A non-invasive bioelectric signal detection method based on flexible sensing technology

Abstract

The present application relates to the technical field of non-invasive bioelectricity detection, in particular to a non-invasive bioelectricity signal detection method based on flexible sensing technology, which collects the original myoelectricity signal of a target area through a flexible electrode, synchronously acquires the skin micro-deformation amplitude parameter captured by a high-frequency micro-strain sensor, the interface impedance change rate parameter generated by dynamic scanning of alternating current impedance spectrum, and the electrode contact pressure attenuation parameter quantified by a distributed thin film pressure sensor array; inputs the three types of parameters into a dynamic coupling analysis model, generates a real-time contact pressure compensation instruction through a pressure compensation algorithm, generates a signal amplification gain adjustment instruction through a gain adaptive algorithm, and generates a periodic pressure maintenance instruction through a pressure attenuation suppression algorithm, drives a micro pressure actuator and a programmable amplifier according to the instructions, aligns the timing to ensure that the adjustment is completed before the peak period of the myoelectricity signal, and outputs the myoelectricity signal with optimized signal-to-noise ratio, which takes into account the detection accuracy and wearing comfort.

Description

Technical Field

[0001] This invention relates to the field of non-invasive bioelectrical detection technology, and more specifically, to a non-invasive bioelectrical signal detection method based on flexible sensing technology. Background Technology

[0002] Non-invasive bioelectrical signal detection is an important technology, specifically applied to the daily monitoring of non-invasive electromyography (EMG) signals on the wrist. Its core principle is to achieve dual optimization of signal-to-noise ratio and wearing comfort by collaboratively processing the coupling effects of multiple factors in dynamic scenarios. This aligns with the core requirements of accuracy and comfort in daily office settings for EMG monitoring. In daily non-invasive EMG monitoring on the wrist, human respiration causes periodic micro-deformations in the wrist skin, sweat evaporation alters the impedance of the skin-electrode interface, and the contact pressure of the flexible electrodes decreases over time. Because these three factors form a dynamic coupling, micro-deformation causes momentary disconnection between the electrode and skin, leading to increased interface impedance and amplification of low-frequency noise. Pressure attenuation further exacerbates contact instability, creating a vicious cycle that results in a decrease in the signal-to-noise ratio of EMG signals in dynamic scenarios. To address this technical problem, we provide a non-invasive bioelectrical signal detection method based on flexible sensing technology. Summary of the Invention

[0003] The purpose of this invention is to provide a non-invasive bioelectrical signal detection method based on flexible sensing technology to solve the problems mentioned in the background art.

[0004] To achieve the above objectives, one objective of this invention is to provide a non-invasive bioelectrical signal detection method based on flexible sensing technology, comprising the following steps:

[0005] S1. Acquire raw electromyographic signals of the target area through flexible electrodes, and simultaneously initialize skin micro-deformation amplitude parameters, interface impedance change rate parameters, and electrode contact pressure attenuation parameters.

[0006] S2. The skin micro-deformation amplitude parameter is captured by a high-frequency micro-strain sensor, the interface impedance change rate parameter is generated by dynamic scanning of AC impedance spectrum, and the electrode contact pressure attenuation parameter is obtained by quantification by a distributed thin-film pressure sensor array.

[0007] S3. Input the three types of parameters—skin micro-deformation amplitude parameter, interface impedance change rate parameter, and electrode contact pressure attenuation parameter—into the dynamic coupling analysis model and perform the following processing, specifically including:

[0008] a. Based on the proportional relationship between the skin micro-deformation amplitude parameter and the predetermined deformation amplitude threshold, a real-time contact pressure compensation command is generated through a pressure compensation algorithm, wherein the pressure compensation algorithm establishes a nonlinear mapping between the unit increment of micro-deformation and the pressure compensation amount, so that the contact pressure increases synchronously when the micro-deformation increases.

[0009] b. Based on the deviation between the interface impedance change rate parameter and the predetermined impedance threshold, a signal amplification gain adjustment command is generated through a gain adaptive algorithm, wherein the gain adaptive algorithm establishes a positive feedback mechanism between the unit impedance increment and the gain up adjustment.

[0010] c. Based on the derivative relationship between the electrode contact pressure attenuation parameter and the time attenuation function, a periodic pressure maintenance command is generated through a pressure attenuation suppression algorithm to block the coupling negative loop of the skin micro-deformation amplitude parameter, the interface impedance change rate parameter, and the electrode contact pressure attenuation parameter.

[0011] S4. Drive the micro pressure actuator according to the real-time contact pressure compensation command, control the programmable amplifier according to the signal amplification gain adjustment command, and output the electromyographic signal with optimized signal-to-noise ratio.

[0012] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0013] This invention acquires raw electromyographic (EMG) signals using flexible electrodes, simultaneously capturing three parameters: the amplitude of skin micro-deformation, the rate of change of interfacial impedance, and the attenuation of electrode contact pressure. Through a dynamic coupling analysis model, these parameters are collaboratively processed to effectively block the negative coupling cycle. The pressure compensation algorithm, based on nonlinear mapping and step compensation, synchronously increases the contact pressure as micro-deformation increases, maintaining stable contact between the electrode and the skin. The gain adaptive algorithm, through impedance deviation tuning and sweat evaporation correction, suppresses gain oscillations, achieving precise matching between signal amplification and impedance changes. The pressure attenuation suppression algorithm generates periodic maintenance commands, dynamically adjusting the output energy to offset pressure attenuation. Simultaneously, the timing of pressure compensation and gain adjustment is strictly synchronized, ensuring optimization is completed before the peak of the EMG signal. This improves the signal-to-noise ratio and acquisition stability of non-invasive EMG signals in dynamic scenarios, balancing detection accuracy and wearing comfort, providing reliable technical support for daily EMG monitoring. Attached Figure Description

[0014] Figure 1 This is a flowchart illustrating the overall workflow of the present invention. Detailed Implementation

[0015] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0016] Please see Figure 1 As shown, this embodiment provides a non-invasive bioelectrical signal detection method based on flexible sensing technology, including the following steps:

[0017] S1. Acquire raw electromyographic signals of the target area using flexible electrodes, and simultaneously initialize skin micro-deformation amplitude parameters, interface impedance change rate parameters, and electrode contact pressure attenuation parameters; wherein, the skin micro-deformation amplitude parameters are captured by a high-frequency micro-strain sensor, the interface impedance change rate parameters are generated by dynamic scanning of AC impedance spectrum, and the electrode contact pressure attenuation parameters are quantized by a distributed thin-film pressure sensor array.

[0018] S2. Input the skin micro-deformation amplitude parameter, interface impedance change rate parameter, and electrode contact pressure attenuation parameter into the dynamic coupling analysis model, and perform the following processing, specifically including:

[0019] a. Based on the proportional relationship between the skin micro-deformation amplitude parameter and the predetermined deformation amplitude threshold, a real-time contact pressure compensation command is generated through a pressure compensation algorithm, wherein the pressure compensation algorithm establishes a nonlinear mapping between the unit increment of micro-deformation and the pressure compensation amount, so that the contact pressure increases synchronously when the micro-deformation increases.

[0020] b. Based on the deviation between the interface impedance change rate parameter and the predetermined impedance threshold, a signal amplification gain adjustment command is generated through a gain adaptive algorithm, wherein the gain adaptive algorithm establishes a positive feedback mechanism between the unit impedance increment and the gain up adjustment.

[0021] c. Based on the derivative relationship between the electrode contact pressure attenuation parameter and the time attenuation function, a periodic pressure maintenance command is generated through a pressure attenuation suppression algorithm to block the coupling negative loop of the skin micro-deformation amplitude parameter, the interface impedance change rate parameter, and the electrode contact pressure attenuation parameter.

[0022] S3. Drive the micro pressure actuator according to the real-time contact pressure compensation command, control the programmable amplifier according to the signal amplification gain adjustment command, and output the electromyographic signal with optimized signal-to-noise ratio.

[0023] Real-time contact pressure compensation commands are generated through a pressure compensation algorithm, specifically including:

[0024] A nonlinear converter is constructed with skin micro-deformation amplitude parameters as input and pressure compensation amount as output. When the real-time skin micro-deformation amplitude parameters captured by the high-frequency micro-strain sensor exceed a predetermined deformation amplitude threshold, a dynamic compensation mechanism based on the deformation and pressure coupling factor is triggered. The dynamic compensation mechanism converts the instantaneous increment of the real-time skin micro-deformation amplitude parameters into an air chamber volume adjustment command of the micro pressure actuator through the feedback channel of the embedded processor, so that the pressure compensation amount increases nonlinearly with the deformation amplitude.

[0025] The pressure compensation algorithm establishes a nonlinear mapping between the unit increment of micro-deformation and the pressure compensation amount, specifically including:

[0026] A deformation-pressure transformation matrix is ​​pre-set in the embedded processor. The deformation-pressure transformation matrix is ​​generated through machine learning training. Its row vectors correspond to the discretized intervals of micro-deformation unit increments, and the column vectors are mapped to the nonlinear increasing gradient of the pressure compensation amount. The transformation weight of each discretized interval is dynamically corrected by the coordinated change rate of the skin micro-deformation amplitude parameter and the contact pressure attenuation parameter.

[0027] This causes the contact pressure to increase synchronously with the increase of micro-deformation, specifically including:

[0028] When the deformation pressure conversion matrix detects a small deformation unit increment transition across intervals, it activates the gradient leap module for pressure compensation. The gradient leap module uses a multi-stage pneumatic valve controller of a micro pressure actuator to superimpose the differential value of the previous command on the current air chamber volume adjustment command, forming a step compensation for the contact pressure. At the same time, it uses real-time feedback data from a distributed thin-film pressure sensor array to calibrate the step amplitude in a closed loop.

[0029] Based on the deviation between the interface impedance change rate parameter and a predetermined impedance threshold, a signal amplification gain adjustment command is generated using a gain adaptive algorithm, specifically including:

[0030] An impedance deviation gain tuning function is established. The impedance deviation gain tuning function uses the algebraic difference between the interface impedance change rate parameter generated by AC impedance spectrum scanning and the predetermined impedance threshold as the independent variable. The output terminal is connected to the gain control register of the programmable amplifier. When the impedance deviation value exceeds the linear response range, the piecewise saturation mechanism of the gain adaptive algorithm is triggered, so that the gain adjustment amount switches from linear growth to logarithmic growth when the impedance deviation value increases.

[0031] The derivative of the sweat evaporation rate parameter is introduced as a correction coefficient for the impedance deviation gain tuning function. The second time derivative of the sweat evaporation rate parameter is calculated by synchronously capturing the skin surface curvature change rate parameter using a high-frequency micro-strain sensor. The second time derivative is then convolved with the impedance deviation value to generate a dynamic damping factor for gain tuning, which suppresses gain oscillations caused by abrupt changes in sweat evaporation.

[0032] The gain adaptive algorithm establishes a positive feedback mechanism between the unit impedance increment and the gain upscaling amount, specifically including:

[0033] An integrated gain-impedance coupler is used in the programmable amplifier. The gain-impedance coupler decomposes the unit impedance increment into the fundamental frequency component and the harmonic component. The fundamental frequency component is directly input to the gain control register to achieve linear up-adjustment, while the harmonic component is injected into the reference voltage terminal of the impedance detection circuit through a positive feedback loop, forming a self-reinforcing loop between the unit impedance increment and the gain up-adjustment.

[0034] Periodic pressure maintenance commands are generated using a pressure decay suppression algorithm, specifically including:

[0035] A pressure maintenance function is constructed with the first derivative of the electrode contact pressure decay parameter with respect to time as input. The output of the pressure maintenance function is connected to the pulse trigger module of the micro pressure actuator. When the derivative value of the pressure decay parameter exceeds the decay rate threshold of the time decay function, a pulse sequence generator based on pressure decay phase compensation is activated to generate a periodic pressure maintenance command with an exponential decay envelope.

[0036] During the pressure maintenance command execution cycle, the spectral characteristics of the skin micro-deformation amplitude parameter and the fluctuation variance of the interface impedance change rate parameter are simultaneously collected. Through the reverse feedback channel of the dynamic coupling analysis model, the spectral characteristics and fluctuation variance are input into the recursive filter of the pressure attenuation suppression algorithm to generate a negative loop blocking coefficient. The negative loop blocking coefficient adjusts the envelope attenuation slope of the pulse sequence generator so that the output energy of the periodic pressure maintenance command is inversely proportional to the strength of the coupled negative loop.

[0037] The real-time contact pressure compensation command is encoded into a duty cycle control signal for a pneumatic valve, which is then converted into a stepwise change in the air chamber volume by a piezoelectric ceramic actuator of a miniature pressure actuator. Simultaneously, the signal amplification gain adjustment command is compiled into a gain control word and loaded into the digital-to-analog converter of a programmable amplifier. The timing of the air chamber volume change and the loading of the gain control word is strictly aligned by the clock synchronization module of the dynamic coupling analysis model to ensure that pressure compensation and gain adjustment are completed before the peak period of the electromyographic signal action potential.

[0038] Further explanation is needed: After the dynamic coupling analysis model receives parameters such as the amplitude of skin micro-deformation, the rate of change of interfacial impedance, and the attenuation of electrode contact pressure, the first step, addressing the potential instability in electrode-skin contact caused by skin micro-deformation, involves generating a real-time contact pressure compensation command using a pressure compensation algorithm. This command must accurately match the deformation change pattern to ensure a balance between contact stability and wearing comfort. The specific implementation method is as follows:

[0039] First, a nonlinear converter is constructed, taking the skin micro-deformation amplitude parameter as input and the pressure compensation amount as output. The nonlinear converter is a data-driven hardware logic module whose core function is to establish a non-proportional mapping relationship between skin micro-deformation and contact pressure compensation, adapting to the nonlinear physical correlation between the two caused by skin elasticity and electrode material characteristics. During the construction process, contact pressure data of wrist skin from different groups within the range of 0 to maximum safe deformation is collected. Combined with the elastic modulus and fit of the flexible electrode, a basic mapping model is generated through machine learning algorithms. Subsequently, this model is solidified into a logic circuit that can be directly executed by an embedded processor. The circuit responds rapidly according to the input-operation-output process, ensuring that after the input of the skin micro-deformation amplitude parameter, the corresponding pressure compensation amount can be output within 1 millisecond to meet the real-time compensation requirements. Specifically, when the real-time skin micro-deformation amplitude parameter captured by the micro-strain sensor exceeds the predetermined deformation amplitude threshold, the micro-strain sensor—a high-frequency micro-strain sensor installed on the side of the flexible electrode that contacts the skin—captures the minute deformations of the wrist skin caused by respiration and limb micro-movements at a frequency of 1000 Hz. It converts the physical deformation into a voltage signal, which is then converted from analog to digital to obtain the quantified skin micro-deformation amplitude parameter. The predetermined deformation amplitude threshold is based on extensive experimental calibration. The critical value corresponds to the point at which unstable deformation will occur when the electrode contacts the skin. This threshold is stored in the configuration register of the embedded processor and can be finely adjusted according to the skin sensitivity of different wearers. During real-time detection, the embedded processor continuously reads the output signal of the micro-strain sensor, converts it into a skin micro-deformation amplitude parameter, and compares it frame by frame with the predetermined deformation amplitude threshold. When the parameter value exceeds the threshold, a subsequent dynamic compensation mechanism is immediately triggered. This mechanism is based on the deformation and pressure coupling factor, which is a key parameter for quantifying the degree of interaction between skin micro-deformation and contact pressure. Its value is determined by the interaction between the flexible electrode and the skin. The coupling relationship under different conditions is determined by factors such as the material of the skin and the ambient temperature. It is then solidified into a constant through experimental measurement and used to correct the compensation calculation, avoiding deviations caused by compensation based solely on deformation. The dynamic compensation mechanism is a closed-loop system composed of an embedded processor, a feedback channel, and actuator control logic. Its core is to respond quickly to deformation changes and adjust the pressure in real time. After the mechanism is activated, it first obtains the instantaneous increment of the real-time skin micro-deformation amplitude parameter through the feedback channel of the embedded processor. The feedback channel is a high-speed serial communication link between the embedded processor and the micro-strain sensor, with the transmission delay controlled within 1 millisecond to ensure data real-time performance.The instantaneous increment is the difference between the current and previous moments in the skin micro-deformation amplitude parameter, reflecting the rate of deformation change. A larger difference indicates faster deformation growth and a more rapid compensation response is required. The dynamic compensation mechanism converts the instantaneous increment of the real-time skin micro-deformation amplitude parameter into an air chamber volume adjustment command for a micro-pressure actuator via the feedback channel of the embedded processor. The micro-pressure actuator is a miniature pneumatic device mounted on the back of a flexible electrode, containing a retractable air chamber and a piezoelectric ceramic drive unit. Adjusting the air chamber volume changes the contact pressure on the skin; a smaller volume increases the pressure, and vice versa. The air chamber volume adjustment command is a standardized digital command controlling the micro-pressure actuator's movement, containing information on the direction, amplitude, and rate of volume adjustment. During the conversion process, the embedded processor first calculates the pressure compensation amount required to offset the deformation based on the magnitude of the instantaneous increment and the deformation-pressure coupling factor, and then adjusts the compensation amount using a preset... The pressure-volume conversion relationship transforms the pressure compensation amount into a corresponding air chamber volume adjustment value. This adjustment value is then encoded into an 8-bit digital instruction and sent to the drive module of the miniature pressure actuator via a control interface. This causes the pressure compensation amount to increase non-linearly with the deformation amplitude. Since the physical relationship between skin micro-deformation and contact pressure is not non-linear, the pressure compensation amount adopts a non-linear increasing mode. Specifically, when the skin micro-deformation amplitude parameter is within a threshold, the pressure compensation amount increases with a gentle gradient. When the deformation amplitude exceeds 0.2 mm, the growth rate of the compensation amount accelerates, ensuring that sufficient pressure is quickly provided when deformation is significant, maintaining stable contact between the electrode and the skin. This non-linear relationship is achieved through the non-linear converter constructed earlier. The mapping logic inside the converter is preset to a piecewise increasing gradient, avoiding both insufficient compensation leading to increased contact impedance and excessive compensation compressing the skin and affecting comfort, ultimately achieving dynamic adaptation between deformation and pressure.

[0040] After the pressure compensation algorithm triggers the dynamic compensation mechanism, the nonlinear mapping between the micro-deformation unit increment and the pressure compensation amount is the core to ensure the accuracy of compensation. This mapping is achieved through a preset deformation-pressure transformation matrix, combined with real-time parameter dynamic correction, so that the compensation amount always adapts to the coupled changes of skin deformation and pressure attenuation. The specific implementation method is as follows:

[0041] The pressure compensation algorithm establishes a nonlinear mapping between the unit increment of micro-deformation and the pressure compensation amount. The unit increment of micro-deformation refers to the increase in the amplitude parameter of skin micro-deformation per unit time, quantifying the rate of deformation change. The pressure compensation amount is the contact pressure value that the micro-pressure actuator needs to adjust to counteract contact instability caused by deformation. The nonlinear mapping means that the two are not correlated in a fixed proportion. Due to skin elasticity and electrode adhesion characteristics, the faster the deformation rate, the greater the increase in the required compensation amount. This non-proportional relationship needs to be solidified and dynamically adjusted through a specific matrix structure. Specifically, this involves pre-setting a deformation-pressure conversion matrix in the embedded processor. The embedded processor is the core computing unit of the system, responsible for real-time processing of sensor data and execution of algorithm logic. Its computational latency is controlled at the microsecond level to meet the real-time compensation requirements. The deformation-pressure conversion matrix is ​​a two-dimensional structured data table that stores the mapping relationship between the unit increment of micro-deformation and the pressure compensation amount. The pre-setting process involves training... The trained matrix data is written to the processor's non-volatile memory, and can be directly read and called after power-on without recalculation, ensuring mapping response speed. The deformation-pressure conversion matrix is ​​generated through machine learning training. Before training, multi-dimensional sample data is collected, covering the wrist skin of users of different ages and skin types within the micro-deformation unit increment range of 0.01 to 0.2 mm / ms, and the corresponding measured values ​​of pressure compensation. At the same time, the correlation parameters such as the skin micro-deformation amplitude and contact pressure attenuation under each sample are recorded. The gradient boosting tree algorithm is used as the training model, with the micro-deformation unit increment as the input feature and the pressure compensation as the output label. The sample data is substituted for iterative training to continuously optimize the values ​​of matrix elements, so that the error between the predicted value and the measured value of the compensation output by the matrix is ​​controlled within 5%. After training, the matrix is ​​discretized to adapt to the storage and computing characteristics of the embedded processor, and finally, structured matrix data that can be directly called is generated.Its row vectors correspond to the discretized intervals of the unit increment of micro-deformation. The discretized interval is a non-overlapping numerical range divided by a fixed step size of 0.01 mm / ms, covering the common deformation rate range of 0.01 to 0.2 mm / ms, forming a total of 20 intervals. The row vectors of the matrix correspond one-to-one with these intervals, with each row vector representing a deformation rate interval. For example, the first row corresponds to the 0.01-0.02 mm / ms interval, the second row corresponds to the 0.02-0.03 mm / ms interval, and so on. Through discretization, continuous rate values ​​are transformed into indexable row identifiers of the matrix. To simplify the calculation logic, column vectors are mapped to a non-linear increasing gradient of pressure compensation. This non-linear increasing gradient refers to the accelerating growth trend of pressure compensation as the unit increment of micro-deformation increases. Each element of the column vector corresponds to the baseline value of the compensation in the row vector interval. The gradient change is reflected by the difference between the column vector elements. For example, in the low deformation rate interval (0.01-0.05 mm / ms), the difference between the column vector elements is small, and the compensation increases slowly. In the medium-high deformation rate interval (0.05-0.2 mm / ms), the difference between the column vector elements gradually increases, and the compensation increases rapidly. This gradient distribution perfectly matches the material properties of skin deformation and contact pressure. The coupling principle is used to ensure that the faster the deformation, the timely compensation can offset the contact gap. The transformation weight of each discretized interval is dynamically corrected by the coordinated change rate of the skin micro-deformation amplitude parameter and the contact pressure attenuation parameter. The transformation weight is a coefficient that adjusts the influence of the mapping relationship in each discretized interval. The initial value is set to 1. The larger the weight, the more significant the impact of the compensation benchmark value of that interval on the final compensation command. The coordinated change rate is an index that quantifies the rate of synchronous change of the skin micro-deformation amplitude parameter and the contact pressure attenuation parameter. It is calculated as the ratio of the changes in the two parameters within the same time window. The larger the ratio, the tighter the coupling between the two, i.e., the faster the deformation. As the deformation increases, the pressure decays faster. During the dynamic correction process, the embedded processor calculates the coordinated change rate at a frequency of 1 millisecond per calculation. If the coordinated change rate is higher than the preset threshold under the deformation rate corresponding to a certain discretized interval, it indicates that the coupling effect between the two is strong in that interval, and the conversion weight needs to be increased to make the compensation benchmark value of that interval more dominant. If the coordinated change rate is lower than the threshold, the weight is appropriately reduced to avoid overcompensation. Through this real-time correction, the mapping relationship of the deformation pressure conversion matrix is ​​always adapted to the parameter coupling state under the current working condition, so that the pressure compensation amount can accurately offset the deformation effect without affecting the wearing comfort due to excessive pressure.

[0042] After the deformation-pressure conversion matrix completes dynamic weight correction and clarifies the nonlinear mapping relationship between the unit increment of micro-deformation and the pressure compensation amount, in order to achieve the core objective of synchronously increasing contact pressure when micro-deformation increases, a precise step-like compensation mechanism needs to be designed for scenarios with abrupt changes in deformation rate. This ensures that the pressure increase can promptly offset the contact instability caused by the intensified deformation. The specific implementation method is as follows:

[0043] The core logic behind the synchronous increase in contact pressure as micro-deformation increases is to rapidly strengthen the compensation force when the deformation rate shows a significant jump. Specifically, this involves the deformation pressure transformation matrix detecting a cross-interval transition in micro-deformation unit increment. This cross-interval transition refers to the real-time acquired micro-deformation unit increment (the increase in deformation per unit time) jumping from its current discretization interval to a higher-level discretization interval, such as from 0.02-0.03 mm / ms to 0.03-0.04 mm / ms. This phenomenon indicates a sudden acceleration in the deformation rate. If the compensation gradient is adjusted according to the original interval, a compensation lag will occur. During the measurement process, the deformation-pressure conversion matrix receives micro-deformation unit increment data transmitted by the high-frequency micro-strain sensor in real time. Every 1 millisecond, it is compared with the boundary value of the discretized interval. When the increment data of three consecutive sampling periods (3 milliseconds) all stably fall within a higher-level interval, it is determined to be a cross-interval transition to avoid false triggering caused by instantaneous fluctuations. Then, the gradient leap module of pressure compensation is activated. The gradient leap module is a dedicated function module pre-built in the embedded processor, which is specifically responsible for handling the compensation enhancement under the scenario of sudden change in deformation rate. Its core function is to break the conventional linear compensation rhythm during cross-interval transitions and realize the rapid increase of compensation. The activation process is as follows:

[0044] The transition detection signal emitted by the deformation pressure conversion matrix triggers the enable pin inside the module, and the module immediately switches from low-power standby mode to working mode. Simultaneously, it reads the specific value of the current micro-deformation unit increment, the corresponding pressure compensation reference value, and historical compensation command data from the previous 5 sampling cycles through the internal data bus, providing complete data support for subsequent calculations. The gradient transition module uses a multi-stage pneumatic valve controller of the micro-pressure actuator to superimpose the differential value of the previous command onto the current air chamber volume adjustment command. The multi-stage pneumatic valve controller is the core control component of the micro-pressure actuator, consisting of three stages of pneumatic valves with different orifice diameters and corresponding drive circuits. The larger the orifice diameter, the faster the air chamber volume adjustment rate. Through multi-stage combination, it can achieve… For volume control with varying precision, the differential value of the preceding command refers to the rate of change of the air chamber volume adjustment command in the previous sampling period, reflecting the rate trend of the previous compensation round. Its function is to allow the current compensation command to inherit the change inertia of the preceding command, avoiding pressure fluctuations caused by abrupt compensation changes. In specific implementation, the gradient leap module first retrieves the corresponding air chamber volume adjustment command from the deformation pressure conversion matrix based on the micro-deformation unit increment after the current cross-interval, i.e., the current command, to achieve basic compensation. Then, it extracts the differential value of the preceding command through numerical calculation. This differential value is obtained by dividing the volume change of the preceding command by the sampling period (1 millisecond). Subsequently, the current command is superimposed with this differential value. For example, if the current command is to reduce the air chamber volume by 0.15 cubic millimeters... The volume, with a pre-command differential value of 0.1 cubic millimeters per millisecond, is superimposed to obtain a strengthened adjustment command reducing the volume by 0.25 cubic millimeters. This is rapidly executed through the first-stage large-diameter valve of the multi-stage pneumatic valve controller, improving volume adjustment efficiency. This superposition operation forms a step-like compensation for contact pressure. Step-like compensation means that the contact pressure does not increase gradually, but rather experiences a rapid, step-like increase after a transition between intervals. This compensation mode can accurately adapt to scenarios with sudden changes in deformation rate. When a unit increment of micro-deformation transitions between intervals, the contact gap between the skin and the electrode increases rapidly. Gradual compensation cannot fill the gap in time, while step-like compensation can instantly offset the increased gap through a strengthened pressure increase, maintaining stable contact. For example... The initial contact pressure before crossing the interval is 0.3 Newtons, which is directly increased to 0.5 Newtons after step compensation, quickly filling the pressure gap and avoiding increased interface impedance due to contact instability. This ensures the continuity of electromyography signal acquisition. Simultaneously, the step amplitude is calibrated using real-time feedback data from a distributed thin-film pressure sensor array. This array consists of 16 uniformly distributed micro-thin-film pressure sensors embedded on the side of the flexible electrode that contacts the skin, with each sensor spaced 2 mm apart. This allows for simultaneous acquisition of contact pressure data from different locations, avoiding deviations from single-point measurements. Closed-loop calibration refers to forming a cyclic mechanism of command output - pressure detection - deviation correction, ensuring accurate and controllable step compensation amplitude. The specific process is as follows:

[0045] After the step compensation command is executed, the sensor array collects contact pressure data at a frequency of 1000 Hz. The embedded processor averages the 16 collected pressure values ​​to obtain the actual contact pressure. The actual pressure is compared with the target pressure (determined by the standard pressure value corresponding to the current pressure compensation). If the actual pressure is lower than the target pressure, the gradient step module generates a supplementary adjustment command to control the secondary medium-diameter valve of the multi-stage pneumatic valve controller to further reduce the air chamber volume until the deviation between the actual pressure and the target value is less than ±0.02 Newtons. If the actual pressure is higher than the target pressure, the air chamber volume is appropriately increased through the tertiary small-diameter valve to avoid excessive pressure compressing the skin and affecting comfort. Through this real-time closed-loop calibration, it is ensured that the step compensation after each cross-range transition can accurately match the current micro-deformation state, truly realizing that the contact pressure increases synchronously with the increase of micro-deformation, which not only ensures the stability of signal acquisition but also takes into account the wearing experience.

[0046] While ensuring the stability of electrode-skin contact through a pressure compensation algorithm, the dynamic changes in interface impedance still directly affect the transmission efficiency of electromyographic (EMG) signals. Increased impedance at the skin-electrode interface leads to attenuation of the original EMG signal and amplification of low-frequency noise, thereby reducing the signal-to-noise ratio. Therefore, it is necessary to generate precise signal amplification gain adjustment commands based on the deviation between the interface impedance change rate parameter and a predetermined impedance threshold, using a gain adaptive algorithm to ensure that the EMG signal remains clear and discernible at all times. The specific implementation method is as follows:

[0047] Based on the deviation between the interface impedance change rate parameter and a predetermined impedance threshold, the interface impedance change rate parameter is a quantitative indicator reflecting how quickly the interface impedance between the skin and the flexible electrode changes over time. It is generated through dynamic scanning of AC impedance spectroscopy. During the scanning process, the system applies an AC signal with a frequency range of 10 Hz to 100 kHz to the electrode, and acquires impedance values ​​at different frequencies in real time. The ratio of the difference between two adjacent scan results to the time interval is calculated, which is the interface impedance change rate parameter. A larger value indicates a more drastic impedance change. The predetermined impedance threshold is a critical impedance value calibrated based on numerous non-invasive electromyography (EMG) detection experiments. This value corresponds to a signal transmission efficiency that meets the detection requirements. The maximum impedance limit required for measurement is determined by statistically analyzing the average interface impedance under normal skin conditions, taking 1.2 times this as the predetermined impedance threshold, and storing it in the configuration storage area of ​​the embedded processor. This threshold can be fine-tuned according to different skin types. The deviation value is the algebraic difference between the current interface impedance change rate parameter and the predetermined impedance threshold. When the interface impedance increases, the deviation value is positive, and the larger the value, the further the impedance deviates from the ideal range, requiring a higher signal amplification gain. A gain adaptive algorithm generates signal amplification gain adjustment instructions. This algorithm is an intelligent algorithm that dynamically adjusts the signal amplification factor based on the input impedance deviation value. Its core is through real-time... Feedback enables precise matching of gain and impedance changes. The signal amplification gain adjustment command is a digital command that controls the amplification factor of the programmable amplifier. The command format is a 16-bit binary code, with different code values ​​corresponding to different gain factors, ensuring that the amplifier can accurately adjust the amplification intensity according to the command. An impedance deviation gain tuning function is established. The impedance deviation gain tuning function is the core function connecting the impedance deviation value and the gain adjustment amount. Its role is to transform the abstract impedance deviation into specific gain adjustment command parameters. The establishment process requires a large amount of experimental data. First, it simulates interface impedance change scenarios under different skin conditions, and collects the impedance deviation values ​​and corresponding maximum values ​​under each scenario. The optimal gain adjustment refers to the amplification factor that enables the electromyography signal to achieve the best signal-to-noise ratio. A nonlinear fitting algorithm is then used to fit the acquired impedance deviation value-optimal gain adjustment data pair to generate an initial function model. The function parameters are then optimized through actual wear testing, controlling the error between the predicted and actual optimal values ​​to within 3%. Finally, a stable and reliable impedance deviation gain tuning function is formed and stored in the algorithm storage area of ​​the embedded processor. The impedance deviation gain tuning function uses the algebraic difference between the interface impedance change rate parameter generated by AC impedance spectroscopy scanning and a predetermined impedance threshold as its independent variable. The calculation of the function's independent variable requires two steps:

[0048] The first step involves acquiring real-time interface impedance values ​​through dynamic scanning of the AC impedance spectrum. The scanning frequency is set to 100 milliseconds per scan to ensure timely capture of dynamic impedance changes. After each scan, the impedance difference between two adjacent scans is calculated and divided by the scan time interval (100 milliseconds) to obtain the interface impedance change rate parameter. The second step involves reading a predetermined impedance threshold from the configuration storage area and performing an algebraic operation (real-time parameter minus predetermined threshold) on the real-time interface impedance change rate parameter. The result is the independent variable of the function, directly reflecting the degree and direction of impedance deviation from the ideal state, providing a core basis for gain adjustment. The function output is connected to the gain control register of the programmable amplifier. The programmable amplifier is the core component in the system responsible for amplifying the raw electromyographic signals. Its amplification factor can be flexibly adjusted via digital instructions to adapt to raw signals of different intensities. The gain control register is a dedicated storage unit inside the programmable amplifier that stores gain control parameters. The register value directly determines the amplifier's amplification factor. The connection process is implemented through the embedded processor's SPI interface. The output value of the impedance deviation gain tuning function (i.e., the target gain adjustment amount) is encoded and converted into a 16-bit binary control code, which is then used to... The SPI interface writes to the gain control register of the programmable amplifier in real time. After the register value is updated, the amplifier immediately operates at the new gain multiple, ensuring real-time gain adjustment. When the impedance deviation exceeds the linear response range, which is the range where the impedance deviation and gain adjustment are linearly related, the linear response range is determined based on the performance parameters of the programmable amplifier and the dynamic range of the electromyography (EMG) signal. The linear amplification capability of the amplifier at different gains is tested experimentally, and combined with the amplitude range of the original EMG signal, the linear response range of the impedance deviation is calibrated. For example, when the deviation is set between 0 and 500 ohms / millisecond, the gain adjustment is linearly related to the deviation. This range parameter is stored in the algorithm configuration file. During the detection process, the embedded processor compares the current impedance deviation value with the boundary value of the linear response range in real time. When the deviation value is greater than the upper limit of the range, it is determined to exceed the linear response range, triggering the segmented saturation mechanism of the gain adaptive algorithm. The segmented saturation mechanism is a protective mechanism in the gain adaptive algorithm used to limit excessive gain growth. Its core is to prevent the gain from being adjusted indefinitely due to excessive impedance deviation, which could lead to EMG signal distortion or amplifier overload. The triggering process is as follows:

[0049] When the impedance deviation value exceeds the linear response range, the embedded processor sends a trigger signal to the gain adaptive algorithm. The algorithm immediately switches its working mode from linear gain adjustment mode to piecewise saturation mode. At this time, the gain adjustment amount no longer increases linearly with the deviation value, but enters a phase of gradual increase. The gain adjustment switches from linear to logarithmic growth as the impedance deviation increases. Within the linear response range, the gain adjustment and impedance deviation increase linearly; that is, for every unit increase in the deviation, the gain adjustment increases proportionally, ensuring a rapid gain response to slight impedance changes. When the deviation exceeds the linear response range, it switches to logarithmic growth mode. At this point, as the impedance deviation continues to increase, the rate of increase in gain gradually slows down and eventually stabilizes. The core reason for this switching design is that when the impedance deviation is too large, excessive gain adjustment amplifies noise rather than the effective signal, and may even lead to amplifier saturation distortion. Logarithmic growth mode limits the upper limit of gain while ensuring signal strength, balancing the relationship between signal amplification and noise suppression. The specific switching process is automatically implemented by the impedance deviation gain tuning function. The function has preset parameters for the connection between the linear and logarithmic segments. When the independent variable (deviation value) exceeds the boundary of the linear range, the function automatically calls the logarithmic operation logic to calculate the output value, ensuring a smooth switching of the gain adjustment and avoiding sudden signal changes.

[0050] After the gain adaptive algorithm balances signal amplification and noise suppression through a piecewise saturation mechanism, considering that abrupt changes in sweat evaporation can cause sharp fluctuations in interface impedance, leading to frequent gain adjustments and oscillations, an additional correction mechanism is needed to stabilize the gain output. Therefore, a dynamic damping factor is generated by correlating sweat evaporation-related parameters. The specific implementation method is as follows:

[0051] First, the derivative of the sweat evaporation rate parameter is introduced as a correction coefficient for the impedance deviation gain tuning function. The sweat evaporation rate parameter is a physical quantity that quantifies the rate of sweat evaporation on the skin surface. Its value is related to skin humidity, ambient temperature, etc. Abrupt changes in sweat evaporation can directly cause a sudden increase or decrease in interfacial impedance, leading to drastic fluctuations in the output of the gain tuning function. The derivative of the sweat evaporation rate parameter reflects the rate of change of this rate; the larger the absolute value of the derivative, the more drastic the change in evaporation rate. Using it as a correction coefficient allows the gain tuning function to dynamically adapt to abrupt changes in the evaporation rate. The correction coefficient is a parameter used to adjust the output of the impedance deviation gain tuning function. By multiplying it by the original output of the tuning function, the gain adjustment amount can be weakened or strengthened, avoiding over-response of gain caused by abrupt changes in evaporation. In the process, the effective range of the correction coefficient is first preset in the embedded processor to ensure that the correction intensity is within a reasonable range, neither weakening the normal gain adjustment nor suppressing the oscillation caused by sudden changes. Then, the second time derivative of the sweat evaporation rate parameter is calculated by synchronously capturing the skin surface curvature change rate parameter using a high-frequency micro-strain sensor. The skin surface curvature change rate parameter is an indicator reflecting the change of the skin surface curvature over time. Sweat evaporation leads to a decrease in local skin humidity and contraction, thereby changing the skin surface curvature. The two are strongly correlated, so the sweat evaporation situation can be indirectly inferred through this parameter. The high-frequency micro-strain sensor has been previously used to capture skin micro-deformation, and its curvature detection function is activated here. The acquisition frequency is maintained at 1000 Hz to ensure real-time capture of subtle changes in curvature. The specific calculation process is as follows:

[0052] A high-frequency micro-strain sensor converts changes in skin surface curvature into a voltage signal. This signal is then converted from an analog-to-digital converter to obtain a quantified curvature change rate parameter. An embedded processor uses a pre-set mapping model between curvature change rate and sweat evaporation rate (this model is generated through extensive experimental data collection on curvature and evaporation rates under varying humidity levels) to convert the curvature change rate parameter into a corresponding sweat evaporation rate parameter. Subsequently, the first derivative of the sweat evaporation rate parameter is calculated at a 1-millisecond sampling period to reflect the changing trend of the evaporation rate. The same calculation is repeated on the first derivative to obtain the second derivative. The second derivative accurately captures abrupt changes in the sweat evaporation rate; if the evaporation rate suddenly accelerates or decelerates, the second derivative will... A significant peak value appears, providing a core basis for the subsequent generation of the damping factor. The second derivative of time is convolved with the impedance deviation value. Convolution is a mathematical operation that fuses the characteristics of two signals; here, it is used to combine the abrupt changes in sweat evaporation with impedance deviation information, ensuring that the generated damping factor reflects both the abrupt changes in evaporation and the current impedance deviation state. Specifically, the second derivative of time and the impedance deviation value are first standardized, mapping their values ​​to the 0-1 range to eliminate computational bias caused by dimensional differences. Then, a sliding window with a length of 5 sampling periods (5 milliseconds) is set, and convolution is performed on the two standardized parameters in units of this window. During the operation… By using weighted summaries to highlight the influence of recent data (e.g., the most recent data within the window has the highest weight, decreasing sequentially towards previous times), the calculation results are ensured to respond to parameter changes in real time. This calculation merges the characteristics of two independent parameters into a single comprehensive feature value, preserving both the instantaneous characteristics of sweat evaporation mutations and the magnitude of the current impedance deviation. This series of calculations generates a gain-tuned dynamic damping factor. The dynamic damping factor is a suppression parameter that changes in real time with the sweat evaporation state and impedance deviation; its value is positively correlated with the intensity of sweat evaporation mutations—the more severe the mutation, the larger the damping factor. During generation, the result of the convolution operation, after nonlinear transformation, becomes the dynamic damping factor. This conversion ensures that the damping factor remains within a range that effectively suppresses oscillations without affecting normal gain adjustment. If sweat evaporation is stable (second derivative close to 0), the damping factor approaches 0.8, resulting in a weaker correction effect on the gain tuning function, ensuring that the gain can respond normally to impedance changes. If sweat evaporation is abrupt (second derivative shows a peak), the damping factor rapidly drops to the 0.3 to 0.5 range, strengthening the correction effect. Ultimately, this dynamic damping factor suppresses gain oscillations caused by abrupt sweat evaporation. Gain oscillations refer to the phenomenon where abrupt sweat evaporation causes a sudden change in interface impedance, leading to frequent switching of gain adjustment commands, drastic fluctuations in amplification, and ultimately distortion of electromyographic signals. The suppression process is as follows:

[0053] The dynamic damping factor is multiplied by the original output value of the impedance deviation gain tuning function to obtain the corrected gain adjustment. When sweat evaporates abruptly, the damping factor decreases, weakening the change amplitude of the corrected gain adjustment and preventing large jumps in gain. At the same time, the embedded processor monitors the frequency of change of the gain adjustment in real time. If slight oscillations still exist, the damping factor value is further fine-tuned until the gain output tends to be stable. Through this mechanism, the sensitivity of the gain adaptive algorithm to normal impedance changes is ensured, and the interference caused by abrupt sweat evaporation is effectively resisted, ensuring the stability and continuity of the electromyography signal amplification process.

[0054] After stabilizing the gain output and suppressing oscillations caused by sudden changes in sweat evaporation through dynamic damping factors, in order to ensure that the gain adjustment can more accurately follow impedance changes and that the gain can continuously and efficiently adapt to compensate for signal attenuation when impedance increases, the gain adaptive algorithm further establishes a positive feedback mechanism between the unit impedance increment and the gain adjustment. This mechanism forms a self-reinforcing loop by splitting the impedance change components and processing them specifically. The specific implementation method is as follows:

[0055] The gain adaptive algorithm establishes a positive feedback mechanism between impedance unit increment and gain adjustment. Its core principle is to make impedance change detection and gain adjustment mutually reinforcing. That is, the larger the impedance unit increment, the more precise the gain adjustment. Conversely, precise gain adjustment improves the sensitivity of impedance detection, thus more clearly capturing impedance changes, forming a virtuous cycle. Specifically, this involves integrating a gain-impedance coupler inside the programmable amplifier. As the core component of signal amplification, the programmable amplifier has a dedicated area for integrating functional modules. The gain-impedance coupler is a specially designed analog signal processing module used to correlate impedance changes with gain adjustment. During integration, the coupler is connected to the amplifier's signal link via internal metal wiring, with one end connected to the output of the impedance detection circuit. The other end connects to the gain control register and the positive feedback loop respectively, ensuring efficient transmission of the impedance signal and the gain control signal. The coupler employs a low-noise design to avoid introducing interference that could affect signal quality. The gain-impedance coupler decomposes the unit impedance increment into fundamental frequency and harmonic components. The unit impedance increment is the change in interface impedance per unit time, containing both a stable core trend and instantaneous fluctuations. The fundamental frequency component is the main stable component of the impedance change, corresponding to the low-frequency signal portion and reflecting the core trend of the impedance change. The harmonic components are the high-frequency fluctuation portion of the impedance change, reflecting the detailed characteristics of the impedance change. This decomposition process is achieved through the coupler's built-in signal separation unit, which includes a set of precise bandpass and high-pass filters. The frequency range of the filter is adapted to the main frequency of impedance change to extract the fundamental frequency component. The high-pass filter filters out high-frequency signals above 100 Hz as harmonic components. The cutoff frequencies of both types of filters are experimentally calibrated to ensure that the fundamental frequency and harmonic components do not overlap or omit any, fully preserving all characteristics of the unit impedance increment. The fundamental frequency component is directly input to the gain control register for linear up-adjustment. As the core of impedance change, the fundamental frequency component has strong stability and high reliability, and is the main basis for determining the amount of gain up-adjustment. Its direct input method is achieved through a dedicated signal line between the coupler and the gain control register, without additional conversion processing, ensuring no signal delay and distortion. The gain control register is the core unit for storing gain parameters inside the programmable amplifier. The value stored in the register directly determines the amplifier's gain. When the fundamental frequency component is input, the register adjusts the stored gain parameter linearly according to the preset fundamental frequency component and gain adjustment ratio. For example, for every unit increase in the fundamental frequency component, the gain parameter is increased by a fixed ratio, achieving linear gain growth and ensuring that the gain adjustment smoothly follows the core change trend of impedance. Harmonic components are injected into the reference voltage terminal of the impedance detection circuit through a positive feedback loop. Although harmonic components are fluctuating, they can reflect subtle dynamics of impedance changes. The positive feedback loop is a closed-loop circuit composed of a signal amplification submodule, a delay calibration submodule, and a coupling injection submodule. Its function is to inject the processed harmonic components in reverse into the impedance detection circuit, enhancing the impedance change detection signal.The reference voltage terminal serves as the reference voltage source for the impedance detection circuit. Its voltage stability directly affects the accuracy of the impedance detection. In practice, the harmonic components are first amplified to an amplitude suitable for the reference voltage by the signal amplification submodule in the positive feedback loop. Then, the signal transmission delay is corrected by the delay calibration submodule to ensure timing synchronization with the fundamental frequency component. Finally, the harmonic components are smoothly injected into the reference voltage terminal via capacitive coupling by the coupling injection submodule. This causes the reference voltage to fluctuate slightly with the harmonic components, allowing the impedance detection circuit to more sensitively detect minute changes in impedance. This design creates a self-reinforcing loop of impedance increment and gain upscaling. When the interface impedance increases, the impedance... The fundamental frequency component, which resists unit increments, drives the gain control register to linearly increase the gain, resulting in clearer amplification of the electromyographic signal and reduced noise interference with impedance detection. Simultaneously, harmonic components are injected into the reference voltage terminal via a positive feedback loop, enhancing the sensitivity of the impedance detection circuit to impedance changes. This makes subsequent impedance increment detection more accurate, and the precise impedance increment, in turn, drives a more precise gain increase. This cycle repeats, achieving self-reinforcement between impedance increments and gain increases. This ensures both rapid gain adjustment and improved accuracy, guaranteeing timely and accurate gain adaptation even to complex dynamic impedance changes, maintaining a high signal-to-noise ratio for the electromyographic signal.

[0056] After the periodic pressure maintenance command begins execution, in order to accurately adapt to the real-time intensity of the coupled negative cycle and avoid insufficient maintenance command energy to stop the cycle, or excessive energy to compress the skin and affect comfort, the command parameters need to be dynamically adjusted through real-time feedback. The specific implementation method is as follows:

[0057] During the pressure maintenance command execution cycle, the spectral characteristics of the skin micro-deformation amplitude parameter and the fluctuation variance of the interface impedance change rate parameter are simultaneously acquired. The pressure maintenance command execution cycle refers to the time from the start to the completion of a single periodic pressure maintenance command, calibrated to 1 second based on wearing comfort and compensation effect. Each cycle contains multiple pulse signals. Synchronous acquisition ensures that the timestamps of the two types of parameters are perfectly aligned. The acquisition frequency is maintained at 1000 Hz (skin micro-deformation) and 100 milliseconds / time (interface impedance), consistent with the sensor, to avoid analysis bias caused by timing misalignment. The spectral characteristics of the skin micro-deformation amplitude parameter are frequency domain information reflecting the periodic law of micro-deformation. After acquisition, the time-domain micro-deformation sequence is converted into frequency domain data through Fast Fourier Transform to extract the main... Frequency components, peak frequencies, and harmonic distributions—for example, deformation caused by respiration corresponds to low-frequency components, while limb micro-movements correspond to mid-frequency components—can intuitively reflect the intensity and periodicity of micro-deformation, thereby determining the trigger frequency of negative cycles. The variance of the interface impedance change rate parameter is an indicator of the stability of impedance changes. The dispersion of the impedance change rate parameter within each execution cycle is calculated; a larger variance indicates more severe impedance fluctuations, meaning a stronger coupling negative cycle between skin micro-deformation, impedance changes, and pressure attenuation. During data acquisition, the high-frequency micro-strain sensor and the AC impedance spectrum scanning module synchronously output data according to a preset time sequence. The embedded processor stores the two types of parameters separately through dual buffers to ensure no data loss or overlap. This is achieved through the inverse dynamic coupling analysis model. The dynamic coupling analysis model has previously been used to handle the forward coupling operations of three types of parameters in the feedback channel. The reverse feedback channel is a high-speed data link specifically designed to transmit real-time monitoring data from the execution end back to the model. It works independently yet collaboratively with the forward operation channel, with transmission latency controlled within 1 millisecond to ensure real-time feedback adjustments. This channel uses a serial communication protocol, compressing and encoding data before transmission to reduce bandwidth consumption, and adding a checksum to ensure data integrity and prevent adjustment deviations due to transmission errors. Its core function is to build a closed-loop link of acquisition-analysis-adjustment, allowing the pressure maintenance command to be dynamically optimized based on the real-time changes in the negative loop. The spectral characteristics and fluctuation variance are input into the recursive filter of the pressure attenuation suppression algorithm. The recursive filter is the pressure... The core component of the force attenuation suppression algorithm, which processes time-series feedback data, employs a fourth-order infinite impulse response filter structure. This structure smooths noise while preserving the dynamic trend of the data, avoiding response lag caused by over-filtering. Before input, the spectral characteristics and fluctuation variance are standardized, mapping the values ​​to the 0-1 range to eliminate the influence of dimensional differences on the filtering effect. During filtering, the filter selects effective signals according to a preset cutoff frequency, removing high-frequency instantaneous noise. Simultaneously, it retains the trend information of historical data through recursive calculations, such as the increasing trend of fluctuation variance over multiple consecutive periods. This ensures that the output data accurately reflects the cumulative strength of the coupled negative loop, providing a stable and reliable input basis for the subsequent generation of the blocking coefficient.The negative cycle blocking coefficient is a core parameter for quantifying the energy intensity required to block the coupling of a negative cycle. Its value ranges from 0.2 to 0.8; a smaller value corresponds to stronger blocking energy, and a larger value to weaker energy. During the generation process, the standardized data output by the recursive filter is processed by a nonlinear mapping model. If the spectral characteristics of skin microdeformation show a high proportion of low-frequency components (indicating that periodic deformations such as respiration dominate, and the negative cycle is continuous), and the variance of the interface impedance fluctuation is large (indicating severe impedance fluctuation and high cycle intensity), then the mapping model outputs a smaller blocking coefficient. Conversely, if the spectral characteristics show a low proportion of high-frequency components and a small variance of fluctuation (indicating a gentle negative cycle), then a larger blocking coefficient is output. This mapping model is trained and generated through extensive experiments collecting feedback data under different cycle intensities to ensure that the coefficient corresponds to the cycle intensity. Precise matching of the degree of interference ensures that energy is not wasted while effectively blocking the cycle. The negative cycle blocking coefficient adjusts the envelope attenuation slope of the pulse sequence generator. The envelope attenuation slope is a parameter describing how quickly the pulse signal energy decays in the periodic pressure maintenance command. The larger the slope, the faster the pulse energy decays and the shorter the effective duration of a single pulse; the smaller the slope, the slower the energy decays and the longer the duration. The adjustment process is achieved by the embedded processor sending control signals to the pulse sequence generator. When the blocking coefficient is small (corresponding to a strong negative cycle), the generator is controlled to reduce the envelope attenuation slope, causing the pulse signal energy to decay slowly, extending the effective duration, and increasing the overall output energy. When the blocking coefficient is large (corresponding to a weak negative cycle), the attenuation slope is increased, shortening the energy duration and reducing the output energy. Ultimately, the output energy of the periodic pressure maintenance command is inversely proportional to the strength of the coupled negative cycle. This inverse relationship ensures the accuracy of the compensation: the higher the strength of the coupled negative cycle, the greater the command output energy, which can quickly suppress cycle deterioration; the lower the cycle strength, the smaller the output energy, avoiding excessive pressure on the skin and affecting wearing comfort. For example, when the negative circulation intensity increases to its peak, the blocking coefficient drops to 0.2, the envelope attenuation slope is minimized, and the pulse energy remains at a high level, effectively counteracting the coupling effect of pressure attenuation and micro-deformation. When the circulation intensity weakens to a plateau state, the blocking coefficient rises to 0.8, the attenuation slope increases, and energy is released gently, maintaining the baseline contact pressure without causing skin discomfort. Through this dynamic adjustment closed loop, the coupling negative circulation of the three parameters is continuously blocked, providing a stable contact environment for electromyography signal acquisition.

[0058] After the dynamic coupling analysis model generates real-time contact pressure compensation commands and signal amplification gain adjustment commands, in order to ensure the accurate execution of these two types of commands and achieve coordinated optimization of pressure compensation and signal amplification, it is necessary to convert the commands into control signals that the actuator can recognize and strictly align the execution timing. The specific implementation method is as follows:

[0059] First, the real-time contact pressure compensation command is encoded into a duty cycle control signal for the pneumatic valve. This duty cycle control signal is a pulse width modulation (PWM) signal. By adjusting the duty cycle (the ratio of the high-level time to the period) of the high-level signal within a cycle, the opening and closing degree of the pneumatic valve is controlled. A larger duty cycle results in a longer valve opening time and a more significant change in the air chamber volume. During encoding, the embedded processor first reads the air chamber volume adjustment amount from the real-time contact pressure compensation command. Based on a preset volume adjustment-duty cycle mapping table (which is experimentally calibrated to determine the optimal duty cycle for different volume changes, ensuring adjustment accuracy), the volume adjustment amount is converted into the corresponding duty cycle value. Then, a pulse width modulation module generates a pulse signal with a fixed frequency (100 Hz, adapted to the response speed of the pneumatic valve). The high-level duration is adjusted according to the calculated duty cycle to form the duty cycle control signal. This signal is amplified by the drive circuit to the 5-12 volt range of the pneumatic valve's response voltage, ensuring that the valve can accurately receive the control command.

[0060] The piezoelectric ceramic actuator of the miniature pressure actuator transforms the gas chamber volume into a step-like change. The miniature pressure actuator is the core actuator for adjusting the contact pressure of the flexible electrode. It contains a sealed gas chamber and a piezoelectric ceramic actuator. The piezoelectric ceramic actuator uses the piezoelectric effect (voltage change causes mechanical deformation) to achieve precise drive with deformation accuracy down to the micrometer level. It can accurately control the gas chamber volume. The step-like change means that the volume is adjusted step by step (0.05 cubic millimeters) to avoid pressure fluctuations caused by sudden changes. During the conversion process, after receiving the duty cycle control signal, the piezoelectric ceramic actuator generates corresponding mechanical deformation according to the signal duty cycle. When the duty cycle increases, the actuator extends and squeezes the gas chamber, the volume decreases and the pressure increases. When the duty cycle decreases, the actuator contracts, the volume increases and the pressure decreases. After each adjustment, it pauses for 1 millisecond. The current contact pressure is fed back in real time through a distributed thin-film pressure sensor array. The next step is only taken after confirming that the expected value has been reached, ensuring that the contact pressure changes smoothly. Simultaneously, the signal amplification gain adjustment instruction is compiled into a gain control word. The gain control word is a digital control code that the programmable amplifier can recognize. Different code combinations correspond to different amplification factors. The 16-bit control word can achieve gain adjustment from 0 to 65535 levels. During the compilation process, the embedded processor parses the target gain factor in the instruction and, according to the gain configuration table of the programmable amplifier (clarifying the correspondence between the gain factor and the control word, which has been verified by experiments), converts the target gain into a 16-bit binary control word, adds a check bit to avoid transmission errors, and ensures accurate gain adjustment.The gain control word is loaded onto the digital-to-analog converter (DAC) of the programmable amplifier. The DAC converts the digital gain control word into an analog voltage signal. A 16-bit precision model is selected, with conversion error controlled within 0.1%. The loading process is implemented via a serial peripheral interface bus. The embedded processor transmits the control word to the DAC input port in real time. The converter converts it into an analog control voltage within 1 microsecond, which is then input to the gain control terminal of the programmable amplifier. This drives the internal amplification circuit to adjust the feedback resistor ratio, achieving the target gain switching. The entire loading delay is controlled within 5 microseconds to meet the real-time signal amplification requirements. The air cavity volume change and the loading timing of the gain control word are strictly aligned using the clock synchronization module of the dynamic coupling analysis model. The clock synchronization module is a built-in timing control unit in the dynamic coupling analysis model, using a 10 MHz high-precision crystal oscillator (nanosecond-level timing accuracy) as the clock source. It is responsible for coordinating the instruction execution time. During the alignment process, the clock synchronization module first acquires electromyography (EMG) signal sampling data, extracts timing features to determine the sampling period, and then sets the synchronization clock signal, adjusting the duty cycle... The timing of the control signal output and the timing of the gain control word loading are bound within the same cycle, with a starting time deviation of no more than 100 nanoseconds. Simultaneously, the transmission delay is monitored in real time. If the deviation exceeds 50 nanoseconds, the signal output buffer delay time is adjusted to compensate, ensuring that pressure compensation and gain adjustment start and complete synchronously. Ultimately, this ensures that pressure compensation and gain adjustment are completed before the peak period of the electromyographic signal action potential. The peak period of the electromyographic signal action potential is the time when the signal amplitude reaches its peak value, during which the signal is most susceptible to distortion due to adjustment actions. To ensure this, the embedded processor monitors the signal amplitude in real time through the electromyographic signal preprocessing module and predicts the peak period start time according to preset rules (a peak period is determined when the amplitude exceeds three times the average amplitude). Combining the execution delays of the two types of adjustments (20 milliseconds for pressure compensation and 5 milliseconds for gain adjustment), the instruction loading deadline is set to ensure that the adjustment is completed at least 5 milliseconds before the peak period begins. If the peak period is earlier, the clock synchronization module shortens the signal transmission buffer and execution waiting time to ensure that the adjustment action is completed before the peak period, ultimately outputting an electromyographic signal with optimized signal-to-noise ratio.

[0061] After the dynamic coupling analysis model generates real-time contact pressure compensation commands and signal amplification gain adjustment commands, in order to ensure that the two types of commands are accurately implemented and to achieve synergistic optimization of pressure compensation and signal amplification, the commands need to be converted into control signals that the actuator can recognize, and the execution timing must be strictly aligned to avoid affecting the quality of electromyographic signal acquisition due to asynchronous movements. The specific implementation method is as follows:

[0062] First, the real-time contact pressure compensation command is encoded into a duty cycle control signal for the pneumatic valve. This duty cycle control signal is a pulse width modulation (PWM) signal. By adjusting the ratio of the high-level duration to the period within a cycle, the opening and closing degree of the pneumatic valve is controlled. A larger duty cycle results in a longer valve opening time and a more significant change in the air chamber volume, thus achieving precise adjustment of the contact pressure. During the encoding process, the embedded processor first reads the air chamber volume adjustment amount contained in the real-time contact pressure compensation command. This adjustment amount is the target volume change value calculated based on the skin micro-deformation amplitude parameter. Then, it retrieves the preset volume adjustment... The volume-duty cycle mapping table is calibrated through numerous experiments to determine the optimal duty cycle corresponding to different volume changes (e.g., a 0.1 cubic millimeter volume reduction corresponds to a 50% duty cycle) to ensure adjustment accuracy. Then, the processor's built-in pulse width modulation module generates a pulse signal with a fixed frequency (100 Hz, adapted to the mechanical response speed of the pneumatic valve). The high-level duration is adjusted according to the duty cycle determined by the mapping table to form a duty cycle control signal. Finally, this signal is amplified by the drive circuit to the 5-12 volt range suitable for the pneumatic valve to avoid valve lag or failure due to insufficient voltage. The piezoelectric ceramic actuator of the miniature pressure actuator transforms the change in air chamber volume into a step-like change. The miniature pressure actuator is the core actuating component for adjusting the contact pressure of the flexible electrode. It contains a sealed air chamber and a piezoelectric ceramic actuator. The change in air chamber volume is directly related to the contact pressure (a decrease in volume increases the pressure, and vice versa). The piezoelectric ceramic actuator uses the piezoelectric effect (micrometer-level mechanical deformation after applying voltage) to achieve precise actuation with a deformation accuracy of 0.1 micrometers, which can meet the needs of fine adjustment of air chamber volume. The step-like change refers to the gradual adjustment of volume in fixed steps (0.05 cubic millimeters). To avoid sudden changes that cause drastic fluctuations in contact pressure and to balance stability and wearing comfort, during the conversion process, the piezoelectric ceramic actuator receives the duty cycle control signal and generates corresponding mechanical deformation according to the signal duty cycle. When the duty cycle increases, the actuator extends and compresses the air chamber, gradually reducing the volume. When the duty cycle decreases, the actuator contracts, gradually increasing the volume. After each adjustment step is completed, it pauses for 1 millisecond and collects the current contact pressure data in real time through a distributed thin-film pressure sensor array. The data is compared with the target pressure value, and the next adjustment is only performed when the deviation is confirmed to be less than ±0.02 Newtons, ensuring that the change in air chamber volume is stable and precise.

[0063] Simultaneously, the signal amplification gain adjustment instruction is compiled into a gain control word. The gain control word is a digital control code recognizable by the programmable amplifier, using a 16-bit binary format. Different code combinations correspond to different signal amplification factors. The 16-bit length allows for 0-65535 levels of gain adjustment, meeting the dynamic amplification requirements of electromyographic signals. During compilation, the embedded processor first parses the target gain factor in the signal amplification gain adjustment instruction. This factor is the optimal value calculated by the gain adaptive algorithm based on the interface impedance change rate parameter. Then, it retrieves the gain configuration table of the programmable amplifier, which clearly defines the one-to-one correspondence between the gain factor and the 16-bit control word. After converting the target gain factor into the corresponding 16-bit binary control word, two check bits are added to the end of the control word to verify the data integrity during transmission and avoid gain adjustment errors due to bit errors. The gain control word is then loaded into the digital-to-analog converter of the programmable amplifier. The digital-to-analog converter is the core component that converts the digital gain control word into an analog voltage signal. A 16-bit high-precision model is selected, with conversion errors controlled within 0.Within 1%, ensuring the accuracy of gain adjustment, the programmable amplifier adjusts the feedback resistor ratio of its internal amplification circuit by receiving an analog control voltage, thereby switching the signal amplification factor. The loading process is implemented through a serial peripheral interface bus. The embedded processor transmits the gain control word with a check bit to the input port of the digital-to-analog converter in real time. The converter converts the digital control word into the corresponding analog control voltage within 1 microsecond and inputs this voltage to the gain control terminal of the programmable amplifier. The entire loading process delay is controlled within 5 microseconds to meet the requirements of real-time amplification of electromyographic signals and avoid signal distortion caused by delay. The change in air cavity volume is related to the gain control... The loading timing of the gain control word is strictly aligned using the clock synchronization module of the dynamic coupling analysis model. This clock synchronization module, a built-in timing control unit within the model, uses a 10 MHz high-precision crystal oscillator as its clock source and coordinates the execution time of pressure compensation and gain adjustment to avoid signal acquisition deviations caused by timing misalignment. During alignment, the clock synchronization module first acquires the sampling data of the electromyography signal, extracts the signal timing characteristics to determine the sampling period, and then generates a synchronization clock signal. This binds the output timing of the duty cycle control signal and the loading timing of the gain control word to the same sampling period, with the time deviation between their starting moments strictly controlled within 100 nanoseconds. Simultaneously, the transmission delay of both types of signals is monitored in real time. If the delay deviation exceeds 50 nanoseconds, compensation is made by adjusting the delay time of the signal output buffer to ensure that pressure compensation and gain adjustment are started and completed synchronously. This achieves synergistic optimization of contact stability and signal amplification effect, ultimately ensuring that pressure compensation and gain adjustment are completed before the peak period of the electromyographic signal action potential. The peak period of the electromyographic signal action potential refers to the time period during which the amplitude of the electromyographic signal reaches its peak. During this period, the signal is most susceptible to distortion due to external interference. Therefore, all adjustments must be completed before the peak period. During the process, the embedded processor monitors the amplitude changes of the original signal in real time through the electromyographic signal preprocessing module. The system predicts the start time of the peak period according to preset rules (when the signal amplitude exceeds three times the average amplitude of the past 10 sampling periods, it is determined that the peak period is about to begin). Combining this with the execution delays of two types of adjustment actions (approximately 20 milliseconds for pressure compensation and approximately 5 milliseconds for gain adjustment), a deadline for instruction loading is set to ensure that all adjustment actions are completed at least 5 milliseconds before the peak period begins. If the peak period is predicted to arrive earlier than expected, the clock synchronization module immediately shortens the signal transmission buffer time and execution waiting time, prioritizing the completion of adjustment actions before the peak period. The final output is an EMG signal with optimized signal-to-noise ratio, providing high-quality data support for subsequent signal analysis.

[0064] This invention acquires raw electromyographic (EMG) signals from a target area using flexible electrodes, simultaneously obtaining parameters such as skin micro-deformation amplitude captured by a high-frequency micro-strain sensor, interface impedance change rate generated by dynamic scanning of AC impedance spectroscopy, and electrode contact pressure attenuation quantified by a distributed thin-film pressure sensor array. These three parameters are input into a dynamic coupling analysis model. A pressure compensation algorithm generates real-time contact pressure compensation commands, a gain adaptive algorithm generates signal amplification gain adjustment commands, and a pressure attenuation suppression algorithm generates periodic pressure maintenance commands. These commands drive a micro-pressure actuator and a programmable amplifier, aligning the timing to ensure adjustment is completed before the peak of the EMG signal, outputting an EMG signal with optimized signal-to-noise ratio, balancing detection accuracy and wearing comfort.

[0065] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely preferred examples and are not intended to limit the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. A non-invasive bioelectrical signal detection method based on flexible sensing technology, characterized in that: Includes the following steps: S1. Acquire raw electromyographic signals of the target area using flexible electrodes, and simultaneously initialize skin micro-deformation amplitude parameters, interface impedance change rate parameters, and electrode contact pressure attenuation parameters; wherein, the skin micro-deformation amplitude parameters are captured by a high-frequency micro-strain sensor, the interface impedance change rate parameters are generated by dynamic scanning of AC impedance spectrum, and the electrode contact pressure attenuation parameters are quantized by a distributed thin-film pressure sensor array. S2. Input the three types of parameters—skin micro-deformation amplitude parameter, interface impedance change rate parameter, and electrode contact pressure attenuation parameter—into the dynamic coupling analysis model, and perform the following processing, specifically including: a. Based on the proportional relationship between the skin micro-deformation amplitude parameter and the predetermined deformation amplitude threshold, a real-time contact pressure compensation command is generated through a pressure compensation algorithm, wherein the pressure compensation algorithm establishes a nonlinear mapping between the unit increment of micro-deformation and the pressure compensation amount, so that the contact pressure increases synchronously when the micro-deformation increases. b. Based on the deviation between the interface impedance change rate parameter and the predetermined impedance threshold, a signal amplification gain adjustment command is generated through a gain adaptive algorithm, wherein the gain adaptive algorithm establishes a positive feedback mechanism between the unit impedance increment and the gain up adjustment. c. Based on the derivative relationship between the electrode contact pressure attenuation parameter and the time attenuation function, a periodic pressure maintenance command is generated through a pressure attenuation suppression algorithm to block the coupling negative loop of the skin micro-deformation amplitude parameter, the interface impedance change rate parameter, and the electrode contact pressure attenuation parameter. S3. Drive the micro pressure actuator according to the real-time contact pressure compensation command, control the programmable amplifier according to the signal amplification gain adjustment command, and output the electromyographic signal with optimized signal-to-noise ratio.

2. The non-invasive bioelectrical signal detection method based on flexible sensing technology according to claim 1, characterized in that: The generation of real-time contact pressure compensation commands through a pressure compensation algorithm specifically includes: A nonlinear converter is constructed with skin micro-deformation amplitude parameters as input and pressure compensation amount as output. When the real-time skin micro-deformation amplitude parameters captured by the micro-strain sensor exceed a predetermined deformation amplitude threshold, a dynamic compensation mechanism based on the deformation and pressure coupling factor is triggered. The dynamic compensation mechanism converts the instantaneous increment of the real-time skin micro-deformation amplitude parameters into an air chamber volume adjustment command of the micro pressure actuator through the feedback channel of the embedded processor, so that the pressure compensation amount increases nonlinearly with the deformation amplitude.

3. The non-invasive bioelectrical signal detection method based on flexible sensing technology according to claim 2, characterized in that: The pressure compensation algorithm establishes a nonlinear mapping between the unit increment of micro-deformation and the pressure compensation amount, specifically including: A deformation-pressure transformation matrix is ​​pre-set in the embedded processor. The deformation-pressure transformation matrix is ​​generated through machine learning training. Its row vectors correspond to the discretized intervals of micro-deformation unit increments, and the column vectors are mapped to the nonlinear increasing gradient of the pressure compensation amount. The transformation weight of each discretized interval is dynamically corrected by the coordinated change rate of the skin micro-deformation amplitude parameter and the contact pressure attenuation parameter.

4. The non-invasive bioelectrical signal detection method based on flexible sensing technology according to claim 3, characterized in that: The method that causes the contact pressure to increase synchronously with the increase of micro-deformation specifically includes: When the deformation pressure conversion matrix detects a small deformation unit increment transition across intervals, it activates the gradient leap module for pressure compensation. The gradient leap module uses a multi-stage pneumatic valve controller of a micro pressure actuator to superimpose the differential value of the previous command on the current air chamber volume adjustment command, forming a step compensation for the contact pressure. At the same time, it uses real-time feedback data from a distributed thin-film pressure sensor array to calibrate the step amplitude in a closed loop.

5. The non-invasive bioelectrical signal detection method based on flexible sensing technology according to claim 1, characterized in that: The method of generating signal amplification gain adjustment instructions based on the deviation between the interface impedance change rate parameter and a predetermined impedance threshold using a gain adaptive algorithm specifically includes: An impedance deviation gain tuning function is established. The impedance deviation gain tuning function uses the algebraic difference between the interface impedance change rate parameter generated by AC impedance spectrum scanning and the predetermined impedance threshold as the independent variable. The output terminal is connected to the gain control register of the programmable amplifier. When the impedance deviation value exceeds the linear response range, the piecewise saturation mechanism of the gain adaptive algorithm is triggered, so that the gain adjustment amount switches from linear growth to logarithmic growth when the impedance deviation value increases.

6. The non-invasive bioelectrical signal detection method based on flexible sensing technology according to claim 5, characterized in that: The derivative of the sweat evaporation rate parameter is introduced as a correction coefficient for the impedance deviation gain tuning function. The second time derivative of the sweat evaporation rate parameter is calculated by synchronously capturing the skin surface curvature change rate parameter using a high-frequency micro-strain sensor. The second time derivative is then convolved with the impedance deviation value to generate a dynamic damping factor for gain tuning, which suppresses gain oscillations caused by abrupt changes in sweat evaporation.

7. The non-invasive bioelectrical signal detection method based on flexible sensing technology according to claim 6, characterized in that: The gain adaptive algorithm establishes a positive feedback mechanism between the unit impedance increment and the gain upscaling amount, specifically including: An integrated gain-impedance coupler is used in the programmable amplifier. The gain-impedance coupler decomposes the unit impedance increment into the fundamental frequency component and the harmonic component. The fundamental frequency component is directly input to the gain control register to achieve linear up-adjustment, while the harmonic component is injected into the reference voltage terminal of the impedance detection circuit through a positive feedback loop, forming a self-reinforcing loop between the unit impedance increment and the gain up-adjustment.

8. The non-invasive bioelectrical signal detection method based on flexible sensing technology according to claim 1, characterized in that: The generation of periodic pressure maintenance commands through the pressure decay suppression algorithm specifically includes: A pressure maintenance function is constructed with the first derivative of the electrode contact pressure decay parameter with respect to time as input. The output of the pressure maintenance function is connected to the pulse trigger module of the micro pressure actuator. When the derivative value of the pressure decay parameter exceeds the decay rate threshold of the time decay function, a pulse sequence generator based on pressure decay phase compensation is activated to generate a periodic pressure maintenance command with an exponential decay envelope.

9. The non-invasive bioelectrical signal detection method based on flexible sensing technology according to claim 8, characterized in that: During the pressure maintenance command execution cycle, the spectral characteristics of the skin micro-deformation amplitude parameter and the fluctuation variance of the interface impedance change rate parameter are simultaneously collected. Through the reverse feedback channel of the dynamic coupling analysis model, the spectral characteristics and fluctuation variance are input into the recursive filter of the pressure attenuation suppression algorithm to generate a negative loop blocking coefficient. The negative loop blocking coefficient adjusts the envelope attenuation slope of the pulse sequence generator so that the output energy of the periodic pressure maintenance command is inversely proportional to the strength of the coupled negative loop.

10. The non-invasive bioelectrical signal detection method based on flexible sensing technology according to claim 8, characterized in that: The real-time contact pressure compensation command is encoded into a duty cycle control signal for a pneumatic valve, which is then converted into a stepwise change in the air chamber volume by a piezoelectric ceramic actuator of a miniature pressure actuator. Simultaneously, the signal amplification gain adjustment command is compiled into a gain control word and loaded into the digital-to-analog converter of a programmable amplifier. The timing of the air chamber volume change and the loading of the gain control word is strictly aligned by the clock synchronization module of the dynamic coupling analysis model to ensure that pressure compensation and gain adjustment are completed before the peak period of the electromyographic signal action potential.

Images