Cortical contact adaptive impedance control method and system for electroencephalogram acquisition

Abstract

The invention discloses a cortical contact self-adaptive impedance control method and system for electroencephalogram acquisition, the system comprises an electrode, a pressure sensor, an impedance measurement module, an air bag, an air pump, an electromagnetic valve and a main control unit, when a patient has micro-motion conditions such as cough, swallowing and breathing, an emergency compensation mode is triggered, and electrode displacement is rapidly compensated about 5 ms; an electrode is divided into N CAN domains, a transfer function of electrode contact impedance and pressure is established, regional pressure equalization and hierarchical regulation are realized by adopting an ADMM algorithm, and the pressure equalization CV value of the electrode is improved.

Description

Technical Field

[0001] This invention relates to the field of medical instrument technology, and in particular to a cortical contact adaptive impedance control method and system for electroencephalogram (EEG) acquisition. Background Technology

[0002] Current electrode adjustment structure of EEG caps:

[0003] 1. Manual adjustment: Electrode depth is physically adjusted via a knob or strap. This results in time-consuming adjustments, requiring operation on each electrode individually, with an average time exceeding 5 minutes.

[0004] 2. Use elastic materials: employ silicone or memory foam materials to passively adapt to the head shape;

[0005] 3. Pre-formed rigid bracket: Rigid electrode brackets sized according to head circumference;

[0006] None of the above structures can guarantee the uniformity of pressure distribution, with pressure differences between electrodes >40% (pressure uniformity value CV value >30%).

[0007] Furthermore, when the patient makes slight head movements (such as breathing, swallowing, or coughing), the system response delay is >500ms, the electrodes are prone to detach from contact, and the electrode displacement cannot be adjusted in a timely and rapid manner to restore contact. Summary of the Invention

[0008] The purpose of this invention is to provide a cortical contact adaptive impedance control method and system for electroencephalogram (EEG) acquisition, so as to solve one or more of the above-mentioned technical problems.

[0009] To achieve this objective, the present invention adopts the following technical solution:

[0010] A cortical contact adaptive impedance control method for electroencephalogram (EEG) acquisition, comprising the following steps:

[0011] S1: Real-time data acquisition:

[0012] Set the target pressure value P of the airbag set Read the real-time pressure value P of each electrode in contact with the scalp. real[ i ] Calculate the pressure deviation value ΔP for each electrode. [i] Calculate the dynamic rate of change ΔP of the pressure at each electrode. t[i] ;

[0013] Read the impedance value Z of each electrode. [i] Input electrode spacing d, generate impedance matrix Z [i] / d;

[0014] S2: Dynamic Deviation Analysis

[0015] Calculate the pressure deviation matrix E p[i] ;

[0016] The initial pressure regulation amount is generated by combining the ADMM algorithm with the logical rule table;

[0017] Establish an impedance voltage drop model;

[0018] Generate dynamic impedance pressure regulation amount ΔP total[i] ΔP total[i] =ΔP [i] +ΔP diaphragm[i] ;

[0019] S3: Multi-objective optimization decision-making:

[0020] According to the dynamic impedance pressure adjustment amount ΔP total[i] Generate the penalty coefficient matrix ρ and the maximum pressure difference threshold P. max Then input it into the ADMM algorithm engine;

[0021] S4: Distributed pressure balancing

[0022] The ADMM algorithm engine divides the electrode into N CAN domains, performs step-by-step iteration and correction on the CAN domains, and generates the final pressure regulation amount P. final[ i ] ;

[0023] S5: Hierarchical execution control:

[0024] Based on the final pressure adjustment amount P final[ i ] The system generates a PWM signal for the solenoid valve and a current signal for the air pump, thereby controlling the opening degree of the solenoid valve and the speed of the air pump respectively, performing hierarchical control, and achieving dynamic compensation.

[0025] In some implementations, in step S1, when ΔP [i] When the pressure exceeds 50 Pa, the emergency compensation mode is triggered.

[0026] In some embodiments, step S1 further includes: calculating the gradient value of the pressure region.

[0027] In some implementations, in step S3:

[0028] Input ΔP total[i] Load the constraint parameter table, generate the penalty coefficient matrix ρ and the maximum pressure difference threshold P. max .

[0029] In some implementations, in step S4:

[0030] Iterate through each CAN field;

[0031] A weighted average of the stress values ​​of N CAN domains is used to achieve a globally consistent average stress value.

[0032] After iteration, Lagrange multiplier correction is performed to achieve local correction;

[0033] When the convergence condition is met, the final pressure regulation quantity P is generated. final[i] .

[0034] In some implementations, in step S5:

[0035] Based on the final pressure adjustment amount P final[ i ] Generate the PWM signal for the solenoid valve and map the weights of the PWM signal using the COG centroid method;

[0036] Based on the final pressure adjustment amount P final[ i ] Generate the current signal for the air pump.

[0037] In some implementations, in step S5:

[0038] The hierarchical control mode is as follows:

[0039] When P final[ i ] When the pressure is ≤10Pa, in fine-tuning mode, the solenoid valve opening is ±2%, and the response time is 15ms.

[0040] When 10 <P final[ i ] At ≤50Pa, in rapid adjustment mode, the air pump speed increases by 30% to 50%, and the response time is 40ms;

[0041] When P final[ i ] When the pressure is >50Pa, the emergency compensation mode is activated, with the air pump operating at full power and redundant valves opening, resulting in a response time of 5ms.

[0042] In some implementations, step S6 is also included;

[0043] S6: Real-time security monitoring and fault tolerance:

[0044] Input feedback status;

[0045] Calculate the variance of pressure σ in the sliding window 2 ;

[0046] When the variance σ 2 When the pressure is >50Pa, automatic compensation for adjacent areas is performed.

[0047] The beneficial effects of this invention are:

[0048] 1. When the patient experiences micro-movements such as coughing, swallowing, or breathing (pressure deviation > 50 Pa), the emergency compensation mode is triggered, and the electrode displacement is quickly compensated in about 5 ms;

[0049] 2. Divide the electrode into N CAN domains, establish the transfer function of electrode contact impedance and pressure, and use the ADMM algorithm to achieve regional pressure equalization, hierarchical control, and improve the pressure equalization value (CV value) of the electrode.

[0050] 3. Set up a monitoring and feedback mechanism to control lag and trigger a fault tolerance mechanism when pressure fluctuations occur. Attached Figure Description

[0051] Figure 1 This is one of the flowcharts for a cortical contact adaptive impedance control method for electroencephalogram (EEG) acquisition according to the present invention;

[0052] Figure 2 This is the second flowchart of a cortical contact adaptive impedance control method for electroencephalogram (EEG) acquisition according to the present invention.

[0053] Figure 3 This is a flowchart of the hierarchical execution mode of the present invention;

[0054] Figure 4 This is a flowchart illustrating the real-time security monitoring and fault tolerance mechanism of the present invention.

[0055] Figure 5 This is a graph showing the impedance-pressure relationship of the present invention.

[0056] Figure 6 This is a structural diagram of the EEG cap of the present invention;

[0057] Figure 7 This is a structural diagram of the electrical stimulation unit of the present invention;

[0058] Figure 8 This is an exploded view of the electrical stimulation unit of the present invention;

[0059] Figure 9 This is a cross-sectional view of the electrical stimulation unit of the present invention;

[0060] Figure 10 This is a connection structure diagram of the main control unit of the present invention. Detailed Implementation

[0061] The present invention will now be described in further detail with reference to the accompanying drawings.

[0062] First embodiment:

[0063] refer to Figures 1 to 4 A cortical contact adaptive impedance control method for electroencephalogram (EEG) acquisition, comprising the following steps:

[0064] S1: Real-time data acquisition:

[0065] S11: Set the target pressure value P set Read the real-time pressure value P of each electrode 120 in contact with the scalp. real[i] ; Calculate the pressure deviation value ΔP for each electrode at 120°. [i] ΔP [i] =P set -P real[i] ;

[0066] S12: Calculate the dynamic rate of change ΔP of the pressure at each electrode 120. t[i] :

[0067] S13: Read the impedance value Z of each electrode at 120°. [i] Generate impedance matrix Z [i] / d, where d: electrode spacing, i.e., the center distance between adjacent electrodes 120;

[0068] S14: Output data: Target pressure value P set Real-time pressure value P real[ i ] Impedance matrix Z [i] / d, electrode spacing d; the above data are transmitted to step S2;

[0069] Among them, the pressure deviation value ΔP [i] With dynamic rate of change ΔP t[i] Used for real-time detection of scalp micro-movements (such as coughing, swallowing, breathing, etc.):

[0070] Pressure deviation value ΔP [i] Used to trigger tiered compensation (fine-tuning / emergency mode).

[0071] Dynamic rate of change ΔP t[i] The rate of change of pressure deviation over time is used to calculate the time derivative of the difference between the target pressure and the actual pressure; it reflects the instantaneous pressure fluctuations caused by scalp micro-movements (such as coughing); it is used to quantify the instantaneous displacement amplitude and input into the dynamic model to calculate the compensation amount.

[0072] Both serve as core inputs to the dynamic deviation analysis step and can trigger a graded compensation mode.

[0073] When ΔP [i] When the pressure exceeds 50Pa, the emergency compensation mode is triggered. For example, if coughing, swallowing, shaking, or breathing occurs, causing electrode 120 to shift or loosen, the contact pressure will change accordingly. At this time, the emergency compensation mode is triggered to achieve rapid adjustment, compensate for the displacement, and restore normal contact of electrode 120.

[0074] Furthermore, step S1 also includes step S15: calculating the pressure region gradient value.

[0075]

[0076] The rate of change of pressure in the x-direction (lateral direction) reflects the pressure difference between the left and right sides of the electrode 120 array;

[0077] The rate of change of pressure in the y-direction (longitudinal direction) reflects the pressure difference between the front and rear sides of the electrode 120 array;

[0078] For example, in a 64-channel electrode 120 structure, the regional gradient calculation involves mapping the pressure data into an 8×8 grid matrix.

[0079] This achieves multi-dimensional sensing fusion: simultaneous acquisition of three parameters—pressure, impedance, and displacement—overcoming the limitations of traditional systems that rely solely on pressure feedback. Displacement compensation refers to dynamically adjusting the pressure of the airbag 4 via the air pump 3 to offset the relative displacement (>0.5mm) between the electrode 120 and the scalp caused by minor head movements (such as coughing, breathing, swallowing, etc.). When the preset value is exceeded, the emergency compensation mode can be triggered; or the pressure deviation ΔP can be monitored in real time. When ΔP>50Pa, the emergency compensation mode is triggered, and the electrode 120 contact is restored in about 5ms.

[0080] S2: Dynamic Deviation Analysis

[0081] S20: Input the data from step S14;

[0082] S21: Calculate the pressure deviation matrix: E p[i] =P set[i] -P real[i] (For example, when setting up a 64-channel electrode 120 structure, i = 1 to 64);

[0083] S22: Combining the ADMM algorithm and the logic rule table, generate the initial pressure adjustment amount for each electrode 120; based on the initial pressure adjustment amount, derive the corresponding action (speed, opening, etc.) trend of the air pump, solenoid valve, etc.

[0084] The logical rule table is as follows:

[0085] ΔP range Fuzzy set Initial pressure regulation -5kPa to -3kPa Large burden Rapid pressure relief from 0 to 20% -3kPa to 0kPa Negative small 20% to 40% 0 kPa to +3 kPa Just small 60 to 80% +3kPa to +5kPa Zhengda 80% to 100%

[0086] S23: Establish the impedance voltage drop model:

[0087] Calculate the dynamic impedance value:

[0088] Calculate the dynamic pressure drop value: (Establish a connection between dynamic impedance, air pump current, and scalp biomechanical properties;)

[0089] Generation: ΔP total[i] =ΔP [i] +ΔP diaphragm[i] ;

[0090] Among them, establishing real-time pressure values ​​(P) real[ i ] ) and contact resistance value (Z [i] Transfer function between ) Where K is the scalp elasticity coefficient. β is the fitting parameter (contact resistance when the real-time pressure value is 0). The mapping between scalp hardness grading and parameters is shown in the table below:

[0091]

[0092]

[0093] refer to Figure 5 This is an impedance-pressure relationship curve, which shows the relationship between different scalp hardness and real-time pressure values.

[0094] The effect of dynamic impedance: extremely soft scalp, Z diaphragm The value is smaller, so air pump 3 requires a lower current to achieve the same ΔP (to avoid excessive pressure).

[0095] Extremely hard scalp, Z diaphragm The value is relatively large, so air pump 3 requires a higher current to drive it and ensure rapid compensation.

[0096] Among them, based on transfer function By limiting Z diaphragm[i] With P real[ i ] The K value is obtained by measuring the pressure-resistance relationship through a stepped pressure test (0.1 kPa → 5.0 kPa). In this way, a personalized K value can be measured for different patients, thus achieving personalized pressure optimization.

[0097] The meanings of the symbols in the dynamic pressure drop formula are as follows:

[0098]

[0099] S3: Multi-objective optimization decision-making:

[0100] S31: Input ΔP total[i] Load the constraint parameter table, generate the penalty coefficient matrix ρ and the maximum pressure difference threshold P. max ;

[0101] The constraint parameter table is as follows:

[0102]

[0103] S32: Penalty coefficient matrix ρ and maximum pressure difference threshold P max Input into the ADMM algorithm engine (step S4);

[0104] S4: Distributed Pressure Balancing (ADMM Algorithm Engine)

[0105] S41: Divide the electrodes into N CAN domains; for example, when setting 64 channel electrodes, N=8, each region has 8 electrodes, forming 8 CAN domains;

[0106]

[0107] This represents the parameter or state variable associated with the i-th element at iteration step k+1;

[0108] ΔP [i] : Represents the pressure change of the i-th element;

[0109] Z K : Global or auxiliary variables at iteration step k;

[0110] The Lagrange multiplier or local adjustment term of the i-th element at iteration step k;

[0111] ρ: Penalty coefficient;

[0112] Therefore, local iterations are performed on each region.

[0113] S42: Weighted average of pressure values ​​across N CAN domains:

[0114]

[0115] Z k+1 : Represents a global variable at iteration step k+1;

[0116] For example, when setting up 64-channel electrodes 120, N=8, i=64; step-by-step iteration (average 8 convergences) is achieved, rapid pressure equalization is achieved, and a globally consistent average pressure is achieved.

[0117] S43: Lagrange multiplier correction:

[0118]

[0119] The Lagrange multiplier (dual variable) of the i-th subproblem at iteration step k+1 is used to measure the deviation between the local solution and the global objective.

[0120] in, Adjust the weights between local optimization and global consistency. If the pressure deviation in a certain region remains consistently large... The increase forces subsequent iterations to tilt towards this region, accelerating global equilibrium. As a global benchmark, it coordinates the pressure across domains, resolving the control lag problem of traditional systems. The average of the optimization results from the eight CAN domains is taken as the benchmark value Z. k+1 The difference between the pressure of each domain and the global benchmark is limited to a preset threshold (e.g., 30 Pa for extremely soft scalp). By using the mean benchmark and multiplier correction, the pressure of each domain is forced to converge toward global equilibrium, while retaining local fast response capability; thus, the weights that correct local optimization and global consistency are adjusted.

[0121] S44: Convergence judgment:

[0122] Convergence criteria: Reaching a preset value (default 50 times) or all electrodes meeting the 120 condition. At the convergence threshold (5 Pa), the final pressure regulation value P is generated. final[ i ] ;

[0123] S45: Output final pressure regulation value P final[ i ] (Target compensation value);

[0124] Thus, the final pressure regulation vector generated after ADMM iteration contains the target compensation value (unit: kPa) for 64 electrodes at 120. This vector is transmitted to the hierarchical execution control step, which drives the air pump 3 and solenoid valve 5 via PWM signals to achieve dynamic pressure compensation. This balances the contradiction between local optimization and global consistency, accelerates algorithm convergence, and reduces the number of iterations (average convergence in 8 iterations).

[0125] S5: Hierarchical execution control:

[0126] S51: Based on the final pressure adjustment amount P final[i] This generates PWM (Pulse Width Modulation) signals for solenoid valve 5. For example, each solenoid valve corresponds to 4 PWM signals: PWM1: intake positive pressure valve opening (controls the charging rate), PWM2: intake negative pressure valve opening (adjusts airflow stability), PWM3: exhaust pressure relief valve opening (controls the venting rate), and PWM4: redundant valve opening and closing (fully open in emergency compensation mode). Each electrode corresponds to a solenoid valve or solenoid valve group, and every 4 PWM signals control the solenoid valve or solenoid valve group of the corresponding electrode. For example, if 64 channels of electrodes are set, 256 PWM signals will be generated.

[0127] Weights of the PWM signal are mapped using the COG centroid method:

[0128]

[0129] PWM j : The weighted percentage mean of the i-th element;

[0130] W ji This represents the weighting coefficient of the j-th solenoid valve to the i-th electrode;

[0131] Therefore, the opening degree of the solenoid valve is adjusted according to the overall contribution after weight adjustment, thereby controlling the inflation and deflation speed of the airbag 4, and thus changing the contact pressure of the electrode 120.

[0132] S52: Based on the final pressure adjustment amount P final[ i ] Generate the current signal for air pump 3;

[0133] For example: extremely soft scalp, Z diaphragm The value is relatively low, so the air pump 3 requires a lower current to drive it (to avoid excessive pressure); extremely hard scalp, Z diaphragm The value is relatively large, so air pump 3 requires a higher current to ensure rapid compensation; the magnitude of the current signal depends on P. final[ i ] Generate graded current signals;

[0134] S53: The opening degree of solenoid valve 5 and the speed of air pump 3 are controlled by PWM signal and current signal respectively to perform hierarchical control and realize dynamic compensation; the PWM signal can be sent through CAN bus.

[0135] The specific pattern is as follows:

[0136] When P final[ i ] When the pressure is ≤10Pa, in fine-tuning mode, the solenoid valve opening is ±2% (precision priority), and the response time is about 15ms.

[0137] When 10 <P final[ i ] When the pressure is ≤50Pa, in the rapid adjustment mode, the speed of air pump 3 increases by 30% to 50% (balanced), and the response time is about 40ms.

[0138] When P final[ i ] When the pressure is >50Pa, in emergency compensation mode, the air pump 3 is at full power and the redundant valve is opened (speed priority), with a response time of about 5ms.

[0139] S53: Feedback: Balancing speed and energy consumption for different response modes.

[0140] S6: Real-time security monitoring and fault tolerance:

[0141] S61: Input feedback status; for example, feedback of real-time pressure value, impedance value, displacement and other parameters of each electrode 120;

[0142] S62: Calculate the sliding window pressure variance σ 2 :

[0143] Where N = 10-second window, the time span of the sliding window;

[0144] P t : Real-time pressure value at second t.

[0145] μ: Average pressure within the window,

[0146] Sliding window pressure variance σ 2 It refers to the average of the squares of the deviations of pressure values ​​from the average pressure value in pressure sample data within a certain time range, and is used to detect abnormal pressure fluctuations;

[0147] refer to Figure 4 When pressure fluctuates (pressure fluctuation > 15Pa), the fault tolerance mechanism is triggered, and the backup channel is switched, such as switching the redundant channel. When the main channel fails, it automatically switches to the backup channel such as the mirror channel or the safety channel; it can also perform automatic compensation for adjacent areas.

[0148] When the air pump times out, detects air leakage, or the sensor malfunctions (variance > 50Pa), the fault tolerance mechanism is triggered. The fault tolerance mechanism automatically compensates for the adjacent area when a single point of failure occurs. That is, when the physical displacement of the electrode (not signal abnormality) is caused by air leakage or a single point of failure, the airbags 4 of the adjacent electrodes 120 are triggered to inflate and deflate in tandem. The displacement of the fault point is compensated through mechanical linkage. For example, if the electrode 120 at position A1 fails, the airbags 4 of the adjacent electrodes 120 at positions A2 and A3 will expand to fill the contact gap.

[0149] When impedance changes abruptly, the fault tolerance mechanism is triggered, and the re-touch procedure is executed;

[0150] The re-touch procedure is used for automated recovery processes in case of impedance anomalies. For example, when |ΔZ|>5kΩ (impedance surge) or Z... diaphragm [i ] When the threshold range is exceeded, it is determined to be a contact failure. At this time, the re-contact procedure is executed, which can be divided into three steps:

[0151] 1: The airbag 4 of the faulty electrode 120 is completely depressurized (PWM3 = 100%);

[0152] 2: The air pump 3 inflates at full power (emergency compensation mode) to make the electrode 120 re-contact the scalp;

[0153] 3: Based on transfer function Dynamic calibration pressure. Recovery verification: The fault-tolerant state is exited when the pressure deviation value stabilizes at |ΔP|≤10Pa and the impedance fluctuation is <5%.

[0154] Optionally, in step S2, it can also be based on Reverse derivation of pressure compensation amount:

[0155] For example, if ΔZ = 0.5kΩ, k = 0.6, then ΔP total[i] =(0.5 / 0.6)×(0.5 / 0.6)≈0.69kPa.

[0156] Then, ΔP total[i] The ADMM algorithm engine in step S3 is input to generate 64 pressure adjustment commands for electrode 120 to compensate for electrode 120 displacement and adjust the contact pressure of electrode 120. The calculations are performed in parallel across 8 CAN domains to achieve global pressure coordination and balance.

[0157] Pressure-resistance mapping table:

[0158] Pressure (kPa) Impedance (kΩ) Gradient threshold action 0 to 1.2 5 to 8 15% / cm Regional coordinated regulation 1.2 to 3.0 3 to 5 12% / cm Single-point adjustment 3.0 to 5.0 1 to 3 18% / cm Pressure relief protection

[0159] refer to Figure 5 This is an impedance-pressure relationship curve, which shows the relationship between different scalp hardness and real-time pressure values.

[0160] The technical effects achieved by the method in this application are as follows:

[0161] direct effects chain reaction Deep Value Contact pressure CV < 8% Preparation time reduced by 82% Epilepsy localization accuracy increased by 40%. Impedance fluctuation <5% Conductive paste usage ↓90% Brain-computer interface bit error rate ↓10^3 Displacement compensation delay <100ms Shortened operation training cycle Establish a scalp biomechanical database

[0162] By establishing a pressure-impedance transfer function, dynamic compensation displacement of the electrode is achieved, impedance stability is improved, and contact impedance fluctuation is reduced from >30% in the traditional scheme to <5%. When the electrode displacement causes a change in contact pressure, the approximate change in contact impedance can be predicted in advance, so as to adjust the electrode position or apply appropriate pressure in time to keep the impedance within the ideal range and ensure the quality of signal acquisition.

[0163] Based on the transfer function and dynamic compensation, the electrode contact impedance can be monitored in real time. Once the impedance deviates from the normal value, the corresponding pressure is automatically calculated according to the transfer function to restore good contact between the electrode and the scalp, thus realizing automated and precise compensation of the contact state.

[0164] Stable electrode contact impedance helps reduce interference from external factors on EEG signals and improves the signal-to-noise ratio of EEG signals.

[0165] Through domain-based collaborative computing (8 CAN domains), the pressure distribution uniformity (pressure equilibrium value CV) was improved from 32% to 8% (indicating a 75% improvement in pressure distribution uniformity).

[0166] Among them, the pressure equilibrium value σ: Standard deviation of pressure distribution (reflecting the dispersion of pressure values), μ: Average pressure (overall pressure level).

[0167] Real-time performance guarantee: Calculation time <20ms, meets 200Hz control frequency, and solves the control lag of traditional systems.

[0168] Dynamic compensation capability: It can quickly compensate for scalp micro-movements (displacement > 0.5 mm) caused by coughing, swallowing, breathing, etc. within 5 ms.

[0169] Energy consumption optimization: Intelligent scheduling of air pump power reduces power consumption by more than 3W.

[0170] Sliding window variance detection (2-second window pressure fluctuation analysis).

[0171] Triple redundancy valve control: When the main channel fails, it automatically switches to a backup channel such as a mirror channel or a safety channel.

[0172] The air leakage fault detection rate is 99.7% (compared to 85% for traditional systems).

[0173] When pressure waves or single-point failures occur, adjacent areas automatically compensate to achieve a "distributed fault-tolerant architecture," ensuring continuous system operation.

[0174] Second embodiment:

[0175] refer to Figures 6 to 10 A cortical contact adaptive impedance control system for electroencephalogram (EEG) acquisition, comprising:

[0176] EEG cap 1 includes a soft fabric layer 11 and an electrical stimulation unit 12. Multiple electrical stimulation units 12 are assembled on the soft fabric layer 11 to form an array. The electrical stimulation unit 12 is provided with electrodes 120.

[0177] Pressure sensor 2 is used to detect the pressure value of each electrode 120 in contact with the scalp. The number of pressure sensors 2 corresponds one-to-one with the electrode 120, and the pressure sensors 2 are arranged in an array.

[0178] Impedance measurement module 6 is used to detect the impedance value of each electrode 120; it has multiple channels, and the number of channels corresponds one-to-one with the number of electrodes 120.

[0179] Airbag 4 is used to adjust the contact pressure between each electrode 120 and the scalp; the airbag 4 is connected to the air pump 3 via the solenoid valve 5; the number of airbags 4 corresponds one-to-one with the electrode 120; the airbag 4 adjusts the position of the electrode 120 by inflating and deflating, thereby adjusting the contact pressure between the electrode 120 and the scalp.

[0180] as well as

[0181] The main control unit 8 is used to execute the above-mentioned cortical contact adaptive impedance control method for EEG acquisition; the main control unit 8 has a single-chip microcomputer and other structures.

[0182] The main control unit 8 is connected to the pressure sensor 2 via an SPI line;

[0183] The main control unit 8 is connected to the air pump 3 and the solenoid valve 5 via a CAN line;

[0184] The main control unit 8 is connected to the redundancy control module 7 via a CAN bus. The redundancy control module 7 is used to perform redundancy control.

[0185] Furthermore, pressure sensor 2 is a MEMS pressure sensor 2, model: Honeywell ASDXRRX010PDAA5, range: 0 to 10 kPa, accuracy: ±0.1%; pressure sensor 2 is set with 64 sensors, divided into 8 areas, each area is connected through CAN lines to form 8 CAN domains.

[0186] Impedance measurement module 6 is implemented using a Keysight E4980AL precision LCR meter (accuracy: ±0.1kΩ) combined with a custom circuit. It acquires 64 channels of contact impedance in real time at a sampling frequency of 10kHz and eliminates drift through temperature compensation.

[0187] Pressure sensor 2, airbag 4 and solenoid valve 5 can all be located inside the electrical stimulation unit 12; pressure sensor 2 can be located below airbag 4, for example, pressure sensor 2 can be located between airbag 4 and electrode 120.

[0188] The airbag 4 is inflated and deflated by the air pump 3, and the solenoid valve 5 controls the airflow direction. The opening degree of the solenoid valve 5 can control the airflow speed and flow rate, thereby realizing the displacement compensation of the electrode 120 and the dynamic compensation of the contact pressure between the electrode 120 and the scalp.

[0189] The air pump 3 can be a diaphragm pump, and the size and speed of the airflow can be adjusted by controlling the rotation speed of the air pump 3.

[0190] Furthermore, the soft fabric layer 11 of the EEG cap 1 is composed of an outer coating fabric, a middle flexible circuit, and an inner antibacterial fabric; the soft fabric layer 11 has a certain degree of elasticity.

[0191] Outer coating fabric: High-elasticity polyurethane (TPU) coated fabric is used, which is tear-resistant and adaptable to rigidity. It is beneficial to fix the geometric position of the electrode 120 and airbag 4 array, maintain the shape of the cap, and prevent the airbag 4 from expanding and causing overall deformation. It is also beneficial to hide the air passage pipes and circuit wiring, and provide dust protection.

[0192] Middle layer flexible circuit: used for signal transmission, with circuit trace structure;

[0193] Inner antibacterial fabric: Made of nano-silver coated polyester fiber fabric, it inhibits bacterial growth; it directly contacts the scalp, providing a stable conductive interface, breathable and sweat-wicking, reducing the stuffiness of long-term wear; the hydrogel at the bottom of the electrical stimulation unit 12 can seep out electrolyte through micropores, permeate the inner antibacterial fabric, and dynamically maintain low impedance.

[0194] Further reference Figure 8 and Figure 9 The electrical stimulation unit 12 includes an electrode 120, a fixing cap 121, an upper fixing member 122 and a lower fixing member 123;

[0195] The upper fixing member 122 and the lower fixing member 123 clamp and fix the outer coating fabric in the middle by means of hot pressing or ultrasonic welding. The electrode 120 is inserted into the mounting hole of the EEG cap 1, forming a movable connection to the mounting hole. The fixing cap 121 is screwed to the upper fixing member 122 to fix the electrode 120 to the outer coating fabric. The air bladder 4 is placed inside or above the electrode 120, and the electrode 120 can move within a certain range along its axis. When inflated, the air bladder 4 expands and pushes the electrode 120 towards the scalp, so that the electrode 120 contacts the scalp; when deflated, the air bladder 4 contracts, and the electrode 120 returns to its original position under the action of the elastic soft fabric layer 11, etc.

[0196] The above description only discloses some embodiments of the present invention. For those skilled in the art, various modifications and improvements can be made without departing from the inventive concept of the present invention, and these all fall within the scope of protection of the invention.

Claims

1. A cortical contact adaptive impedance control method for electroencephalogram (EEG) acquisition, characterized in that, The method includes the following steps: S1: Real-time data acquisition: Set target pressure value P set Read the real-time pressure value P of each electrode in contact with the scalp. real[ i ] Calculate the pressure deviation value ΔP for each electrode. [i] Calculate the dynamic rate of change ΔP of the pressure at each electrode. t[i] ; Read the impedance value Z of each electrode. [i] Input electrode spacing d, generate impedance matrix Z [i] / d; S2: Dynamic Deviation Analysis Calculate the pressure deviation matrix E p[i] ; The initial pressure regulation amount is generated by combining the ADMM algorithm with the logical rule table; Establish an impedance voltage drop model: Calculate the dynamic impedance value: Calculate the dynamic pressure drop value: Generate dynamic impedance pressure regulation amount ΔP total[i] ΔP total[i] =ΔP [i] +ΔP diaphragm[i] ; S3: Multi-objective optimization decision-making: According to the dynamic impedance pressure adjustment amount ΔP total[i] Generate the penalty coefficient matrix ρ and the maximum pressure difference threshold P. max Then input it into the ADMM algorithm engine; S4: Distributed pressure balancing The ADMM algorithm engine divides the electrode into N CAN domains, performs step-by-step iteration and correction on the CAN domains, and generates the final pressure regulation amount P. final[ i ] ; S5: Hierarchical execution control: Based on the final pressure adjustment amount P final[ i ] The system generates a PWM signal for the solenoid valve and a current signal for the air pump, thereby controlling the opening degree of the solenoid valve and the speed of the air pump respectively, performing hierarchical control, and achieving dynamic compensation.

2. The cortical contact adaptive impedance control method for electroencephalogram (EEG) acquisition according to claim 1, characterized in that, In step S1, when ΔP [i] When the pressure exceeds 50 Pa, the emergency compensation mode is triggered.

3. The cortical contact adaptive impedance control method for electroencephalogram (EEG) acquisition according to claim 2, characterized in that, Step S1 also includes: calculating the gradient value of the pressure region.

4. The cortical contact adaptive impedance control method for electroencephalogram (EEG) acquisition according to claim 1, characterized in that, In step S3: Input dynamic impedance pressure regulation ΔP total[i] Load the constraint parameter table, generate the penalty coefficient matrix ρ and the maximum pressure difference threshold P. max .

5. The cortical contact adaptive impedance control method for electroencephalogram (EEG) acquisition according to claim 1, characterized in that, In step S4: Iterate through each CAN field; A weighted average of the stress values ​​of N CAN domains is used to achieve a globally consistent average stress value. After iteration, Lagrange multiplier correction is performed to achieve local correction; When the convergence condition is met, the final pressure regulation quantity P is generated. final[i] .

6. The cortical contact adaptive impedance control method for electroencephalogram (EEG) acquisition according to claim 1, characterized in that, In step S5: Based on the final pressure adjustment amount P final[ i ] Generate the PWM signal for the solenoid valve and map the weights of the PWM signal using the COG centroid method; Based on the final pressure adjustment amount P final[ i ] Generate the current signal for the air pump.

7. The cortical contact adaptive impedance control method for electroencephalogram (EEG) acquisition according to claim 1, characterized in that, In step S5: The hierarchical control mode is as follows: When P final[ i ] When the pressure is ≤10Pa, in fine-tuning mode, the solenoid valve opening is ±2%, and the response time is 15ms. When 10 <P final[ i ] At ≤50Pa, in rapid adjustment mode, the air pump speed increases by 30% to 50%, and the response time is 40ms; When P final[ i ] When the pressure is >50Pa, the emergency compensation mode is activated, with the air pump operating at full power and redundant valves opening, resulting in a response time of 5ms.

8. The cortical contact adaptive impedance control method for electroencephalogram (EEG) acquisition according to claim 1, characterized in that, It also includes step S6; S6: Real-time security monitoring and fault tolerance: Input feedback status; Calculate the variance of pressure σ in the sliding window 2 ; P t The real-time pressure value at second t; μ: Average pressure within the window, When the variance σ 2 When the pressure is >50Pa, automatic compensation for adjacent areas is performed.

9. A cortical contact adaptive impedance control system for electroencephalogram (EEG) acquisition, characterized in that, include: An EEG cap includes a soft fabric layer and electrical stimulation units, with multiple electrical stimulation units assembled in the soft fabric layer, and each electrical stimulation unit is equipped with electrodes; Airbags are used to adjust the contact pressure between each electrode and the scalp; the airbags are connected to an air pump via solenoid valves. Pressure sensor used to detect the pressure value when the electrode contacts the scalp; Impedance measurement module, used to detect the impedance value of each electrode; as well as The main control unit is used to execute the cortical contact adaptive impedance control method for electroencephalogram (EEG) acquisition as described in claim 1. The main control unit is connected to the pressure sensor. The main control unit is connected to the air pump and the solenoid valve; The main control unit is connected to the redundant control module.

Images