Robot navigation path planning method and system

Abstract

The invention provides a robot navigation path planning method and system which are applied to the technical field of robot navigation, and the method comprises the steps: obtaining a preset navigation path to carry out surface interaction operation (such as ultrasonic scanning, physiotherapy massage and other scenes) on a target muscle group, and extracting a real-time muscle group activation degree index; the deviation between the current surface interaction operation effect and the expected target is quantified through the error signal, and the navigation path parameters of the surface interaction robot are dynamically adjusted by calculating the control quantity to obtain a corrected navigation path instruction; by executing the corrected navigation path instruction, the robot can adjust its behavior according to real-time physiological feedback, so that the surface interaction operation process can be continuously optimized towards the preset target muscle group activation degree target value, thereby improving the effectiveness and individuation degree of physiotherapy. The method has the advantages that the dynamic adjustment of the navigation path is realized through real-time electromyographic signal feedback, and the individuation, the self-adaptability and the effect of the surface interaction operation are improved.

Description

Technical Field

[0001] The present application relates to the field of robot navigation technology, and in particular to a robot navigation path planning method and system. Background Art

[0002] When performing physical therapy, a surface-interactive robot typically moves across the body's surface according to a pre-defined navigation path, performing these operations. Physical therapy aims to physically stimulate specific muscle groups to alleviate fatigue, reduce pain, or promote recovery. The surface-interactive robot uses its end effector to act on the human skin along a pre-planned path, which determines the robot's positional sequence and movement trajectory.

[0003] Existing surface-interactive robots primarily rely on pre-set programs and paths. However, in actual physical therapy, the human body's response to surface-interactive stimulation varies significantly from person to person, and the same user's response to surface-interactive stimulation can also vary at different times and in different physical states. Therefore, existing technologies lack the ability to dynamically adjust based on real-time muscle physiological responses. Relying solely on pre-set paths is unable to effectively and real-timely correct the path based on collected electromyographic signal feedback.

[0004] In summary, when the surface interactive robot performs physiotherapy interaction on users with physiological reactions whose reactions have individual differences and real-time changes, the preset navigation path cannot fully adapt to the real-time physiological state of the target muscle group and the deviation of the physiotherapy target. How to dynamically and adaptively adjust the robot's navigation path parameters according to the real-time collected electromyographic signal feedback of the target muscle group while executing the preset navigation path, so as to effectively optimize the physiotherapy effect on the target muscle group and ensure the smoothness and safety of the robot's movement. Summary of the Invention

[0005] In view of the above-mentioned shortcomings of the existing technology, the present application provides a robot navigation path planning method and system, which is applied to the field of robot navigation technology. It has the advantages of realizing dynamic adjustment of the navigation path through real-time electromyographic signal feedback, thereby improving the personalization, adaptability and effect of surface interactive operations.

[0006] In a first aspect, a robot navigation path planning method is provided, the method comprising the steps of: S1: Obtain a preset navigation path, and control the surface interactive robot to perform interactive operations on the target muscle group according to the preset navigation path; S2: collecting the electromyographic signals of the target muscle group during the interactive operation in real time, processing the electromyographic signals, and extracting a real-time muscle group activation index; S3: Compare the real-time muscle group activation index with the preset target muscle group activation target value and calculate an error signal; S4: inputting the error signal into a PID controller to calculate a control quantity, and adjusting the navigation path parameters of the surface interactive robot according to the control quantity to obtain a corrected navigation path instruction; S5: The control surface interactive robot performs interactive operations on the target muscle groups according to the corrected navigation path instructions.

[0007] A robot navigation path planning method proposed in the present application performs surface interactive operations on target muscle groups by obtaining a preset navigation path, providing a starting point for subsequent feedback adjustments; by extracting real-time muscle group activation indicators, it can reflect the real-time response of muscles to surface interactive operation stimuli, providing a data basis for subsequent path adjustments; the deviation between the current surface interactive operation effect and the expected target is quantified through an error signal, providing a basis for driving subsequent path adjustments, and by calculating the control quantity, the navigation path parameters of the surface interactive robot are dynamically adjusted to obtain corrected navigation path instructions; by executing the corrected navigation path instructions, the robot can adjust its behavior according to real-time physiological feedback, so that the surface interactive operation process can be continuously optimized towards the preset target muscle group activation target value, forming a closed loop of continuous feedback and adjustment, thereby improving the effectiveness and personalization of physical therapy.

[0008] Furthermore, step S1 includes: S11: Collect the user's body shape data, muscle distribution data, and pain sensitivity data; S12: adjusting the coverage of the preset navigation path according to the collected body shape data and muscle distribution data to obtain an individualized coverage; S13: adjusting the surface interaction force of the preset navigation path according to the collected pain sensitivity data to obtain an individualized interaction force; S14: Control the surface interactive robot to perform interactive operation on the target muscle group according to the individualized coverage range and the individualized interactive operation strength.

[0009] A robot navigation path planning method proposed in this application realizes individualized adjustment of the preset navigation path by introducing the collection and application of user individual data, thereby improving the pertinence and comfort of initial interactive operations.

[0010] Furthermore, step S12 includes: S121: Establishing a first mapping relationship between the body shape data and the coverage range, and a second mapping relationship between the muscle distribution data and the coverage range; S122: Query the first mapping relationship according to the body shape data to obtain an initial coverage range corresponding to the body shape; S123: querying the second mapping relationship according to the muscle distribution data to obtain a coverage adjustment parameter corresponding to the muscle distribution; S124: Adjust the initial coverage according to the coverage adjustment parameter to obtain the personalized coverage.

[0011] This application proposes a robot navigation path planning method that adjusts the preset navigation path coverage range according to body shape data and muscle distribution data, and then systematically determines the individualized coverage range by establishing and utilizing a mapping relationship, thereby solving the problem of how to specifically implement coverage range adjustment based on the user's body characteristics.

[0012] Furthermore, step S2 includes: S21: using the electromyographic signal acquisition unit to collect raw electromyographic signals of the target muscle group during the operation in real time at a preset sampling frequency; S22: Filtering the original electromyographic signal to remove noise interference and power frequency interference to obtain a filtered electromyographic signal; S23: Perform time domain analysis on the filtered electromyographic signal and calculate the root mean square (RMS) value as an indicator of the degree of muscle group activation; S24: Normalize the RMS value to obtain a real-time muscle group activation index.

[0013] A robot navigation path planning method proposed in this application, by elaborating on a method for obtaining indicators reflecting the real-time activation degree of the target muscle group, provides high-quality feedback signals for subsequent adaptive path adjustment, thereby solving the problem that the original electromyographic signal is susceptible to interference and it is difficult to accurately extract effective indicators.

[0014] Furthermore, step S23 includes: S231: Construct an EMG signal power spectrum model, analyze the frequency distribution of the filtered EMG signal, and determine the dominant frequency range; S232: Designing a bandpass filter according to the dominant frequency range to extract the myoelectric signal component within the dominant frequency range; S233: Performing square integration processing on the electromyographic signal component to obtain an electromyographic signal energy value; S234: Perform time sliding average filtering and square root processing on the electromyographic signal energy value to obtain an RMS value reflecting the degree of muscle group activation.

[0015] Furthermore, step S3 includes: S31: Construct a multivariate regression model that includes robot pressure parameters, temperature parameters, and real-time physiological state parameters of the target muscle group, and quantify the influence weight of each parameter on the electromyographic signal; S32: Correcting the real-time muscle group activation index according to the multivariate regression model to obtain a corrected muscle group activation index; S33: Compare the corrected muscle group activation index with the preset target muscle group activation target value to calculate an error signal.

[0016] Furthermore, step S4 includes: S41: inputting the error signal into a PID controller, and calculating a preliminary control variable according to the error signal; S42: Obtain the current configuration parameters of the surface interactive robot and determine whether the current configuration satisfies the singular configuration neighborhood condition; S43: If the current configuration satisfies the singular configuration neighborhood condition, the preliminary control amount is adjusted based on the current configuration parameters according to the singular configuration avoidance algorithm to obtain an adjusted control amount that avoids the singular configuration; S44: If the current configuration does not satisfy the singular configuration neighborhood condition, the navigation path parameters of the surface interactive robot are adjusted according to the adjusted control amount or the preliminary control amount to obtain a corrected navigation path instruction.

[0017] Furthermore, step S41 includes: S411: Obtain the proportional coefficient, integral coefficient and differential coefficient of the PID controller; S412: Calculate a proportional term based on the error signal and the proportional coefficient; calculate an integral term based on the error signal and the integral coefficient; calculate a differential term based on the error signal and the differential coefficient; S413: Add the proportional term, the integral term, and the differential term to obtain the preliminary control variable.

[0018] Furthermore, step S42 includes: S421: Storing multiple singular configuration neighborhood conditions, each singular configuration neighborhood condition corresponds to a joint angle combination and a safety threshold; S422: Obtain the current joint angle values ​​of the robot and calculate the current configuration parameters based on the joint angle values; S423: Compare the current configuration parameters with multiple stored singular configuration neighborhood conditions. If the current configuration parameters meet any of the singular configuration neighborhood conditions and exceed the safety threshold corresponding to the singular configuration neighborhood condition, then it is determined that the current configuration meets the singular configuration neighborhood condition.

[0019] In a second aspect, a robot navigation path planning system is applied to the steps of any of the above methods, the system comprising: Preset navigation module: obtains a preset navigation path and controls the surface interactive robot to perform interactive operations on the target muscle group according to the preset navigation path; Real-time acquisition module: collects the electromyographic signals of the target muscle group during the interactive operation in real time, processes the electromyographic signals, and extracts real-time muscle group activation indicators; Comparison module: compares the real-time muscle group activation index with the preset target muscle group activation target value and calculates the error signal; Adjustment module: inputs the error signal into the PID controller, calculates the control amount, and adjusts the navigation path parameters of the surface interactive robot according to the control amount to obtain a corrected navigation path instruction; Execution control module: controls the surface interactive robot to perform interactive operations on the target muscle groups according to the corrected navigation path instructions.

[0020] Beneficial effects: The present application proposes a robot navigation path planning method and system. By acquiring a preset navigation path, the system performs surface interaction on the target muscle group, providing a starting point for subsequent feedback adjustment. By extracting real-time muscle group activation indicators, the system can reflect the real-time response of the muscle to the stimulation of the surface interaction operation, providing a data basis for subsequent path adjustment. The error signal quantifies the deviation between the current surface interaction operation effect and the desired target, providing a basis for driving subsequent path adjustment. By calculating the control quantity, the navigation path parameters of the surface interaction robot are dynamically adjusted to obtain a corrected navigation path instruction. By executing the corrected navigation path instruction, the robot can adjust its behavior according to real-time physiological feedback, so that the surface interaction operation process can be continuously optimized towards the preset target muscle group activation target value, forming a closed loop of continuous feedback and adjustment, thereby improving the effectiveness and personalization of physical therapy. The system has the advantages of realizing dynamic adjustment of the navigation path through real-time electromyographic signal feedback, and improving the personalization, adaptability and effect of surface interaction operation. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 This is a flowchart of a robot navigation path planning method proposed in this application.

[0022] Figure 2 This is a structural diagram of a robot navigation path planning system proposed in this application.

[0023] Figure 3 This is a framework diagram of a robot navigation path planning system proposed in this application.

[0024] Description of reference numerals: 201, preset navigation module; 202, real-time acquisition module; 203, comparison module; 204, adjustment module; 205, execution control module. DETAILED DESCRIPTION

[0025] The technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all of the embodiments. The components of the embodiments of the present application generally described and marked in the drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the application for protection, but merely represents the selected embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without making creative work fall within the scope of protection of the present application.

[0026] It should be noted that similar reference numerals and letters represent similar items in the following figures. Therefore, once an item is defined in one figure, it does not need to be further defined or explained in subsequent figures. At the same time, in the description of this application, the terms "first", "second", etc. are only used to distinguish the description and should not be understood as indicating or implying relative importance.

[0027] In a first aspect, a robot navigation path planning method comprises the steps of: S1: Obtain a preset navigation path and control the surface interactive robot to perform interactive operations on the target muscle group according to the preset navigation path; S2: Real-time collection of the electromyographic signals of the target muscle groups during the interactive task, processing the electromyographic signals, and extracting real-time muscle group activation indicators; S3: Compare the real-time muscle group activation index with the preset target muscle group activation target value and calculate an error signal; S4: Input the error signal into the PID controller to calculate the control quantity, and adjust the navigation path parameters of the surface interactive robot according to the control quantity to obtain the corrected navigation path instruction; S5: The control surface interactive robot performs interactive operations on the target muscle groups according to the corrected navigation path instructions.

[0028] Among them, electromyographic signals refer to physiological electrical signals that reflect the contraction or activation state of muscles. They can be obtained using surface electromyographic sensors, and are mainly used to obtain the physiological responses of muscle groups to interactive work stimuli.

[0029] The muscle activation index refers to an indicator that quantifies the strength of the electromyographic signal. It can be calculated based on the amplitude, frequency or energy of the electromyographic signal. It is mainly used to objectively evaluate the degree of muscle activation.

[0030] Navigation path parameters refer to the attributes that describe the robot's surface interactive operation path, which may include the position, velocity, acceleration, force or action angle of the robot's end effector. It is mainly used to define the trajectory and method of the robot's actual execution of interactive operation operations.

[0031] Specifically, this method constructs a control system based on physiological feedback of the target muscle group.

[0032] First, a preset navigation path is acquired and the surface interactive robot is controlled to perform interactive tasks on the target muscle group along this path. During the interactive tasks, the system collects the target muscle group's electromyographic signals in real time and processes them to extract an indicator reflecting the muscle group's activation status. This real-time indicator is then compared with the preset target activation value for the target muscle group, and the error signal between the two is calculated.

[0033] This error signal is fed into a PID controller, which calculates a control variable based on the error. This control variable is used to adjust the surface-interactive robot's navigation path parameters, generating a revised navigation path instruction. The robot then continues to perform interactive tasks on the target muscle group according to the revised instruction.

[0034] In this way, the system can adjust the robot's interactive work behavior according to the physiological response of the muscle group, so that the interactive work process can be optimized towards the preset physical therapy goals, forming a feedback and adjustment cycle.

[0035] As a preferred embodiment, the solution of this application is specifically implemented as follows: Preset navigation paths can be stored in the robot's control system, for example, as waypoints or parameterized curves. The surface-interactive robot can use its motion control system to drive the end effector along these waypoints or curves and apply interactive force. Myoelectric signals can be collected using surface electromyography sensors worn on the surface of the target muscle groups, which transmit the collected electrical signals to a processing unit.

[0036] The processing unit can process the signal and calculate the root mean square value of the electromyographic signal as a muscle group activation index. The preset target muscle group activation target value can be set according to the user's condition or physical therapy needs.

[0037] The processing unit subtracts the calculated muscle group activation index from the target value to generate an error signal. This error signal is fed into the PID control algorithm module, which calculates the control output value based on the proportional, integral, and differential coefficients. This control output value is used to adjust the robot's navigation path parameters. For example, the robot's speed along the current path can be adjusted based on the control variable, or the local shape of the path can be adjusted to adjust the stimulation of specific areas. The adjusted path instructions are then sent to the robot's motion control system to guide the robot in performing subsequent interactive tasks.

[0038] Through the above-mentioned solution, this application achieves adjustment of the navigation path of a surface-interactive robot. This solution can adapt to the physiological state of the target muscle group based on feedback, overcoming the limitation of the preset navigation path's inability to adapt to changes in the user's physiological state. By incorporating electromyographic signal feedback into the control system, the robot can modify its path and parameters, optimizing the interactive operation process to achieve the preset physical therapy goals, thereby improving the adaptability and effectiveness of physical therapy.

[0039] Furthermore, step S1 includes: S11: Collect the user's body shape data, muscle distribution data, and pain sensitivity data; S12: adjusting the coverage of the preset navigation path according to the collected body shape data and muscle distribution data to obtain an individualized coverage; S13: adjusting the surface interaction force of the preset navigation path according to the collected pain sensitivity data to obtain an individualized interaction force; S14: Controlling the surface interactive robot to perform interactive operations on the target muscle groups according to the individualized coverage range and individualized interactive operation strength.

[0040] Among them, body shape data refers to data that reflects the external morphological characteristics of the user's body, such as height, weight, shoulder width, limb length or the circumference of specific parts. It can be collected non-contact using three-dimensional scanning equipment, or it can be manually input by the user or read from the user's health record.

[0041] Muscle distribution data refers to data that reflects characteristics such as the shape, size, position or tension of the user's target muscle group, which can be obtained by scanning using ultrasonic imaging equipment.

[0042] Pain sensitivity data refers to data reflecting the user's tolerance or discomfort reaction to pressure and temperature, which can be evaluated using a user feedback scale.

[0043] The coverage of a preset navigation path refers to the size of the area or spatial range that the surface interactive robot acts on when executing the preset path, which can be expressed as the density of path points, the boundary definition of the path, or the area covered by the path.

[0044] Individualized coverage refers to the interactive work area or path range that is adjusted to better suit a specific user based on their body shape and muscle distribution. The interactive work force of a preset navigation path refers to the pressure or force applied by the surface interactive robot on the target muscle group when executing the preset path. This can be controlled by the robot's end-effector output force or joint torque.

[0045] Individualized interactive task intensity refers to the interactive task pressure or force that is adjusted according to the user's pain sensitivity characteristics and is more suitable for a specific user.

[0046] A surface-interactive robot is a robotic system capable of autonomously or semi-autonomously performing surface-interactive tasks. It typically includes a robotic arm, end effector, sensors, and a control system. Target muscle groups are specific muscle areas on the user's body that require interactive occupational therapy or relaxation, such as the neck, waist, or leg muscles.

[0047] The improved solution of this application is able to achieve personalized adjustments to the initial interactive task because it first collects key data reflecting the user's individual physiological characteristics, including body shape data, muscle distribution data, and pain sensitivity data, before starting the interactive task. This data is directly related to the actual physical characteristics of the target muscle group and the user's response to the interactive task stimulation.

[0048] Based on the collected body shape data and muscle distribution data, the system can more accurately understand the specific position, size and shape of the target muscle group on the current user's body, so that it can make spatial adjustments to the preset general navigation path, such as expanding or shrinking the boundaries of the path, or adjusting the density of path points to ensure that the range of action of the surface interactive robot can accurately cover the target muscle group, avoiding omissions or excesses, thereby obtaining personalized coverage.

[0049] At the same time, based on the collected pain sensitivity data, the system can evaluate the user's tolerance to pressure, so that it can adjust the size of the preset general interactive operation force, such as increasing or decreasing the force, to ensure that the interactive operation force can achieve a therapeutic effect without causing excessive discomfort or pain to the user, thereby obtaining personalized interactive operation force.

[0050] Finally, the surface interactive robot is controlled to perform initial interactive operations according to the individualized coverage and individualized interactive operation strength adjusted by body shape, muscle distribution and pain sensitivity data.

[0051] In this way, the initial interaction is no longer a one-size-fits-all solution, but is tailored to the specific circumstances of the current user, providing a more optimal and stable starting state for subsequent real-time feedback control based on electromyographic signals. This personalized initial interaction not only improves the effectiveness and comfort of the interaction, but also enables subsequent real-time electromyographic signal feedback to be performed under conditions more consistent with the user's physiological basis, thereby enhancing the adaptability and optimization of the entire navigation path planning method.

[0052] For example, in one specific embodiment, the user's body shape data, muscle distribution data, and pain sensitivity data can be collected by manually entering them through a user registration interface, or automatically measured by connecting to an external sensor device. For example, the body shape data can be obtained by the user entering their height and weight, the muscle distribution data can be obtained by the user selecting a preset muscle model template and fine-tuning it, and the pain sensitivity data can be obtained by the user selecting a pain level from 1 to 10. When adjusting the coverage of a preset navigation path based on the collected body shape data and muscle distribution data, the system can preset a number of first mapping relationships between body shape and coverage range, and a second mapping relationship between muscle distribution and coverage range adjustment parameters.

[0053] For example, for taller users, the preset path can be stretched vertically; for users with well-developed muscle groups, the path density in that area can be increased. When adjusting the interaction force of the preset navigation path based on the collected pain sensitivity data, the system can preset a mapping table between pain sensitivity level and interaction force coefficient. For example, the higher the sensitivity level, the smaller the interaction force coefficient, and the final interaction force is equal to the preset force multiplied by the coefficient.

[0054] When the control surface interactive robot performs interactive operations on the target muscle groups according to the individualized coverage range and individualized interactive operation force, the adjusted path point coordinate sequence and the corresponding force instructions can be sent to the robot's motion controller and force controller, which drive the joint motor and end effector to perform corresponding actions and apply force.

[0055] Furthermore, step S12 includes: S121: Establishing a first mapping relationship between body shape data and coverage, and a second mapping relationship between muscle distribution data and coverage; S122: Query the first mapping relationship according to the body shape data to obtain an initial coverage range corresponding to the body shape; S123: querying a second mapping relationship according to the muscle distribution data to obtain a coverage adjustment parameter corresponding to the muscle distribution; S124: Adjust the initial coverage range according to the coverage adjustment parameters to obtain the personalized coverage range.

[0056] Taking the target muscle group as the back muscles as an example, in step S121, a first mapping relationship is established. For example, the body mass index (BMI) is associated with the initial coverage area of ​​the back interactive task. The higher the BMI value, the larger the initial coverage area.

[0057] A second mapping relationship is also established. For example, it associates the thickness or stiffness data of the major back muscle groups (such as the trapezius, latissimus dorsi, and erector spinae) with coverage adjustment parameters. The greater the thickness or stiffness of a muscle group, the corresponding adjustment parameter may indicate an increase in the coverage weight or boundary expansion of the muscle group area. These mapping relationships can be stored in a lookup table.

[0058] In step S122, the user's height of 175 cm and weight of 70 kg are collected, and the BMI is calculated to be 22.86. According to the first mapping relationship, the corresponding initial back coverage is obtained as a rectangular area with an area set to 0.5 square meters.

[0059] In step S123, the user's back muscle distribution data is collected. For example, ultrasonic measurement reveals that the average erector spinae thickness is 2.5 cm, and the average trapezius thickness is 1.8 cm. According to the second mapping relationship, an adjustment parameter corresponding to an erector spinae thickness of 2.5 cm indicates a 20% increase in coverage weight for the erector spinae region, and an adjustment parameter corresponding to a trapezius thickness of 1.8 cm indicates an 1 cm expansion in coverage weight for the trapezius region.

[0060] In step S124, these adjustment parameters are applied to the initial coverage area. Within the initial 0.5-square-meter rectangular coverage area, the coverage weight of the area corresponding to the erector spinae muscles is increased by 20%, while the boundary of the trapezius muscle area is expanded outward by 1 cm. This results in a personalized back interaction task coverage area adjusted to the user's specific body shape and muscle distribution. Subsequent path planning and execution by the surface-interactive robot will be performed within this personalized coverage area.

[0061] Through the above technical solution, this application solves the problem of how to specifically and systematically adjust the coverage of the preset navigation path according to the user's body shape data and muscle distribution data during surface interactive robot physiotherapy to ensure the accuracy and effectiveness of the adjustment.

[0062] Furthermore, step S2 includes: S21: using the electromyographic signal acquisition unit to collect raw electromyographic signals of the target muscle group during the interactive operation in real time at a preset sampling frequency; S22: Filtering the original electromyographic signal to remove noise interference and power frequency interference, thereby obtaining a filtered electromyographic signal; S23: Perform time domain analysis on the filtered electromyographic signal and calculate the root mean square (RMS) value as an indicator of the degree of muscle group activation; S24: Normalize the RMS value to obtain a real-time muscle group activation index.

[0063] Among them, the electromyographic signal acquisition unit refers to the sensor and related circuits used to obtain the electrical activity of human muscles, and a surface electromyographic sensor can be attached to the surface of the skin.

[0064] The preset sampling frequency refers to the number of myoelectric signal samples collected per unit time, and can be set according to the frequency characteristics of the myoelectric signal and the requirements of subsequent processing, for example, to 1000 Hz or 2000 Hz.

[0065] Time domain analysis refers to the analysis of signals on the time axis, which can include calculating the signal's amplitude, duration, frequency and other time domain characteristics.

[0066] The root mean square (RMS) value refers to the square root of the average value of the square of the signal. It is a way to measure the effective value of the signal and is often used to quantify the amplitude or power of the electromyographic signal.

[0067] In one embodiment, the electromyographic signal acquisition unit may use a surface electromyographic sensor, such as an Ag / AgCl electrode, connected to the signal acquisition circuit via a wire. The preset sampling frequency may be set to 1000 Hz. Filtering may use a digital filter, such as a Butterworth bandpass filter to filter out 0-20 Hz motion artifacts and high-frequency noise above 500 Hz, and a notch filter to filter out 50 Hz power frequency interference.

[0068] When calculating the RMS value in time-domain analysis, the filtered EMG signal samples can be squared, averaged, and then square-rooted within a fixed time window (e.g., 200ms). Normalization can be performed using maximum-minimum normalization. This involves scaling the real-time RMS value to a value between 0 and 1, based on the previously measured RMS value of the target muscle group during maximal voluntary contraction (as the maximum value) and the RMS value at rest (as the maximum value).

[0069] Through the above technical solution, the present application can effectively solve the problem that the original electromyographic signal is easily affected by noise and power frequency interference, and it is difficult to accurately extract effective indicators. By filtering the original electromyographic signal, the noise and power frequency interference are effectively removed, and the signal purity is improved. By calculating the root mean square RMS value through time domain analysis, a reliable method for quantifying the degree of muscle group activation is provided. Through normalization processing, individual differences and amplitude changes are eliminated, making the indicators stable and comparable. Finally, an accurate and reliable real-time muscle group activation index is obtained, which provides a high-quality feedback signal for subsequent adaptive path adjustment, thereby improving the effectiveness and accuracy of the path adjustment of the surface interactive robot and optimizing the physical therapy effect.

[0070] Furthermore, step S23 includes: S231: Construct an EMG signal power spectrum model, analyze the frequency distribution of the filtered EMG signal, and determine the dominant frequency range; S232: Design a bandpass filter based on the dominant frequency range to extract the electromyographic signal components within the dominant frequency range; S233: Performing square integration processing on the electromyographic signal component to obtain an electromyographic signal energy value; S234: Perform time sliding average filtering and square root processing on the electromyographic signal energy value to obtain an RMS value reflecting the degree of muscle group activation.

[0071] The EMG signal power spectrum model refers to a model that describes the energy distribution of the EMG signal at different frequencies through mathematical methods, which can be implemented by calculating the power spectrum density using fast Fourier transform.

[0072] The dominant frequency range refers to the frequency range in which the energy in the electromyographic signal is relatively concentrated and has a high correlation with the activation state of the muscle group. It can be determined based on the results of power spectrum model analysis by setting energy thresholds or identifying energy peaks.

[0073] Time sliding average filtering refers to a processing method that calculates the average of signal values ​​within a set time window and slides the window along the time axis for repeated calculation. It can be achieved by setting a fixed-length sampling point window and calculating the average value within the window.

[0074] This solution performs more refined frequency and time domain processing on the filtered electromyographic signals, aiming to calculate the RMS value that can more accurately reflect the actual activation degree of the muscle group, thereby providing a more reliable basis for subsequent navigation path adjustments.

[0075] Specifically, by first constructing a power spectrum model of the EMG signal and analyzing its frequency distribution, we can identify the dominant frequency range within the EMG signal where energy is most concentrated and best reflects the activation state of the muscle group. This analysis forms the basis for subsequent processing, ensuring focus on key signal components.

[0076] Next, based on the determined dominant frequency range, a bandpass filter is designed and applied to accurately extract the signal components within the dominant frequency range from the filtered electromyographic signal, effectively filtering out non-dominant frequency components that are weakly associated with muscle group activation or are noise, so that subsequent calculations focus on the most relevant signal parts.

[0077] Then, the EMG signal components within the extracted dominant frequency range are squared and integrated to calculate the energy value of the EMG signal. The energy of the EMG signal is directly related to the intensity of muscle contraction. Square integration is an effective method to measure signal energy. The energy value obtained preliminarily quantifies the activation level of the muscle group. Finally, the EMG signal energy value is filtered with a time sliding average and square rooted. Due to the instantaneous fluctuations in the EMG signal and its energy value, directly using the instantaneous energy value as an activation indicator is not stable enough. The time sliding average filter can smooth out short-term fluctuations in the energy value, obtaining a value that is more stable in time and better reflects the average activation state over a period of time. The square root of this value is then taken to obtain the RMS value that ultimately reflects the degree of muscle group activation.

[0078] This RMS value, derived through frequency-domain focusing and time-domain smoothing, more accurately and stably reflects the true degree of muscle activation than the RMS value calculated directly from the original filtered signal. Inputting this more accurate and stable muscle activation indicator into subsequent comparisons, error calculations, and PID control steps enables more precise and timely adjustments to the robot's navigation path, thereby optimizing the therapeutic effect on the target muscle group.

[0079] Furthermore, step S3 includes: S31: Construct a multivariate regression model that includes robot pressure parameters, temperature parameters, and real-time physiological state parameters of the target muscle group, and quantify the influence weight of each parameter on the electromyographic signal; S32: Correcting the real-time muscle group activation index according to the multivariate regression model to obtain a corrected muscle group activation index; S33: Compare the corrected muscle group activation index with the preset target muscle group activation target value to calculate an error signal.

[0080] Among them, in order to build a multivariate regression model to describe the quantitative relationship between factors such as robot pressure, temperature, and the real-time physiological state of the target muscle group and the electromyographic signal, a linear regression method can be used. The specific steps are as follows: Assuming that there is a linear relationship between the electromyographic signal (EMG_signal) and pressure (Pressure), temperature (Temperature) and physiological state (Physiological_State), the following linear regression model can be established: ,in, is the intercept term; 、 、 are regression coefficients, which respectively represent the influence weights of pressure, temperature and physiological state on electromyographic signals, and can be obtained by training the model using statistical methods such as the least squares method; is the error term.

[0081] Correcting the real-time muscle activation index involves using a constructed multivariate regression model to remove or mitigate the influence of inactive factors from the real-time muscle activation index. This can be achieved by first collecting the robot's pressure, temperature, and the real-time physiological state parameters of the target muscle group in real time. These parameters are then input into the multivariate regression model to calculate the predicted impact of these inactive factors on the current EMG signal. Finally, this predicted impact is subtracted from the original real-time muscle activation index to obtain the corrected muscle activation index. For example, if the multivariate regression model predicts that the current pressure, temperature, and physiological state will cause the EMG signal to increase by a specific value, this value will be subtracted from the original index during correction.

[0082] Comparing the corrected muscle group activation index with the preset target muscle group activation value to calculate the error signal involves comparing the corrected index, which better reflects the actual muscle group activation state, with the desired target value. This can be achieved using a simple subtraction operation: the error signal equals the corrected muscle group activation index minus the preset target muscle group activation value. This error signal reflects the deviation between the current muscle group activation state and the target state.

[0083] This approach combines a holistic approach that involves acquiring a preset navigation path, collecting electromyographic signals in real time and extracting metrics, inputting an error signal into a PID controller to adjust the path, and then executing the interactive task according to the corrected path. By introducing a correction step before comparing and calculating the error signal, the accuracy of the error signal is improved. An accurate error signal can more realistically reflect the gap between the current interactive task effect and the target, making subsequent PID control and path adjustment based on this error signal more precise and effective. This combination enables the entire interactive task process to better adapt to the real-time physiological state of the target muscle group, optimizing the therapeutic effect.

[0084] Furthermore, step S4 includes: S41: Input the error signal into the PID controller and calculate the preliminary control quantity according to the error signal; S42: Obtain the current configuration parameters of the surface interactive robot and determine whether the current configuration satisfies the singular configuration neighborhood condition; S43: If the current configuration satisfies the singular configuration neighborhood condition, the preliminary control amount is adjusted based on the current configuration parameters according to the singular configuration avoidance algorithm to obtain an adjusted control amount that avoids the singular configuration; S44: If the current configuration does not satisfy the singular configuration neighborhood condition, the navigation path parameters of the surface interactive robot are adjusted according to the adjusted control amount or the preliminary control amount to obtain a corrected navigation path instruction.

[0085] Among them, the current configuration parameters refer to the parameters that describe the current posture and position of the surface interactive robot. Specifically, they can be expressed by the angle values ​​of each joint of the robot, the length of the connecting rod, the posture of the end effector in space, etc.

[0086] The singular configuration neighborhood condition refers to the condition for judging whether the current configuration of the robot is close to or in a singular configuration. Specifically, the judgment can be made by setting a threshold or function relationship based on the rank, determinant value, joint angle limit, joint speed limit, etc. of the robot's Jacobian matrix.

[0087] The singular configuration avoidance algorithm refers to an algorithm used to adjust the control quantity to avoid entering a singular configuration when the robot is close to or in a singular configuration. Specifically, it can be implemented by using methods such as damped least squares method, adding zero space velocity component, and modifying task priority.

[0088] The adjusted control quantity refers to the control quantity obtained by correcting the initial control quantity according to the singular configuration avoidance algorithm when the current configuration satisfies the singular configuration neighborhood condition. This control quantity takes into account the robot kinematic constraints.

[0089] As a preferred embodiment, the solution of this application is specifically implemented as follows: When the surface interactive robot performs interactive tasks with a user, it collects the electromyographic signals of the user's target muscle groups in real time. After filtering, rectification, and smoothing, it calculates the RMS value of the real-time muscle group activation. This RMS value is compared with the preset target RMS value to calculate the error signal e(t).

[0090] e(t) = target RMS value − real-time RMS value.

[0091] The error signal e(t) is input into a digital PID controller, which calculates the preliminary control variable according to the discretized PID algorithm: ,in, 、 、 are proportional, integral and differential coefficients respectively; Indicates the error signal of the i-th time step in the discrete time series. At the same time, the system obtains the current joint angle value of the robot . Calculate the Jacobian matrix J corresponding to the current configuration of the robot and calculate its determinant value det(J).

[0092] A singular configuration neighborhood threshold ε is preset. If , then the current configuration satisfies the singular configuration neighborhood condition. If the singular configuration neighborhood condition is met, the damped least squares method is used to calculate the initial control quantity. For example, if the initial control volume represents the desired end effector velocity , then by solving Get joint velocity increment , where λ is the damping coefficient and I is the unit matrix. As the adjusted control .

[0093] If the singular configuration neighborhood condition is not met, the preliminary control amount is used directly . According to the final control quantity ( or ), for example, converting it into the desired displacement of the end effector, and then updating the coordinates of the next target point on the navigation path or adjusting the parameters of the path curve to generate a corrected navigation path instruction.

[0094] Furthermore, step S41 includes: S411: Obtain the proportional coefficient, integral coefficient and differential coefficient of the PID controller; S412: Calculate a proportional term based on the error signal and the proportional coefficient; calculate an integral term based on the error signal and the integral coefficient; calculate a differential term based on the error signal and the differential coefficient; S413: Add the proportional term, integral term and differential term to obtain a preliminary control variable.

[0095] In one embodiment, step S411 may include reading a preset scaling factor from a configuration file stored in the robot control system. , integral coefficient and differential coefficients .

[0096] Step S412 may include executing a standard discrete PID control algorithm. For example, the proportional term may be calculated as the current value of the error signal multiplied by The integral term can be calculated as the sum of the historical values ​​of the error signal multiplied by , accumulation can be achieved by maintaining an error queue and summing it. The differential term can be calculated as the difference between the current error signal value and the error signal value at the previous moment multiplied by .

[0097] Step S413 may include summing the calculated values ​​of the proportional term, integral term, and differential term to obtain a preliminary control variable, which can then be used to adjust navigation path parameters such as the robot's speed, direction, or force.

[0098] By defining the specific process by which the PID controller calculates the initial control variable based on the error signal, this solution provides operational technical details, resolving the issue of unclear technical details in this step. Using a standard PID control algorithm, it is possible to comprehensively consider the current error, historical error, and error rate of change to generate a more accurate and stable initial control variable, providing a more reliable basis for subsequent adjustments to navigation path parameters. This helps improve the surface interactive robot's adaptability to the real-time physiological state of the target muscle group and optimize the therapeutic effect.

[0099] Furthermore, step S42 includes: S421: Storing multiple singular configuration neighborhood conditions, each singular configuration neighborhood condition corresponds to a joint angle combination and a safety threshold; S422: Obtain the current joint angle values ​​of the robot and calculate the current configuration parameters based on the joint angle values; S423: Compare the current configuration parameters with multiple stored singular configuration neighborhood conditions. If the current configuration parameters meet any of the singular configuration neighborhood conditions and exceed the safety threshold corresponding to the singular configuration neighborhood condition, then it is determined that the current configuration meets the singular configuration neighborhood condition.

[0100] Among them, the singular configuration neighborhood condition refers to a predefined description of the area close to the singular configuration in the robot joint space, which can be implemented using a geometric description of the joint angle range, joint angle combination, or a mathematical expression based on the robot Jacobian matrix characteristics.

[0101] A joint angle combination refers to a set of joint angle values ​​that describes a specific posture of the robot, which can be represented by a set of rotation or translation angles of each joint of the robot.

[0102] The safety threshold refers to the numerical limit used to define the danger level of the singular configuration neighborhood, which can be set using a certain metric value of the joint space distance, task space distance, or Jacobian matrix.

[0103] Configuration parameters refer to parameters that can describe the current posture and position of the robot. They can be represented by the real-time angle values ​​of each joint of the robot, the coordinates and posture of the robot's end effector in the task space, or other parameters that can reflect the robot's configuration.

[0104] As a preferred embodiment, the solution of this application is specifically implemented as follows: Assume that the surface-interactive robot is a six-degree-of-freedom serial robot. Consider a typical singular configuration, such as when the robot's wrist center is collinear with the shoulder joint. In this configuration, the robot's end-effector's motion capabilities in certain directions are restricted. In this embodiment, the singular configuration neighborhood condition can be defined based on the joint angles and the Jacobian matrix determinant.

[0105] S421: Store two singular configuration neighborhood conditions. Condition 1: When the angle of joint 5 is close to 0 degrees or 180 degrees, the Jacobian matrix determinant approaches zero. Condition 1 is set as follows: When the angle of joint 5 is within the range of [-5 degrees, 5 degrees] or [175 degrees, 185 degrees], the corresponding safety threshold is that the absolute value of the Jacobian matrix determinant is less than 0.01. Condition 2: When the axes of joints 1 and 6 are aligned, the Jacobian matrix determinant approaches zero. Condition 3 is set as follows: The absolute value of the angle difference between joints 1 and 6 is less than 5 degrees, and the corresponding safety threshold is that the absolute value of the Jacobian matrix determinant is less than 0.005. These conditions are stored in the storage unit of the robot controller.

[0106] S422: Obtain the angle values ​​of the six joints in real time through the encoders installed on each joint of the robot Based on these angle values, calculate the Jacobian matrix J of the robot in the current configuration and calculate its determinant value D=det(J).

[0107] S423: First check condition 1. Is it within the range of [-5 degrees, 5 degrees] or [175 degrees, 185 degrees]. If so, check whether |D| is less than 0.01. If both conditions are met, the current configuration is determined to meet the singular configuration neighborhood condition. If condition 1 is not met, or the joint angle range of condition 1 is met but |D| is not less than 0.01, check condition 2. Is |D| less than 5 degrees? If so, check whether |D| is less than 0.005. If both conditions are met, the current configuration is determined to meet the singular configuration neighborhood condition. If both conditions are not met, the current configuration is determined to not meet the singular configuration neighborhood condition.

[0108] Through the above technical solution, the present application can determine whether the robot's current configuration is close to a singular configuration based on preset conditions and the robot's real-time configuration parameters. This provides a basis for subsequent avoidance measures, helping to reduce motion instability or control difficulties that may occur when the robot moves near a singular configuration, thereby maintaining the robot's motion performance.

[0109] Please refer to Figure 2 、 Figure 3 A robot navigation path planning system is applied to the steps of any of the above methods, and the system includes: Preset navigation module 201: obtains a preset navigation path and controls the surface interactive robot to perform interactive operations on the target muscle group according to the preset navigation path; Real-time acquisition module 202: collects the electromyographic signals of the target muscle group during the interactive operation in real time, processes the electromyographic signals, and extracts the real-time muscle group activation index; Comparison module 203: compares the real-time muscle group activation index with the preset target muscle group activation target value, and calculates an error signal; Adjustment module 204: inputs the error signal into the PID controller, calculates the control quantity, and adjusts the navigation path parameters of the surface interactive robot according to the control quantity to obtain a corrected navigation path instruction; Execution control module 205: controls the surface interactive robot to perform interactive operations on the target muscle group according to the corrected navigation path instructions.

[0110] The preset navigation module 201 refers to a functional unit responsible for initializing the interactive operation process, which can be implemented by a software program module, a dedicated processing chip, or a hardware unit including a storage medium and a processor.

[0111] The real-time acquisition module 202 refers to a functional unit responsible for acquiring and preliminarily processing physiological signals, which can be implemented by a combination of an integrated electromyographic sensor, a signal amplification circuit, an analog-to-digital converter, and a data preprocessing software module.

[0112] The comparison module 203 refers to a functional unit responsible for quantifying the difference between the current state and the desired target, and can be implemented using a software algorithm module or a hardware comparison circuit.

[0113] The adjustment module 204 is a functional unit responsible for calculating control instructions and correcting path parameters according to the error signal, and can be implemented by a software module including a PID control algorithm, a digital signal processor, or a dedicated control chip.

[0114] The execution control module 205 is a functional unit responsible for converting the corrected instructions into actual motion control signals of the robot, which can be implemented by a robot controller interface, a motor driver or a motion control software module.

[0115] This system provides a structural framework for implementing a robot navigation path planning method based on real-time electromyographic signal feedback. The preset navigation module 201 in the system first obtains an initial preset navigation path and controls the surface interactive robot to begin interactive work on the target muscle group according to this path, providing a basic starting point for subsequent dynamic adjustments.

[0116] During the interactive task execution process, the real-time acquisition module 202 continuously collects the electromyographic signals of the target muscle group and performs necessary processing on these raw signals to extract a real-time muscle group activation index that can reflect the current activation state of the muscle group. This real-time index is sent to the comparison module 203, which compares it with the preset target muscle group activation target value, calculates the difference between the two, and generates an error signal. This error signal quantifies the deviation between the current interactive task effect and the desired target. Subsequently, the adjustment module 204 receives the error signal, inputs it into the internal PID controller, and calculates the corresponding control variable according to the control algorithm. Based on this control variable, the adjustment module 204 corrects the navigation path parameters of the surface interactive robot and generates a new navigation path instruction. Finally, the execution control module 205 receives the corrected navigation path instruction and controls the surface interactive robot to perform the interactive task operation according to the new instruction.

[0117] In this way, the system forms a real-time closed-loop feedback control loop, enabling the surface interactive robot's navigation path to be dynamically adjusted based on the actual physiological responses of the target muscle groups. This systematic structure enables efficient and accurate collaborative execution of the various steps in the method, ensuring that real-time physiological feedback can promptly and effectively influence the robot's movements. This overcomes the problem of relying solely on preset paths that cannot adapt to real-time physiological conditions and achieves adaptive optimization of interactive work paths.

[0118] In this document, relational terms such as first and second, etc. are used merely to distinguish one entity or operation from another entity or operation, but do not necessarily require or imply any actual relationship or order between these entities or operations.

[0119] The foregoing is merely an embodiment of the present application and is not intended to limit the scope of protection of the present application. Persons skilled in the art will readily appreciate that the present application may be modified and altered in various ways. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present application shall be included within the scope of protection of the present application.

Claims

1. A robot navigation path planning method, characterized in that: The method comprises the steps of: S1: Obtain a preset navigation path, and control the surface interactive robot to perform interactive operations on the target muscle group according to the preset navigation path; S2: collecting the electromyographic signals of the target muscle group during the interactive operation in real time, processing the electromyographic signals, and extracting a real-time muscle group activation index; S3: Compare the real-time muscle group activation index with the preset target muscle group activation target value and calculate an error signal; S4: inputting the error signal into a PID controller to calculate a control quantity, and adjusting the navigation path parameters of the surface interactive robot according to the control quantity to obtain a corrected navigation path instruction; S5: The control surface interactive robot performs interactive operations on the target muscle groups according to the corrected navigation path instructions.

2. A robot navigation path planning method according to claim 1, characterized in that: Step S1 includes: S11: Collect the user's body shape data, muscle distribution data, and pain sensitivity data; S12: adjusting the coverage of the preset navigation path according to the collected body shape data and muscle distribution data to obtain an individualized coverage; S13: adjusting the surface interaction force of the preset navigation path according to the collected pain sensitivity data to obtain an individualized interaction force; S14: Control the surface interactive robot to perform interactive operation on the target muscle group according to the individualized coverage range and the individualized interactive operation strength.

3. A robot navigation path planning method according to claim 2, characterized in that: Step S12 includes: S121: Establishing a first mapping relationship between the body shape data and the coverage range, and a second mapping relationship between the muscle distribution data and the coverage range; S122: Query the first mapping relationship according to the body shape data to obtain an initial coverage range corresponding to the body shape; S123: querying the second mapping relationship according to the muscle distribution data to obtain a coverage adjustment parameter corresponding to the muscle distribution; S124: Adjust the initial coverage according to the coverage adjustment parameter to obtain the personalized coverage.

4. A robot navigation path planning method according to claim 1, characterized in that: Step S2 includes: S21: using the electromyographic signal acquisition unit to collect raw electromyographic signals of the target muscle group during the interactive operation in real time at a preset sampling frequency; S22: Filtering the original electromyographic signal to remove noise interference and power frequency interference to obtain a filtered electromyographic signal; S23: Perform time domain analysis on the filtered electromyographic signal and calculate the root mean square (RMS) value as an indicator of the degree of muscle group activation; S24: Normalize the RMS value to obtain a real-time muscle group activation index.

5. A robot navigation path planning method according to claim 4, characterized in that: Step S23 includes: S231: Construct an EMG signal power spectrum model, analyze the frequency distribution of the filtered EMG signal, and determine the dominant frequency range; S232: Designing a bandpass filter according to the dominant frequency range to extract the myoelectric signal component within the dominant frequency range; S233: Performing square integration processing on the electromyographic signal component to obtain an electromyographic signal energy value; S234: Perform time sliding average filtering and square root processing on the electromyographic signal energy value to obtain an RMS value reflecting the degree of muscle group activation.

6. A robot navigation path planning method according to claim 1, characterized in that: Step S3 includes: S31: Construct a multivariate regression model that includes robot pressure parameters, temperature parameters, and real-time physiological state parameters of the target muscle group, and quantify the influence weight of each parameter on the electromyographic signal; S32: Correcting the real-time muscle group activation index according to the multivariate regression model to obtain a corrected muscle group activation index; S33: Compare the corrected muscle group activation index with the preset target muscle group activation target value to calculate an error signal.

7. A robot navigation path planning method according to claim 1, characterized in that: Step S4 includes: S41: inputting the error signal into a PID controller, and calculating a preliminary control variable according to the error signal; S42: Obtain the current configuration parameters of the surface interactive robot and determine whether the current configuration satisfies the singular configuration neighborhood condition; S43: If the current configuration satisfies the singular configuration neighborhood condition, the preliminary control amount is adjusted based on the current configuration parameters according to the singular configuration avoidance algorithm to obtain an adjusted control amount that avoids the singular configuration; S44: If the current configuration does not satisfy the singular configuration neighborhood condition, the navigation path parameters of the surface interactive robot are adjusted according to the adjusted control amount or the preliminary control amount to obtain a corrected navigation path instruction.

8. A robot navigation path planning method according to claim 7, characterized in that: Step S41 includes: S411: Obtain the proportional coefficient, integral coefficient and differential coefficient of the PID controller; S412: Calculate a proportional term based on the error signal and the proportional coefficient; calculate an integral term based on the error signal and the integral coefficient; calculate a differential term based on the error signal and the differential coefficient; S413: Add the proportional term, the integral term, and the differential term to obtain the preliminary control variable.

9. A robot navigation path planning method according to claim 7, characterized in that: Step S42 includes: S421: Storing multiple singular configuration neighborhood conditions, each singular configuration neighborhood condition corresponds to a joint angle combination and a safety threshold; S422: Obtain the current joint angle values ​​of the robot and calculate the current configuration parameters based on the joint angle values; S423: Compare the current configuration parameters with multiple stored singular configuration neighborhood conditions. If the current configuration parameters meet any of the singular configuration neighborhood conditions and exceed the safety threshold corresponding to the singular configuration neighborhood condition, then it is determined that the current configuration meets the singular configuration neighborhood condition.

10. A robot navigation path planning system, characterized in that: In the steps of the method according to any one of claims 1 to 9, the system comprises: Preset navigation module: obtains a preset navigation path and controls the surface interactive robot to perform interactive operations on the target muscle group according to the preset navigation path; Real-time acquisition module: collects the electromyographic signals of the target muscle group during the interactive operation in real time, processes the electromyographic signals, and extracts real-time muscle group activation indicators; Comparison module: compares the real-time muscle group activation index with the preset target muscle group activation target value and calculates the error signal; Adjustment module: inputs the error signal into the PID controller, calculates the control amount, and adjusts the navigation path parameters of the surface interactive robot according to the control amount to obtain a corrected navigation path instruction; Execution control module: controls the surface interactive robot to perform interactive operations on the target muscle groups according to the corrected navigation path instructions.

Images