of parameters. The spatial parameter adjustment is based on the independent component analysis (ICA) results to optimize the weight coefficient of the spatial filter. For example, when the user uses headphones to switch tasks in an office environment, the method adjusts the parameters in real time according to the preset configuration of the scene mode to ensure the continuity of signal processing. All parameter adjustment strategies and execution processes form a complete parameter configuration solution.

[0083] According to the adjusted parameter configuration scheme and the adjustment record of each configuration parameter, the method generates the final adaptation result. First, a multi-dimensional verification system is established according to the parameter configuration scheme: frequency domain verification analyzes the filtering effect, tests the passband ripple (<0.5dB), stopband attenuation (>40dB) and phase characteristics; time domain verification evaluates the signal quality, including signal-to-noise ratio (>20dB), baseline stability (drift <5μV) and transient response; spatial verification checks the correlation between channels and signal independence. Then an online evaluation is performed: a sliding window (256 points) is used to analyze the processed signal in real time and calculate the stability index of the characteristic parameters. The method also establishes a parameter self-correction mechanism. When the processing effect is detected to be unsatisfactory, the parameter configuration is automatically optimized through feedback adjustment. In the actual application of headphones, such as when the user performs computer interactive control, the method continuously evaluates the recognition accuracy of the control command and dynamically optimizes the parameter configuration; in multimedia scenarios, the method adjusts the signal processing parameters in real time according to the user's attention level and interaction needs. Through multiple rounds of iterative optimization and actual verification, the method outputs a complete adaptation data structure containing the final parameter configuration, adjustment records and verification results.

[0084] Step S107, determining a characteristic region according to the adaptation result, converting the characteristic region into a quality expression, generating an optimized signal according to the quality expression, and completing the quality optimization of the original signal.

[0085] Specifically, the adaptation data structure output by the receiving step contains the final parameter configuration, adjustment records and verification results. Based on the parameter configuration information, the frequency processing range (δ: 0.5-4Hz, θ: 4-8Hz, α: 8-13Hz, β: 13-30Hz, γ: >30Hz), amplitude processing range (baseline drift ±5μV, effective signal 0.5-50μV) and time processing window (20-200ms) of the signal are determined. Combined with the adjustment records, the method analyzes the parameter change law: marking the parameter stable area (signal segment with a change rate of <5%) and the parameter fluctuation area (dynamically adjusted transition segment), and calculating the adjustment frequency and amplitude distribution of each area. According to the verification results, the signal quality area is divided into: high-quality area (signal-to-noise ratio> 20dB, passband ripple <0.5dB), transition area (signal-to-noise ratio 15-20dB, passband ripple 0.5-1dB) and area to be optimized (signal-to-noise ratio <15dB, passband ripple>1dB). In daily office scenarios, when users wear headphones to process documents, the method identifies the dominant area of ​​the beta band (13-30Hz) and monitors the stability of the prefrontal electrodes. In the music appreciation scenario, the method focuses on the characteristic area of ​​the alpha band (8-13Hz) and performs synchronous analysis in combination with the headphone audio input. In the human-computer interaction scenario, the method dynamically tracks the changing characteristics of the μ rhythm (8-13Hz) and establishes a precise control mapping. Through multi-dimensional analysis, a regional description set is formed that includes regional boundaries (time axis positioning), feature attributes (frequency features, amplitude features, spatial features) and quality indicators (signal-to-noise ratio, stability, consistency).

[0086] Based on the obtained regional description set, the method converts the feature region into quality expression. First, a hierarchical quality evaluation system is established: in the frequency dimension, the spectrum quality indicators {band energy distribution (weight 0.4), spectrum peak clarity (weight 0.3), spectrum stability (weight 0.3)} are designed; in the amplitude dimension, the amplitude quality indicators {signal dynamic range (weight 0.35), baseline stability (weight 0.35), amplitude stability (weight 0.3)} are constructed; in the time dimension, the timing quality indicators {signal continuity (weight 0.4), mutation characteristics (weight 0.3), trend change (weight 0.3)} are formulated. Then, the regional characteristics are quantitatively evaluated: the spectrum quality uses wavelet analysis to calculate the energy concentration, the amplitude quality evaluates the signal distribution through statistical characteristics, and the timing quality determines the signal stability based on autocorrelation analysis. In practical applications, when users control smart devices through headphones, the method evaluates the spectral purity (>90% is excellent) and temporal stability (fluctuation <10% is excellent) of the control signal in real time; in video viewing scenarios, the method evaluates the frequency band energy proportion of α waves (>40% indicates relaxation) and the instantaneous change of β waves (mutation <20% is stable); during voice interaction, the method simultaneously evaluates the quality of EEG signals and voice signals, and establishes a multimodal quality evaluation mechanism. All evaluation indicators form a standardized quality expression, including the score of each dimension (0-100 points) and the comprehensive weight distribution.

[0087] Using the generated quality expression data, the method carried out the final signal optimization. The reconstruction strategy was configured according to the quality score: the excellent area (90-100 points) used fidelity reconstruction, using the least squares filter to maintain the original characteristics; the good area (75-89 points) used enhanced reconstruction, and the signal quality was improved by Wiener filtering; the qualified area (60-74 points) performed compensation reconstruction, and the Kalman filter was used to improve the signal characteristics. The method established an adaptive parameter control mechanism: the excellent area maintained the original parameter configuration, and the processing amplitude was suppressed within the range of ±5%; the good area dynamically adjusted the parameters, and the processing amplitude could reach ±15%; the qualified area increased the adjustment intensity, and the processing amplitude was up to ±30%. In the signal reconstruction process, the method used a processing window of 100ms, the window overlap rate was set to 50%, and a smooth transition was performed through the Hanning window. The method configures different reconstruction parameters for different application scenarios: when processing online documents, high-precision reconstruction (fidelity>95%) is used for β-band signals; when enjoying music, mild reconstruction (distortion<10%) is used for α-band signals; when playing games, fast reconstruction (delay<50ms) is implemented for μ rhythm signals. The method also establishes a real-time monitoring mechanism for signal quality. When the quality of the reconstructed signal is lower than expected, the parameter optimization process is automatically triggered. During long-term use, the method regularly performs calibration of the reconstruction parameters through cumulative error analysis to ensure the sustainability of the optimization effect. The final output optimized signal has enhanced target characteristics, stable signal quality and precise time synchronization.

[0088] The method provided in this application has at least the following beneficial effects:

[0089] 1The present invention combines time domain, frequency domain and space features to evaluate signal quality by constructing a multi-dimensional quality evaluation system. First, the original signal is collected for spectrum marking and converted into signal strength for quality level evaluation; then interference features are extracted based on quality data, and a feature library is established for type matching; finally, a comprehensive evaluation of signal quality is achieved through signal tracking and quality monitoring of processing channels, which effectively improves the accuracy and reliability of signal quality evaluation.

[0090] 2 The present invention adopts a scenario-aware parameter optimization strategy, extracts rules based on monitoring results and constructs rule chains, determines key nodes through feature adaptation and maps them to scenario modes, and then dynamically adjusts processing parameters based on scenario modes. This adaptive parameter configuration mechanism enables the method to adjust the processing strategy in a timely manner according to the actual use environment and user status, significantly improving the adaptability and real-time performance of signal processing.

[0091] 3 The present invention proposes a signal enhancement method with hierarchical processing, and constructs a multi-level signal channel network by dividing and subdividing the signal processing interval. The segmented organization and channel selection strategy are adopted to realize the progressive processing from signal conditioning to feature enhancement. This hierarchical and progressive processing mechanism effectively improves the accuracy and stability of signal enhancement while ensuring signal continuity.

[0092] In order to perform the signal quality optimization method of the intelligent brain-controlled headset corresponding to the above method embodiment, so as to achieve the corresponding functions and technical effects. Figure 2 , Figure 2 The structure block diagram of a signal quality optimization device 200 provided in an embodiment of the present application is shown. For the convenience of description, only the parts related to the present embodiment are shown. The signal quality optimization device 200 provided in an embodiment of the present application includes:

[0093] The signal acquisition module 201 is used to acquire the original signal collected by the intelligent brain-controlled headset, and acquire multiple quality levels corresponding to the original signal and quality data corresponding to each quality level;

[0094] A fragment acquisition module 202 is used to acquire interference fragments from the quality data, generate interference features according to the interference fragments, acquire a processing interval corresponding to the interference features and a processing signal corresponding to the processing interval, perform signal enhancement on the processed signal, and acquire an enhanced signal;

[0095] A sequence acquisition module 203, configured to organize the enhanced signal in sections, acquire multiple processing sequences, determine resource allocation information according to the processing sequences, and determine an optimized channel according to the resource allocation information;

[0096] An inflection point acquisition module 204 is used to acquire a signal inflection point corresponding to the optimization channel, determine a monitoring process corresponding to the signal tracking according to the signal inflection point, and acquire a quality monitoring result corresponding to the processing sequence according to the monitoring process;

[0097] A parameter determination module 205 is used to obtain a rule chain corresponding to the quality monitoring result, and determine a processing scheme and an optimization parameter corresponding to the processing scheme according to the rule chain;

[0098] A result acquisition module 206, configured to acquire key nodes corresponding to the optimization parameters and scene modes corresponding to the key nodes, adjust the optimization parameters based on the scene modes, and acquire adaptation results corresponding to the optimization parameter adjustments;

[0099] The optimization completion module 207 is used to determine a characteristic region according to the adaptation result, convert the characteristic region into a quality expression, generate an optimization signal according to the quality expression, and complete the quality optimization of the original signal.

[0100] The above-mentioned signal quality optimization device 200 can implement the signal quality optimization method of the intelligent brain-controlled headset of the above-mentioned method embodiment. The options in the above-mentioned method embodiment are also applicable to this embodiment and will not be described in detail here. The rest of the contents of the embodiment of this application can refer to the contents of the above-mentioned method embodiment, and will not be repeated in this embodiment.

[0101] Figure 3 This is a schematic diagram of the structure of a computer device provided in one embodiment of the present application. Figure 3 As shown, the computer device 3 of this embodiment includes: at least one processor 30 ( Figure 3 Only one is shown in the figure), a memory 31 and a computer program 32 stored in the memory 31 and executable on the at least one processor 30, wherein the processor 30 implements the steps of any of the above method embodiments when executing the computer program 32.

[0102] The computer device 3 may be a computing device such as a smart phone, a tablet computer, a desktop computer, a cloud server, etc. The computer device may include but is not limited to a processor 30 and a memory 31. Those skilled in the art will understand that Figure 3 It is only an example of computer device 3 and does not constitute a limitation on computer device 3. It may include more or fewer components than shown in the figure, or a combination of certain components, or different components, for example, it may also include input and output devices, network access devices, etc.

[0103] The processor 30 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc.

[0104] In some embodiments, the memory 31 may be an internal storage unit of the computer device 3, such as a hard disk or memory of the computer device 3. In other embodiments, the memory 31 may also be an external storage device of the computer device 3, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. equipped on the computer device 3. Further, the memory 31 may also include both an internal storage unit and an external storage device of the computer device 3. The memory 31 is used to store an operating system, an application program, a boot loader (BootLoader), data, and other programs, such as the program code of the computer program. The memory 31 may also be used to temporarily store data that has been output or is to be output.

[0105] In addition, an embodiment of the present application further provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps in any of the above method embodiments are implemented.

[0106] An embodiment of the present application provides a computer program product. When the computer program product is run on a computer device, the computer device implements the steps in the above-mentioned method embodiments when executing the computer device.

[0107] In several embodiments provided in the present application, it is understood that each box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, a program segment or a part of a code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order from that marked in the accompanying drawings. For example, two consecutive boxes can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved.

[0108] If the functions are implemented in the form of software function modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium and includes several instructions for a computer device to execute all or part of the steps of the methods described in the various embodiments of the present application. The aforementioned storage media include: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, and other media that can store program codes.

[0109] The specific embodiments described above further describe the purpose, technical solutions and beneficial effects of the present application in detail. It should be understood that the above description is only a specific embodiment of the present application and is not intended to limit the scope of protection of the present application. It is particularly pointed out that for those skilled in the art, any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the scope of protection of the present application.

Claims

1. A method for optimizing the signal quality of an intelligent brain-controlled headset, characterized in that: include: Acquire an original signal collected by the intelligent brain-controlled headset, and acquire multiple quality levels corresponding to the original signal and quality data corresponding to each quality level, including: acquiring a spectrum mark of the original signal; acquiring a signal strength corresponding to the spectrum mark; calculating a quality level according to the signal strength, and acquiring a quality level and quality data corresponding to each spectrum mark; Obtaining an interference segment from the quality data, generating an interference feature according to the interference segment, obtaining a processing interval corresponding to the interference feature and a processing signal corresponding to the processing interval, performing signal enhancement on the processed signal, and obtaining an enhanced signal; Organizing the enhanced signal in sections to obtain multiple processing sequences, determining resource allocation information according to the processing sequences, and determining an optimized channel according to the resource allocation information; Obtaining a signal inflection point corresponding to the optimization channel, determining a monitoring process corresponding to signal tracking according to the signal inflection point, and obtaining a quality monitoring result corresponding to the processing sequence according to the monitoring process; Obtain a rule chain corresponding to the quality monitoring result, and determine a processing scheme and optimization parameters corresponding to the processing scheme according to the rule chain; the quality monitoring result includes basic state information, quality diagnosis information and trend analysis information; the obtaining of the rule chain corresponding to the quality monitoring result includes: extracting rules according to the basic state information; mapping rules according to the quality diagnosis information; constructing rule prediction according to the trend analysis information; forming a rule data set according to the results corresponding to the rule extraction, rule mapping and rule prediction; the rule data set includes the rule type, trigger condition, execution action and priority information corresponding to each rule; constructing the rule chain according to the rule data set, including: rearranging each rule in the rule data set according to a preset signal processing logic order, and obtaining a linear chain structure with the arrangement order of pre-processing rules, main processing rules and post-processing rules; constructing a time association mechanism and conflict handling strategy corresponding to the linear chain structure, and obtaining the rule chain; Acquire key nodes corresponding to the optimization parameters and scene modes corresponding to the key nodes, adjust the optimization parameters based on the scene modes, and acquire adaptation results corresponding to the optimization parameter adjustments; A characteristic region is determined according to the adaptation result, the characteristic region is converted into a quality expression, an optimization signal is generated according to the quality expression, and quality optimization of the original signal is completed.

2. The method according to claim 1, characterized in that Acquiring interference segments from the quality data and generating interference features according to the interference segments includes: Performing feature analysis on the quality data to obtain interference fragments corresponding to the quality data; Perform feature matching between the interference segment and a preset feature library to obtain the interference type corresponding to the interference segment; Obtain an interference feature corresponding to the interference type.

3. The method according to claim 1, characterized in that The step of performing signal enhancement on the processed signal to obtain an enhanced signal includes: Acquire an enhanced position corresponding to the processed signal; Obtaining an enhancement chain corresponding to the enhancement position; Signal reconstruction is completed according to the enhancement chain to obtain the enhanced signal.

4. The method according to claim 1, characterized in that: The obtaining of the signal inflection point corresponding to the optimization channel and determining the monitoring process corresponding to the signal tracking according to the signal inflection point include: Performing signal tracking on the optimized channel to obtain the signal inflection point; the signal inflection point at least includes an amplitude inflection point, a frequency inflection point and a space inflection point; Generate an inflection point data set according to a plurality of said signal inflection points; The monitoring process is generated according to the inflection point data set.

5. A signal quality optimization device, characterized in that: include: A signal acquisition module is used to acquire the original signal collected by the intelligent brain-controlled headset, and acquire multiple quality levels corresponding to the original signal and quality data corresponding to each quality level, including: acquiring a spectrum mark of the original signal; acquiring a signal strength corresponding to the spectrum mark; calculating the quality level according to the signal strength, and acquiring a quality level and quality data corresponding to each spectrum mark; a fragment acquisition module, configured to acquire interference fragments from the quality data, generate interference features according to the interference fragments, acquire a processing interval corresponding to the interference features and a processing signal corresponding to the processing interval, perform signal enhancement on the processing signal, and acquire an enhanced signal; A sequence acquisition module, used to organize the enhanced signal in sections, acquire multiple processing sequences, determine resource allocation information according to the processing sequences, and determine an optimized channel according to the resource allocation information; An inflection point acquisition module, used to acquire the signal inflection point corresponding to the optimization channel, determine the monitoring process corresponding to the signal tracking according to the signal inflection point, and acquire the quality monitoring result corresponding to the processing sequence according to the monitoring process; A parameter determination module, used to obtain a rule chain corresponding to the quality monitoring result, and determine a processing scheme and optimization parameters corresponding to the processing scheme according to the rule chain; the quality monitoring result includes basic state information, quality diagnosis information and trend analysis information; the obtaining of the rule chain corresponding to the quality monitoring result includes: extracting rules according to the basic state information; mapping rules according to the quality diagnosis information; constructing rule prediction according to the trend analysis information; forming a rule data set according to the results corresponding to the rule extraction, rule mapping and rule prediction; the rule data set includes the rule type, trigger condition, execution action and priority information corresponding to each rule; constructing the rule chain according to the rule data set, including: rearranging each rule in the rule data set according to a preset signal processing logic order, and obtaining a linear chain structure in which the arrangement order is pre-processing rule, main processing rule and post-processing rule; constructing a time association mechanism and conflict handling strategy corresponding to the linear chain structure, and obtaining the rule chain; A result acquisition module, used to acquire key nodes corresponding to the optimization parameters and scene modes corresponding to the key nodes, adjust the optimization parameters based on the scene modes, and acquire adaptation results corresponding to the optimization parameter adjustments; The optimization completion module is used to determine the characteristic region according to the adaptation result, convert the characteristic region into a quality expression, generate an optimization signal according to the quality expression, and complete the quality optimization of the original signal.

6. A computer device, characterized in that: The method comprises a memory and a processor; the memory is used to store a computer program; the processor is used to execute the computer program and implement the method according to any one of claims 1 to 4 when executing the computer program.

7. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, which, when executed by a processor, enables the processor to implement the method according to any one of claims 1 to 4.

Images