Joint torque estimation method, device and equipment based on electromyographic signals of lower limb exoskeleton and storage medium

Abstract

The embodiment of the invention provides a joint torque estimation method and device based on electromyographic signals of lower limb exoskeletons, computer equipment, a storage medium and a computer program product, and relates to the technical field of exoskeletons. The method comprises the following steps: acquiring an initial filtering electromyographic signal of a target object; obtaining an electromyographic signal envelope corresponding to the initial filtering electromyographic signal according to the initial filtering electromyographic signal; performing normalization processing on the electromyographic signal envelope to obtain an electromyographic signal value corresponding to the electromyographic signal envelope; according to the electromyographic signal value, obtaining an independent source signal corresponding to the initial filtering electromyographic signal; and obtaining a joint torque value of a target exoskeleton matched with the target object according to the independent source signal. According to the method, the accuracy and efficiency of joint torque estimation based on the electromyographic signals are improved.

Description

Technical Field

[0001] The present application relates to the field of exoskeleton technology, and in particular to a method, device, computer equipment, storage medium and computer program product for estimating joint torque based on electromyographic signals of a lower limb exoskeleton. Background Art

[0002] With the widespread use of exoskeleton devices, the demand for control accuracy of exoskeleton devices is getting higher and higher. Usually, the target object's electromyographic signal can be collected, and the target object's movement intention can be estimated based on the electromyographic signal, and then the target object's joint torque can be estimated.

[0003] Current technologies for estimating joint torque based on electromyographic signals often rely on empirical formulas or preset gait patterns for control, which cannot achieve precise control of specific individuals and have the problem of low accuracy in joint torque estimation. Summary of the invention

[0004] Based on this, it is necessary to provide a method, device, computer equipment, storage medium and computer program product for estimating joint torque based on electromyographic signals of a lower limb exoskeleton to address the above technical problems.

[0005] In a first aspect, the present application provides a method for estimating joint torque based on electromyographic signals of a lower limb exoskeleton. The method comprises:

[0006] Obtaining an initial filtered electromyographic signal of the target object;

[0007] According to the initial filtered electromyographic signal, obtaining an electromyographic signal envelope corresponding to the initial filtered electromyographic signal;

[0008] Normalizing the electromyographic signal envelope to obtain an electromyographic signal value corresponding to the electromyographic signal envelope;

[0009] According to the electromyographic signal value, an independent source signal corresponding to the initial filtered electromyographic signal is obtained;

[0010] A joint torque value of a target exoskeleton matching the target object is obtained according to the independent source signal.

[0011] In one of the embodiments, obtaining the independent source signal corresponding to the initial filtered electromyographic signal according to the electromyographic signal value includes: obtaining an unmixing matrix for the electromyographic signal value; and obtaining the independent source signal corresponding to the initial filtered electromyographic signal based on the electromyographic signal value using the unmixing matrix.

[0012] In one embodiment, the method of obtaining the initial filtered electromyographic signal of the target object includes: obtaining electromyographic signal acquisition data of the target object; using a Butterworth bandpass filter to perform a primary filtering on the electromyographic signal acquisition data to obtain a primary filtered electromyographic signal corresponding to the electromyographic signal acquisition data; using a notch filter to perform a secondary filtering on the primary filtered electromyographic signal to obtain the initial filtered electromyographic signal corresponding to the electromyographic signal acquisition data.

[0013] In one embodiment, the electromyographic signal acquisition data is filtered once using a Butterworth bandpass filter to obtain a filtered electromyographic signal corresponding to the electromyographic signal acquisition data, including: determining the cutoff frequency of the Butterworth bandpass filter according to the electromyographic frequency range of the electromyographic signal contained in the electromyographic signal acquisition data; inputting the electromyographic signal acquisition data into the Butterworth bandpass filter, and the Butterworth bandpass filter filters the electromyographic signal acquisition data based on the cutoff frequency to determine and output the filtered electromyographic signal.

[0014] In one embodiment, the use of a notch filter to perform secondary filtering on the once-filtered electromyographic signal to obtain the initial filtered electromyographic signal corresponding to the electromyographic signal acquisition data includes: determining the notch frequency of the notch filter based on the signal frequency range of the electrical interference signal of the electromyographic signal; inputting the once-filtered electromyographic signal into the notch filter to determine and output the initial filtered electromyographic signal; wherein the notch filter, based on the notch frequency, filters out the electrical interference signal in the once-filtered electromyographic signal to obtain the initial filtered electromyographic signal.

[0015] In one of the embodiments, obtaining the electromyographic signal envelope corresponding to the initial filtered electromyographic signal based on the initial filtered electromyographic signal includes: rectifying the initial filtered electromyographic signal to obtain the absolute value electromyographic signal of the initial filtered electromyographic signal; obtaining the electromyographic signal envelope based on the absolute value electromyographic signal.

[0016] In a second aspect, the present application provides a joint torque estimation device based on electromyographic signals of a lower limb exoskeleton. The device comprises:

[0017] A signal acquisition module, used to acquire an initial filtered electromyographic signal of a target object;

[0018] A signal envelope acquisition module, used for obtaining the electromyographic signal envelope corresponding to the initial filtered electromyographic signal according to the initial filtered electromyographic signal;

[0019] A mapping module, used for normalizing the electromyographic signal envelope to obtain an electromyographic signal value corresponding to the electromyographic signal envelope;

[0020] An independent analysis module, used for obtaining an independent source signal corresponding to the initial filtered electromyographic signal according to the electromyographic signal value;

[0021] The torque estimation module is used to obtain the joint torque value of the target exoskeleton matching the target object according to the independent source signal.

[0022] In a third aspect, the present application further provides a computer device. The computer device includes a memory and a processor, the memory stores a computer program, and the processor implements the following steps when executing the computer program:

[0023] Obtaining an initial filtered electromyographic signal of the target object;

[0024] According to the initial filtered electromyographic signal, obtaining an electromyographic signal envelope corresponding to the initial filtered electromyographic signal;

[0025] Normalizing the electromyographic signal envelope to obtain an electromyographic signal value corresponding to the electromyographic signal envelope;

[0026] According to the electromyographic signal value, an independent source signal corresponding to the initial filtered electromyographic signal is obtained;

[0027] A joint torque value of a target exoskeleton matching the target object is obtained according to the independent source signal.

[0028] In a fourth aspect, the present application further provides a computer-readable storage medium. The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the following steps are implemented:

[0029] Obtaining an initial filtered electromyographic signal of the target object;

[0030] According to the initial filtered electromyographic signal, obtaining an electromyographic signal envelope corresponding to the initial filtered electromyographic signal;

[0031] Normalizing the electromyographic signal envelope to obtain an electromyographic signal value corresponding to the electromyographic signal envelope;

[0032] According to the electromyographic signal value, an independent source signal corresponding to the initial filtered electromyographic signal is obtained;

[0033] A joint torque value of a target exoskeleton matching the target object is obtained according to the independent source signal.

[0034] In a fifth aspect, the present application further provides a computer program product. The computer program product includes a computer program, and when the computer program is executed by a processor, the following steps are implemented:

[0035] Obtaining an initial filtered electromyographic signal of the target object;

[0036] According to the initial filtered electromyographic signal, obtaining an electromyographic signal envelope corresponding to the initial filtered electromyographic signal;

[0037] Normalizing the electromyographic signal envelope to obtain an electromyographic signal value corresponding to the electromyographic signal envelope;

[0038] According to the electromyographic signal value, an independent source signal corresponding to the initial filtered electromyographic signal is obtained;

[0039] A joint torque value of a target exoskeleton matching the target object is obtained according to the independent source signal.

[0040] In the above-mentioned joint torque estimation method, device, computer equipment, storage medium and computer program product based on the electromyographic signal of the lower limb exoskeleton, the initial filtered electromyographic signal of the target object is obtained; based on the initial filtered electromyographic signal, the electromyographic signal envelope corresponding to the initial filtered electromyographic signal is obtained; the electromyographic signal envelope is normalized to obtain the electromyographic signal value corresponding to the electromyographic signal envelope; based on the electromyographic signal value, the independent source signal corresponding to the initial filtered electromyographic signal is obtained; based on the independent source signal, the joint torque value of the target exoskeleton matching the target object is obtained. In the method provided in the embodiment of the present application, the electromyographic signal acquisition data of the target object can be collected in real time, and the electromyographic signal acquisition data can be amplified and processed, and multiple filtering processes can be performed to improve the accuracy of the electromyographic signal, and further, the accuracy of the joint torque estimation based on the electromyographic signal can be improved, and the independent component analysis technology is introduced, the electromyographic signal can be deeply analyzed, and the accuracy of the recognition of the target object's movement intention can be improved, and further, the accuracy and efficiency of the joint torque estimation based on the electromyographic signal can be improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] In order to more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the drawings required for use in the embodiments of the present application or related technical descriptions will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0042] Figure 1 A schematic flow chart of a joint torque estimation method based on electromyographic signals of a lower limb exoskeleton provided in an embodiment of the present application;

[0043] Figure 2 A schematic diagram of a process for obtaining an independent source signal provided in an embodiment of the present application;

[0044] Figure 3A schematic diagram of a process for obtaining an initial filtered electromyographic signal provided in an embodiment of the present application;

[0045] Figure 4 A schematic diagram of a process of filtering using a Butterworth bandpass filter provided in an embodiment of the present application;

[0046] Figure 5 A schematic diagram of a process of filtering using a notch filter provided in an embodiment of the present application;

[0047] Figure 6 A schematic flow chart of another joint torque estimation method based on electromyographic signals of a lower limb exoskeleton provided in an embodiment of the present application;

[0048] Figure 7 A structural block diagram of a joint torque estimation device based on electromyographic signals of a lower limb exoskeleton provided in an embodiment of the present application;

[0049] Figure 8 An internal structure diagram of a computer device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0050] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0051] In an exemplary embodiment, Figure 1 As shown, a joint torque estimation method based on electromyographic signals of a lower limb exoskeleton is provided. This embodiment uses the method applied to a server as an example. It can be understood that the method can also be applied to a terminal, and can also be applied to a system including a terminal and a server, and is implemented through the interaction between the terminal and the server. In this embodiment, the method includes the following steps:

[0052] Step 102: Acquire the initial filtered electromyographic signal of the target object.

[0053] The target object may be any object equipped with a target exoskeleton, for example, a worker equipped with a lower limb exoskeleton. The target exoskeleton may be a human exoskeleton and / or a powered exoskeleton, for example, a lower limb exoskeleton. The human exoskeleton and / or powered exoskeleton may be a machine device that uses mechanical devices to enhance the structural strength, motor strength and endurance of human limbs. The working principle is to detect the movement intention and state of the human body through sensors, and then the power unit provides energy, and transmits the power to the execution unit through the transmission device, thereby assisting or enhancing the movement of the human body. The joint torque estimation method based on the electromyographic signal of the lower limb exoskeleton provided in the embodiment of the present application can be applied to the exoskeleton system of the target exoskeleton. The exoskeleton system may include an electromyographic sensor. The working principle of the electromyographic sensor is the principle of bioelectric conduction. Muscles generate bioelectric signals during contraction and relaxation. The electromyographic sensor can convert the sensed bioelectric signals into measurable electrical signals by contacting the skin surface. The electromyographic sensor has the characteristics of biocompatibility and high signal-to-noise ratio. The electromyographic sensor can be directly installed on the muscle belly or the largest cross-section. The signal strength of the electromyographic signal at the muscle belly or the largest cross-section is relatively high, which is conducive to improving the accuracy of electromyographic signal acquisition. In a possible implementation method, the electromyographic sensor can be installed on the most active muscle groups around joints such as legs and arms, so as to effectively obtain electromyographic signals reflecting the muscle activity state, and provide accurate data for subsequent motion control of the target exoskeleton. Next, the electromyographic sensor can be used to collect electromyographic signals of the target object to obtain electromyographic signal collection data of the target object in the current collection cycle, and then the electromyographic signal collection data can be processed to obtain the initial filtered electromyographic signal, and the data processing includes but is not limited to signal amplification and filtering processing. The initial filtered electromyographic signal can be an electromyographic signal that has been amplified and filtered.

[0054] Step 104: obtaining an electromyographic signal envelope corresponding to the initial filtered electromyographic signal according to the initial filtered electromyographic signal.

[0055] Wherein, in this step, the initial filtered electromyographic signal can be subjected to full-wave rectification and low-pass filtering operations in sequence to obtain the electromyographic signal envelope corresponding to the initial filtered electromyographic signal. The electromyographic signal envelope can intuitively reflect the muscle activity state. The full-wave rectification can reverse all negative signals in the initial filtered electromyographic signal to positive signals, thereby obtaining the absolute value electromyographic signal of the initial filtered electromyographic signal, and the absolute value electromyographic signal can reflect the contraction strength of the muscle. Furthermore, the low-pass filtering can be used to filter out the high-frequency part of the absolute value electromyographic signal, and the low-frequency part can be better retained, thereby obtaining a smoothed estimate of the signal amplitude of the absolute value electromyographic signal, and obtaining the electromyographic signal envelope corresponding to the initial filtered electromyographic signal.

[0056] Step 106, normalizing the electromyographic signal envelope to obtain an electromyographic signal value corresponding to the electromyographic signal envelope.

[0057] In this step, a normalization process can be used to construct a mapping relationship between the target object's muscle force and the target object's electromyographic signal. The normalization process can include linear normalization and nonlinear normalization. Usually, there is a nonlinear relationship between the muscle force and the electromyographic signal envelope. In the first step, the electromyographic signal envelope can be linearly normalized to map the electromyographic signal envelope to a corresponding linear electromyographic signal mapping value. For example, the electromyographic signal mapping value can range from 0 to 100. The linear normalization process can refer to formula (1):

[0058]

[0059] in, It can be the EMG signal envelope; It can be the maximum signal value of the electromyographic signal envelope; It can be the minimum signal value of the electromyographic signal envelope; It can be the linear electromyographic signal mapping value corresponding to the electromyographic signal envelope.

[0060] In the first step, the linear electromyographic signal mapping value can be subjected to nonlinear normalization processing to obtain the electromyographic signal value corresponding to the electromyographic signal envelope. The nonlinear normalization processing can be referred to in formula (2):

[0061]

[0062] in, The curvature exponent constant of the predefined nonlinear curve; It may be a linear electromyographic signal mapping value corresponding to the electromyographic signal envelope; It may be the electromyographic signal value corresponding to the electromyographic signal envelope.

[0063] Step 108: Obtain an independent source signal corresponding to the initial filtered electromyographic signal according to the electromyographic signal value.

[0064] In this step, the independent component analysis technique can be used to obtain the independent source signal corresponding to the initial filtered electromyographic signal according to the electromyographic signal value. Independent component analysis is a blind source separation technique that can recover independent source signals from multiple mixed signals in the case of an unknown mixing process. For example, in the electromyographic signal joint torque estimation scenario of a weight-bearing lower limb exoskeleton, the independent component analysis can be used to extract multiple independent independent source signals from the electromyographic signal value containing multiple mixed signals. The electromyographic signal value may include the observation signal value of multiple observation signals, and the independent source signal may be a physiological signal such as a neuron discharge event behind the electromyographic signal generation. The relationship between the observation signal and the independent source signal can be seen in equation (3):

[0065]

[0066] in, It can be the observed signal value of multiple observed signals; It can be the independent source signal value of multiple independent source signals; A is an N×N mixing matrix. Then, the inverse matrix of A can be used to unmix the matrix, and the independent source signal corresponding to the initial filtered electromyographic signal can be obtained according to the observation signal values ​​of multiple observation signals contained in the electromyographic signal value.

[0067] Step 110: Obtain joint torque values ​​of a target exoskeleton that matches the target object based on the independent source signal.

[0068] In this step, a mathematical model from independent source signals to joint torque can be established based on physiological and kinematic principles. Considering that muscle contraction force is the direct cause of joint torque, according to the mechanical properties of muscles and the structural characteristics of joints, factors such as muscle force-length relationship, force-velocity relationship and joint geometric parameters are included in the model. For example, by analyzing the force generated by muscles under different contraction states, combined with parameters such as the moment of inertia of joints, a quantitative relationship between torque and muscle contraction force is established. The specific form of the model may be more complex, may contain multiple parameters and variables, and need to be fitted and optimized through experimental data. Then, the separated independent source signal can be input into the constructed torque estimation model. According to the algorithm and formula of the model, combined with relevant physiological and kinematic parameters, calculation is performed. For example, if the model involves a linear relationship between muscle contraction force and independent source signal, and a relationship between torque and muscle contraction force and joint moment of inertia, the estimated value of joint torque is calculated by substituting the corresponding parameter values ​​and independent source signal data. During the calculation process, it is necessary to ensure the accuracy of the parameters and the effectiveness of the model, and the model may need to be continuously verified and adjusted to improve the accuracy of torque estimation.

[0069] In addition, in a possible implementation, a pre-built exoskeleton motion control model can be used to obtain the desired angular velocity of the joint of the target exoskeleton based on the joint torque value of the target exoskeleton. The relationship between the joint torque value and the desired angular velocity can be seen in equation (4):

[0070]

[0071] in, is the moment of inertia of the joint; is the static friction torque of the joint.

[0072] Furthermore, a proportional-integral-derivative controller (PID) can be used to calculate the desired angular velocity and the current angular velocity of the target exoskeleton. and the current angle , get the control signal of the target exoskeleton , see formula (5):

[0073]

[0074] in, and are the desired joint angles and angular velocities; , and They are the proportional, integral and differential coefficients of the PID controller respectively; It can be a control signal of the target exoskeleton. Then, the control signal is sent to the driver of the target exoskeleton to generate a corresponding torque to drive the movement of the target exoskeleton.

[0075] Next, during the movement of the target exoskeleton, the motion data of the target object, such as joint angle, angular velocity and electromyographic signal, etc., can be collected and analyzed in real time, and the estimated joint torque value can be adjusted based on the motion data feedback.

[0076] In the method of this embodiment, the initial filtered electromyographic signal of the target object is obtained; based on the initial filtered electromyographic signal, the electromyographic signal envelope corresponding to the initial filtered electromyographic signal is obtained; the electromyographic signal envelope is normalized to obtain the electromyographic signal value corresponding to the electromyographic signal envelope; based on the electromyographic signal value, the independent source signal corresponding to the initial filtered electromyographic signal is obtained; based on the independent source signal, the joint torque value of the target exoskeleton matching the target object is obtained. In the method provided by the embodiment of the present application, the electromyographic signal acquisition data of the target object can be collected in real time, and the electromyographic signal acquisition data can be amplified and processed, and multiple filtering processes can be performed to improve the accuracy of the electromyographic signal, and further, the accuracy of the joint torque estimation based on the electromyographic signal can be improved, and the independent component analysis technology is introduced to deeply analyze the electromyographic signal, improve the accuracy of the recognition of the target object's movement intention, and further, improve the accuracy and efficiency of the joint torque estimation based on the electromyographic signal.

[0077] In an exemplary embodiment, Figure 2 As shown, step 108 may include steps 202 to 204. Among them:

[0078] Step 202: Obtain an unmixing matrix for the electromyographic signal value.

[0079] Among them, in the process of obtaining the independent source signal corresponding to the initial filtered electromyographic signal according to the electromyographic signal value using the independent component analysis technology, its goal is to iteratively optimize the unmixing matrix to obtain a converged unmixing matrix. First, a criterion for measuring signal independence can be determined, which can be called an independence measurement criterion, such as negative entropy maximization and maximum likelihood estimation. Taking negative entropy maximization as an example, negative entropy is an indicator to measure the degree of difference between a random variable and a Gaussian distribution. The negative entropy value of the independent source signal is usually large. By maximizing the negative entropy of the estimated source signal, the separated estimated signal can be made as close to the independent source signal as possible; next, the unmixing matrix can be initialized, for example, a random initialization method can be used; further, the unmixing matrix can be iteratively optimized using the independence measurement criterion. If the change of the unmixing matrix after multiple iterations is less than the preset change threshold, it indicates that the unmixing matrix has converged, and the converged unmixing matrix can be used to obtain the independent source signal corresponding to the initial filtered electromyographic signal based on the electromyographic signal value.

[0080] Step 204: using the unmixing matrix, based on the electromyographic signal value, obtain an independent source signal corresponding to the initial filtered electromyographic signal.

[0081] Among them, the observation signal values ​​of multiple observation signals contained in the electromyographic signal value are used to obtain the independent source signal corresponding to the initial filtered electromyographic signal, which can be seen from formula (6):

[0082]

[0083] in, Can be The inverse matrix of is the unmixing matrix of the independent component analysis process; It can be the observed signal value of multiple observed signals;

[0084] It can be an independent source signal value of multiple independent source signals; Can be independent source signal values ​​of independent source signals.

[0085] In the method of this embodiment, independent component analysis technology is introduced, which can deeply analyze the electromyographic signal and improve the accuracy of identifying the target object's movement intention, thereby improving the accuracy and efficiency of joint torque estimation based on the electromyographic signal.

[0086] In an exemplary embodiment, Figure 3 As shown, step 102 may include steps 302 to 306. Among them:

[0087] Step 302: Acquire electromyographic signal data of the target object.

[0088] The myoelectric sensor of the exoskeleton system of the target exoskeleton can be used to collect the myoelectric signal of the target object to obtain the myoelectric signal collection data of the target object in the current collection cycle. The myoelectric signal collection data includes the myoelectric signal of the target object, the low-frequency interference signal and high-frequency invalid signal generated by the electrode and muscle movement, and the alternating current signal.

[0089] Step 304: Use a Butterworth bandpass filter to filter the electromyographic acquisition signal data once to obtain a primary filtered electromyographic signal corresponding to the electromyographic acquisition signal data.

[0090] Among them, the low-frequency interference signals and high-frequency invalid signals generated by the electrodes and muscle movements contained in the electromyographic acquisition signal data, as well as the alternating current signals, can be filtered out to obtain the electromyographic signal of the target object. In this step, the electromyographic acquisition signal data can be filtered once using a Butterworth bandpass filter to obtain a filtered electromyographic signal corresponding to the electromyographic acquisition signal data. The primary filtering can filter out the low-frequency interference signals and high-frequency invalid signals generated by the electrodes and muscle movements contained in the electromyographic acquisition signal data. The obtained filtered electromyographic signal contains the electromyographic signal of the target object and the alternating current signal. The Butterworth filter is a filter with a smooth response characteristic within the passband, which can filter out the low-frequency interference signals and high-frequency invalid signals generated by the electrodes and muscle movements contained in the electromyographic acquisition signal data, as shown in formula (7):

[0091]

[0092] in, is the frequency response of the filter, is the angular frequency, is the cut-off frequency, is the filter order.

[0093] Step 306: perform secondary filtering on the primary filtered electromyographic signal using a notch filter to obtain an initial filtered electromyographic signal corresponding to the electromyographic acquisition signal data.

[0094] In this step, the primary filtering can filter out the AC signal contained in the electromyographic acquisition signal data, and the obtained initial filtered electromyographic signal can be the electromyographic signal used to estimate the joint torque of the target object. The notch filter is a filter that can accurately remove specific frequency signals and can be used to filter out power frequency interference caused by 50Hz AC. The transfer function of the notch filter can be seen in formula (8):

[0095]

[0096] in, is a complex variable, is the notch frequency, For bandwidth, set is 2π, performing a 50Hz notch filter.

[0097] In the method of this embodiment, the electromyographic signal acquisition data of the target object can be collected in real time, and the electromyographic signal acquisition data can be amplified and subjected to multiple filtering processes, which can improve the accuracy of the electromyographic signal, and further improve the accuracy of joint torque estimation based on the electromyographic signal.

[0098] In an exemplary embodiment, Figure 4 As shown, step 304 may include steps 402 to 404. Among them:

[0099] Step 402: Determine the cutoff frequency of the Butterworth bandpass filter according to the electromyographic frequency range of the electromyographic signal contained in the electromyographic acquisition signal data.

[0100] Among them, the Butterworth bandpass filter has a smooth response characteristic within the passband, and its frequency response With angular frequency , Cutoff frequency and filter order Related. The Butterworth bandpass filter can allow signals in a specific frequency range to pass, that is, the electromyographic signal, while effectively attenuating signals outside the range, that is, low-frequency interference signals and high-frequency invalid signals generated by electrodes and muscle movements, so as to achieve the purpose of filtering signals. In this application scenario, its main function is to remove interference signals generated by electrodes and muscle movements. In this step, when the Butterworth bandpass filter is used for filtering, the passband frequency of the bandpass filter can be set according to the frequency range of the electromyographic signal, and the passband frequency can be a cutoff frequency, which includes an upper cutoff frequency and a lower cutoff frequency. In general, electromyographic signals are mainly concentrated in 10Hz to 500Hz, that is, the electromyographic frequency range of electromyographic signals is 10Hz to 500Hz. Therefore, the lower cutoff frequency of the Butterworth bandpass filter is set to 10Hz, and the upper cutoff frequency is set to 500Hz. Such a setting enables the filter to only allow signals of 10Hz-500Hz to pass, while low-frequency interference signals below 10Hz and high-frequency invalid signals above 500Hz will be attenuated.

[0101] Step 404: input the EMG signal data into a Butterworth bandpass filter. The Butterworth bandpass filter performs filtering processing on the EMG signal data based on a cutoff frequency, and determines and outputs a primary filtered EMG signal.

[0102] Among them, the collected electromyographic acquisition signal data can be input into a Butterworth bandpass filter with set parameters (lower cutoff frequency 10Hz, upper cutoff frequency 500Hz). The Butterworth bandpass filter processes the input electromyographic acquisition signal data according to its characteristics and set parameters. During the processing, low-frequency interference below 10Hz, such as slow-changing signals generated by electrode movement, and high-frequency invalid signals above 500Hz, such as stray high-frequency signals generated by electronic equipment, will be effectively attenuated. After the filtering operation, the output signal is a relatively pure primary filtered electromyographic signal that removes low-frequency interference and high-frequency invalid signals. The primary filtered electromyographic signal can more truly reflect the electrical activity of the muscle and provide a high-quality data basis for subsequent signal processing and joint torque estimation.

[0103] In the method of this embodiment, a Butterworth bandpass filter can be used to perform a filtering process on the electromyographic acquisition signal data based on the cutoff frequency, and a filtered electromyographic signal can be determined and output, which can improve the accuracy of the electromyographic signal, and further improve the accuracy of joint torque estimation based on the electromyographic signal.

[0104] In an exemplary embodiment, Figure 5 As shown, step 306 may include steps 502 to 504. Among them:

[0105] Step 502: Determine the notch frequency of the notch filter according to the signal frequency range of the electrical interference signal of the electromyographic signal.

[0106] Among them, the notch filter can be used to accurately remove specific frequency signals. Since the electromyographic signal is mainly interfered by 50Hz AC, the notch frequency of the notch filter is set to 50Hz. This setting ensures that the filter only works on the 50Hz power frequency interference signal and will not affect the useful signals of other frequencies.

[0107] Step 504: input the first filtered electromyographic signal into a notch filter, and determine and output an initial filtered electromyographic signal.

[0108] Among them, the notch filter, based on the notch frequency, filters out the electrical interference signal in the first-filtered electromyographic signal to obtain the initial filtered electromyographic signal. The first-filtered electromyographic signal is input into the notch filter with the set notch frequency (50Hz). The notch filter processes the input first-filtered electromyographic signal according to its characteristics and set parameters, and effectively attenuates the 50Hz power frequency interference signal therein. After this step, the output signal is the electromyographic signal with the 50Hz power frequency interference removed. Next, although the 50Hz power frequency interference is removed, there may still be other random noises in the signal. At this time, a sliding average filter or other methods can be used for further processing. The sliding average filter smoothes the data and performs point-by-point average calculation on the signal by setting a suitable window size. The data in the window is averaged and output, which can effectively remove the random noise, leaving a relatively stable and effective electromyographic signal containing only random signals, that is, the initial filtered electromyographic signal. The initial filtered electromyographic signal reflects the muscle contraction state of the target object to a certain extent, and provides more accurate data for subsequent steps such as signal envelope extraction, normalization and torque estimation.

[0109] In the method of this embodiment, a notch filter can be used to perform secondary filtering on the primary filtered electromyographic signal based on the notch frequency, and the initial filtered electromyographic signal can be determined and output, which can improve the accuracy of electromyographic signal acquisition and, further, improve the accuracy of joint torque estimation based on the electromyographic signal.

[0110] In an exemplary embodiment, step 104 may include:

[0111] The initial filtered electromyographic signal is rectified to obtain an absolute value electromyographic signal of the initial filtered electromyographic signal; and the electromyographic signal envelope is obtained according to the absolute value electromyographic signal.

[0112] Among them, in this step, the initial filtered electromyographic signal can be subjected to full-wave rectification and low-pass filtering operations in sequence to obtain the electromyographic signal envelope corresponding to the initial filtered electromyographic signal. The electromyographic signal envelope can intuitively reflect the muscle activity state. The full-wave rectification can invert all negative signals in the initial filtered electromyographic signal to positive signals, thereby obtaining the absolute value electromyographic signal of the initial filtered electromyographic signal, which can be seen in formula (9). The absolute value electromyographic signal can reflect the contraction strength of the muscle. Furthermore, the low-pass filter, for example, a first-order Butterworth low-pass filter, can be used to filter out the high-frequency part of the absolute value electromyographic signal and better retain the low-frequency part, thereby obtaining a smooth estimate of the signal amplitude of the absolute value electromyographic signal and obtaining the electromyographic signal envelope corresponding to the initial filtered electromyographic signal.

[0113]

[0114] in, It can be the initial filtered EMG signal; It can be an absolute value electromyographic signal.

[0115] In the method of this embodiment, the initial filtered electromyographic signal can be subjected to full-wave rectification and low-pass filtering operations in sequence to obtain the electromyographic signal envelope corresponding to the initial filtered electromyographic signal, which can more intuitively reflect the muscle activity state and improve the accuracy of muscle activity recognition.

[0116] In another exemplary embodiment, Figure 6 As shown, a method for estimating joint torque based on electromyographic signals of a lower limb exoskeleton is provided, and the method may include:

[0117] Step 602: collect the electromyographic signal of the target object using an electromyographic signal sensor to obtain initial electromyographic signal collection data.

[0118] Step 604: using an amplifier to amplify the collected initial electromyographic signal data to obtain electromyographic signal data.

[0119] Step 606: filter out low-frequency interference signals and high-frequency invalid signals generated by the electrode and muscle movement through a Butterworth bandpass filter to obtain a primary filtered electromyographic signal.

[0120] Step 608: Use notch filtering to remove power frequency interference and random noise caused by the 50 Hz alternating current.

[0121] Step 610 , full-wave rectification and a first-order Butterworth low-pass filter are used to perform processing, thereby obtaining an electromyographic signal envelope of the surface electromyographic signal that can reflect changes in muscle force amplitude.

[0122] Among them, full-wave rectification: full-wave rectification is to invert all negative signals to positive values, so as to obtain the absolute value of the signal. The rectified electromyographic signal can better reflect the contraction strength of the muscle, as shown in formula (9); low-pass filtering (envelope detection): low-pass filtering can delete the high-frequency part of the signal, better retain the low-frequency part, and obtain a smooth estimate of the signal amplitude. This filter must be able to effectively retain the fundamental frequency information of the signal and filter out as many sharp high-frequency fluctuations as possible. In this way, after full-wave rectification and low-pass filtering, the output is the envelope of the electromyographic signal, which roughly reflects the changes in the degree of activation and muscle strength. The full-wave rectified electromyographic signal is input into a first-order Butterworth low-pass filter, which is filtered and the envelope of the electromyographic signal is obtained. After obtaining the signal envelope, this envelope signal can be used to reflect the degree of muscle activity. For example, when the muscle is in a tense state, the signal strength is large and the amplitude of the signal envelope is also large; conversely, when the muscle is in a relaxed state, the signal strength is small and the amplitude of the signal envelope is also small. Therefore, this signal envelope can be used as an indicator to judge the muscle contraction state and thus to estimate the joint torque.

[0123] Step 612: perform linear normalization and nonlinear normalization on the electromyographic signal to obtain an electromyographic signal value corresponding to the electromyographic signal envelope.

[0124] In order to establish the mapping relationship between the muscle force and the surface electromyographic signal envelope, the electromyographic signal needs to be linearly normalized and nonlinearly normalized.

[0125] Since the relationship between muscle force and the envelope of the surface electromyographic signal is relatively nonlinear, the surface electromyographic signal is first mapped to the range of 0-100 through linear normalization. Then, according to the nonlinear relationship between muscle force and the surface electromyographic signal, equation (2) is introduced to obtain the electromyographic signal value corresponding to the electromyographic signal envelope.

[0126] Step 614 , independent component analysis is performed to obtain deeper information from the electromyographic signal, and obtain the joint torque value of the target exoskeleton that matches the target object.

[0127] Among them, independent component analysis (ICA) is used to obtain deeper information from the electromyographic signal, such as neuronal discharge events, and then perform higher-precision torque estimation to generate more accurate joint torque estimates. ICA is a powerful and widely used blind source separation technique. It allows independent source signals to be recovered from mixtures and noise without any prior knowledge about the mixing process. This is very useful in estimating complex physiological signals (such as electromyographic signals), the generation of which may be affected by many factors. It can handle many types of problems and is applicable to many situations. Compared with sparse coding, ICA has a deeper theoretical foundation. ICA's ability to recover independent source signals from multiple mixed signals is irreplaceable in many applications, which can play an important role in the problem of torque estimation.

[0128] Step 616, based on the exoskeleton motion control model driven by the electromyographic signal, the calculated torque value is directly input into the control logic of the exoskeleton bone marrow as a control parameter, and the obtained torque value is used to calculate the expected angular velocity of the joint.

[0129] The relationship between the joint torque value and the desired angular velocity can be seen in equation (10):

[0130]

[0131] in, is the moment of inertia of the joint; is the angular velocity of the joint; is the damping coefficient; is the static friction torque of the joint.

[0132] External bone marrow reappearance of joint damping and joint static friction torque The effect on the joints can be ignored, so equation (10) is simplified to equation (4).

[0133] Step 618, based on the calculated desired joint angular acceleration and the current angular velocity and angle, the PID control strategy is used to calculate the control signal required to control the exoskeleton actuator.

[0134] Step 620, collect and analyze the user's motion data in real time, such as joint angle, angular velocity, electromyographic signal, etc., and use feedback to adjust the estimated torque value.

[0135] In the method of this embodiment, the initial filtered electromyographic signal of the target object is obtained; based on the initial filtered electromyographic signal, the electromyographic signal envelope corresponding to the initial filtered electromyographic signal is obtained; the electromyographic signal envelope is normalized to obtain the electromyographic signal value corresponding to the electromyographic signal envelope; based on the electromyographic signal value, the independent source signal corresponding to the initial filtered electromyographic signal is obtained; based on the independent source signal, the joint torque value of the target exoskeleton matching the target object is obtained. In the method provided by the embodiment of the present application, the electromyographic signal acquisition data of the target object can be collected in real time, and the electromyographic signal acquisition data can be amplified and processed, and multiple filtering processes can be performed to improve the accuracy of the electromyographic signal, and further, the accuracy of the joint torque estimation based on the electromyographic signal can be improved, and the independent component analysis technology is introduced to deeply analyze the electromyographic signal, improve the accuracy of the recognition of the target object's movement intention, and further, improve the accuracy and efficiency of the joint torque estimation based on the electromyographic signal.

[0136] It should be understood that, although the various steps in the flowcharts involved in the above-mentioned embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-mentioned embodiments can include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.

[0137] Based on the same inventive concept, the embodiment of the present application also provides a joint torque estimation device based on electromyographic signals of a lower limb exoskeleton for implementing the above-mentioned joint torque estimation method based on electromyographic signals of a lower limb exoskeleton. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme recorded in the above-mentioned method, so the specific limitations in the embodiments of one or more joint torque estimation devices based on electromyographic signals of a lower limb exoskeleton provided below can refer to the limitations of the joint torque estimation method based on electromyographic signals of a lower limb exoskeleton above, and will not be repeated here.

[0138] In one embodiment, Figure 7 As shown, a joint torque estimation device based on the electromyographic signal of a lower limb exoskeleton is provided, comprising: a signal acquisition module 702, a signal envelope acquisition module 704, a mapping module 706, an independent analysis module 708 and a torque estimation module 710, wherein:

[0139] A signal acquisition module 702 is used to acquire an initial filtered electromyographic signal of a target object;

[0140] A signal envelope acquisition module 704 is used to obtain the electromyographic signal envelope corresponding to the initial filtered electromyographic signal according to the initial filtered electromyographic signal;

[0141] A mapping module 706 is used to perform normalization processing on the electromyographic signal envelope to obtain an electromyographic signal value corresponding to the electromyographic signal envelope;

[0142] An independent analysis module 708, used to obtain an independent source signal corresponding to the initial filtered electromyographic signal according to the electromyographic signal value;

[0143] The torque estimation module 710 is used to obtain the joint torque value of the target exoskeleton matching the target object according to the independent source signal.

[0144] In one embodiment, the independent analysis module 708 is further used to: obtain an unmixing matrix for the electromyographic signal value; and obtain the independent source signal corresponding to the initial filtered electromyographic signal based on the electromyographic signal value by using the unmixing matrix.

[0145] In one embodiment, the signal acquisition module 702 is also used to: acquire the electromyography acquisition signal data of the target object; use a Butterworth bandpass filter to filter the electromyography acquisition signal data once to obtain a first filtered electromyography signal corresponding to the electromyography acquisition signal data; use a notch filter to filter the first filtered electromyography signal twice to obtain the initial filtered electromyography signal corresponding to the electromyography acquisition signal data.

[0146] In one embodiment, the signal acquisition module 702 is also used to: determine the cutoff frequency of the Butterworth bandpass filter according to the electromyographic frequency range of the electromyographic signal contained in the electromyographic acquisition signal data; input the electromyographic acquisition signal data into the Butterworth bandpass filter, and the Butterworth bandpass filter filters the electromyographic acquisition signal data based on the cutoff frequency to determine and output the first filtered electromyographic signal.

[0147] In one embodiment, the signal acquisition module 702 is also used to: determine the notch frequency of the notch filter according to the signal frequency range of the electrical interference signal of the electromyographic signal; input the once filtered electromyographic signal into the notch filter, determine and output the initial filtered electromyographic signal; wherein the notch filter, based on the notch frequency, filters out the electrical interference signal in the once filtered electromyographic signal to obtain the initial filtered electromyographic signal.

[0148] In one embodiment, the signal envelope acquisition module 704 is further used to: perform rectification processing on the initial filtered electromyographic signal to obtain an absolute value electromyographic signal of the initial filtered electromyographic signal; and obtain the electromyographic signal envelope according to the absolute value electromyographic signal.

[0149] Each module in the joint torque estimation device based on the electromyographic signal of the lower limb exoskeleton can be implemented in whole or in part by software, hardware and a combination thereof. Each module can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory in the computer device in the form of software, so that the processor can call and execute the operations corresponding to each module.

[0150] In one embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as follows: Figure 8 As shown. The computer device includes a processor, a memory and a network interface connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store data related to joint torque estimation based on electromyographic signals of a lower limb exoskeleton. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a joint torque estimation method based on electromyographic signals of a lower limb exoskeleton is implemented.

[0151] Those skilled in the art will understand that Figure 8 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0152] In one embodiment, a computer device is further provided, including a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the steps in the above method embodiments when executing the computer program.

[0153] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps in the above-mentioned method embodiments are implemented.

[0154] In one embodiment, a computer program product is provided, including a computer program, which implements the steps in the above method embodiments when executed by a processor.

[0155] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties.

[0156] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited to this. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., but are not limited to this.

[0157] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0158] The above-described embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. A joint torque estimation method based on electromyographic signals of a lower limb exoskeleton, characterized in that: The method comprises: Obtaining an initial filtered electromyographic signal of a target object; According to the initial filtered electromyographic signal, obtaining an electromyographic signal envelope corresponding to the initial filtered electromyographic signal; Normalizing the electromyographic signal envelope to obtain an electromyographic signal value corresponding to the electromyographic signal envelope; According to the electromyographic signal value, an independent source signal corresponding to the initial filtered electromyographic signal is obtained; A joint torque value of a target exoskeleton matching the target object is obtained according to the independent source signal.

2. The method according to claim 1, characterized in that: The step of obtaining the independent source signal corresponding to the initial filtered electromyographic signal according to the electromyographic signal value includes: Obtaining an unmixing matrix for the electromyographic signal value; The independent source signal corresponding to the initial filtered electromyographic signal is obtained based on the electromyographic signal value by using an unmixing matrix.

3. The method according to claim 1, characterized in that The step of obtaining an initial filtered electromyographic signal of the target object comprises: Acquire electromyographic signal acquisition data of the target object; Using a Butterworth bandpass filter, the electromyographic signal acquisition data is filtered once to obtain a primary filtered electromyographic signal corresponding to the electromyographic signal acquisition data; The primary filtered electromyographic signal is secondary filtered by using a notch filter to obtain the initial filtered electromyographic signal corresponding to the electromyographic signal acquisition data.

4. The method according to claim 3, characterized in that The method of filtering the electromyographic signal acquisition data once using a Butterworth bandpass filter to obtain a filtered electromyographic signal corresponding to the electromyographic signal acquisition data includes: Determining the cutoff frequency of the Butterworth bandpass filter according to the electromyographic frequency range of the electromyographic signal contained in the electromyographic signal acquisition data; The electromyographic signal acquisition data is input into the Butterworth bandpass filter, and the Butterworth bandpass filter performs filtering processing on the electromyographic signal acquisition data based on the cutoff frequency, and determines and outputs the primary filtered electromyographic signal.

5. The method according to claim 3, characterized in that: The method of performing secondary filtering on the primary filtered electromyographic signal by using a notch filter to obtain the initial filtered electromyographic signal corresponding to the electromyographic signal acquisition data includes: Determining a notch frequency of the notch filter according to a signal frequency range of the electrical interference signal of the electromyographic signal; Inputting the primary filtered electromyographic signal into the notch filter, determining and outputting the initial filtered electromyographic signal; The notch filter filters out the electrical interference signal in the primary filtered electromyographic signal based on the notch frequency to obtain the initial filtered electromyographic signal.

6. The method according to claim 1, characterized in that The step of obtaining an electromyographic signal envelope corresponding to the initial filtered electromyographic signal according to the initial filtered electromyographic signal comprises: Performing rectification processing on the initial filtered electromyographic signal to obtain an absolute value electromyographic signal of the initial filtered electromyographic signal; The electromyographic signal envelope is obtained according to the absolute value electromyographic signal.

7. A joint torque estimation device based on electromyographic signals of a lower limb exoskeleton, characterized in that: The device comprises: A signal acquisition module, used to acquire an initial filtered electromyographic signal of a target object; A signal envelope acquisition module, used for obtaining the electromyographic signal envelope corresponding to the initial filtered electromyographic signal according to the initial filtered electromyographic signal; A mapping module, used for normalizing the electromyographic signal envelope to obtain an electromyographic signal value corresponding to the electromyographic signal envelope; An independent analysis module, used for obtaining an independent source signal corresponding to the initial filtered electromyographic signal according to the electromyographic signal value; The torque estimation module is used to obtain the joint torque value of the target exoskeleton matching the target object according to the independent source signal.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 6 are implemented.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images