Biomedical electrode skin contact intelligent regulation and control method, device and equipment and medium

Abstract

The invention relates to a biomedical electrode skin contact intelligent regulation and control method, device and equipment and a medium, and the method comprises the steps: collecting skin humidity, thickness and elasticity data in real time through a multi-modal sensor, and constructing a dynamic skin characteristic description matrix; an initial electrode control scheme adaptive to the individual skin state is generated based on a classification algorithm and a preset rule base, and the problem that a traditional electrode is poor in adaptability to infants, old people and pathological skin is solved; by dynamically adjusting pressure and gel component parameters, collaborative optimization of contact impedance and signal quality is realized; in combination with sliding window analysis and an adaptive learning mechanism, poor contact or parameter drift is detected in real time, and a skin model and electrode configuration are iteratively updated; according to the method, the limitations of standardized design, dynamic response lag and parameter mismatch in the prior art are broken through, the signal stability and long-term wearing safety in a complex scene can be improved, and accurate and reliable technical support is provided for medical diagnosis and health monitoring.

Description

Technical Field

[0001] The present invention relates to the field of intelligent medical technology, and in particular to a method, device, equipment and medium for intelligent control of biomedical electrode skin contact. Background Art

[0002] In the fields of medical diagnosis, rehabilitation therapy, and wearable health monitoring, biomedical electrodes serve as the core carriers for the collection and transmission of human bioelectrical signals. The stability of their contact with the skin directly affects the signal quality and device reliability. Traditional electrode systems achieve signal acquisition through fixed material properties and contact pressure, relying on conductive gels or mechanical bonding to maintain interface impedance. This approach can meet basic needs in static scenarios. However, as a dynamic biological interface, the skin's properties such as humidity, elasticity, and thickness can vary significantly due to individual differences (such as age, pathological conditions) and physiological activities (such as exercise and perspiration), resulting in electrode contact failure or signal distortion, which in turn affects the accuracy and continuity of medical data.

[0003] Existing biomedical electrode skin contact control technology has three major limitations: First, standardized designs are difficult to adapt to individual skin characteristics, such as the fragility of infant skin, the laxity of elderly skin, and abnormal microcirculation in diabetic patients. Fixed pressure and material parameters can easily lead to poor contact or skin damage. Second, the lack of a dynamic response mechanism makes it impossible to sense and compensate for changes in skin condition in real time (such as a sudden increase in humidity caused by sweating during exercise), and signal quality significantly degrades over time. Third, there is a lack of coordinated optimization of pressure, material properties, and electrical parameters. Adjustment in a single dimension can easily lead to an imbalance in the coupling of multiple physical fields, resulting in fluctuations in the electrode-skin interface impedance. These shortcomings have severely restricted the application of biomedical electrodes in precision medicine and long-term health monitoring. Summary of the Invention

[0004] Based on this, the purpose of the present invention is to provide a biomedical electrode skin contact intelligent control method, device, equipment and medium that can achieve dynamic personalized adaptation.

[0005] The purpose of the present invention is achieved by the following scheme:

[0006] In a first aspect, the present invention provides a method for intelligently controlling biomedical electrode skin contact, comprising the following steps:

[0007] S1: Collecting humidity data, thickness data, and elasticity data of the user's skin surface based on the sensor array, and normalizing the humidity data, thickness data, and elasticity data to generate a skin characteristic description matrix;

[0008] S2: Based on the pre-established classification algorithm and rule base, the skin characteristic description matrix is classified and identified to generate the initial electrode control scheme;

[0009] S3: Adjusting the parameters of the electrode end based on the initial electrode control scheme, generating and sending the initialized electrode working configuration;

[0010] S4: Based on the detection data set containing signal quality feedback data and skin dynamic change data fed back by the electrode tip, the sliding window processing technology is analyzed and processed to generate the electrode tip judgment result, which is used to indicate whether the electrode tip has poor contact or parameter drift;

[0011] S5: Based on the nonlinear mapping formula and iterative optimization algorithm, the parameters of the judgment result and the initial electrode control scheme are adjusted and optimized to generate an updated electrode control scheme;

[0012] S6: Adjust the parameters of the electrode end based on the updated electrode control scheme, and generate and send the adjusted optimized electrode configuration.

[0013] In one embodiment, S2 of a biomedical electrode skin contact intelligent control method provided by the present invention specifically includes the following steps:

[0014] S21: extracting features from the skin characteristic description matrix to generate skin statistical features including the mean humidity, mean thickness, mean elasticity, humidity variance, and elasticity gradient change rate;

[0015] S22: Classify the skin statistical features based on the pre-trained random forest classification model to generate skin condition category labels;

[0016] S23: Matching skin condition category labels based on a preset rule library, and calling corresponding pressure adjustment coefficients and gel component adaptation strategies;

[0017] S24: Parameter calculation is performed on the preset default pressure value and the basic gel formula based on the pressure adjustment coefficient and the gel component adaptation strategy to generate an initial motor control scheme containing target pressure parameters and target gel parameters.

[0018] In one embodiment, S3 of a biomedical electrode skin contact intelligent control method provided by the present invention specifically includes the following steps:

[0019] S31: Based on the target pressure parameter in the initial electrode control scheme, the pressure adjustment parameter is calculated and processed to generate pressure parameter configuration data, where the pressure parameter configuration data is used to control the pressure adjustment of the electrode array at the electrode end;

[0020] S32: Based on the target gel parameters in the initial electrode control scheme, the gel component ratio is optimized to generate gel parameter configuration data, and the gel parameter configuration data is used to control the mixing and injection of the conductive gel in the electrode end;

[0021] S33: Based on the gel parameter configuration data and the pressure parameter configuration data, a decision is made on the electrode activation state, an initialized electrode working configuration is generated, and the electrode working configuration is sent to the electrode terminal.

[0022] In one embodiment, S4 of the biomedical electrode skin contact intelligent control method provided by the present invention specifically includes the following steps:

[0023] S41: collecting and preprocessing the signal quality feedback data and the skin dynamic change data fed back by the electrode end, and generating a detection data set including the signal quality feedback data and the skin dynamic change data and having a time series;

[0024] S42: Analyze and process the signal quality change trend in the detection data set based on the sliding window processing technology to generate a sliding window analysis result;

[0025] S43: Based on the sliding window analysis result, threshold judgment processing is performed on the contact impedance abnormality or the signal attenuation rate to generate a judgment result of the electrode end.

[0026] In one embodiment, S5 of the biomedical electrode skin contact intelligent control method provided by the present invention specifically includes the following steps:

[0027] S51: Based on the nonlinear dynamic mapping formula and the signal quality degradation range of the judgment result, the target pressure parameter and the target gel parameter in the initial electrode control scheme are adjusted to generate the optimized pressure parameter and the optimized gel parameter. The calculation formulas of the optimized pressure parameter and the optimized gel parameter are as follows:

[0028]

[0029] Among them, P' is the optimized pressure parameter, G' is the optimized gel parameter, P0 is the target pressure parameter, G0 is the target gel parameter, k p (t), k g (t) is the adaptive proportional coefficient based on the skin moisture change rate, ΔQ is the signal quality degradation, τ p , τ g is the time decay constant of pressure and gel parameters;

[0030] S52: Based on the multi-objective optimization algorithm, the optimized pressure parameters and the optimized gel parameters are jointly optimized to generate a joint optimized parameter set. The calculation formula of the joint optimized parameter set is:

[0031] L(P,G)=α*Z(P,G)+β*||P0-P'|| 2 +γ*||G0-G'|| 2

[0032] Where L(P, G) is the joint optimization parameter set, α, β, and γ are weight coefficients that are dynamically allocated according to the skin elastic gradient change rate, and Z(P, G) is the contact impedance of the optimized pressure parameters and the optimized gel parameters;

[0033] S53: Analyze and process the joint optimization parameter set to generate an updated electrode control scheme.

[0034] In one embodiment, after step S6 of the biomedical electrode skin contact intelligent control method provided by the present invention, the following steps are further included:

[0035] S61: Based on the real-time feedback data of the optimized electrode configuration, the skin characteristic description matrix is weighted fused and updated through an adaptive learning algorithm to generate a fused description matrix. The update formula of the fused description matrix is:

[0036] M new =λ*M+(1-λ)*LSTM(M,C opt )

[0037] Among them, M new is the fusion description matrix, M is the skin characteristic description matrix, C opt To optimize the electrode configuration, λ is the forgetting factor, and LSTM(·) is the long short-term memory network, which is used to predict the evolution trend of skin condition;

[0038] S62: Retrain the fusion description matrix based on the pre-trained random forest classification model to generate an optimized personalized skin feature description matrix.

[0039] In one embodiment, S1 of a biomedical electrode skin contact intelligent control method provided by the present invention specifically includes the following steps:

[0040] S11: Collecting humidity data on the user's skin surface based on a capacitive humidity sensor to generate a humidity vector;

[0041] S12: collecting skin thickness data of the user based on a high-frequency ultrasonic thickness sensor to generate a thickness vector;

[0042] S13: collecting user skin elasticity data based on a piezoresistive elasticity sensor to generate an elasticity vector;

[0043] S14: Normalize the humidity vector, thickness vector, and elasticity vector and merge them into a skin characteristic description matrix.

[0044] In a second aspect, the present invention provides a biomedical electrode skin contact intelligent control device, which is configured with the following modules:

[0045] A data acquisition and processing module is used to collect moisture data, thickness data, and elasticity data of the user's skin surface based on the sensor array, and normalize the moisture data, thickness data, and elasticity data to generate a skin characteristic description matrix;

[0046] A classification and identification module is used to classify and identify the skin characteristic description matrix based on a pre-established classification algorithm and rule base, and generate an initial electrode control plan;

[0047] An initial configuration generation module, configured to adjust the parameters of the electrode end based on the initial electrode control scheme, and generate and send an initialized electrode working configuration;

[0048] A data detection and analysis module is used to analyze and process the detection data set containing signal quality feedback data and skin dynamic change data based on the electrode tip feedback through sliding window processing technology to generate a judgment result of the electrode tip. The judgment result is used to indicate whether the electrode tip has poor contact or parameter drift;

[0049] The scheme mapping optimization module is used to adjust and optimize the parameters of the judgment results and the initial electrode control scheme based on the nonlinear mapping formula and iterative optimization algorithm to generate an updated electrode control scheme;

[0050] The optimized configuration generation module is used to adjust the parameters of the electrode end based on the updated electrode control scheme, and generate and send the adjusted optimized electrode configuration.

[0051] In a third aspect, the present application provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, any one of the above-mentioned biomedical electrode skin contact intelligent control methods is implemented.

[0052] In a fourth aspect, the present application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements any of the above-mentioned biomedical electrode skin contact intelligent control methods.

[0053] In summary, the present invention provides a method for intelligently controlling the contact of biomedical electrodes with skin. This method collects and analyzes skin characteristics such as moisture, thickness, and elasticity in real time, accurately identifies skin types using a classification algorithm and rule base, and generates an initial electrode control scheme based on the identification results. The method dynamically adjusts and optimizes the electrode control scheme by performing sliding window processing and judgment on the detection data fed back by the electrode tip, combined with a nonlinear mapping formula and an iterative optimization algorithm, thereby enabling intelligent control of electrode-skin contact. This method breaks through the limitations of standardized design, dynamic response lag, and parameter mismatch in the prior art, effectively improving the signal acquisition quality of biomedical electrodes under different skin types and dynamic environments, reducing problems such as poor contact and parameter drift, and providing reliable technical support for precision medicine and long-term health monitoring.

[0054] For better understanding and implementation, the present invention is described in detail below with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 A schematic flow chart of a method for intelligently controlling biomedical electrode skin contact provided in an embodiment of the present application;

[0056] Figure 2 A schematic diagram of a process for generating an initial electrode control scheme according to an embodiment of the present application;

[0057] Figure 3 A schematic structural diagram of a biomedical electrode skin contact intelligent control system provided in another embodiment of the present application. DETAILED DESCRIPTION

[0058] To facilitate understanding of the present invention, the present invention will be described more fully below with reference to the accompanying drawings. The drawings illustrate preferred embodiments of the present invention. However, the present invention may be implemented in many different forms and is not limited to the embodiments described herein. Rather, these embodiments are provided to provide a more thorough and comprehensive understanding of the disclosure.

[0059] Unless otherwise defined, all technical and scientific terms used herein have the same meanings as those commonly understood by those skilled in the art to which this invention pertains. The terms used in this specification are for the purpose of describing specific embodiments only and are not intended to limit the invention. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0060] In one embodiment, Figure 1As shown, a method for intelligently controlling skin contact of a biomedical electrode is provided. This embodiment uses the method applied to a terminal as an example for illustration. It is understandable that the method can also be applied to a server, or to a system including a terminal and a server, and implemented through interaction between the terminal and the server. In this embodiment, the method includes the following steps:

[0061] S1: Collect humidity data, thickness data, and elasticity data of the user's skin surface based on the sensor array, and normalize the humidity data, thickness data, and elasticity data to generate a skin characteristic description matrix.

[0062] Specifically, a high-density, highly sensitive flexible sensor array can be used, adhered to the surface of human skin. This flexible sensor array can sense the skin's physical properties, such as humidity, thickness, and elasticity, in real time. Humidity sensors can acquire humidity data by detecting changes in moisture content on the skin's surface based on capacitive or resistive principles. Thickness sensors utilize piezoelectric or optical principles to measure changes in skin thickness under different conditions. Elasticity sensors reflect the elastic properties of the skin by detecting the degree of deformation under external forces. The sensor array transmits the collected raw data to the system in real time through multiplexing technology and a high-speed transmission interface. Multiplexing technology can effectively reduce the amount of wiring, improve data acquisition efficiency, and reduce signal interference.

[0063] After receiving the collected data from the sensor array, the system normalizes the collected moisture, thickness, and elasticity data, mapping them to the same numerical range (e.g., 0-1) to eliminate differences between different physical dimensions and improve the accuracy and consistency of subsequent data processing. Preferably, the normalization formula can use methods such as linear normalization or Z-score normalization. The normalized data constitutes a skin characteristic description matrix, providing the basis for subsequent classification, identification, and control solution generation.

[0064] S2: Based on the pre-established classification algorithm and rule base, the skin characteristic description matrix is classified and identified to generate an initial electrode control plan.

[0065] Specifically, the system can pre-establish a machine learning-based classification algorithm, such as a support vector machine (SVM), decision tree, or neural network, to analyze the skin characteristic description matrix. Simultaneously, a rule library containing characteristic parameters of different skin types (such as infant skin, elderly skin, and diabetic skin) is constructed to provide a reference for classification and identification. In this embodiment, the system inputs the skin characteristic description matrix into the classification algorithm, matches and compares it with the characteristic parameters in the rule library, and determines the current skin type. For example, based on the skin's thickness, elasticity, and moisture, it identifies whether it is the fragile skin of an infant or the loose skin of the elderly. Based on the identified skin type, the system retrieves the corresponding initial electrode control scheme from a pre-set electrode control scheme library. This scheme includes electrode pressure parameters, material property parameters, and electrical parameters to adapt to the characteristics of different skin types and provide an initial configuration for subsequent electrode parameter adjustments.

[0066] S3: Adjust the parameters of the electrode end based on the initial electrode control scheme, generate and send the initialized electrode working configuration.

[0067] Specifically, the system adjusts the parameters of the electrode tip based on the initial electrode control scheme, generates an electrode operating configuration based on the adjusted electrode parameters, and transmits this to the electrode tip via a communication module, either wirelessly or wired, to enable it to begin operating according to the new configuration. Electrode tip parameter adjustments can be made by adjusting properties of the electrode material, such as its elastic modulus and thickness, to change the contact pressure between the electrode and the skin; adjusting the electrode's electrical conductivity and optimizing electrical parameters to improve the stability and quality of signal acquisition; and electrode operating configurations, including parameters such as the electrode's operating mode, sampling frequency, and signal amplification factor, to ensure the electrode can stably and efficiently acquire human bioelectrical signals.

[0068] Preferably, the communication protocol can adopt Bluetooth Low Energy (BLE) or ZigBee protocol (a low-power wireless communication protocol) to ensure the stability and low latency of data transmission. After receiving the configuration, the electrode end enters the working state according to the parameters of the motor working configuration.

[0069] S4: A detection dataset containing signal quality feedback data and skin dynamic change data based on electrode tip feedback is analyzed and processed using sliding window processing technology to generate a judgment result of the electrode tip. The judgment result is used to indicate whether the electrode tip has poor contact or parameter drift.

[0070] Specifically, after the electrode tip enters the operating state, the multi-dimensional sensor at the tip collects real-time signal quality feedback data from the motor and dynamic skin change data during operation, forming a detection data set and feeding it back to the control end. Signal quality feedback data includes indicators such as the signal-to-noise ratio and stability, while dynamic skin change data covers changes in moisture, thickness, and elasticity during physiological activities such as exercise and perspiration. The system uses sliding window processing technology to analyze and process the detection data set. Sliding window technology slides a fixed-length window on the timeline, performing statistical analysis and feature extraction on the data within the window. This technology can capture dynamic data trends in real time and effectively identify fluctuations in signal quality and sudden changes in skin condition.

[0071] Based on the data processed by the sliding window, the system generates a judgment result of the electrode end. The judgment result is expressed in the form of binary or quantized values, and the judgment result includes two states: poor contact and parameter drift. The judgment basis for poor contact is that the signal quality index is lower than the preset threshold or the skin dynamic change index exceeds the normal range; the judgment basis for parameter drift is that the electrode parameter deviates from the initial configuration value by more than the preset deviation range. Preferably, the judgment result is generated by an anomaly detection algorithm, and a single-class classification method based on support vector data description (SVDD) is used to identify situations where the signal quality is abnormal and the skin dynamic change exceeds the preset threshold. The judgment result is stored in a state identification vector, which is used to indicate whether the electrode end has poor contact (such as abnormal impedance increase) or parameter drift (such as operating frequency offset).

[0072] S5: Based on the nonlinear mapping formula and iterative optimization algorithm, the judgment results and the initial electrode control scheme are parameter adjusted and optimized to generate an updated electrode control scheme.

[0073] Specifically, the system establishes a nonlinear mapping formula to associate the judgment results of the electrode end with the parameters in the initial electrode control scheme. The nonlinear mapping formula can reflect the complex relationship between the electrode parameters and the judgment results, providing a theoretical basis for parameter adjustment. For example, by analyzing the nonlinear relationship between the judgment results and the electrode pressure, material properties and electrical parameters, the direction and amplitude of parameter adjustment are determined. At the same time, the system uses iterative optimization algorithms, such as particle swarm optimization (PSO), genetic algorithm (GA) or simulated annealing algorithm (SA), to adjust and optimize the parameters of the judgment results and the initial electrode control scheme. The iterative optimization algorithm continuously searches and iterates to find the optimal electrode parameter combination to minimize abnormal indicators in the judgment results, improve the contact quality between the electrode and the skin and the stability of signal acquisition.

[0074] Based on the results of the nonlinear mapping formula and iterative optimization algorithm, the system generates an updated electrode control scheme. The updated electrode control scheme includes optimized electrode pressure, material properties, and electrical parameters, which can better adapt to the dynamic changes of the skin and improve the performance and reliability of the electrode.

[0075] S6: Adjust the parameters of the electrode end based on the updated electrode control scheme, and generate and send the adjusted optimized electrode configuration.

[0076] Specifically, the system optimizes and adjusts the parameters of the electrode tip according to the parameter settings in the updated electrode control scheme. For example, it adjusts the properties of the electrode material, such as elasticity and conductivity, to adapt to the dynamic changes of the skin; and optimizes the pressure distribution of the electrode to ensure good contact between the electrode and the skin. After completing the parameter optimization, the system generates a new electrode working configuration based on the adjusted electrode parameters and sends it to the electrode tip via the communication interface, allowing it to continue operating according to the updated configuration. The updated electrode working configuration can further improve the quality and stability of signal acquisition, ensuring the reliability and accuracy of biomedical electrodes in dynamic environments.

[0077] In summary, the present invention provides a method for intelligently controlling the contact of biomedical electrodes with skin. This method collects and analyzes skin characteristics such as moisture, thickness, and elasticity in real time, accurately identifies skin types using a classification algorithm and rule base, and generates an initial electrode control scheme based on the identification results. The method dynamically adjusts and optimizes the electrode control scheme by performing sliding window processing and judgment on the detection data fed back by the electrode tip, combined with a nonlinear mapping formula and an iterative optimization algorithm, thereby enabling intelligent control of electrode-skin contact. This method breaks through the limitations of standardized design, dynamic response lag, and parameter mismatch in the prior art, effectively improving the signal acquisition quality of biomedical electrodes under different skin types and dynamic environments, reducing problems such as poor contact and parameter drift, and providing reliable technical support for precision medicine and long-term health monitoring.

[0078] In one embodiment, Figure 2 As shown, S2 of the biomedical electrode skin contact intelligent control method provided by the present invention specifically includes the following steps:

[0079] S21: Extract features from the skin characteristic description matrix to generate skin statistical features including the mean humidity, mean thickness, mean elasticity, humidity variance, and elasticity gradient change rate.

[0080] Specifically, feature extraction is performed on the skin characteristic description matrix to generate skin statistical features containing multiple skin parameter features. These skin parameter features can fully reflect the physical properties of the skin and its dynamic change trends. Specifically, the extracted skin parameter features include:

[0081] 1. Mean Humidity: Calculates the average value of the humidity data in the matrix, reflecting the average humidity level on the skin surface.

[0082] 2. Mean Thickness: Calculates the average thickness data in the matrix, reflecting the average thickness of the skin.

[0083] 3. Mean Elasticity: Calculates the average value of the elastic data in the matrix, reflecting the average elastic modulus of the skin.

[0084] 4. Variance of Humidity: Calculate the variance of humidity data to reflect the degree of discreteness of humidity changes.

[0085] 5. Gradient Change Rate of Elasticity: Calculates the gradient change rate of elasticity data, reflecting the rate of elasticity change.

[0086] The above statistical characteristics are calculated using mathematical formulas to ensure data accuracy and repeatability. The mean moisture, mean thickness, and mean elasticity are calculated using the following formulas:

[0087]

[0088] Among them, μ humidity 、μ thickness and μ elasticity are the mean moisture, thickness and elasticity of the skin, respectively, h i , t i 、e i are the i-th sample values of humidity, thickness and elasticity data respectively, N is the total number of samples, humidity variance Calculated by the following formula:

[0089]

[0090] The elastic gradient change rate is obtained by calculating the gradient of the elastic data and finding its change rate. The calculation formula is as follows:

[0091]

[0092] S22: Classify the skin statistical features based on the pre-trained random forest classification model to generate skin condition category labels.

[0093] Specifically, the system uses a pre-trained random forest classification model to classify the feature vector. The random forest model is based on the integrated learning of multiple decision trees, and each decision tree is trained by Bootstrap sampling and random feature selection. The model input is a feature vector, and the output is a skin state category label. Skin state categories include dry type, sensitive type, loose type and normal type, etc. Specifically, the model input is a skin statistical feature vector, and the output is a skin state category label. The classification process can adopt a majority voting mechanism to ensure the stability and accuracy of the classification results. In an embodiment of the present application, the skin state categories generated include but are not limited to dry type (Dry), moist type (Moist), loose type (Loose) and firm type (Firm), and the generated skin state category labels serve as the basis for subsequent parameter adjustment.

[0094] S23: Match the skin condition category labels based on the preset rule library, and call the corresponding pressure adjustment coefficient and gel component adaptation strategy.

[0095] Specifically, the preset rule base contains the mapping relationship between different skin condition category labels and electrode parameter adjustment strategies. The construction of the rule base is based on a large amount of experimental data and clinical experience to ensure that each skin condition category has a corresponding optimal electrode parameter configuration. During specific implementation, the system matches the skin condition category labels generated above in the rule base and calls the corresponding pressure adjustment coefficient and gel component adaptation strategy. For example: for dry skin (Dry), a higher pressure adjustment coefficient is called to ensure good contact between the electrode and the skin, and a gel component with high moisturizing properties is adapted; for moist skin (Moist), a moderate pressure adjustment coefficient is called to avoid electrode sliding due to high skin humidity, and a gel component with good adhesion is adapted; for loose skin (Loose), a lower pressure adjustment coefficient is called to avoid excessive pressure on the skin, and a gel component with good elasticity is adapted; for firm skin (Firm), a moderate pressure adjustment coefficient is called to ensure stable contact between the electrode and the skin; and a gel component with moderate viscosity is adapted.

[0096] S24: Parameter calculation is performed on the preset default pressure value and the basic gel formula based on the pressure adjustment coefficient and the gel component adaptation strategy to generate an initial motor control scheme containing target pressure parameters and target gel parameters.

[0097] Specifically, the preset default pressure value and basic gel formula are the initial configuration parameters of the electrode system. According to the pressure adjustment coefficient and gel composition adaptation strategy called in the above steps, the default pressure value and basic gel formula are calculated to generate the initial electrode control scheme. Specifically, the system adjusts the default pressure value according to the pressure adjustment coefficient and calculates the target pressure parameter. The calculation formula is as follows:

[0098] P target =P default *α

[0099] Among them, P target is the target pressure parameter, P default is the default pressure value, and α is the pressure adjustment coefficient. Simultaneously, the system adjusts the base gel formula based on the gel component adaptation strategy to generate target gel parameters. Gel parameters include viscosity, elasticity, and moisturizing properties. This adjustment method is based on a linear combination or nonlinear mapping of gel components to ensure that gel properties meet the needs of specific skin conditions.

[0100] The above-mentioned embodiment provides a method for intelligently controlling biomedical electrode skin contact, which, through feature extraction, random forest classification, rule base matching, and parameter calculation, enables dynamic identification of skin conditions and personalized adaptation of electrode parameters. This method can effectively improve the stability of electrode-skin contact, thereby reducing signal distortion and providing reliable technical support for precision medicine and long-term health monitoring. By adjusting electrode parameters in real time, this technology can adapt to different skin types and dynamically changing physiological states, significantly improving the signal acquisition quality and device reliability of biomedical electrodes.

[0101] In one embodiment, S3 of a biomedical electrode skin contact intelligent control method provided by the present invention specifically includes the following steps:

[0102] S31: Based on the target pressure parameters in the initial electrode control scheme, the pressure adjustment parameters are calculated and processed to generate pressure parameter configuration data, and the pressure parameter configuration data is used to control the pressure adjustment of the electrode array at the electrode end.

[0103] Specifically, the target pressure parameter is calculated by the system based on the skin condition category label and the pressure adjustment coefficient. In order to achieve the pressure adjustment of the electrode array at the electrode end, it is necessary to perform detailed calculation and processing on the pressure adjustment parameter to generate the pressure parameter configuration data. Specifically, the system calculates the target pressure parameter P according to the target pressure parameter P. target and the actual pressure value P at the electrode end actual , calculate the pressure regulation parameter, the pressure regulation parameter ΔP is calculated by the following formula:

[0104] ΔP=P target -P actual

[0105] Among them, the actual pressure value P at the electrode end actualThe actual pressure values between the electrode array and the skin are monitored and collected in real time by multiple pressure sensors at the electrode tips, and transmitted to the system in real time via a communication interface. The optimized pressure adjustment parameters are then integrated into pressure parameter configuration data and stored in the system configuration database. This pressure parameter configuration data includes parameters such as target pressure value, adjustment rate, and stabilization time.

[0106] Preferably, after the pressure regulation parameters are calculated, a PID control algorithm can be used to optimize the pressure regulation parameters to ensure the stability and speed of the pressure regulation process. The PID control algorithm dynamically adjusts the pressure output through a combination of proportional, integral, and differential terms to minimize pressure deviation.

[0107] S32: Based on the target gel parameters in the initial electrode control scheme, the gel component ratio is optimized to generate gel parameter configuration data, and the gel parameter configuration data is used to control the mixing and injection of the conductive gel in the electrode end.

[0108] Specifically, the system optimizes the gel component ratio based on the target gel parameters in the initial electrode control scheme. The optimization process for the gel component ratio is as follows: First, the system extracts the target gel parameters from the initial electrode control scheme, including conductivity, viscosity, and moisturizer content; then, based on the target gel parameters, the system optimizes the component ratio in the basic gel formula. The optimization process uses a linear programming method to ensure that the gel component ratio meets the target parameter requirements. The optimization formula is:

[0109]

[0110] Among them, w i is the weight of the i-th component, c i The system integrates the optimized gel component ratios into gel parameter configuration data and stores them in the system configuration database. The gel parameter configuration data includes parameters such as the ratio of each component, mixing order, and injection rate.

[0111] S33: Based on the gel parameter configuration data and the pressure parameter configuration data, a decision is made on the electrode activation state, an initialized electrode working configuration is generated, and the electrode working configuration is sent to the electrode terminal.

[0112] Specifically, the decision-making process is based on the gel parameter configuration data and the pressure parameter configuration data, combined with preset activation conditions (such as pressure threshold, gel injection completion flag, etc.), and determines the activation state of the electrode through fuzzy logic or decision tree algorithms. The generated initialization electrode working configuration contains the pressure parameters, gel parameters, and electrical parameters (such as operating frequency, impedance range, etc.) of the electrode array. The electrode working configuration file is generated by the embedded controller and sent to the electrode end via the wireless communication module, causing it to enter the initialization working state.

[0113] In one embodiment, S4 of the biomedical electrode skin contact intelligent control method provided by the present invention specifically includes the following steps:

[0114] S41: collecting and preprocessing the signal quality feedback data and the skin dynamic change data fed back by the electrode end, and generating a detection data set including the signal quality feedback data and the skin dynamic change data and having a time series.

[0115] Specifically, the system uses a high-precision sensor array to collect real-time signal quality feedback data from the electrode tip and skin dynamic change data. Signal quality feedback data, including signal-to-noise ratio (SNR), impedance fluctuation coefficient, and signal stability indicators, is acquired through an embedded signal processing module. Skin dynamic change data, including humidity change rate, elastic modulus change, and thickness change rate, is acquired through a skin characteristic sensor array. The collected data is filtered and denoised, and a Kalman filter algorithm can be used to eliminate random noise interference to ensure data accuracy and reliability. The preprocessed data is organized into a time series to generate a timestamped detection dataset for subsequent sliding window analysis and processing.

[0116] S42: Analyze and process the signal quality change trend in the detection data set based on the sliding window processing technology to generate a sliding window analysis result.

[0117] Specifically, the system uses sliding window processing technology to dynamically analyze the signal quality change trend in the detection data set. The length of the sliding window is dynamically adjusted according to the signal sampling rate and system response requirements to ensure the time domain continuity and statistical stability of the data. Within each sliding window, the system can extract the trend characteristics of signal quality changes through time series analysis algorithms (such as moving average method, exponential smoothing method). Trend characteristics include short-term change rate of signal quality, long-term change trend and fluctuation amplitude, etc. The sliding window analysis results contain trend feature vectors and corresponding timestamps.

[0118] S43: Based on the sliding window analysis result, threshold judgment processing is performed on the contact impedance abnormality or the signal attenuation rate to generate a judgment result of the electrode end.

[0119] Specifically, based on the sliding window analysis results, the system uses a preset threshold judgment model to perform real-time evaluation of contact impedance anomalies or signal attenuation rates. The threshold judgment model is determined based on a large amount of experimental data and includes a contact impedance anomaly threshold (such as an impedance mutation exceeding a preset percentage) and a signal attenuation rate threshold (such as a signal strength decrease rate exceeding a preset value). The system generates a judgment result for the electrode end by comparing the analysis results with the threshold. The judgment result includes an indication of poor contact (such as an impedance anomaly exceeding a threshold) and an indication of signal quality degradation (such as a signal attenuation rate exceeding a threshold), which is used for subsequent optimization and adjustment processing.

[0120] In one embodiment, S5 of the biomedical electrode skin contact intelligent control method provided by the present invention specifically includes the following steps:

[0121] S51: Based on the nonlinear dynamic mapping formula and the signal quality degradation range of the judgment result, the target pressure parameters and the target gel parameters in the initial electrode control scheme are adjusted to generate optimized pressure parameters and optimized gel parameters.

[0122] Specifically, the calculation formulas for optimizing pressure parameters and optimizing gel parameters are:

[0123]

[0124] ΔQ=Q initial -Q current

[0125]

[0126] Among them, P' is the optimized pressure parameter, G' is the optimized gel parameter, P0 is the target pressure parameter, G0 is the target gel parameter, k p (t), k g (t) is the adaptive proportional coefficient based on the skin moisture change rate, ΔQ is the signal quality degradation, τ p , τ g is the time decay constant of pressure and gel parameters, Q initial is the initial signal quality, Q current is the current signal quality, k p0 and k g0 is the initial proportional coefficient, η is the weight coefficient of humidity change rate, is the skin moisture change rate.

[0127] S52: Jointly optimizing the optimized pressure parameters and the optimized gel parameters based on a multi-objective optimization algorithm to generate a joint optimization parameter set.

[0128] Preferably, a genetic algorithm or a particle swarm optimization algorithm can be used to solve the joint optimization parameter set to ensure that the optimal solution is found while satisfying multiple objective constraints. The optimization objectives include minimizing the contact impedance, the pressure parameter adjustment range, and the gel parameter adjustment range. Specifically, the calculation formula for the joint optimization parameter set is:

[0129] L(P,G)=α*Z(P,G)+β*||P0-P'|| 2 +γ*||G0-G'|| 2

[0130] Where L(P, G) is the joint optimization parameter set, α, β, and γ are weight coefficients that are dynamically allocated according to the skin elastic gradient change rate, and Z(P, G) is the contact impedance of the optimized pressure parameters and the optimized gel parameters;

[0131] S53: Analyze and process the joint optimization parameter set to generate an updated electrode control scheme.

[0132] Specifically, the system analyzes the joint optimization parameter set L(P, G) to evaluate the effect of optimizing pressure and gel parameters on signal quality. The analysis metrics include contact impedance, signal stability, and skin adaptability. Based on the analysis results, the system uses a decision-making algorithm to determine whether further adjustment of the optimization parameters is necessary. This decision-making algorithm can employ a state machine model to determine whether update conditions are met based on the current optimization parameters and system performance indicators. If so, an updated electrode control scheme is generated. The updated electrode control scheme includes the optimized pressure parameters, gel parameters, operating frequency, and signal strength. After the update, the system stores the updated electrode control scheme in the system configuration database and sends it to the electrode tip via the communication module. Upon receiving the updated scheme, the electrode tip adjusts its operating state according to the new parameters to ensure contact stability between the electrode and the skin and signal quality.

[0133] The above-mentioned intelligent control method for biomedical electrode skin contact can achieve dynamic adjustment and joint optimization of electrode pressure parameters and gel parameters through nonlinear dynamic mapping formulas and multi-objective optimization algorithms; this method can effectively improve the stability of electrode-skin contact and the accuracy of signal acquisition, and reduce signal distortion and poor contact problems caused by dynamic changes in the skin; by adjusting and optimizing electrode parameters in real time, it can improve the adaptability and reliability of biomedical electrodes in different skin types and dynamic environments, providing strong technical support for precision medicine and long-term health monitoring. In addition, the dynamic response mechanism and multi-physical field collaborative optimization capabilities of this method can ensure the efficient operation of electrodes in complex physiological environments, further promoting the application and development of biomedical electrode technology in the field of intelligent medicine.

[0134] In one embodiment, after step S6 of the biomedical electrode skin contact intelligent control method provided by the present invention, the following steps are further included:

[0135] S61: Based on the real-time feedback data of the optimized electrode configuration, the skin characteristic description matrix is weighted fused and updated through an adaptive learning algorithm to generate a fused description matrix.

[0136] Specifically, the real-time feedback data for optimizing electrode configuration includes the actual contact pressure between the electrode and the skin, the distribution of the conductive gel, and signal quality indicators. This data is collected in real time by the sensor array and signal processing module at the electrode tip. The system uses an adaptive learning algorithm to perform a weighted fusion update of the skin characteristic description matrix to generate a fused description matrix. This algorithm dynamically adjusts weights based on real-time feedback data, ensuring that the fused description matrix accurately reflects the current skin condition.

[0137] Preferably, a long short-term memory network (LSTM) can be used to predict the evolution trend of skin condition. LSTM learns the historical skin characteristic description matrix M and optimizes the electrode configuration C. opt , predicting future skin condition changes, the LSTM input is the historical data sequence, and the output is the predicted skin condition evolution trend. Among them, the update formula of the fusion description matrix is:

[0138] M new =λ*M+(1-λ)*LSTM(M,C opt )

[0139] Among them, M new is the fusion description matrix, M is the skin characteristic description matrix, C opt To optimize the electrode configuration, λ is the forgetting factor, ranging from 0 to 1, which is used to control the weight of historical data and new prediction data. A larger λ value indicates a higher emphasis on historical data, while a smaller λ value indicates a higher emphasis on new prediction data. LSTM(·) is a long short-term memory network used to predict the evolution trend of skin condition; the generated fusion description matrix M new It contains historical skin characteristic data and predicted skin condition evolution trends, and can more accurately reflect the current skin condition and its future change trends.

[0140] S62: Retrain the fusion description matrix based on the pre-trained random forest classification model to generate an optimized personalized skin feature description matrix.

[0141] Specifically, random forest is an ensemble learning algorithm that constructs multiple decision trees and uses a majority voting mechanism for classification, resulting in high classification accuracy and robustness. In this embodiment, the system first feeds the fused description matrix as input into a pre-trained random forest classification model. This model evaluates the input data using multiple internal decision trees, each of which classifies the data based on its internal node splitting rules. During this evaluation process, the model calculates an importance score for each feature to determine which features contribute most to skin condition classification. These feature importance scores are used to guide subsequent model retraining. Next, the system retrains the random forest model using the new data from the fused description matrix. The goal of retraining is to better adapt the model to the new skin characteristic data and improve classification accuracy and robustness. During the retraining process, the system adjusts the decision tree splitting rules and model parameters to optimize model performance. This includes updating the decision tree structure, recalculating the splitting points, and adjusting hyperparameters such as the number and depth of trees in the random forest. Finally, the retrained model generates an optimized, personalized skin characteristic description matrix. This matrix not only contains the original skin characteristic data but also incorporates information updated through an adaptive learning algorithm, enabling it to more accurately reflect the current state and changing trends of the user's skin. The optimized description matrix is then used to generate electrode control solutions to ensure stable electrode-skin contact and signal quality. Through this continuous model update mechanism, the system can dynamically adapt to changes in skin characteristics and provide personalized electrode control solutions suitable for applications such as medical diagnosis, rehabilitation therapy, and wearable health monitoring.

[0142] The above-mentioned intelligent control method for biomedical electrode-skin contact uses an adaptive learning algorithm and a long-short-term memory network to dynamically update and optimize the skin property description matrix. By fusing historical data with real-time feedback data, the system can more accurately reflect the skin's current state and future trends through the fused description matrix. Furthermore, the system retrains the fused description matrix using a pre-trained random forest classification model to generate an optimized, personalized skin property description matrix, significantly improving the accuracy and adaptability of skin state classification. This method can effectively improve the stability of electrode-skin contact and the accuracy of signal acquisition, reducing signal distortion and poor contact caused by dynamic skin changes. Through real-time feedback and adaptive learning, this technology significantly improves the adaptability and reliability of biomedical electrodes across different skin types and dynamic environments, providing strong technical support for precision medicine and long-term health monitoring. Its dynamic response mechanism and multi-physics field collaborative optimization capabilities ensure the efficient operation of electrodes in complex physiological environments, further promoting the application and development of biomedical electrode technology in the field of intelligent medicine.

[0143] In one embodiment, S1 of a biomedical electrode skin contact intelligent control method provided by the present invention specifically includes the following steps:

[0144] S11: Collect humidity data on the user's skin surface based on the capacitive humidity sensor to generate a humidity vector.

[0145] Specifically, the system uses a capacitive humidity sensor to collect humidity data on the user's skin surface. This sensor indirectly reflects the moisture content of the skin surface by measuring the change in capacitance between two electrodes. The sensor undergoes initialization and calibration before operation, which includes zero-point and span calibration. After calibration, the sensor collects moisture on the skin surface in real time at a preset sampling frequency (e.g., 10 Hz). The humidity data at each sampling point is stored as a capacitance value, which is proportional to the moisture content of the skin surface. The collected capacitance value is converted to a digital signal using an analog-to-digital converter (ADC) and filtered to remove high-frequency noise and environmental interference. Finally, the filtered humidity data is organized into a time series to generate a humidity vector H. Each element of the vector represents the humidity value at a specific moment in time. The vector dimension is N × 1, where N is the number of sampling points.

[0146] S12: Collecting the user's skin thickness data based on a high-frequency ultrasonic thickness sensor to generate a thickness vector.

[0147] Specifically, the system uses a high-frequency ultrasonic thickness sensor to collect the user's skin thickness data. The high-frequency ultrasonic sensor transmits and receives ultrasonic waves, measures the reflection time difference of ultrasonic waves in the skin tissue, and thus calculates the skin thickness. The sensor is initialized and calibrated before operation to ensure measurement accuracy and stability. The calibration process includes zero-point calibration and time delay calibration. The sensor collects skin thickness in real time at a preset sampling frequency (such as 5Hz). The thickness data of each sampling point is calculated by measuring the reflection time difference of ultrasonic waves. The calculation formula is:

[0148]

[0149] Where Δt is the time difference between ultrasound reflections, and v is the propagation velocity of ultrasound in skin tissue. The collected thickness data is converted to a digital signal using an analog-to-digital converter (ADC) and filtered to remove noise and interference. Ultimately, the filtered thickness data is organized into a time series to generate a thickness vector T. Each element of the vector represents the skin thickness value at a specific moment, and the vector dimension is N × 1, where N is the number of sampling points.

[0150] S13: Collecting user skin elasticity data based on the piezoresistive elastic sensor to generate an elastic vector.

[0151] Specifically, the system uses a piezoresistive elasticity sensor to collect the user's skin elasticity data. The piezoresistive elasticity sensor indirectly reflects the elastic modulus of the skin by measuring the deformation of the skin when it is under pressure. The sensor is initialized and calibrated before operation to ensure measurement accuracy and stability. The calibration process includes zero-point calibration and sensitivity calibration. The sensor collects skin elasticity in real time at a preset sampling frequency (such as 8Hz). The elasticity data of each sampling point is obtained by measuring the resistance change of the sensor. The resistance change is proportional to the deformation of the skin. The collected resistance change data is converted into a digital signal by an analog-to-digital converter (ADC) and filtered to remove noise and interference. Finally, the filtered elasticity data is organized in time series to generate an elasticity vector E. Each element of the vector represents the skin elasticity value at a certain moment. The vector dimension is N×1, where N is the number of sampling points.

[0152] S14: Normalize the humidity vector, thickness vector, and elasticity vector and merge them into a skin characteristic description matrix.

[0153] Specifically, the system normalizes the humidity vector H, thickness vector T, and elasticity vector E and combines them into the skin characteristic description matrix M. Normalization is achieved by mapping the data in each vector to the 0, 1 interval, as shown in the following formula:

[0154]

[0155] Among them, x is the original data, x min and x max are the minimum and maximum values of the data, respectively, x norm The normalized data is represented by the moisture, thickness, and elasticity vectors. The normalized moisture, thickness, and elasticity vectors are merged column by column to generate the skin characteristic description matrix M. Each row of the matrix represents skin characteristic data at a specific moment, and each column represents the temporal variation of a characteristic. The matrix dimensions are N × 3, where N is the number of sampling points. Finally, the generated skin characteristic description matrix M is stored in the system database for subsequent feature extraction and classification.

[0156] Preferably, if Figure 3 As shown, the present invention provides a biomedical electrode skin contact intelligent control device 700, which is configured with the following modules:

[0157] The data acquisition and processing module 710 is used to collect moisture data, thickness data, and elasticity data of the user's skin surface based on the sensor array, and normalize the moisture data, thickness data, and elasticity data to generate a skin characteristic description matrix;

[0158] The classification and identification module 720 is used to perform type classification and identification processing on the skin characteristic description matrix based on a pre-established classification algorithm and rule base, and generate an initial electrode control plan;

[0159] An initial configuration generating module 730 is configured to adjust the parameters of the electrode tip based on the initial electrode control scheme, and to generate and send an initialized electrode working configuration;

[0160] Data detection and analysis module 740 is used to analyze and process the detection data set containing signal quality feedback data and skin dynamic change data fed back by the electrode tip using sliding window processing technology to generate a judgment result of the electrode tip. The judgment result is used to indicate whether the electrode tip has poor contact or parameter drift;

[0161] A scheme mapping optimization module 750 is used to adjust and optimize the parameters of the judgment result and the initial electrode control scheme based on a nonlinear mapping formula and an iterative optimization algorithm to generate an updated electrode control scheme;

[0162] The optimized configuration generation module 760 is used to adjust the parameters of the electrode end based on the updated electrode control scheme, and generate and send the adjusted optimized electrode configuration.

[0163] In summary, the biomedical electrode skin contact intelligent control device provided by the present invention realizes the following steps through the cooperation of various modules: real-time collection and analysis of characteristic data such as skin moisture, thickness and elasticity, accurate identification of skin type using classification algorithms and rule bases, and generation of an initial electrode control scheme based on the identification results; dynamic adjustment and optimization of the electrode control scheme by sliding window processing and judgment of the detection data fed back by the electrode end, combined with nonlinear mapping formulas and iterative optimization algorithms, to achieve intelligent control of electrode-skin contact. This device breaks through the limitations of standardized design, dynamic response lag and parameter mismatch in the existing technology, can effectively improve the signal acquisition quality of biomedical electrodes under different skin types and dynamic environments, reduce problems such as poor contact and parameter drift, and provide reliable technical support for precision medicine and long-term health monitoring.

[0164] Preferably, the data acquisition and processing module 710 is configured with the following units:

[0165] The humidity data acquisition unit 711 is used to collect humidity data of the user's skin surface based on the capacitive humidity sensor and generate a humidity vector;

[0166] a thickness data acquisition unit 712 for acquiring skin thickness data of the user based on a high-frequency ultrasonic thickness sensor and generating a thickness vector;

[0167] The elasticity data acquisition unit 713 is used to collect the user's skin elasticity data based on the piezoresistive elasticity sensor and generate an elasticity vector;

[0168] The matrix generating unit 714 is used to normalize the humidity vector, thickness vector and elasticity vector and combine them into a skin characteristic description matrix.

[0169] Preferably, the classification and identification module 720 is configured with the following units:

[0170] The feature extraction unit 721 is used to extract features from the skin characteristic description matrix to generate skin statistical features including moisture mean, thickness mean, elasticity mean, moisture variance, and elasticity gradient change rate;

[0171] A classification unit 722 is configured to classify the skin statistical features based on a pre-trained random forest classification model to generate skin condition category labels;

[0172] Strategy matching unit 723, used to match skin condition category labels based on a preset rule library and call corresponding pressure adjustment coefficient and gel composition adaptation strategy;

[0173] The scheme generating unit 724 is used to perform parameter calculation on the preset default pressure value and the basic gel formula based on the pressure adjustment coefficient and the gel component adaptation strategy, and generate an initial electrode control scheme containing target pressure parameters and target gel parameters.

[0174] Preferably, the initial configuration generation module 730 is configured with the following units:

[0175] A pressure configuration generating unit 731 is configured to calculate and process pressure adjustment parameters based on target pressure parameters in the initial electrode control scheme to generate pressure parameter configuration data, which is used to control pressure adjustment of the electrode array at the electrode end;

[0176] A gel configuration generating unit 732 is configured to optimize the gel component ratio based on the target gel parameters in the initial electrode control scheme and generate gel parameter configuration data, which is used to control the mixing and injection of the conductive gel in the electrode end;

[0177] The working configuration decision unit 733 is used to make a decision on the electrode activation state based on the gel parameter configuration data and the pressure parameter configuration data, generate an initialized electrode working configuration, and send the electrode working configuration to the electrode terminal.

[0178] Preferably, the data detection and analysis module 740 is configured with the following units:

[0179] Feedback collection and processing unit 741 is used to collect and pre-process the signal quality feedback data and skin dynamic change data fed back by the electrode end, and generate a detection data set containing the signal quality feedback data and skin dynamic change data and having a time series;

[0180] A sliding window analysis unit 742 is configured to analyze and process the signal quality change trend in the detection data set based on a sliding window processing technique to generate a sliding window analysis result;

[0181] The judgment result generating unit 743 is used to perform threshold judgment processing on the contact impedance anomaly or the signal attenuation rate based on the sliding window analysis result, and generate a judgment result of the electrode end.

[0182] Preferably, the solution mapping optimization module 750 is configured with the following units:

[0183] A parameter adjustment unit 751 is configured to adjust the target pressure parameter and the target gel parameter in the initial electrode control scheme based on the nonlinear dynamic mapping formula and the signal quality degradation amplitude of the judgment result, and generate an optimized pressure parameter and an optimized gel parameter;

[0184] A joint optimization unit 752 is configured to jointly optimize the optimized pressure parameters and the optimized gel parameters based on a multi-objective optimization algorithm to generate a joint optimization parameter set;

[0185] The scheme generating unit 753 is used to analyze and process the joint optimization parameter set to generate an updated electrode control scheme.

[0186] Preferably, the biomedical electrode skin contact intelligent control device provided by the present invention is further configured with the following units:

[0187] A matrix updating unit 761 is configured to perform weighted fusion updating on the skin characteristic description matrix using an adaptive learning algorithm based on the real-time feedback data of the optimized electrode configuration to generate a fused description matrix;

[0188] The model retraining unit 762 is used to retrain the fusion description matrix based on the pre-trained random forest classification model to generate an optimized personalized skin characteristic description matrix.

[0189] In one embodiment, the present application also provides a computer device including a memory and a processor, wherein the memory stores a computer program, and the processor implements the above-mentioned biomedical electrode skin contact intelligent control method when executing the computer program.

[0190] In one embodiment, the present application further provides a computer-readable storage medium having a computer program stored thereon, which implements the above-mentioned biomedical electrode skin contact intelligent control method when the computer program is executed by a processor.

[0191] In the description of this specification, the reference terms "one embodiment," "some embodiments," "example," "specific example," or "some examples" mean that the specific features, structures, materials, or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. Moreover, the specific features, structures, materials, or characteristics described may be combined in any appropriate manner in any one or more embodiments or examples. In addition, those skilled in the art may combine and integrate different embodiments or examples described in this specification, as well as features of different embodiments or examples, unless they are mutually inconsistent.

[0192] For the device embodiments, since they basically correspond to the method embodiments, the relevant parts can be referred to the partial description of the method embodiments. The device embodiments described above are merely illustrative, wherein the components described as separate parts may or may not be physically separated, and the parts displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the disclosed solution. A person of ordinary skill in the art can understand and implement it without expending creative work.

[0193] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any person skilled in the art can easily conceive of various modifications or substitutions within the technical scope disclosed in this application, and such modifications or substitutions should be included within the scope of protection of the present application. Therefore, the scope of protection of the present application should be based on the scope of protection of the claims.

Claims

1. A biomedical electrode skin contact intelligent control method, characterized in that: The following steps are involved: S1: collecting humidity data, thickness data, and elasticity data of the user's skin surface based on a sensor array, and normalizing the humidity data, the thickness data, and the elasticity data to generate a skin characteristic description matrix; S2: performing type classification and identification processing on the skin characteristic description matrix based on a pre-established classification algorithm and rule base to generate an initial electrode control scheme; S3: adjusting the parameters of the electrode end based on the initial electrode control scheme, generating and sending an initialized electrode working configuration; S4: Analyzing and processing the detection data set containing signal quality feedback data and skin dynamic change data fed back by the electrode tip using a sliding window processing technique to generate a judgment result of the electrode tip, the judgment result being used to indicate whether the electrode tip has poor contact or parameter drift; S5: adjusting and optimizing parameters of the judgment result and the initial electrode control scheme based on a nonlinear mapping formula and an iterative optimization algorithm to generate an updated electrode control scheme; S6: Adjusting the parameters of the electrode end based on the updated electrode control scheme, generating and sending the adjusted optimized electrode configuration.

2. The method according to claim 1, characterized in that The S2 includes: S21: performing feature extraction on the skin characteristic description matrix to generate skin statistical features including moisture mean, thickness mean, elasticity mean, moisture variance, and elasticity gradient change rate; S22: Classify the skin statistical features based on a pre-trained random forest classification model to generate skin condition category labels; S23: Matching the skin condition category label based on a preset rule library, and calling the corresponding pressure adjustment coefficient and gel component adaptation strategy; S24: Parameter calculation is performed on the preset default pressure value and the basic gel formula based on the pressure adjustment coefficient and the gel component adaptation strategy to generate an initial motor control scheme containing target pressure parameters and target gel parameters.

3. The method according to claim 2, characterized in that The S3 includes: S31: Based on the target pressure parameter in the initial electrode control scheme, calculating and processing the pressure adjustment parameter to generate pressure parameter configuration data, wherein the pressure parameter configuration data is used to control the pressure adjustment of the electrode array at the electrode end; S32: Optimizing the ratio of gel components based on the target gel parameters in the initial electrode control scheme to generate gel parameter configuration data, wherein the gel parameter configuration data is used to control mixing and injection of the conductive gel in the electrode end; S33: Based on the gel parameter configuration data and the pressure parameter configuration data, decision processing is performed on the electrode activation state, an initialized electrode working configuration is generated, and the electrode working configuration is sent to the electrode terminal.

4. The method according to claim 3, characterized in that The S4 includes: S41: collecting and preprocessing the signal quality feedback data and the skin dynamic change data fed back by the electrode end, and generating a detection data set including the signal quality feedback data and the skin dynamic change data and having a time series; S42: Analyze and process the signal quality change trend in the detection data set based on a sliding window processing technology to generate a sliding window analysis result; S43: Based on the sliding window analysis result, threshold judgment processing is performed on the contact impedance abnormality or the signal attenuation rate to generate a judgment result of the electrode end.

5. The method according to claim 4, characterized in that The S5 includes: S51: Based on the nonlinear dynamic mapping formula and the signal quality degradation amplitude of the judgment result, the target pressure parameter and the target gel parameter in the initial electrode control scheme are adjusted to generate the optimized pressure parameter and the optimized gel parameter. The calculation formulas of the optimized pressure parameter and the optimized gel parameter are: Among them, P' is the optimized pressure parameter, G' is the optimized gel parameter, P0 is the target pressure parameter, G0 is the target gel parameter, k p (t), k g (t) is the adaptive proportional coefficient based on the skin moisture change rate, ΔQ is the signal quality degradation, τ p , τ g is the time decay constant of pressure and gel parameters; S52: Jointly optimize the optimized pressure parameter and the optimized gel parameter based on a multi-objective optimization algorithm to generate a joint optimization parameter set. The calculation formula of the joint optimization parameter set is: L(P,G)=α*Z(P,G)+β*||P0-P'|| 2 +γ*||G0-G'|| 2 Where L(P, G) is the joint optimization parameter set, α, β, and γ are weight coefficients that are dynamically allocated according to the skin elastic gradient change rate, and Z(P, G) is the contact impedance of the optimized pressure parameters and the optimized gel parameters; S53: Analyze and process the joint optimization parameter set to generate an updated electrode control scheme.

6. The method according to claim 5, characterized in that After S6, the following is also included: S61: Based on the real-time feedback data of the optimized electrode configuration, the skin characteristic description matrix is weightedly fused and updated by an adaptive learning algorithm to generate a fused description matrix. The update formula of the fused description matrix is: M new =λ*M+(1-λ)*LSTM(M,C opt ) Among them, M new is the fusion description matrix, M is the skin characteristic description matrix, C opt To optimize the electrode configuration, λ is the forgetting factor, and LSTM(·) is the long short-term memory network, which is used to predict the evolution trend of skin condition; S62: Retraining the fusion description matrix based on the pre-trained random forest classification model to generate an optimized personalized skin characteristic description matrix.

7. The method according to any one of claims 1 to 6, characterized in that Said S1 comprises: S11: Collecting humidity data on the user's skin surface based on a capacitive humidity sensor to generate a humidity vector; S12: collecting skin thickness data of the user based on a high-frequency ultrasonic thickness sensor to generate a thickness vector; S13: collecting user skin elasticity data based on a piezoresistive elasticity sensor to generate an elasticity vector; S14: normalizing the humidity vector, the thickness vector, and the elasticity vector, and merging them into a skin characteristic description matrix.

8. A biomedical electrode skin contact intelligent control device, characterized in that: The device comprises: a data acquisition and processing module, configured to collect moisture data, thickness data, and elasticity data of the user's skin surface based on the sensor array, and perform normalization processing on the moisture data, thickness data, and elasticity data to generate a skin characteristic description matrix; A classification and identification module, configured to perform type classification and identification processing on the skin characteristic description matrix based on a pre-established classification algorithm and rule base, and generate an initial electrode control scheme; an initial configuration generating module, configured to adjust the parameters of the electrode end based on the initial electrode control scheme, and generate and send an initialized electrode working configuration; A data detection and analysis module is used to analyze and process the detection data set containing signal quality feedback data and skin dynamic change data fed back by the electrode end through sliding window processing technology to generate a judgment result of the electrode end, which is used to indicate whether the electrode end has poor contact or parameter drift; a scheme mapping optimization module, configured to adjust and optimize parameters of the judgment result and the initial electrode control scheme based on a nonlinear mapping formula and an iterative optimization algorithm, and generate an updated electrode control scheme; The optimized configuration generation module is used to adjust the parameters of the electrode end based on the updated electrode control scheme, and generate and send the adjusted optimized electrode configuration.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the method according to any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Images