Anti-crosstalk biological rehabilitation electric signal safety control method and system under high-density scene

Abstract

The application relates to the technical field of behavior feature recognition, and discloses an anti-crosstalk biological rehabilitation electrical signal safety control method and system in a high-density scene. The method comprises the following steps: calculating a bilateral average distance according to an environmental distance; calculating a surface electromyogram signal root mean square value according to an electromyogram original signal; analyzing a pelvis tilt angle standard deviation according to the pelvis tilt angle; calculating a hip joint adduction auxiliary torque according to the surface electromyogram signal root mean square value and a gait phase; calculating a hip joint coronal plane impedance value according to the gait phase and the pelvis tilt angle; calculating an intention state according to the surface electromyogram signal root mean square value, the pelvis tilt angle and a channel state; calculating a hip joint driving torque according to the hip joint adduction auxiliary torque and the hip joint coronal plane impedance value; and calculating a system mode according to the bilateral average distance, the surface electromyogram signal root mean square value and the pelvis tilt angle standard deviation. The application improves the lower limb rehabilitation training efficiency by adaptively adjusting the gait.

Description

Technical Field

[0001] This application relates to the field of behavioral feature recognition technology, and more specifically, to a method and system for secure control of bio-rehabilitation electrical signals against crosstalk in high-density scenarios. Background Technology

[0002] Current lower limb rehabilitation exoskeleton systems primarily rely on inertial measurement unit (IMU) data for gait monitoring, lacking dynamic fusion analysis with surface electromyography (EMG) signals. This results in an inability to effectively identify user intent in narrow passageways commonly found in home environments. Existing technologies typically use preset fixed thresholds to determine passage status and user intent, failing to dynamically integrate bilateral environmental distances, EMG activation characteristics, and pelvic posture fluctuations. This leads to forced stride width adjustments even when passage width is insufficient, potentially increasing the risk of collisions.

[0003] Chinese patent application CN118266914A discloses a method and device for matching gait for personalized lower limb rehabilitation training based on feature transfer. The method includes: Step 1, measuring the body characteristic parameters of a calibrator, then having the calibrator wear a lower limb rehabilitation exoskeleton and conduct a walking test using rehabilitation training gait patterns from a lower limb exoskeleton rehabilitation training gait database. After the walking test, calculating the body fitness results and gait correction results for each rehabilitation training gait pattern, and calculating the comprehensive utility of the rehabilitation training gait pattern based on the body fitness results and gait correction results; Step 2, repeating Step 1 multiple times... First, acquire the body characteristic parameters of multiple calibrators and the comprehensive utility of the corresponding rehabilitation training gait. After normalizing the comprehensive utility, a training set is formed by the body characteristic parameters and the utility of the corresponding normalized rehabilitation training gait. The utility transfer model of rehabilitation training gait is obtained by fitting this training set. Second, measure the body characteristic parameters of the rehabilitation user and input them into the utility transfer model of rehabilitation training gait. The model calculates the transfer utility corresponding to different rehabilitation training gait in the lower limb exoskeleton rehabilitation training gait database, and then obtains the gait matching result of the rehabilitation user.

[0004] While the above methods can meet the needs of most scenarios, research and practical application of these methods and existing technologies have revealed at least the following drawbacks:

[0005] The aforementioned methods rely solely on inertial measurement unit kinematic data for behavioral feature recognition, which makes it difficult to effectively identify the user's intention to pass through narrow passages at home. This increases the false trigger rate, severely restricts system safety, disrupts the patient's autonomous balance rhythm, and easily leads to problems such as collisions with furniture, increasing the safety risks for patients during home rehabilitation and hindering the smooth progress of home rehabilitation training.

[0006] In view of this, this application proposes a method and system for safe control of crosstalk-resistant biorehabilitation electrical signals in high-density scenarios to solve the above problems. Summary of the Invention

[0007] To address the shortcomings of existing technologies, this application aims to provide a method and system for safe control of bio-rehabilitation electrical signals in high-density scenarios with anti-crosstalk. This method involves real-time acquisition of environmental distance, surface electromyography (EMG) signals, and raw signals from the inertial measurement unit (IMU) in narrow passages within a home environment. The method calculates and analyzes the bilateral average distance, the root mean square value of the EMG signals, and the raw IMU signals, fusing environmental distance, EMG data, and posture data to effectively identify the user's intention to actively adjust the system, thus eliminating false triggers. Based on EMG activity and gait phase, the method dynamically calculates the hip joint adduction assist torque and coronal impedance value to achieve adaptive gait adjustment. By comprehensively analyzing multi-source data to dynamically generate system modes and calculate driving torques, the system can automatically switch to a safe mode when encountering narrow passages or pelvic fluctuations, thereby reducing false trigger rates and gait interruption rates. This effectively improves the safety, continuity, and seamless experience of rehabilitation training in home settings.

[0008] This application provides the following technical solution: a method for secure control of bio-rehabilitation electrical signals against crosstalk in high-density scenarios, including:

[0009] Collect and analyze the raw environmental distance signal to obtain the bilateral average distance and channel status;

[0010] Raw electromyography (EMG) signals were collected and analyzed to obtain the root mean square (RMS) value of the surface EMG signal.

[0011] The raw inertial signals are collected and processed to obtain the inertial state signals;

[0012] Based on the root mean square value of surface electromyography (EMG) signals, channel state, and pelvic tilt angle in inertial state signals, the intention state is obtained through analysis; based on the root mean square value of surface EMG signals and gait phase in inertial state signals, the hip joint adduction assist torque is calculated; based on the intention state, gait phase, and pelvic tilt angle, the coronal impedance value of the hip joint is calculated.

[0013] The hip joint driving torque is calculated based on the inertial state signal, the bilateral average distance, the root mean square value of the surface electromyography signal, the hip joint adduction auxiliary torque, and the hip joint coronal impedance value. The standard deviation of the pelvic tilt angle is obtained by analysis based on the pelvic tilt angle. The system mode is calculated based on the bilateral average distance, the root mean square value of the surface electromyography signal, and the standard deviation of the pelvic tilt angle.

[0014] A high-density environment anti-crosstalk biorehabilitation electrical signal safety control system, comprising the following methods:

[0015] The first analysis module is used to collect and analyze the raw environmental distance signal to obtain the average distance on both sides and the channel status.

[0016] The second analysis module is used to collect and analyze raw electromyographic signals to obtain the root mean square value of surface electromyographic signals.

[0017] The third analysis module is used to collect and process the raw inertial signals to obtain the inertial state signal.

[0018] The stride width adjustment module is used to analyze and obtain the intention state based on the root mean square value of the surface electromyography signal, the channel state, and the pelvic tilt angle in the inertial state signal; to calculate the hip joint adduction auxiliary torque based on the root mean square value of the surface electromyography signal and the gait phase in the inertial state signal; and to calculate the hip joint coronal plane impedance value based on the intention state, gait phase, and pelvic tilt angle.

[0019] The stable switching module is used to calculate the hip joint driving torque based on the inertial state signal, the bilateral average distance, the root mean square value of the surface electromyography signal, the hip joint adduction assist torque, and the hip joint coronal impedance value; to analyze and obtain the standard deviation of the pelvic tilt angle based on the pelvic tilt angle; and to calculate the system mode based on the bilateral average distance, the root mean square value of the surface electromyography signal, and the standard deviation of the pelvic tilt angle.

[0020] The technical effects and advantages of the anti-crosstalk bio-rehabilitation electrical signal security control method and system proposed in this application are as follows:

[0021] By acquiring raw environmental distance signals and performing window filtering, bilateral average distance and channel status are obtained; raw electromyography (EMG) signals are acquired and filtered and feature extracted to obtain the root mean square (RMS) value of surface EMG signals; raw signals from the inertial measurement unit (IMU) are acquired and attitude calculation is performed to obtain hip joint angle, gait phase, and pelvic tilt angle; the scattered sensor data are organized into a unified state vector to avoid the intention misjudgment problem caused by relying on data from a single sensor in existing technologies; based on this, the intention state is generated by fusing the RMS value of surface EMG signals, pelvic tilt angle, and channel status, transforming intention recognition from static threshold judgment to a recognition process based on multi-source dynamic features.

[0022] By calculating the hip adduction assist torque based on the root mean square value of surface electromyography (EMG) signals and gait phase, and calculating the coronal impedance value of the hip joint based on the intention state, gait phase, and pelvic tilt angle, adaptive gait adjustment is achieved. When the intention state is 1 and the gait phase is the swing phase, the root mean square value of EMG signals and the gait phase are co-mapped to generate the adduction assist torque, avoiding interference with the stability of the support phase. When the intention state is 1 and the gait phase is the support phase, the coronal impedance value is dynamically adjusted by the pelvic tilt angle to effectively suppress the risk of body tilt. The step width adjustment amount is obtained by calculating the difference in safe step width thresholds from the average distance between the two sides. A dynamic target angle is generated by combining the support phase reference and the swing phase reference. Then, the hip joint driving torque is calculated by co-integrating the coronal impedance value, adduction assist torque, hip joint angle, and dynamic target angle, achieving real-time matching of environment, posture, and intention.

[0023] By setting a sliding window to remove outliers and calculate the standard deviation of the pelvic tilt angle, the standard deviation of the pelvic tilt angle is obtained. Multi-source data status analysis is performed by combining the bilateral average distance, the root mean square value of surface electromyography signal, and the standard deviation of the pelvic tilt angle to generate a system mode. When the intention status is that there is a desire to pass, the bilateral average distance is not greater than the safe step width threshold, and the standard deviation of the pelvic tilt angle is lower than the stability threshold, the system is adjusted to the step width adjustment mode; otherwise, the normal mode is maintained. This three-dimensional safety judgment criterion enables the system to automatically switch to the safety mode in narrow passages or when the pelvis is fluctuating, reducing the false trigger rate, gait interruption rate, and fall risk.

[0024] By organizing the steps of environmental distance acquisition, electromyography signal acquisition, inertial measurement unit signal acquisition, intention state analysis, auxiliary torque and impedance value calculation, driving torque calculation, and system mode judgment into a continuous data processing chain, the processing results of the previous step serve as the input basis for the next step, realizing a closed-loop flow relationship between each processing link, and providing a complete data traceability basis for rehabilitation training effect analysis and system optimization.

[0025] This application solves the problems of inaccurate intent recognition and improper gait adjustment in narrow passage scenarios in the prior art by constructing a system that enables continuous data transmission, closed-loop verification feedback and clear threshold setting basis for the gait adaptive adjustment process, thereby improving the safety, continuity and seamless experience of lower limb rehabilitation training in the home environment. Attached Figure Description

[0026] Figure 1 This is a schematic diagram of the method for controlling the safety of crosstalk-resistant biorehabilitation electrical signals in high-density scenarios according to this application.

[0027] Figure 2 This is a flowchart of the method for obtaining inertial state signals according to this application;

[0028] Figure 3This is a flowchart illustrating the method for obtaining the hip joint angle in this application;

[0029] Figure 4 This is a schematic diagram of the structure of the anti-crosstalk biorehabilitation electrical signal safety control system for high-density scenarios in this application. Detailed Implementation

[0030] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0031] Example 1:

[0032] Please see Figure 1 As shown in this embodiment, the method for secure control of anti-crosstalk bio-rehabilitation electrical signals in high-density scenarios includes:

[0033] Collect and analyze the original environmental distance signal to obtain the average distance on both sides and the channel status.

[0034] In one specific embodiment, the method for obtaining the bilateral average distance and channel state includes:

[0035] Raw environmental distance signals are collected, including left and right raw distance signals. These signals are acquired by distance sensors located on the left and right sides of the lower limb rehabilitation exoskeleton pelvic fixation component. The left raw distance signal represents the lateral distance from the left exoskeleton boundary of the target user to the left obstacle or passage sidewall, while the right raw distance signal represents the lateral distance from the right exoskeleton boundary of the target user to the right obstacle or passage sidewall. The raw environmental distance signals are filtered by a window of length K. The left and right raw distance signals from K consecutive sampling points prior to the current sampling time are selected according to the window length K, and the distance signals within each window are averaged to obtain the left and right filtered distance sequences. K is set by rounding the product of the sampling frequency of the environmental distance sensors and the duration of the target user's most recent complete gait cycle. If the target user's gait cycle is not obtained, the duration of the target user's most recent complete gait cycle is taken as the average of the durations of three consecutive gait cycles during the system startup calibration phase.

[0036] Calculate the arithmetic mean of the left and right filtering distances respectively to obtain the bilateral average distance, which includes the left average distance and the right average distance.

[0037] The channel status is determined by judging the safe step width threshold conditions of the average distance on the left and the average distance on the right. When the average distance on the left and the average distance on the right are not greater than the safe step width threshold, the channel status is judged as a narrow channel; otherwise, the channel status is judged as a normal channel. The safe step width threshold is used to characterize the minimum lateral safety distance that the target user needs to maintain to safely pass through the channel in the current exoskeleton wearing state. The safe step width threshold is set by statistical settings based on gait parameters of healthy people.

[0038] The raw electromyographic signals of the target user's hip adductor muscle group corresponding to the acquisition channel were collected and analyzed to obtain the root mean square value of the surface electromyographic signal.

[0039] In one specific embodiment, the method for obtaining the root mean square value of the surface electromyography signal includes:

[0040] Raw signals from the hip adductor muscle group were collected as raw electromyographic signals.

[0041] The raw electromyography (EMG) signal is filtered using a bandpass filter with a frequency range of [O, P] to obtain a filtered signal. The frequency parameters O and P are set based on the following: the effective energy of the surface EMG signal is concentrated within a specific frequency band; signals below the lower limit of this band are mainly motion artifacts, while signals above the upper limit are mainly high-frequency noise. The frequency parameters O and P are determined by the power spectrum of the hip adductor muscle group EMG signal acquired during the calibration phase. O is the frequency corresponding to 5% of the cumulative energy, and P is the frequency corresponding to 95% of the cumulative energy, with O less than P. The adjustment rules for the frequency parameters O and P are as follows: based on the effective energy frequency band of the surface EMG signal, combined with the electrode type and acquisition environment, the lower cutoff frequency O and upper cutoff frequency P of the bandpass filter are set. Adaptive adjustments should be made. When using dry electrodes or acquiring data in environments with strong motion artifacts, the O value can be appropriately increased to suppress low-frequency motion interference. Motion artifacts refer to low-frequency non-myopathic interference components introduced by relative displacement between the electrode and the skin, wire swaying, or exoskeleton vibration. When using wet electrodes or acquiring data in environments with good signal quality, the O value can be appropriately decreased to retain more low-frequency electromyographic components. When there is high-frequency electromagnetic interference or significant high-frequency noise in the electromyographic signal in the acquisition environment, the P value can be appropriately decreased to filter out high-frequency noise. When it is necessary to retain high-frequency electromyographic details, the P value can be appropriately increased, but it must not exceed the theoretical upper limit of the effective energy band. During the adjustment process, the O value should always be lower than the P value, and the passband range should cover the main energy concentration range of the surface electromyographic signal.

[0042] The root mean square value of the surface electromyography (EMG) signal is obtained by statistically calculating the filtered signal based on a window of length T. The method for setting the window length T is as follows: the product of the EMG sampling frequency and 10% of the duration of the target user's most recent complete gait cycle is rounded to the nearest integer. When the calculation result is less than the number of sampling points corresponding to 50ms, the value is taken as the number of sampling points corresponding to 50ms.

[0043] The raw inertial signals are collected and processed to obtain the inertial state signals.

[0044] In one specific implementation, refer to Figure 2 Methods for obtaining inertial state signals include:

[0045] Inertial state signals include hip joint angle, gait phase, and pelvic tilt angle.

[0046] Reference Figure 3 Methods for obtaining hip joint angles include:

[0047] The static acceleration of the pelvic inertial measurement unit is collected and normalized to obtain the normalized gravity vector. Based on the normalized gravity vector, the initial quaternion of the pelvis is calculated, such as the initial quaternion of the pelvis. , , , ,in, , and These are the components of the normalized gravity vector; The scalar part of the initial quaternion of the pelvis; The vector part of the initial quaternion of the pelvis has three components in the x and y axes; These are the three components of the vector part of the initial quaternion of the pelvis along the y-axis. The vector component of the initial pelvic quaternion has three components along the z-axis. In this embodiment, the initial pelvic quaternion is determined based on the normalized gravity vector measured by the pelvic inertial measurement unit in a static state. Since the initial attitude estimation mainly establishes pitch and roll attitudes based on the gravity direction, and the yaw direction is not used as the corresponding determination basis, the corresponding component is set to 0. The initial pelvic quaternion is used to characterize the initial attitude of the pelvic coordinate system relative to the reference coordinate system. The initial pelvic quaternion is transposed to obtain the pelvic attitude quaternion. Euler angle transformation is performed on the pelvic attitude quaternion and the thigh attitude quaternion to obtain the hip joint angle, such as the hip joint angle. ,in, To implement Euler angle transformation in the order of Z-axis, Y-axis, and X-axis, when the pelvic coordinate system has the anteroposterior direction as the X-axis, the lateral direction as the Y-axis, and the vertical direction as the Z-axis, the hip joint angle is taken as the coronal adduction or abduction angle component around the anteroposterior axis in the relative Euler angle. For pelvic posture quaternions, For the inverse of the thigh posture quaternion, This is quaternion multiplication.

[0048] A dual discriminant analysis was performed on the hip joint angle and hip joint angular velocity.

[0049] Methods for obtaining gait phase include:

[0050] The ratio of the difference in hip joint angle between two consecutive sampling points to the sampling interval is calculated to obtain the hip joint angular velocity. When the hip joint angle is within the range of the support phase angle and the hip joint angular velocity is less than 0, the gait phase is determined to be the support phase. When the hip joint angle is within the range of the swing phase angle and the hip joint angular velocity is greater than 0, the gait phase is determined to be the swing phase. The upper limit of the support phase angle range is less than the lower limit of the swing phase angle range. The support phase angle range and the swing phase angle range are statistically set by statistically analyzing the hip joint angle trajectories of healthy adults under normal gait.

[0051] Methods for obtaining pelvic tilt angle include:

[0052] The arctangent function of the acceleration of the pelvic inertial measurement unit is calculated to obtain the pelvic tilt angle, such as the pelvic tilt angle. ,in, It is the arctangent function. For pelvic inertial measurement unit Axial acceleration, The z-axis acceleration of the pelvic inertial measurement unit. The conversion coefficient for radian to angle; real-time monitoring of hip joint angles allows the system to distinguish between the support phase and the swing phase, facilitating dynamic adjustment of stride width during the swing phase and avoiding interference with stability during the support phase; calculation of pelvic tilt angle enables continuous assessment of body tilt status. When the pelvic tilt angle increases, the system correspondingly increases the coronal impedance value of the hip joint to improve lateral stability during the support phase and reduce the risk of lateral collisions during passage through narrow passages; dynamic calculation of bilateral average distances allows the system to perceive the distance between the user's left and right sides and the passage sidewalls in real time; when both the left and right average distances are not greater than the safe stride width threshold, the system determines that it is currently in a narrow passage state and triggers stride width adjustment control; the root mean square value of surface electromyography (EMG) signals reflects the activation intensity of the hip adductor muscle group. Combined with passage status and pelvic posture information, the system can recognize the user's intention to actively adjust stride width and perform human-machine collaborative adjustment accordingly; the fusion analysis of gait phase and pelvic tilt angle enables the system to achieve adaptive stride width adjustment in narrow passage environments, improving passage safety and the continuity of rehabilitation training.

[0053] The intention state is obtained by analyzing the root mean square value of the surface electromyography (EMG) signal, the channel state, and the pelvic tilt angle in the inertial state signal; the hip joint adduction assist torque is calculated based on the root mean square value of the EMG signal and the gait phase in the inertial state signal; and the coronal impedance value of the hip joint is calculated based on the intention state, gait phase, and pelvic tilt angle.

[0054] In one specific implementation, the method for obtaining the intent state includes:

[0055] The root mean square (RMS) value of surface electromyography (EMG) signals, pelvic tilt angle, and channel status are fused and analyzed to determine the intent state. Specifically, if the channel status is determined to be a narrow channel, the RMS value of the surface EMG signal is not less than the EMG signal activation threshold, and the absolute value of the pelvic angular velocity is not less than the pelvic lateral displacement threshold, then the intent to pass is determined; otherwise, the intent to pass is determined. The pelvic angular velocity is calculated as the ratio of the difference in pelvic tilt angle between two consecutive sampling times to the sampling interval. The EMG signal activation threshold is used to determine whether the muscles are in an effective activated state, avoiding misinterpretation of noise signals or non-EMG signals as muscle activity signals. The method for setting the EMG signal activation threshold is as follows: by collecting the RMS value of the surface EMG signals of the subject's hip adductor muscle group in a relaxed sitting position. The system statistically sets values; the pelvic lateral displacement threshold is used to quantitatively determine whether lateral displacement of the pelvis has occurred; the method for setting the pelvic lateral displacement threshold is: statistically setting it by collecting pelvic angular velocity signals from healthy adults during normal walking; real-time monitoring of the bilateral average distance enables the system to identify narrow passage environments, and triggers step width adjustment when the distance between the left and right legs and the sidewall of the passage is not greater than the safe step width threshold; the root mean square value of the surface electromyography signal reflects the activation intensity of the hip adductor muscle group, and confirms the user's intention to actively pass when the activation intensity exceeds the threshold; the pelvic tilt angle and its rate of change dynamically assess the body's lateral tilt state, and combined with the passage state and electromyographic intention, achieves adaptive step width adjustment, reduces the risk of collisions in narrow passages, and improves the safety of home rehabilitation training and the smoothness of user passage.

[0056] In one specific embodiment, the method for obtaining the hip joint adduction assist torque includes:

[0057] When the intention status is "intent to pass," it indicates that the patient is preparing to pass through the narrow passage and gait adjustment is required. Co-mapping analysis is performed on the root mean square value of the surface electromyography (EMG) signal and the gait phase to obtain the hip adduction assist torque. Specifically, the torque proportionality coefficient, the product of the root mean square value of the EMG signal and the gait phase, is calculated to obtain the hip adduction assist torque. The torque proportionality coefficient is only effective when the gait phase is the swing phase. The torque proportionality coefficient is set based on the ratio between the root mean square value of the EMG signal generated during the maximum voluntary contraction of the hip adductor muscles in a healthy adult and the clinically required assist torque.

[0058] In one specific embodiment, the method for obtaining the coronal plane impedance value of the hip joint includes:

[0059] Adaptive state analysis was performed on gait phase and pelvic tilt angle to obtain the coronal impedance value of the hip joint, such as the coronal impedance value of the hip joint. ,in, Based on the fundamental impedance, The stability coefficient is only used during the gait phase, i.e., the gait phase. When the value is 0, the system takes effect. The method for setting the base impedance is as follows: it is set according to the typical value of the coronal plane stiffness of the hip joint during normal walking in a healthy adult. The method for setting the stability coefficient is as follows: under the known pelvic tilt angle offset, the additional impedance increment required to maintain body balance is measured, and the stability coefficient is obtained by fitting. The fusion of the root mean square value of surface electromyography signal and gait phase enables the system to recognize the user's intention to pass through narrow passages and apply adduction auxiliary torque only in the swing phase to avoid interfering with the stability of the support phase. The dynamic correlation between the pelvic tilt angle and the gait phase enables the system to automatically enhance the coronal plane impedance in the support phase, suppress the risk of body tilt in real time, ensure the accuracy and safety of adaptive step width adjustment in narrow passages, and improve the passage efficiency and user safety of home rehabilitation training.

[0060] The hip joint driving torque is calculated based on the inertial state signal, the bilateral average distance, the root mean square value of the surface electromyography signal, the hip joint adduction auxiliary torque, and the hip joint coronal impedance value. The standard deviation of the pelvic tilt angle is obtained by analysis based on the pelvic tilt angle. The system mode is calculated based on the bilateral average distance, the root mean square value of the surface electromyography signal, and the standard deviation of the pelvic tilt angle.

[0061] In one specific embodiment, the method for obtaining the hip joint driving torque includes:

[0062] The step width adjustment amount is obtained by calculating and analyzing the difference in the safe step width threshold between the two sides' average distances. ;in, The step width adjustment coefficient is set by converting insufficient environmental distance into hip joint adduction angle compensation through geometric relationships. Users can make adaptive adjustments based on the exoskeleton's mechanical structure parameters.

[0063] Binary classification analysis of gait phases is performed to obtain support phase data points and oscillation phase data points, such as support phase data points. Swing phase data points ,in, This is the set of sampling times corresponding to the reference time windows used for statistical support phase reference and oscillating phase reference.

[0064] Statistical mean analysis is performed on the support phase data points, gait phase, and hip joint angles to obtain the support phase reference, such as the support phase reference. ,in, Sampling time The angle of the hip joint.

[0065] Statistical mean analysis is performed on the swing phase data points, gait phase, and hip joint angle to obtain the swing phase reference, such as the swing phase reference. .

[0066] Adaptive analysis is performed on the swing phase reference and the support phase reference to obtain the hip joint reference angle, such as the hip joint reference angle. ;in, The gait angular frequency is set as follows: it is calculated based on the average gait frequency in the user's historical gait data. If there is no historical data, the average gait frequency of healthy adults is used as the default value; it is obtained from historical data statistics. The initial phase is determined based on the starting point of the gait cycle.

[0067] The dynamic target angle is obtained by calculating the difference between the hip joint reference angle and the step width adjustment.

[0068] The hip joint driving torque is obtained by synergistically fusing the hip joint coronal impedance value, hip joint adduction auxiliary torque, hip joint angle, and dynamic target angle. Specifically, the angle difference between the hip joint angle and the dynamic target angle is calculated, the product of the angle difference and the hip joint coronal impedance value is calculated, and the sum of the product and the hip joint adduction auxiliary torque is calculated to obtain the hip joint driving torque.

[0069] In one specific embodiment, the method for obtaining the standard deviation of the pelvic tilt angle includes:

[0070] Set a window duration as Data is collected using a sliding window with a sampling frequency of U, and the window duration is... The method for setting the sampling frequency U is as follows: It needs to cover at least two gait cycles to ensure statistical stability while also considering real-time response; U is set based on the Nyquist frequency of the inertial measurement unit and electromyography signals; data slides into a sliding window in real time, removing the oldest data and always maintaining a consistent flow. A series of data, This indicates that the product within the parentheses is rounded to the nearest integer, and the set of pelvic tilt angle data points within the sliding window is obtained based on the window's sliding update.

[0071] Calculate the mean and standard deviation of the pelvic tilt angle data set; when the absolute value of the difference between any pelvic tilt angle data point and the mean is greater than three times the standard deviation, the corresponding data point is identified as an outlier and replaced with an adjacent valid data point to obtain a purified pelvic tilt angle dataset; calculate the standard deviation of the purified pelvic tilt angle dataset as the pelvic tilt angle standard deviation.

[0072] In one specific implementation, the method for obtaining the system pattern includes:

[0073] Set the system mode to step width adjustment mode and normal mode.

[0074] When the intention status is "there is a passable intention", the average distance on both sides is not greater than the safe step width threshold, and the standard deviation of the pelvic tilt angle is lower than the stability threshold, the mode is adjusted to step width adjustment mode; otherwise, the normal mode is maintained. The stability threshold is set by collecting the sliding window standard deviation of the pelvic tilt angle under normal operating conditions.

[0075] By innovatively integrating surface electromyography (SEMG) activity and posture data from inertial measurement units, a dynamic closed loop of environmental perception, intention recognition, and safety control is constructed: Step width adjustment is calculated in real-time based on the bilateral average distance, enabling the system to dynamically adapt to changes in channel width; precise statistical analysis of support and swing phase references using support / swing phase data points helps ensure the hip joint reference angle accurately reflects the user's natural gait characteristics; real-time calculation of the standard deviation of the pelvic tilt angle serves as a core indicator of posture stability, and together with the bilateral channel safety status and SEMG signal activation level, constitutes a three-dimensional safety judgment criterion, allowing the system to distinguish between step width adjustment mode and conventional mode; the synergistic effect of these intermediate data ensures that hip joint driving torque calculation not only considers the dynamic target angle but also incorporates environmental safety and user status, achieving real-time matching of environment, posture, and intention. This reduces false triggering rate, gait interruption rate, and fall risk, solving the safety risks and human-machine conflict problems of traditional systems in complex environments, and upgrading rehabilitation training from passive execution to adaptive collaborative control.

[0076] Example 2:

[0077] Please see Figure 4 As shown, a high-density scenario anti-crosstalk biorehabilitation electrical signal safety control system includes:

[0078] The first analysis module is used to collect and analyze the raw environmental distance signal to obtain the average distance on both sides and the channel status.

[0079] The second analysis module is used to collect and analyze raw electromyographic signals to obtain the root mean square value of the surface electromyographic signal.

[0080] The third analysis module is used to collect and process the raw inertial signals to obtain inertial state signals, which include hip joint angle, gait phase, and pelvic tilt angle.

[0081] The stride width adjustment module is used to analyze and obtain the intention state based on the root mean square value of surface electromyography (EMG) signal, pelvic tilt angle, and channel status; calculate the hip joint adduction assist torque based on the root mean square value of surface EMG signal and gait phase; and calculate the hip joint coronal impedance value based on the intention state, gait phase, and pelvic tilt angle.

[0082] The stable switching module is used to calculate the hip joint driving torque based on the inertial state signal, the bilateral average distance, the root mean square value of the surface electromyography signal, the hip joint adduction assist torque, and the hip joint coronal impedance value; to analyze and obtain the standard deviation of the pelvic tilt angle based on the pelvic tilt angle; and to calculate the system mode based on the bilateral average distance, the root mean square value of the surface electromyography signal, and the standard deviation of the pelvic tilt angle.

[0083] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0084] Finally: The above description is only a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

Claims

1. A method for secure control of bio-rehabilitation electrical signals against crosstalk in high-density scenarios, characterized in that, include: Collect and analyze the raw environmental distance signal to obtain the bilateral average distance and channel status; Raw electromyography (EMG) signals were collected and analyzed to obtain the root mean square (RMS) value of the surface EMG signal. The raw inertial signals are collected and processed to obtain the inertial state signals; Based on the root mean square value of surface electromyography (EMG) signals, the pelvic tilt angle in the channel state and inertial state signals, the intention state is obtained by analysis; based on the root mean square value of surface EMG signals and the gait phase in the inertial state signals, the hip joint adduction auxiliary torque is calculated. The coronal impedance value of the hip joint is calculated based on the intention state, gait phase, and pelvic tilt angle. The hip joint driving torque is calculated based on the inertial state signal, the bilateral average distance, the root mean square value of the surface electromyography signal, the hip joint adduction auxiliary torque, and the hip joint coronal impedance value. Based on the pelvic tilt angle, the standard deviation of the pelvic tilt angle is obtained; based on the bilateral average distance, the root mean square value of surface electromyography signal, and the standard deviation of the pelvic tilt angle, the system pattern is calculated.

2. The method for secure control of anti-crosstalk biorehabilitation electrical signals in high-density scenarios according to claim 1, characterized in that, Methods for obtaining hip joint driving torque include: The step width adjustment amount is obtained by calculating the safety step width threshold difference based on the average distance between the two sides. Gait phases are classified into two categories to obtain support phase and swing phase data points; Statistical mean analysis was performed on the data points of the support phase and the swing phase and the hip joint angle to obtain the support phase reference and the swing phase reference. Adaptive analysis is performed on the swing phase reference and the support phase reference to obtain the hip joint reference angle; The dynamic target angle is calculated based on the difference between the stride width adjustment and the hip joint reference angle. The hip joint coronal impedance value, hip joint adduction auxiliary torque, hip joint angle, and dynamic target angle are synergistically fused and calculated to obtain the hip joint driving torque.

3. The method for secure control of bio-rehabilitation electrical signals in high-density scenarios according to claim 2, characterized in that, Methods for obtaining coronal impedance values ​​of the hip joint include: When the intention state is "there is a passing intention" and the gait phase is the support phase, adaptive state analysis is performed on the gait phase and pelvic tilt angle to calculate the coronal impedance value of the hip joint.

4. The method for secure control of anti-crosstalk biorehabilitation electrical signals in high-density scenarios according to claim 2, characterized in that, Methods for obtaining hip joint angle include: The static acceleration of the pelvic inertial measurement unit was collected and normalized to obtain the normalized gravity vector; Calculate the initial quaternion of the pelvis based on the normalized gravity vector; Transpose the initial pelvic quaternion to obtain the pelvic posture quaternion; The stationary angular velocity of the thigh inertial measurement unit is collected and attitude is calculated to obtain the initial quaternion of the thigh. Transpose the initial quaternion of the thigh to obtain the thigh pose quaternion; Euler angle transformation is performed on the pelvic posture quaternion and the thigh posture quaternion to obtain the hip joint angle.

5. The method for secure control of anti-crosstalk biorehabilitation electrical signals in high-density scenarios according to claim 2, characterized in that, Methods for obtaining hip adduction assist torque include: When the intention state is "to pass" and the gait phase is the swing phase, the root mean square value of the surface electromyography signal and the gait phase are co-mapped to calculate the hip joint adduction auxiliary torque.

6. The method for secure control of anti-crosstalk biorehabilitation electrical signals in high-density scenarios according to claim 5, characterized in that, Methods for obtaining gait phase include: The hip joint angular velocity is calculated based on the difference between the hip joint angle at the current sampling time and the hip joint angle at the previous sampling time, as well as the sampling interval between the current sampling time and the previous sampling time. When the hip joint angle is within the range of the support phase angle and the hip joint angular velocity is less than 0, it is determined to be the support phase. When the hip joint angle is within the range of the swing phase angle and the hip joint angular velocity is greater than 0, it is determined to be the swing phase.

7. The method for secure control of anti-crosstalk biorehabilitation electrical signals in high-density scenarios according to claim 1, characterized in that, Methods for obtaining intent state include: When the channel state is determined to be a narrow channel, the root mean square value of the surface electromyography signal is not less than the electromyography signal activation threshold, and the absolute value of the pelvic angular velocity is not less than the pelvic lateral displacement threshold, then the intention state is determined to be an intention to pass; otherwise, the intention state is determined to be an intention not to pass. The pelvic angular velocity is calculated by the ratio of the difference in pelvic tilt angle between two consecutive sampling times to the sampling interval.

8. The method for secure control of bio-rehabilitation electrical signals in high-density scenarios with anti-crosstalk as described in claim 7, characterized in that, Methods for obtaining channel status include: The pelvic tilt angle is obtained by calculating the arctangent function based on the lateral and vertical acceleration components output by the pelvic inertial measurement unit. When the average distance on the left and the average distance on the right are not greater than the safe step width threshold, the channel is determined to be a narrow channel; otherwise, the channel is determined to be a normal channel.

9. The method for secure control of bio-rehabilitation electrical signals in high-density scenarios with anti-crosstalk as described in claim 1, characterized in that, Methods for obtaining the two-sided mean distance include: The original environmental distance signal is filtered by a window of length K to obtain a left-side filtered distance sequence and a right-side filtered distance sequence. The arithmetic mean of the left-side filtered distance sequence and the right-side filtered distance sequence are calculated respectively to obtain the left-side average distance and the right-side average distance. The left-side average distance and the right-side average distance are used as the bilateral average distance.

10. The method for secure control of anti-crosstalk biorehabilitation electrical signals in high-density scenarios according to claim 1, characterized in that, The system modes include step width adjustment mode and normal mode, and the system modes are obtained in the following ways: A set of pelvic tilt angle data points is collected through a sliding window; the mean and standard deviation of the pelvic tilt angle data points are calculated, and outliers are filtered and removed based on the mean and standard deviation to obtain a purified pelvic tilt angle dataset; the standard deviation of the pelvic tilt angle is calculated based on the purified pelvic tilt angle dataset; when the intent state is "passing intent", the average distance on the left and the average distance on the right are not greater than the safe step width threshold, and the standard deviation of the pelvic tilt angle is lower than the stability threshold, the step width adjustment mode is activated; otherwise, the normal mode is maintained.

11. A high-density scene anti-crosstalk biorehabilitation electrical signal safety control system, implementing the high-density scene anti-crosstalk biorehabilitation electrical signal safety control method according to any one of claims 1-10, characterized in that, include: The first analysis module is used to collect and analyze the raw environmental distance signal to obtain the average distance on both sides and the channel status. The second analysis module is used to collect and analyze raw electromyographic signals to obtain the root mean square value of surface electromyographic signals. The third analysis module is used to collect and process the raw inertial signals to obtain the inertial state signal. The stride width adjustment module is used to analyze and obtain the intention state based on the root mean square value of the surface electromyography signal, the channel state, and the pelvic tilt angle in the inertial state signal; and to calculate the hip joint adduction auxiliary torque based on the root mean square value of the surface electromyography signal and the gait phase in the inertial state signal. The coronal impedance value of the hip joint is calculated based on the intention state, gait phase, and pelvic tilt angle. The stable switching module is used to calculate the hip joint driving torque based on the inertial state signal, the bilateral average distance, the root mean square value of the surface electromyography signal, the hip joint adduction auxiliary torque, and the hip joint coronal impedance value. Based on the pelvic tilt angle, the standard deviation of the pelvic tilt angle is obtained; based on the bilateral average distance, the root mean square value of surface electromyography signal, and the standard deviation of the pelvic tilt angle, the system pattern is calculated.

Images