Exoskeleton robot impedance adjustment and trajectory planning method, medium and equipment

Abstract

The invention provides an exoskeleton robot impedance adjustment and trajectory planning method, a medium and equipment, and relates to the technical field of exoskeleton robot control, and the method comprises the steps: obtaining a multi-dimensional original signal related to exoskeleton movement, and carrying out the preprocessing of the multi-dimensional original signal; the exoskeleton complex motion trail is decomposed into a basic motion component and a working condition adaptive component; adjusting the Fourier coefficient of the abnormal harmonic wave to suppress the abnormal harmonic wave, and obtaining an anti-interference basic motion component according to the adjusted Fourier coefficient; performing topographic feature matching and gain control on the working condition adaptive component to obtain a reconstructed working condition adaptive component; the anti-interference basic motion component and the reconstruction working condition adaptive component are recombined to obtain a final motion track; according to the error between the final motion track and the expected track of the exoskeleton, adjusting parameters of the impedance model in real time, so that the exoskeleton tracking expected track is optimized in real time. According to the method, the application scene of the exoskeleton can be widened, and the practicability of the exoskeleton in a complex environment is remarkably improved.

Description

Technical Field

[0001] The present invention relates to the technical field of exoskeleton robot control, and particularly relates to a method, medium and device for impedance adjustment and trajectory planning of an exoskeleton robot. Background Art

[0002] As a frontier product in the field of human-computer interaction, exoskeleton robots show broad application prospects in multiple fields. In the field of medical rehabilitation, exoskeletons can help limb-disabled people and rehabilitation patients perform autonomous movement training, accelerating the rehabilitation process; in industrial scenarios, they assist workers in carrying heavy objects, reducing labor intensity and improving work efficiency. However, the complex and changeable actual working conditions pose strict requirements on their trajectory planning and impedance adjustment technologies.

[0003] Currently, the existing exoskeleton trajectory planning algorithms have obvious deficiencies in dealing with complex terrains. Most algorithms adopt fixed trajectory models and are difficult to adjust in real time according to terrain changes. As a result, when the exoskeleton faces rough terrains or obstacles, its motion stability is poor, and even safety accidents such as falling may occur. In terms of impedance adjustment, traditional methods cannot make precise responses according to the real-time motion state of the exoskeleton and the external environment, resulting in a poor interaction experience between the exoskeleton and the user. Not only can it not effectively assist, but it may also cause additional burdens to the user. In addition, the existing algorithms often fail to fully consider the coordination between the motion trajectory and impedance adjustment, making it difficult to improve the overall performance of the exoskeleton under complex working conditions. Summary of the Invention

[0004] The purpose of the present invention is to provide a method, medium and device for impedance adjustment and trajectory planning of an exoskeleton robot, aiming to accurately plan the trajectory to adapt to complex terrains, adjust the impedance in real time, improve the human-computer interaction experience, and ensure the coordinated optimization of the motion trajectory and impedance adjustment, so as to achieve stable and efficient motion control of the exoskeleton robot under complex working conditions. The specific technical solutions are as follows:

[0005] A method for impedance adjustment and trajectory planning of an exoskeleton robot, the method comprising the following steps:

[0006] S100. Obtain multi-dimensional original signals related to the exoskeleton motion and preprocess them;

[0007] S200. Perform Fourier series transformation and wavelet packet transformation on the preprocessed multi-dimensional original signals respectively, and decompose the complex motion trajectory of the exoskeleton into a basic motion component reflecting periodic gait and a working condition adaptation component reflecting terrain features;

[0008] S300. Determine whether there are abnormal harmonics in the basic motion component. If there are abnormal harmonics, adjust the Fourier coefficients of the abnormal harmonics to suppress the abnormal harmonics, and obtain the anti-interference basic motion component according to the adjusted Fourier coefficients; perform terrain feature matching and gain control on the working condition adaptation component to obtain the reconstructed working condition adaptation component;

[0009] S400. Recombine the anti-interference basic motion component and the reconstructed working condition adaptation component to obtain the final motion trajectory;

[0010] S500. According to the error between the final motion trajectory of the exoskeleton and the desired trajectory, adjust the impedance model parameters in real time to enable the exoskeleton to track the desired trajectory and optimize it in real time.

[0011] Further, in the step S100, the multi-dimensional original signals include joint angles, joint torques, trunk accelerations, plantar pressure change rates, gait cycles, and terrain elevation data.

[0012] Further, in the step S100, the preprocessing includes performing normalization, low-pass filtering, and sliding window segmentation processing on the multi-dimensional original signals in sequence.

[0013] Further, the step S200 specifically includes the following steps:

[0014] S201. Extraction of the basic motion component: Receive the preprocessed joint angle signal and gait cycle, fit the joint angle signal with Fourier series, and extract the basic motion component;

[0015] S202. Extraction of the working condition adaptation component: Receive the preprocessed acceleration signal and terrain elevation data, perform wavelet packet decomposition on the acceleration signal, calculate the energy of each node, set an energy threshold, screen out the sub-band signals related to terrain disturbances, reconstruct the selected sub-band signals, and generate the working condition adaptation component;

[0016] S203. Verification of the decomposition effect: Calculate the residual energy ratio of the decomposed signal. If the residual energy ratio does not meet the requirements, increase the wavelet packet decomposition level and the extended Fourier harmonic order, and return to step S201 to perform decomposition and verification again until the residual energy ratio meets the requirements.

[0017] Further, in the step S300, to determine whether there are abnormal harmonics in the basic motion component, if there are abnormal harmonics, adjust the Fourier coefficients of the abnormal harmonics to suppress the abnormal harmonics, and obtain the anti-interference basic motion component according to the adjusted Fourier coefficients, specifically including:

[0018] S301. Detection of abnormal harmonics: Calculate the joint torque fluctuation amplitudes corresponding to each harmonic order, set a joint torque threshold, and determine whether there are abnormal harmonics according to the magnitudes of the joint torque fluctuation amplitudes and the joint torque threshold;

[0019] S302, abnormal harmonic suppression: introduce a suppression coefficient to dynamically attenuate the Fourier coefficients of the abnormal harmonics to obtain the interference-resistant basic motion component;

[0020] S303, characteristic verification: perform frequency domain energy comparison, calculate the energy ratio of the anti-interference basic motion component and the basic motion component, calculate the residual between the joint angle signal and the anti-interference basic motion component. If the energy ratio and residual do not meet the requirements, adjust the parameters and re-optimize until the requirements are met.

[0021] Furthermore, in step S300, terrain feature matching and gain control are performed on the working condition adaptation component to obtain a reconstructed working condition adaptation component, which specifically includes:

[0022] S304, Terrain Matching: Based on the extracted wavelet packet subband energy distribution and terrain elevation data, compare the node energy distribution of the current terrain signal after wavelet packet decomposition with the energy distribution of various terrains in the terrain database to find the terrain type with the highest similarity and determine the optimal subband combination;

[0023] S305 , gain control: dynamically adjusting the amplitude gain of the working condition adaptation component according to the rate of change of the plantar pressure, and obtaining a reconstructed working condition adaptation component after the amplitude adjustment of the working condition adaptation component.

[0024] Furthermore, the step S400 specifically includes:

[0025] S401, weight determination: Determine the weight coefficients of the anti-disturbance basic motion component and the reconstruction working condition adaptation component according to the terrain complexity;

[0026] S402 , weighted superposition: performing weighted superposition on the reconstructed working condition adaptation components according to the weight coefficients to obtain a final motion trajectory.

[0027] Furthermore, the step S500 specifically includes:

[0028] S501, error calculation: Calculate the tracking error between the desired trajectory and the actual joint angle, and divide the frequency band into low-frequency error and high-frequency error;

[0029] S502, parameter adjustment: determine the stiffness coefficient according to the low-frequency error integral, and determine the damping coefficient according to the high-frequency error peak;

[0030] S503, torque calculation: using a second-order impedance model, calculate the expected joint torque according to the stiffness coefficient and the damping coefficient;

[0031] S504, current conversion: converting the desired joint torque into the execution command current of the motor;

[0032] S505. Trajectory tracking control: A PID controller is adopted to output the control current of the motor according to the tracking error, so that the motor tracks the desired joint angle.

[0033] The present invention also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the impedance adjustment and trajectory planning method of the exoskeleton robot as described above are implemented.

[0034] The present invention also provides a computer device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the program, the steps of the impedance adjustment and trajectory planning method of the exoskeleton robot as described above are implemented.

[0035] The impedance adjustment and trajectory planning method, medium, and device of the exoskeleton robot provided by the present invention have the following beneficial effects:

[0036] The impedance adjustment and trajectory planning method of the exoskeleton robot provided by the present invention decomposes a complex motion trajectory into a basic motion component and a working condition adaptation component. Through Fourier series and wavelet packet transform, the periodic gait characteristics and terrain feature frequency components are respectively extracted and recombined in the frequency domain. This enables the exoskeleton to accurately identify and adapt to different terrains. At the same time, according to the terrain complexity, the trajectory weights are dynamically adjusted. In terms of impedance adjustment, the stiffness and damping parameters are adjusted in real time according to the trajectory tracking error, ensuring that the exoskeleton can maintain stable motion under various complex working conditions, greatly expanding the application scenarios of the exoskeleton and significantly enhancing its practicability in complex environments. Description of the Drawings

[0037] Figure 1 is a flowchart of an impedance adjustment and trajectory planning method of an exoskeleton robot provided by the present invention;

[0038] Figure 2 is the overall flowchart of an embodiment of the present invention;

[0039] Figure 3 is a structural block diagram of a computer device in an embodiment of the present invention. Detailed Embodiments

[0040] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the drawings provided by the present invention. According to the following description, the advantages and features of the present invention will be clearer. It should be noted that the drawings are all in a very simplified form and use non-precise scales, only for the purpose of facilitating and clearly assisting in explaining the objectives of the embodiments of the present invention.

[0041] Embodiment 1

[0042] This embodiment provides a method for impedance adjustment and trajectory planning of an exoskeleton robot. Refer to Figure 1 , 2 . The method includes the following steps:

[0043] S100. Obtain multi-dimensional original signals related to exoskeleton movement and preprocess them.

[0044] In one embodiment, the multi-dimensional original signals include joint angle θ(t), joint torque τ(t), trunk acceleration a(t), plantar pressure change rate ΔF, gait cycle T, and terrain elevation data z(x, y).

[0045] Specifically, the multi-dimensional original signals are obtained by an exoskeleton joint encoder, an inertial measurement unit (IMU), a plantar pressure sensor, and a terrain detection radar to obtain human kinematic signals in real time. Among them: the joint angle θ(t) and joint torque τ(t) are obtained through the joint encoder; the trunk acceleration a(t) is collected by the IMU for calculating the position of the human center of mass; the plantar pressure F foot (t) is obtained through the plantar pressure sensor, and the gait phase is identified. The plantar pressure change rate is calculated through the plantar pressure. The gait cycle T is determined through the gait phase, where F foot (t) is the plantar pressure at time t, F foot (t - Δt) is the plantar pressure at time t - Δt, and Δt is the time interval; the terrain elevation data z(x, y) is obtained in real time through the terrain detection radar, where x and y are the horizontal and vertical coordinates.

[0046] In a simulated complex outdoor scenario, the exoskeleton is equipped with multiple sensors to comprehensively obtain motion data and working condition information. When the human body walks, the changes in joint angle and joint torque will cause relative displacement or electrical signal changes of the internal components of the encoder. Through data processing, these physical changes are converted into digital signals, so as to accurately obtain the joint angle θ(t) and joint torque τ(t). These parameters reflect the real-time motion state of the joints and are important bases for analyzing human motion patterns and the power assistance requirements of the exoskeleton. The IMU is built with a high-precision accelerometer and gyroscope, which can sense the acceleration and rotation of the exoskeleton as a whole in real time. The accelerometer measures the acceleration by detecting the displacement of the mass block under the action of inertia force, and the gyroscope measures the angular velocity using the principle of conservation of angular momentum. Through the integration and fusion algorithm of these data, the position of the human center of mass is calculated. This information is crucial for judging the motion posture and stability of the exoskeleton and can assist subsequent trajectory planning and impedance adjustment to ensure the coordination between the exoskeleton and human motion; the plantar pressure sensors are evenly distributed on the sole of the shoe, which can monitor the pressure changes in different areas of the sole in real time and obtain the plantar pressure distribution F foot(t), during walking, as the foot touches the ground, supports the body, and then leaves the ground, the plantar pressure undergoes dynamic changes. By analyzing these pressure data, gait phases can be identified, such as the heel strike, foot flat, and toe off phases. At the same time, the rate of change of plantar pressure is calculated. Among them, F foot (t) is the plantar pressure at time t, and F foot (t - Δt) is the plantar pressure at time t - Δt, where Δt is the time interval. This parameter reflects the speed of change of plantar pressure, can effectively capture sudden changes in ground conditions, and provides key information for the adaptive adjustment of the exoskeleton; the terrain detection radar detects the elevation data z(x, y) of the surrounding terrain by transmitting and receiving electromagnetic waves. The electromagnetic waves emitted by the radar are reflected back after encountering the terrain surface. Based on the time delay and signal strength of the echo, the distance and height information of each point on the terrain surface relative to the exoskeleton are calculated.

[0047] In one embodiment, the preprocessing includes sequentially performing normalization, low-pass filtering, and sliding window segmentation on the multi-dimensional raw signals. Among them, normalization unifies different sensor signals to the same scale, low-pass filtering removes high-frequency noise interference and retains low-frequency signals, and sliding window segmentation divides the continuous signal into fixed-duration segments to extract time-domain features of different time periods.

[0048] Since the measurement principles and ranges of different sensors are different, there are differences in the ranges and units of their output signals. To make subsequent data analysis and processing consistent and comparable, it is necessary to perform normalization processing on the raw signals, map the data of all sensors to the same numerical interval [0, 1], eliminating the influence of dimensional and amplitude differences on subsequent calculations. In actual measurement, sensor signals are inevitably interfered by high-frequency noise, which may originate from electromagnetic interference of electronic devices, environmental noise, etc. The purpose of low-pass filtering is to remove these high-frequency noises and retain the low-frequency effective signals. To better analyze the time characteristics of the signals, the continuous sensor signals are divided into fixed-duration segments to form a series of time-series data. Preferably, the continuous data is divided into fixed-duration segments and a certain window overlap rate is set, and time-domain features are extracted within each window.

[0049] S200. Perform Fourier series transformation and wavelet packet transformation on the preprocessed multi-dimensional raw signals respectively, and decompose the complex motion trajectory of the exoskeleton into a basic motion component reflecting periodic gait and a working condition adaptation component reflecting terrain characteristics.

[0050] In one embodiment, it specifically includes the following steps:

[0051] S201. Extraction of basic motion components: Receive the preprocessed joint angle signals and gait cycles, fit the joint angle signals with Fourier series, and extract the basic motion components;

[0052] Specifically, perform Fourier fitting on the preprocessed joint angle signal θ(t) to obtain the basic motion component x base (t), and the expression is as follows:

[0053]

[0054] where ω0 = 2π / T is the fundamental frequency, a0 is the static offset component of the joint angle, representing the reference angle when standing and bending the knee, a n and b n are Fourier coefficients, which determine the proportion of different frequency harmonic components in the motion signal and the contribution degree of each frequency component to the overall motion. N is the harmonic order, and its value range is 3 - 5, which is used to describe the periodic characteristics of the gait.

[0055] In this embodiment, for the basic motion components, by receiving the preprocessed joint angle signal θ(t) and gait cycle T, regarding it as a periodic signal composed of superposition of different frequency components, and describing the periodic motion of the exoskeleton through the superposition of a finite number of harmonics, the basic motion component x base (t) that can describe the periodic gait characteristics of the exoskeleton is extracted.

[0056] S202. Extraction of working condition adaptation components: Receive the preprocessed acceleration signals and terrain elevation data, perform wavelet packet decomposition on the acceleration signals, calculate the energy of each node, set an energy threshold, screen out the sub-band signals related to terrain disturbances, and reconstruct the selected sub-band signals to generate the working condition adaptation components;

[0057] Specifically, perform wavelet packet decomposition on the preprocessed acceleration signal a(t), and the expression is as follows:

[0058] WPT j,k (t) = ∑ m h j (m)ψ j,k (2 j t - m)

[0059] where WPT j，k (t) is the signal to be analyzed, ψ j，kis the wavelet packet basis function, j represents the decomposition level. For example, when j = 4, wavelet packet decomposition can perform a more refined frequency band division on the signal at different scales. As the decomposition level increases, the signal is decomposed into increasingly narrow frequency bands, enabling a more precise analysis of the signal's frequency components. k is the node index and k ∈ 0 to 15, and each value corresponds to a frequency band. Different node indices represent sub-band signals in different frequency ranges. By analyzing these sub-band signals, the characteristics of the signal in different frequency bands can be obtained. h j (m) is the filtering coefficient corresponding to the j-th layer decomposition. After decomposing the signal collected by the terrain sensor into different frequency bands through wavelet packet transform, the energy threshold method is used to screen the high-frequency sub-band signals related to terrain features, and the energy E j，k = Σ t |WPT j，k (t)| 2 is calculated. An energy threshold E th = 0.2·max(E j，k ) is set. The sub-bands exceeding the threshold are regarded as related to terrain disturbances, and the selected sub-band signals are reconstructed to generate the working condition adaptation component x adapt (t), enabling the exoskeleton movement to adapt to the complex terrain working conditions.

[0060] S203. Decomposition effect verification: Calculate the residual energy ratio of the decomposed signal. If the residual energy ratio does not meet the requirements, increase the wavelet packet decomposition level and the extended Fourier harmonic order, and return to step S201 to re-perform the decomposition and verification until the residual energy ratio meets the requirements.

[0061] Specifically, calculate the residual energy ratio of the decomposed signal This index measures the difference between the reconstructed signal after decomposition and the original joint angle signal. If η is less than the set value, for example, 5%, it indicates that the basic motion component and the working condition adaptation component after decomposition can well reconstruct the original joint angle signal, and the decomposition effect is good; if η is greater than the set value, it indicates that there is a certain error in the decomposition and further optimization is needed. At this time, increase the wavelet packet decomposition level j, for example, j = 5, and the extended Fourier harmonic order N, for example, N = 6, and re-perform the decomposition and verification until the residual energy ratio meets the requirements. This verification and adjustment process ensures the accuracy of the trajectory decomposition and provides a reliable data basis for subsequent trajectory optimization and recombination.

[0062] S300. Determine whether there are abnormal harmonics in the basic motion component. If there are abnormal harmonics, adjust the Fourier coefficients of the abnormal harmonics to suppress the abnormal harmonics, and obtain the anti-disturbance basic motion component according to the adjusted Fourier coefficients; perform terrain feature matching and gain control on the working condition adaptation component to obtain the reconstructed working condition adaptation component.

[0063] After completing the dynamic trajectory decomposition, according to the joint torque feedback during the actual movement of the exoskeleton, under the conditions of complex working conditions, describing the periodic gait characteristics of the exoskeleton based on the extracted basic motion components will generate abnormal joint torque fluctuations, affecting the stability and comfort of the exoskeleton movement. Therefore, the basic motion components are optimized in the frequency domain, and the influence of abnormal harmonics is suppressed by adjusting the weights of the Fourier coefficients.

[0064] In one embodiment, it is determined whether there are abnormal harmonics in the basic motion components. If there are abnormal harmonics, the Fourier coefficients of the abnormal harmonics are adjusted to suppress the abnormal harmonics, and the disturbance-resistant basic motion components are obtained according to the adjusted Fourier coefficients, specifically including:

[0065] S301. Abnormal harmonic detection: Calculate the joint torque fluctuation amplitude corresponding to each harmonic, set the joint torque threshold, and determine whether there are abnormal harmonics according to the magnitude relationship between the joint torque fluctuation amplitude and the joint torque threshold;

[0066] Specifically, define τ n as the joint torque fluctuation amplitude corresponding to the nth harmonic, and the calculation formula is

[0067] , by integrating different frequency components of the joint torque within a gait cycle, the torque fluctuation amplitude corresponding to each harmonic is obtained, thereby quantifying the influence of each harmonic on the joint torque. Set the joint torque threshold τ th . When τ n > τ th , it is determined that the nth harmonic is an abnormal disturbance source.

[0068] S302. Abnormal harmonic suppression: Introduce a suppression coefficient to dynamically attenuate the Fourier coefficients of the abnormal harmonics to obtain the disturbance-resistant basic motion components;

[0069] Specifically, introduce the suppression coefficient γ, and dynamically attenuate the Fourier coefficients a n and b n of the abnormal harmonics to obtain the optimized Fourier coefficients Generate the disturbance-resistant basic component X′ base (t) based on the Fourier coefficients that do not need to be optimized and the optimized Fourier coefficients. In this way, the proportion of abnormal harmonics in the basic motion components is reduced, and the stability of the exoskeleton movement is improved.

[0070] S303. Characteristic verification: Conduct frequency-domain energy comparison, calculate the energy ratio of the disturbance-resistant basic motion components to the basic motion components, calculate the residual between the joint angle signal and the disturbance-resistant basic motion components. If the energy ratio and the residual do not meet the requirements, adjust the parameters and re-optimize until the requirements are met;

[0071] Specifically, calculate the energy ratio of the anti-interference basic motion component to the basic motion component. Require η base to be not lower than a set value, such as 0.8. This is to ensure that while suppressing abnormal harmonics, the periodic characteristics of the basic motion component are not overly damaged. If the energy ratio is lower than the set value, such as 0.8, it indicates that the optimization process may have had a greater impact on the normal motion characteristics, and relevant parameters such as the wavelet packet decomposition layer number and harmonic order need to be adjusted and optimized again; calculate the residual e base (t) = θ(t) - X base (t), and evaluate the optimization effect by observing the magnitude of the residual. When the maximum value of the absolute value of the residual is greater than the set value, such as max(|e base (t)| > 5°, it indicates that there is a large deviation between the optimized anti-interference basic motion component and the original signal, triggering parameter reset. The specific measure is to increase the harmonic order N, reduce the suppression coefficient γ, and then re-optimize and test until the residual meets the requirements.

[0072] In one embodiment, perform terrain feature matching and gain control on the working condition adaptation component to obtain the reconstructed working condition adaptation component, specifically including:

[0073] S304. Terrain matching: According to the extracted wavelet packet subband energy distribution (E j，k ) and the terrain elevation data z(x, y), compare the node energy distribution after wavelet packet decomposition of the current terrain signal with the energy distributions of various terrains in the terrain library, find the terrain type with the highest similarity, and determine the optimal subband combination;

[0074] Specifically, assume that the node energy distribution vector obtained after decomposing the current terrain signal by the wavelet packet subband energy distribution (E j，k ) is E cu , and the energy distribution vector of the i-th terrain in the terrain library is E i . Measure the similarity through the Euclidean distance . Here, j represents the decomposition layer number. This Euclidean distance calculation method can intuitively reflect the difference degree between the two energy distribution vectors. Select the subband combination corresponding to the terrain type with the smallest d i as the optimal subband combination to determine the current terrain type. For example, if the d i value calculated from the energy distribution vector corresponding to the "slope" terrain is the smallest, then it is determined that the current terrain is a slope, and the corresponding subband signal combination is determined to provide accurate terrain information for subsequent gain control

[0075] S305. Gain control: Dynamically adjust the amplitude gain of the working condition adaptation component according to the plantar pressure change rate, and obtain the reconstructed working condition adaptation component after amplitude adjustment of the working condition adaptation component;

[0076] Specifically, the gain control dynamically adjusts the amplitude gain of the working condition adaptation component by using the plantar pressure change rate ΔF. A reference gain K0 is set, and the sensitivity coefficient β ∈ 0.05 - 0.2. According to the formula K ad = K0·(1 + tanh(β·△F)), the amplitude gain K is calculated ad , and the working condition adaptation component x adapt (t) is multiplied by the amplitude gain K ad to obtain the reconstructed working condition adaptation component X′ adapt (t). The characteristics of the tanh function limit the gain range to [K0, 2K0], effectively preventing gain overshoot. When the plantar pressure change rate ΔF is large, it indicates that the ground condition has changed significantly, such as transitioning from a flat ground to a rough road surface. At this time, the value of tanh(β·ΔF) increases, and the amplitude gain K ad also increases accordingly, enhancing the role of the working condition adaptation component and enabling the exoskeleton to better adapt to terrain changes.

[0077] In this embodiment, through terrain feature matching and gain control, the adaptability of the trajectory to rough terrain is enhanced, accurately reflecting the impact of the current terrain on movement, further optimizing the working condition adaptation component, and improving the adaptability of the exoskeleton to complex terrains.

[0078] S400. Recombine the disturbance rejection basic motion component and the reconstructed working condition adaptation component to obtain the final motion trajectory.

[0079] In one embodiment, it specifically includes:

[0080] S401. Weight determination: Determine the weight coefficients of the disturbance rejection basic motion component and the reconstructed working condition adaptation component according to the terrain complexity;

[0081] Specifically, on flat terrains such as stone slab roads, the movement of the exoskeleton mainly follows the periodic gait pattern. Therefore, the weight coefficient λ1 of the disturbance rejection basic motion component is set to a relatively large value to enhance the influence of the basic motion component and make the movement of the exoskeleton more stable and natural; on complex terrains such as slopes, the influence of the terrain on movement is more significant. At this time, the weight coefficient λ2 of the reconstructed working condition adaptation component is set to a relatively large value to highlight the role of the working condition adaptation component.

[0082] S402. Weighted superposition: Perform weighted superposition on the reconstructed working condition adaptation component according to the weight coefficients to obtain the final motion trajectory;

[0083] Specifically, the expression of weighted superposition is as follows:

[0084] X final (t) = λ1X′ base (t) + λ2X′ adapt (t)

[0085] Among them, X final (t) is the final motion trajectory, λ1 is the weight coefficient of the anti-disturbance basic motion component, and λ2 is the weight coefficient of the reconstructed working condition adaptation component.

[0086] In this embodiment, the weights of the anti-disturbance basic motion component and the reconstructed working condition adaptation component are dynamically adjusted according to the complexity of the terrain, and weighted superposition is performed to obtain the motion trajectory that the exoskeleton can actually execute. This ensures that the exoskeleton can adjust the motion trajectory in time according to terrain changes, realizing adaptive trajectory planning for different terrains.

[0087] S500: According to the error between the final motion trajectory of the exoskeleton and the desired trajectory, the impedance model parameters are adjusted in real time to optimize the exoskeleton's tracking of the desired trajectory in real time.

[0088] In one embodiment, it specifically includes:

[0089] S501, error calculation: Calculate the tracking error between the desired trajectory and the actual joint angle, and divide the frequency band into low-frequency error and high-frequency error;

[0090] Specifically, according to the final motion trajectory X final (t) and the actual joint angle x after execution actual (t), calculate the tracking error e(t) = X final (t)-x actual (t), this error accurately reflects the deviation between the current motion state of the exoskeleton and the expected state, and is the key basis for subsequent impedance adjustment. In order to more accurately analyze the impact of the error on the system, it is divided into low-frequency error e low (t), high frequency error e high (t), low-frequency errors mainly reflect the overall trend deviation of the exoskeleton movement, such as the gradual deviation from the expected trajectory during long-term walking; high-frequency errors focus on capturing instantaneous changes and jitters during the movement process.

[0091] S502, parameter adjustment: determine the stiffness coefficient according to the low-frequency error integral, and determine the damping coefficient according to the high-frequency error peak;

[0092] Specifically, the model standard stiffness parameter is set to K nom , the standard damping parameter is B nom , the adaptive coefficients are α and η (both range 0.1-0.5), and the stiffness coefficient is determined according to the low-frequency error integral Determine the damping coefficient based on the high-frequency error peak When the integral value of the low-frequency error is large, it means that the exoskeleton has accumulated a large position deviation over a long period of time. At this time, the stiffness coefficient K is larger than the original K. nomIncreases, which can enhance the ability of the exoskeleton to correct deviations and prompt it to return to the desired trajectory; if the high-frequency error peak is large, it means that the exoskeleton has experienced severe jitter or instantaneous deviation during movement, and the damping coefficient B is larger than the original B nom Increasing can effectively suppress such jitter and ensure the stability of the exoskeleton movement.

[0093] S503. Torque calculation: Use a second-order impedance model to calculate the desired joint torque according to the stiffness coefficient and damping coefficient;

[0094] Specifically, adopt a second-order impedance model to calculate the desired joint torque τ′ for tracking the desired trajectory. Among them, the stiffness coefficient K determines the ability of the exoskeleton to resist position deviations. The larger the K value, the stronger the response of the exoskeleton to position errors, and the greater the force to correct the deviation. B is the damping coefficient used to suppress the oscillation of the exoskeleton movement. When the B value is large, it can effectively slow down the excessive swing of the exoskeleton during adjustment and make the movement smoother. is the rate of change of the error e(t), which reflects how fast the error changes with time.

[0095] S504. Current conversion: Convert the desired joint torque into the execution command current of the motor;

[0096] Specifically, the conversion formula is: where I cmd is the execution command current, and K t is the torque constant, which establishes the conversion relationship between torque and current.

[0097] S505. Trajectory tracking control: Adopt a PID controller to output the control current of the motor according to the tracking error, so that the motor tracks the desired joint angle;

[0098] Specifically, use a PID controller to achieve the mapping from torque to current and trajectory tracking control. Its input is the tracking error, and the output is the control current of the motor where K p is the proportional coefficient, which adjusts the control output proportionally according to the size of the error and can quickly respond to the error change; K d is the differential coefficient, which is adjusted according to the rate of change of the error, helps to predict the change trend of the error, make a response in advance, and suppress the oscillation of the system. The initial K p = 12, K d = 0.6, and finally the output current I = I cmd + I fb is obtained, so that the motor tracks the desired joint angle and the exoskeleton moves along the desired trajectory.

[0099] The impedance adjustment and trajectory planning method for the exoskeleton robot provided by the present invention decomposes complex motion trajectories into basic motion components and working condition adaptation components. Through Fourier series and wavelet packet transform, the periodic gait characteristics and terrain feature frequency components are extracted respectively and recombined in the frequency domain. This enables the exoskeleton to accurately identify and adapt to different terrains. At the same time, according to the terrain complexity, the trajectory weights are dynamically adjusted. In terms of impedance adjustment, the stiffness and damping parameters are adjusted in real time based on the trajectory tracking error to ensure that the exoskeleton can maintain stable motion under various complex working conditions, greatly broadening the application scenarios of the exoskeleton and significantly enhancing its practicability in complex environments.

[0100] Embodiment 2

[0101] This embodiment provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the impedance adjustment and trajectory planning method for the exoskeleton robot described above are implemented.

[0102] Among them, the storage medium can be a magnetic disk, an optical disk, a read-only memory (ROM), a random access memory (RAM), a flash memory, a hard disk drive (HDD), or a solid-state drive (SSD), etc.; the storage medium can also include a combination of the above types of memories.

[0103] Embodiment 3

[0104] This embodiment provides a computer device, which includes: a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the program, the steps of the impedance adjustment and trajectory planning method for the exoskeleton robot described above are implemented.

[0105] Such as Figure 3As shown, the computer device 70 may include: at least one processor 71, such as a CPU (Central Processing Unit), at least one communication interface 73, a memory 74, and at least one communication bus 72. Among them, the communication bus 72 is used to realize the connection and communication between these components. Among them, the communication interface 73 may include a display screen (Display) and a keyboard (Keyboard). Optionally, the communication interface 73 may also include a standard wired interface and a wireless interface. The memory 74 may be a high-speed RAM memory (Random Access Memory, volatile random access memory), or a non-volatile memory, such as at least one disk memory. Optionally, the memory 74 may also be at least one storage device located far from the aforementioned processor 71. Among them, an application program is stored in the memory 74, and the processor 71 calls the program code stored in the memory 74 to execute any of the above method steps.

[0106] Among them, the communication bus 72 may be a peripheral component interconnect (PCI) bus or an extended industry standard architecture (EISA) bus, etc. The communication bus 72 can be divided into an address bus, a data bus, a control bus, etc. For the sake of convenience of representation, Figure 3 only a thick line is used to represent it in the figure, but it does not mean that there is only one bus or one type of bus.

[0107] Among them, the memory 74 may include a volatile memory (English: volatile memory), such as a random access memory (English: random-access memory, abbreviation: RAM); the memory may also include a non-volatile memory (English: non-volatile memory), such as a flash memory (English: flash memory), a hard disk drive (English: hard disk drive, abbreviation: HDD) or a solid-state drive (English: solid-state drive, abbreviation: SSD); the memory 74 may also include a combination of the above types of memories.

[0108] Among them, the processor 71 may be a central processing unit (English: central processing unit, abbreviation: CPU), a network processor (English: network processor, abbreviation: NP), or a combination of a CPU and an NP.

[0109] Among them, the processor 71 may further include a hardware chip. The above-mentioned hardware chip may be an application-specific integrated circuit (ASIC), a programmable logic device (PLD), or a combination thereof. The above-mentioned PLD may be a complex programmable logic device (CPLD), a field-programmable gate array (FPGA), a generic array logic (GAL), or any combination thereof.

[0110] Optionally, the memory 74 is further configured to store program instructions. The processor 71 may call the program instructions to implement the impedance adjustment and trajectory planning method of the exoskeleton robot according to the present invention.

[0111] Those skilled in the art of the present technology should understand that the present invention may be implemented in many other specific forms without departing from the spirit and scope of the present invention. Based on the embodiments of the present invention, any changes and modifications made by those of ordinary skill in the art of the present invention according to the above disclosure shall fall within the protection scope of the claims.

Claims

1. An impedance adjustment and trajectory planning method for an exoskeleton robot, characterized in that, The method includes the following steps: S100. Obtain multi-dimensional original signals related to exoskeleton movement and perform preprocessing on them; S200. Perform Fourier series transformation and wavelet packet transformation on the preprocessed multi-dimensional original signals respectively, and decompose the complex movement trajectory of the exoskeleton into a basic movement component reflecting periodic gait and a working condition adaptation component reflecting terrain features; S300. Determine whether there are abnormal harmonics in the basic movement component. If there are abnormal harmonics, adjust the Fourier coefficients of the abnormal harmonics to suppress the abnormal harmonics, and obtain the disturbance-resistant basic movement component according to the adjusted Fourier coefficients; perform terrain feature matching and gain control on the working condition adaptation component to obtain the reconstructed working condition adaptation component; S400. Recombine the disturbance-resistant basic movement component and the reconstructed working condition adaptation component to obtain the final movement trajectory; S500. According to the error between the final movement trajectory of the exoskeleton and the desired trajectory, adjust the impedance model parameters in real time to enable the exoskeleton to track the desired trajectory and optimize it in real time.

2. The impedance adjustment and trajectory planning method of the exoskeleton robot according to claim 1, characterized in that In the step S100, the multi-dimensional original signals include joint angles, joint torques, trunk accelerations, plantar pressure change rates, gait cycles, and terrain elevation data.

3. The impedance adjustment and trajectory planning method of the exoskeleton robot according to claim 1, characterized in that In the step S100, the preprocessing includes sequentially performing normalization, low-pass filtering, and sliding window segmentation processing on the multi-dimensional original signals.

4. The impedance adjustment and trajectory planning method for the exoskeleton robot according to claim 2, characterized in that The step S200 specifically includes the following steps: S201. Basic movement component extraction: Receive the preprocessed joint angle signal and gait cycle, fit the joint angle signal with Fourier series, and extract the basic movement component; S202. Working condition adaptation component extraction: Receive the preprocessed acceleration signal and terrain elevation data, perform wavelet packet decomposition on the acceleration signal, calculate the energy of each node, set an energy threshold, screen out the sub-band signals related to terrain disturbances, reconstruct the selected sub-band signals, and generate the working condition adaptation component; S203. Decomposition effect verification: Calculate the residual energy ratio of the decomposed signal. If the residual energy ratio does not meet the requirements, increase the wavelet packet decomposition layer number and expand the Fourier harmonic order, and return to step S201 to perform decomposition and verification again until the residual energy ratio meets the requirements.

5. The impedance adjustment and trajectory planning method of the exoskeleton robot according to claim 4, characterized in that In the step S300, determine whether there are abnormal harmonics in the basic movement component. If there are abnormal harmonics, adjust the Fourier coefficients of the abnormal harmonics to suppress the abnormal harmonics, and obtain the disturbance-resistant basic movement component according to the adjusted Fourier coefficients, which specifically includes: S301. Abnormal harmonic detection: Calculate the joint torque fluctuation amplitude corresponding to each order of harmonics, set a joint torque threshold, and determine whether there are abnormal harmonics according to the size relationship between the joint torque fluctuation amplitude and the joint torque threshold; S302. Abnormal harmonic suppression: Introduce a suppression coefficient to dynamically attenuate the Fourier coefficients of the abnormal harmonics to obtain the disturbance-resistant basic movement component; S303. Characteristic verification: Conduct frequency domain energy comparison, calculate the energy ratio between the disturbance-resistant basic movement component and the basic movement component, calculate the residual between the joint angle signal and the disturbance-resistant basic movement component. If the energy ratio and the residual do not meet the requirements, adjust the parameters and optimize again until the requirements are met.

6. The impedance adjustment and trajectory planning method of the exoskeleton robot according to claim 5, characterized in that In the step S300, terrain feature matching and gain control are performed on the working condition adaptation component to obtain a reconstructed working condition adaptation component, which specifically includes: S304. Terrain matching: According to the extracted wavelet packet subband energy distribution and terrain elevation data, compare the node energy distribution after wavelet packet decomposition of the current terrain signal with the energy distributions of various terrains in the terrain library, find the terrain type with the highest similarity, and determine the optimal subband combination; S305. Gain control: Dynamically adjust the amplitude gain of the working condition adaptation component according to the plantar pressure change rate, and obtain a reconstructed working condition adaptation component after amplitude adjustment of the working condition adaptation component.

7. The impedance adjustment and trajectory planning method for the exoskeleton robot according to claim 6, characterized in that The step S400 specifically includes: S401. Weight determination: Determine the weight coefficients of the disturbance-resistant basic motion component and the reconstructed working condition adaptation component according to the terrain complexity; S402. Weighted superposition: Perform weighted superposition on the reconstructed working condition adaptation component according to the weight coefficients to obtain the final motion trajectory.

8. The impedance adjustment and trajectory planning method of the exoskeleton robot according to claim 7, wherein The step S500 specifically includes: S501. Error calculation: Calculate the tracking error between the desired trajectory and the actual joint angle, and divide the frequency band into low-frequency error and high-frequency error; S502. Parameter adjustment: Determine the stiffness coefficient according to the integral of the low-frequency error, and determine the damping coefficient according to the peak value of the high-frequency error; S503. Torque calculation: Use a second-order impedance model to calculate the desired joint torque according to the stiffness coefficient and the damping coefficient; S504. Current conversion: Convert the desired joint torque into the execution command current of the motor; S505. Trajectory tracking control: Adopt a PID controller to output the control current of the motor according to the tracking error, so that the motor tracks the desired joint angle.

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 impedance adjustment and trajectory planning method for an exoskeleton robot as described in any one of claims 1-8 are implemented.

10. A computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, the steps of the impedance adjustment and trajectory planning method for an exoskeleton robot as described in any one of claims 1-8 are implemented.

Images