A triangular grating teleoperation perception method and device based on force signal reconstruction

Abstract

The application provides a triangular grating teleoperation perception method and device based on force signal reconstruction. The method provided by the application comprises: collecting real-time real force signals generated when a slave end mechanical arm contacts a triangular grating surface; performing time domain optimization processing on the real-time real force signals to obtain time domain optimization signals; performing frequency domain optimization processing on the real-time real force signals to extract dominant frequency components and corresponding phase information in the real-time real force signals, and to obtain frequency domain optimization signals containing frequency characteristics; performing time domain and frequency domain hybrid optimization processing on the time domain optimization signals and the frequency domain optimization signals, retaining time domain amplitude characteristics of the time domain optimization signals, fusing the dominant frequency components and the phase information in the frequency domain optimization signals, and reconstructing a three-dimensional force waveform through a trigonometric function method; generating a master end feedback force signal based on the three-dimensional force waveform, driving a master end actuator to generate a physical feedback through the master end feedback force signal, and performing teleoperation perception of the triangular grating surface properties based on the physical feedback.

Description

Technical Field

[0001] This application relates to the field of teleoperation sensing technology, and in particular to a method and apparatus for teleoperation sensing based on force signal reconstruction using a triangular grating. Background Technology

[0002] In the field of teleoperation technology, triangular gratings, due to their precisely controllable periodic textures and spatial geometric features, are often used as a core carrier for verifying the accuracy of teleoperation sensing. The period and amplitude of their surface textures directly reflect the microstructural properties of objects. Accurate sensing of the surface properties of triangular gratings through teleoperation can be widely applied in scenarios such as precision manufacturing inspection, remote medical device operation, and space exploration equipment maintenance. This provides a technological foundation for tactile feedback in subsequent complex environments, thus making research on triangular grating teleoperation sensing of significant practical application value.

[0003] Current mainstream teleoperation sensing technologies are mainly divided into two categories: one is a vibration feedback-based sensing scheme, which tracks and records position, velocity, and acceleration signals from the slave device, models them as force signals with tuning parameters, and uses a DAC to generate control signals to drive a voice coil to produce a Lorentz force, pushing the master device's handle to move linearly, simulating texture perception with high-frequency acceleration feedback; the other is an electro-electro-tactile sensing scheme, which utilizes the capacitance effect between the finger and the electrodes of the tactile display, modulating the electrostatic attraction by controlling the waveform, amplitude, and frequency of the input voltage to change the finger's sliding friction, thereby rendering a virtual grating texture. Both schemes help operators indirectly perceive the surface features of remote objects through physical feedback or tactile simulation at the master device. However, for vibration feedback schemes, the feedback force is calculated based on discrete acceleration measurements, which cannot fully simulate real tactile information such as friction and shear force, and it relies too much on high-frequency acceleration feedback, missing low-frequency force characteristics such as object hardness and elasticity, making it difficult for users to distinguish the feel of different materials; at the same time, most schemes use single-axis linear drive, ignoring lateral friction and shear force, and cannot reproduce the three-dimensional microstructure of the triangular grating surface. For electrostatic tactile solutions, the feedback is based on an independently constructed virtual force field, which is not associated with the physical information of the actual contact at the end. This means that the perceived virtual texture has no direct mapping relationship with the real surface properties of the remote triangular grating, which violates the "transparency" principle of teleoperation and cannot transmit real force information.

[0004] Therefore, there is an urgent need for a method to achieve three-dimensional real force feedback sensing of triangular gratings, thereby improving the accuracy of remote operation sensing of triangular gratings and the user experience. Summary of the Invention

[0005] In view of this, this application provides a method and apparatus for teleoperation sensing of triangular gratings based on force signal reconstruction, so as to realize three-dimensional real force feedback sensing of triangular gratings and improve the accuracy of teleoperation sensing of triangular gratings and user experience.

[0006] Specifically, this application is implemented through the following technical solution:

[0007] The first aspect of this application provides a method for teleoperation sensing of a triangular grating based on force signal reconstruction, the method comprising:

[0008] Real-time force signals generated when the robotic arm contacts the surface of the triangular grating are collected.

[0009] The real-time force signal is subjected to time-domain optimization processing. The signal amplitude is scaled by a position gain variable and a unit conversion factor to obtain a time-domain optimized signal. The position gain variable is dynamically adjusted according to the distance between the end effector and the surface of the triangular grating.

[0010] The real-time real force signal is subjected to frequency domain optimization processing. The dominant frequency component and corresponding phase information in the real-time real force signal are extracted by fast Fourier transform to obtain a frequency domain optimized signal containing frequency features. The dominant frequency component corresponds to the periodic features of the triangular grating surface texture.

[0011] The time-domain optimized signal and the frequency-domain optimized signal are subjected to time-domain and frequency-domain hybrid optimization processing. The time-domain amplitude characteristics of the time-domain optimized signal are preserved, and the dominant frequency components and phase information in the frequency-domain optimized signal are fused. The three-dimensional force waveform is reconstructed by the trigonometric function method. The three-dimensional force waveform corresponds to the spatial geometric properties of the triangular grating surface.

[0012] Based on the three-dimensional force waveform, a master-end feedback force signal is generated. The master-end feedback force signal drives the master-end actuator to generate physical feedback. Based on the physical feedback, the remote sensing of the surface properties of the triangular grating is performed.

[0013] The second aspect of this application provides a triangular grating teleoperation sensing device based on force signal reconstruction, the device comprising an acquisition module, a processing module, a reconstruction module and a sensing module;

[0014] The acquisition module is used to acquire real-time force signals generated when the robotic arm contacts the surface of the triangular grating.

[0015] The processing module is used to perform time-domain optimization processing on the real-time real force signal, and to scale the signal amplitude by using a position gain variable and a unit conversion factor to obtain a time-domain optimized signal; the position gain variable is dynamically adjusted according to the distance between the end effector and the surface of the triangular grating.

[0016] The processing module is also used to perform frequency domain optimization processing on the real-time real force signal, and to extract the dominant frequency component and corresponding phase information in the real-time real force signal by using fast Fourier transform to obtain a frequency domain optimized signal containing frequency features; the dominant frequency component corresponds to the periodic features of the triangular grating surface texture.

[0017] The reconstruction module is used to perform time-domain and frequency-domain hybrid optimization processing on the time-domain optimized signal and the frequency-domain optimized signal, retain the time-domain amplitude characteristics of the time-domain optimized signal, fuse the dominant frequency components and phase information in the frequency-domain optimized signal, and reconstruct the three-dimensional force waveform using the trigonometric function method; the three-dimensional force waveform corresponds to the spatial geometric properties of the triangular grating surface;

[0018] The sensing module is used to generate a main-end feedback force signal based on the three-dimensional force waveform. The main-end feedback force signal drives the main-end actuator to generate physical feedback, and the teleoperation sensing of the surface properties of the triangular grating is performed based on the physical feedback.

[0019] The triangular grating teleoperation sensing method and device based on force signal reconstruction provided in this application have a complete process setup of acquiring real-time force signals → time domain optimization → frequency domain optimization → hybrid optimization to reconstruct three-dimensional force waveforms → generating master feedback. First, real-time force signals from the contact between the robotic arm and the triangular grating are acquired by the slave end, ensuring that the raw data directly reflects the essence of physical contact and avoids the disconnect between the virtual force field and the real surface. Next, time-domain optimization is performed, using position gain variables and unit conversion factors to scale the amplitude. The position gain variable is dynamically adjusted with distance, which enhances the amplitude differentiation of gratings of different sizes and achieves unit standardization, providing accurate time-domain features for subsequent processing. Then, frequency-domain optimization is performed, using Fast Fourier Transform to extract the dominant frequency components and phase information corresponding to the grating texture period, filtering out noise and retaining frequency features, solving the high-frequency dependence and information loss problems in traditional vibration feedback. Subsequently, hybrid optimization is performed, fusing the time-domain amplitude features with the frequency and phase information in the frequency domain, and reconstructing the three-dimensional force waveform using trigonometric functions. This waveform simultaneously carries the amplitude as an intensity feature, the frequency as a periodic feature, and the phase as a spatial position correspondence, corresponding to the spatial geometric properties of the grating. Finally, a master-end feedback force signal is generated based on the three-dimensional waveform, driving the actuator to produce multi-dimensional physical feedback. This end-to-end setup allows the master-end feedback to completely reproduce the three-dimensional force information from the slave end contact during remote operation sensing. Operators can perceive the unevenness of the grating by the intensity difference of the feedback, distinguish the texture period by the frequency change, and judge the spatial shape by the directional features. This solves the problems of incomplete tactile information, single feedback direction and insufficient realism in traditional technology. It realizes a closed loop from real force signal acquisition to multi-feature fusion and then to accurate feedback, which greatly improves the realism, accuracy and spatial three-dimensionality of remote operation perception, and ensures that users can clearly distinguish the surface properties of the triangular grating as if they were touching it directly. Attached Figure Description

[0020] Figure 1 A flowchart of the triangular grating teleoperation sensing method based on force signal reconstruction provided in Embodiment 1 of this application;

[0021] Figure 2 This is a schematic diagram of the structure of the triangular grating teleoperation sensing device based on force signal reconstruction provided in Embodiment 2 of this application. Detailed Implementation

[0022] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application.

[0023] The terminology used in this application is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. The singular forms “a,” “the,” and “the” used herein are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used herein refers to and includes any and all possible combinations of one or more of the associated listed items.

[0024] It should be understood that although the terms first, second, third, etc., may be used in this application to describe various information, such information should not be limited to these terms. These terms are only used to distinguish information of the same type from one another. For example, without departing from the scope of this application, first information may also be referred to as second information, and similarly, second information may also be referred to as first information. Depending on the context, the word "if" as used herein may be interpreted as "when," "when," or "in response to determination."

[0025] The following specific embodiments are given to illustrate the technical solution of this application in detail.

[0026] Figure 1 This is a flowchart of the triangular grating teleoperation sensing method based on force signal reconstruction provided in Embodiment 1 of this application. Please refer to... Figure 1 The method provided in this embodiment may include:

[0027] S101. Collect the real-time force signal generated when the robotic arm contacts the surface of the triangular grating.

[0028] Specifically, the slave robotic arm is a mechanical actuator deployed in a remote working environment within a teleoperation system. These remote working environments include precision inspection stations and hazardous work areas. Its core function is to drive the end effector to make physical contact with a target object based on control signals transmitted from the master end. The end effector includes a gripper and a tactile sensor probe. The target object in this case is a triangular grating. During physical contact, the slave robotic arm simultaneously collects various physical signals. The triangular grating surface refers to a planar structure with periodic triangular protrusions / grooves. Its core characteristics are that the periodic and amplitude parameters of the texture are precisely controllable. The periodic parameter refers to the spacing between adjacent texture units, and the amplitude parameter refers to the protrusion height / groove depth, and the surface geometry is stable. The real-time force signal refers to the raw force data directly collected by the force sensor during the dynamic contact between the end effector of the slave robotic arm and the triangular grating surface, without any signal processing.

[0029] In specific implementation, the slave robotic arm collects the current force signal generated by contact through a force sensor; the difference between the current force signal and a preset contact force threshold is taken as the force deviation; using preset virtual mass, damping, and stiffness parameters as adjustment coefficients, the force deviation is converted into the target position correction amount of the end effector; wherein, the virtual mass parameter controls the response speed of position adjustment, the damping parameter suppresses vibration during the adjustment process, and the stiffness parameter controls the accuracy of position correction; based on the target position correction amount of the end effector and combined with the kinematic model of the slave robotic arm, the required rotation angle of each joint is calculated as the joint adjustment amount; the robotic arm joints are driven to move according to the joint adjustment amount, and the end effector is controlled to adapt to the undulation changes of the triangular grating surface in position and attitude. During the dynamic adjustment of the robotic arm, the force sensor synchronously and continuously collects the real force signal generated by contact.

[0030] Specifically, a six-dimensional force sensor is installed on the end effector of the robotic arm to synchronously collect force signals generated by contact with the surface of the triangular grating, including force components in the x, y, and z directions, forming the current force signal. This signal is transmitted in real time to the processor of the robotic arm control system via a data bus. The control system calls a preset parameter library to read the contact force threshold set for the triangular grating material; the force deviation is obtained through subtraction. The control system loads the admittance control algorithm module and calls the preset virtual parameter group: virtual mass, for example, 0.5 kg, which affects the adjustment speed; the smaller the value, the faster the response; virtual damping, for example, 20 N·s / m, which is used to suppress vibration; the larger the value, the stronger the buffering effect; and virtual stiffness, for example, 500 N / m, which is used for control accuracy; the larger the value, the more sensitive the correction. Taking the force deviation as input, it is transformed into the displacement correction value of the end effector in three-dimensional space through a second-order differential equation. This displacement correction value can be expressed as Δx, Δy, and Δz. The second-order differential equation combines the virtual mass, damping, and stiffness parameters, where the x and y direction corrections are used to compensate for lateral offset, and the z direction correction is used to adjust the vertical distance. The control system calls upon the kinematic model of the slave-end robotic arm, such as the DH parameter model, which pre-stores the coordinate mapping relationship between each joint and the end effector. The target position correction is input into the inverse kinematics algorithm to calculate the required angular increment for each rotary joint, such as the J1-J6 joints of a 6-axis robotic arm, i.e., the joint adjustment amount. Simultaneously, joint limit checks are performed to ensure that the calculated angular increment does not exceed the physical movement range of each joint, such as ±180°; if it exceeds this range, it is automatically truncated to the limit value. The control system converts the joint adjustment amount into pulse signals and sends them to the servo drivers of each joint. The drivers control the motor rotation based on the pulse frequency and quantity, driving the end effector of the robotic arm to move, adapting it in real-time to the undulations of the triangular grating surface. If a grating protrusion is encountered, the z-axis height is automatically fine-tuned to maintain stable contact. During the robotic arm's movement, the force sensor continuously collects contact force signals at the original sampling frequency.

[0031] The method provided in this embodiment constructs a dynamically adapted, precisely contacted, and continuously stable acquisition environment for real-time force signal acquisition. First, by synchronously acquiring distance parameters and force signals, initial data support in both position and force dimensions is provided for subsequent adjustments. Then, the difference between the end effector and the ideal contact state is clarified by calculating the deviation. Next, the coordinated adjustment of three parameters—virtual mass, damping, and stiffness—is utilized. Virtual mass controls the position adjustment response speed to ensure that force signal acquisition does not miss contact details due to adjustment lag. Damping parameters suppress vibration to avoid vibration interference that causes noise and distortion in the force signal. Stiffness parameters control the position correction accuracy to ensure that the end effector can accurately conform to the microscopic undulations of the triangular grating surface. This is then converted into joint adjustment amounts through a kinematic model to drive the robotic arm movement, ensuring that the end effector always adapts to the grating surface undulations in position and orientation. This dynamic adaptation process ensures that the end effector and the triangular grating surface are always in a stable and precise contact state, rather than unstable contact such as detachment or excessive compression. Based on this, the real-time force signal synchronously and continuously acquired by the force sensor will more closely match the actual surface of the triangular grating. The physical properties of the force signal reduce the loss, distortion, or noise caused by unstable contact or delayed adjustment. This high-quality, real-time force signal provides accurate and complete raw data support for the generation of subsequent master feedback force signals, avoiding the disconnect between master feedback and actual slave contact caused by inaccurate raw force signals. This allows the physical feedback generated by the master actuator to be closer to the real tactile sensation of the triangular grating surface, ultimately helping users to perceive the microstructure, hardness, friction characteristics, and other properties of the triangular grating surface more accurately and clearly during remote operation. This improves the realism, accuracy, and stability of remote operation perception, and solves the problems of blurred tactile sensation and disconnect from the real environment caused by poor force signal quality in traditional remote operation perception.

[0032] S102. Perform time-domain optimization processing on the real-time real force signal, and scale the signal amplitude by using a position gain variable and a unit conversion factor to obtain a time-domain optimized signal; the position gain variable is dynamically adjusted according to the distance between the end effector and the surface of the triangular grating.

[0033] Specifically, the position gain variable is a coefficient used to dynamically adjust the amplitude of the real-time force signal. Its value is not fixed but adaptively adjusted based on the real-time distance between the end effector and the triangular grating surface. When the end effector is close to the grating surface, the variable is greater than 1 to amplify the amplitude of the original force signal and enhance the differentiation of amplitude characteristics between grating patterns of different sizes. When the distance is greater, the value is less than 1 to reduce the signal amplitude and avoid signal distortion caused by insufficient contact force. The unit conversion factor is a preset fixed value used to convert the force signal amplitude value after position gain adjustment into a unified standard unit. Optionally, the unit conversion factor is 0.001.

[0034] Furthermore, the time-domain optimized signal is the signal obtained after performing time-domain optimization processing on the real force signal. It retains the changing trend and key features of the original force signal in the time dimension, such as the peak and valley values ​​and the rhythm of change of the force signal. At the same time, the differentiation of different grating patterns is enhanced by amplitude scaling, data consistency is achieved by unit conversion, and signal fluctuations caused by parameter mutations are reduced by smoothing processing.

[0035] In specific implementation, the time-domain optimization processing of the real-time real force signal, which involves scaling the signal amplitude using a position gain variable and a unit conversion factor to obtain a time-domain optimized signal, includes: acquiring the original amplitude value of the real-time real force signal; acquiring the real-time distance parameter between the end effector and the triangular grating surface; introducing a position gain variable and multiplying it by the original amplitude value to obtain an adjusted amplitude value; adjusting the position gain variable inversely with the real-time distance parameter; multiplying the adjusted amplitude value by the unit conversion factor to convert the signal amplitude to a unified standard unit to obtain a standardized amplitude value; and using a continuous scaling function to perform secondary adjustment on the standardized amplitude value, integrating the processed amplitude values ​​to generate the time-domain optimized signal.

[0036] Specifically, the raw amplitude value of the signal is extracted frame by frame from the real-time force signal output by the force sensor, which can be a six-dimensional force sensor. The signal acquisition module reads the force components in the x, y, and z axes of each frame, and uses peak detection to determine the raw amplitude value of a single frame. The maximum absolute value of the force components in the three axes within a single frame is directly taken as the raw amplitude value of that frame. The position sensor integrated into the end effector continuously measures the perpendicular distance to the triangular grating surface at a preset sampling frequency, outputting a real-time distance parameter. This real-time distance parameter is input into a preset distance-gain mapping table. The position sensor can be a laser rangefinder. The mapping table pre-stores multiple distance intervals and corresponding gain values: for example, when the distance is ≤0.1mm, the position gain variable is set to 1.3-1.5; when 0.1mm < distance ≤0.5mm, it is set to 1.0-1.2; and when the distance >0.5mm, it is set to 0.7-0.9. The control system matches the corresponding interval based on the current distance value and automatically calls the position gain variable for that interval. In the signal processing unit, the extracted raw amplitude value of each frame is multiplied by the determined current position gain variable to obtain the adjusted amplitude value. The adjusted amplitude value is then multiplied by a fixed unit conversion factor to ensure that all amplitude values ​​are uniformly converted to standard mechanical units, eliminating unit differences caused by different devices or sensors.

[0037] Optionally, the step of using a continuous scaling function to perform secondary adjustment on the standardized amplitude value, integrating the processed amplitude value, and generating a time-domain optimized signal includes: presetting a reference scaling factor and a smooth adjustment range for the continuous scaling function; the continuous scaling function takes the real-time displacement change rate of the end effector as an input parameter, and the output value of the continuous scaling function is within the smooth adjustment range; when the real-time displacement change rate is positive, the continuous scaling function gradually increases the output value linearly or curvilinearly from the reference scaling factor according to the increase in the real-time displacement change rate; when the real-time displacement change rate is negative, the continuous scaling function gradually decreases the output value linearly or curvilinearly from the reference scaling factor according to the increase in the absolute value of the real-time displacement change rate; when the end effector is stationary or the real-time displacement change rate is close to zero, the output value of the continuous scaling function gradually returns to the reference scaling factor; the output value of the continuous scaling function is multiplied by the standardized amplitude value to obtain the secondary adjusted amplitude value; the secondary adjusted amplitude value is time-series integrated to remove fluctuations caused by function transitions, forming a time-domain optimized signal.

[0038] In practice, the reference scaling factor and smoothing adjustment range of the continuous scaling function are pre-set in the signal processing system. This range is determined based on the surface texture accuracy of the triangular grating and the tactile feedback requirements, ensuring that subsequent output values ​​are always within the effective adjustment range. Real-time position data of the end effector is collected from the position sensor of the robotic arm. The difference between two adjacent frames of position data is calculated according to a preset sampling period, and then divided by the sampling period to obtain the real-time displacement change rate of the end effector. This change rate is used as the input parameter of the continuous scaling function. If the real-time displacement rate of change is positive, meaning the end effector is moving closer to the triangular grating surface, the function gradually increases the output value linearly or curvilinearly, starting from the reference scaling factor, based on the increase in the rate of change, and the output value does not exceed the upper limit of the smooth adjustment range. If the real-time displacement rate of change is negative, meaning the end effector is moving away from the triangular grating surface, the function gradually decreases the output value linearly or curvilinearly, starting from the reference scaling factor, based on the increase in the absolute value of the rate of change, and the output value does not fall below the lower limit of the smooth adjustment range. If the end effector is stationary, meaning the position data has not changed or the real-time displacement rate of change is close to zero, the function uses a progressive regression algorithm to gradually bring the output value closer to the reference scaling factor from the current value until it stabilizes at the reference value. The specific implementation process of the regression algorithm can be found in the description in related technologies. The standardized amplitude value after unit conversion is retrieved and multiplied by the scaling factor currently output by the continuous scaling function to obtain the secondary adjusted amplitude value of each frame of signal. All the secondary adjusted amplitude values ​​are integrated in time sequence according to the timestamp order. A moving average filtering algorithm is used, such as taking the average of the amplitude values ​​of 5 adjacent frames as the final value of the current frame, to remove the small fluctuations caused by the over-adjustment of the scaling function, and finally form a continuous and smooth time-domain optimized signal.

[0039] S103. Perform frequency domain optimization processing on the real-time real force signal, and use Fast Fourier Transform to extract the dominant frequency component and corresponding phase information in the real-time real force signal to obtain a frequency domain optimized signal containing frequency features; the dominant frequency component corresponds to the periodic features of the triangular grating surface texture.

[0040] Specifically, the dominant frequency component refers to the frequency element with the strongest energy and highest proportion after frequency domain decomposition of the real-time real force signal, and this frequency component directly corresponds to the periodic characteristics of the triangular grating surface texture. Phase information refers to the signal phase parameters extracted synchronously with the dominant frequency component during the Fast Fourier Transform process, used to describe the starting position or time offset of this frequency component in the time dimension. The frequency-optimized signal refers to the frequency domain signal after frequency domain processing of the original real-time real force signal, retaining effective frequency characteristics and filtering out invalid noise interference.

[0041] In specific implementation, the real-time real force signal is preprocessed to remove DC components and noise interference; the preprocessed real-time real force signal is segmented according to a preset time window, and a fast Fourier transform is performed on each segment to convert the time-domain signal into a frequency-domain signal, obtaining the energy distribution characteristics of each frequency component; frequency components with energy values ​​exceeding a preset threshold are selected from the frequency-domain signal, and the frequency component with the highest energy is determined as the dominant frequency component; the phase information of the dominant frequency component in the frequency domain is extracted; the continuity of the dominant frequency component and phase information extracted in each time window is verified, and outliers caused by signal mutations are eliminated; the verified dominant frequency component and corresponding phase information are integrated to generate a frequency-domain optimized signal containing frequency characteristics.

[0042] Specifically, the original signal is first processed using a digital high-pass filter to remove DC components, such as the constant bias caused by sensor zero drift. Then, a wavelet denoising algorithm is used to decompose the signal into multiple scales, eliminating high-frequency noise components and retaining the effective fluctuation signals related to the triangular grating contact within the 5-500Hz range, outputting the preprocessed time-domain signal. Preset time window parameters, including window length and window overlap rate, are used in the system. The preprocessed signal is then segmented according to these parameters, generating an independent sub-signal sequence for each window, with a start timestamp added to each sub-signal. A Fast Fourier Transform (FFT) operation is performed on the sub-signals of each time window. Before the operation, the signal is zero-padded, and then the FFT algorithm converts the time-domain sub-signals into frequency-domain signals, obtaining spectral data with a frequency range of 0-10kHz and a frequency resolution of 10Hz. This data includes amplitude information corresponding to each frequency point, forming the energy distribution characteristics of the window. Specifically, the energy distribution characteristics are represented as a spectrum plot with frequency as the horizontal axis and energy as the vertical axis. The energy values ​​of each frequency component are extracted from the output spectral data. A preset energy threshold is applied to filter candidate frequency components whose energy values ​​exceed the threshold. The candidate components are sorted in descending order of energy value, and the frequency component with the highest energy value is selected as the dominant frequency component of the current window. The phase angle corresponding to the dominant frequency component is read from the Fast Fourier Transform result. The phase angle is calculated from the spectral data in complex form, where phase = arctan(imaginary part / real part). The difference between the dominant frequency components of adjacent windows is calculated. If the difference exceeds a preset frequency threshold, the dominant frequency of the current window is determined to be an outlier. The phase difference between adjacent windows is calculated. If the absolute value of the difference exceeds 180°, it is determined to be a phase outlier. Linear interpolation is used to correct outliers, such as interpolating the effective values ​​of the previous and next windows to ensure the continuity of the frequency and phase sequences. The validated dominant frequency components and corresponding phase information of all windows are arranged in timestamp order to form a sequence containing three-dimensional features of time, frequency, and phase, generating a continuous and stable frequency domain optimized signal. For details on the implementation process of the Fast Fourier Transform, please refer to the descriptions in related technologies; they will not be repeated here.

[0043] S104. Perform time-domain and frequency-domain hybrid optimization processing on the time-domain optimized signal and the frequency-domain optimized signal, retain the time-domain amplitude characteristics of the time-domain optimized signal, fuse the dominant frequency components and phase information in the frequency-domain optimized signal, and reconstruct the three-dimensional force waveform using the trigonometric function method; the three-dimensional force waveform corresponds to the spatial geometric properties of the triangular grating surface.

[0044] Specifically, a three-dimensional force waveform refers to a visualized dynamic waveform that quantitatively reflects the force variation when the end effector contacts the surface of a triangular grating in both time and space dimensions, and directly corresponds to the spatial geometric properties of the triangular grating. The spatial geometric properties of the triangular grating include tooth height, tooth pitch, and tilt angle. Its core is to transform abstract force signals into three-dimensional data with spatial physical meaning by fusing three key parameters: time-domain amplitude characteristics, dominant frequency components, and phase information.

[0045] In specific implementation, the time-domain optimized signal and frequency-domain optimized signal undergo time-domain and frequency-domain hybrid optimization processing, retaining the time-domain amplitude characteristics of the time-domain optimized signal and fusing the dominant frequency component and phase information in the frequency-domain optimized signal. The three-dimensional force waveform is reconstructed using the trigonometric function method, including: for any target axis among the x, y, and z axes, extracting the local force signal peaks where the target axis contacts the triangular grating surface from the time-domain optimized signal, calculating the arithmetic mean of the local force signal peaks, and determining the amplitude reference of the force on the target axis; extracting the dominant frequency component and phase information corresponding to the target axis from the frequency-domain optimized signal; constructing a trigonometric function force model of the target axis based on the amplitude reference, dominant frequency component, and phase information, and reconstructing a single-axis force waveform; wherein, the x-axis force waveform corresponds to the frictional force variation characteristics of the triangular grating surface sliding laterally, the y-axis force waveform corresponds to the shear force variation characteristics of the triangular grating surface along the longitudinal direction, and the z-axis force waveform corresponds to the pressure variation characteristics of the triangular grating surface in perpendicular contact; and integrating the single-axis force waveforms of the x, y, and z axes to form a complete three-dimensional force waveform.

[0046] Specifically, processing is performed separately for the x-axis, y-axis, and z-axis. Taking the x-axis as an example: Local force signal peaks when the x-axis contacts the triangular grating surface are extracted from the time-domain optimized signal. A preset number of consecutive peak values ​​are counted, and their arithmetic mean is calculated. This mean is determined as the amplitude reference for the force on the target axis. The dominant frequency component corresponding to the x-axis and its phase information are read from the frequency-domain optimized signal. The dominant frequency component corresponding to the x-axis is, for example, the frequency value corresponding to the lateral sliding of the x-axis, and its phase information is, for example, the initial phase angle. Using the amplitude reference as the amplitude parameter of the trigonometric function, the angular frequency is calculated using the dominant frequency component, where angular frequency = 2π × dominant frequency. Using the phase information as the initial phase parameter, a model is constructed using a sine function. The model outputs the force value corresponding to time, reconstructing the single-axis force waveform of the target axis. Reconstruct the single-axis force waveforms of the y-axis and z-axis respectively according to the above steps, align the single-axis force waveforms of the x-axis, y-axis and z-axis according to the time axis, and integrate them into a complete three-dimensional force waveform containing the force change characteristics of the three axes. Each time point corresponds to the force value data of the three axes.

[0047] Optionally, the step of constructing a trigonometric function force model of the target axis based on the amplitude reference, dominant frequency component, and phase information includes: setting the amplitude reference as the maximum fluctuation amplitude of the trigonometric function; converting the dominant frequency component into the periodic parameter of the trigonometric function; setting the phase information as the starting position parameter of the trigonometric function; and integrating the maximum fluctuation amplitude, periodic parameter, and starting position parameter to construct a sinusoidal continuous force model; wherein the fluctuation amplitude of the waveform is determined by the amplitude reference, the repetition frequency of the waveform is determined by the dominant frequency component, and the initial state of the waveform is determined by the phase information.

[0048] In practice, an amplitude reference is extracted from the force characteristic parameters of the target axis and directly set as the maximum fluctuation amplitude of the trigonometric function, i.e., the amplitude parameter of the sine function. The dominant frequency component corresponding to the target axis is converted, and the period parameter of the trigonometric function is calculated using the formula: period T = 1 / dominant frequency. The extracted phase information is set as the starting position parameter of the trigonometric function. The above parameters are integrated to construct a sinusoidal continuous force model. This model is used to calculate the force values ​​at different time points, generating continuous waveform data.

[0049] For example, in one embodiment, when the target axis is the x-axis, the trigonometric function force model can be expressed as:

[0050] ;

[0051] Among them, the The force model is a trigonometric function; As the amplitude reference; the The dominant frequency component; For phase information; the For time.

[0052] The method provided in this embodiment first determines the amplitude benchmark by extracting the mean of the time-domain peaks on each axis. This accurately anchors the strength benchmark of the contact force on each axis, avoiding amplitude deviations caused by local fluctuations in the single-axis force signal and ensuring that the subsequent waveform closely matches the actual contact situation in terms of force strength. Next, by extracting the dominant frequency components and phase information in the frequency domain on each axis, the period of each axis waveform is made consistent with the texture period in the corresponding direction of the triangular grating, and the initial state is precisely matched with the contact position. For example, the dominant frequency corresponding to the transverse texture on the x-axis allows the repetition rhythm of the transverse friction force waveform to match the texture sliding period, while the phase information ensures that the waveform's starting point is aligned with the texture contact starting point. Then, a trigonometric function model is constructed based on these parameters to obtain the single-axis waveform, clearly defining the transverse friction force on the x-axis, the longitudinal pressure on the y-axis, and the vertical contact pressure on the z-axis. This allows each axis waveform to carry the physical contact characteristics of the triangular grating in different directions, avoiding dimensional confusion. Finally, the three-axis waveforms are integrated to form a three-dimensional force waveform, which transforms the spatial geometric properties of the triangular grating surface into a multi-dimensional, quantifiable force signal waveform. Spatial geometric properties include transverse texture, longitudinal undulations, and vertical concavity / convexity. This allows the feedback force signal received by the master end to completely reconstruct the three-dimensional force information of the slave end during remote operation sensing, rather than a single-dimensional force change. Through the physical feedback of the master end actuator, the user can simultaneously perceive the frictional differences in the lateral direction of the triangular grating, the pressure fluctuations in the longitudinal direction, and the changes in the vertical contact intensity, and accurately distinguish the texture period, spatial morphology, and other attributes of the grating surface.

[0053] S105. Generate a main-end feedback force signal based on the three-dimensional force waveform. The main-end feedback force signal drives the main-end actuator to generate physical feedback. Based on the physical feedback, perform remote operation sensing of the surface properties of the triangular grating.

[0054] Specifically, the master-end feedback force signal is a force command signal generated based on a three-dimensional force waveform, capable of driving the master-end actuator. Essentially, it transforms the three-dimensional force characteristics generated by the contact between the triangular grating surface and the slave device into electrical or control signals that conform to the driving logic of the master-end actuator. These signals carry information on the intensity, frequency of change, and direction of the force, corresponding one-to-one with the three-dimensional force waveform. Teleoperation sensing refers to the process in which, in a teleoperation scenario, the operator indirectly obtains information about the physical properties of the slave device in contact with the triangular grating surface through the physical feedback generated by the master-end actuator based on the master-end feedback force signal.

[0055] In practice, the reconstructed three-dimensional force waveform undergoes master-end adaptation processing. Based on the force output range and dynamic response characteristics of the master-end actuator, the waveform amplitude is proportionally mapped and adjusted. The adapted three-dimensional force waveform is decomposed into single-axis feedback components in the x, y, and z axes, each corresponding to the driving force parameters of the master-end actuator in the corresponding spatial direction. A master-end feedback force signal in pulse width modulation (PWM) form is generated based on the single-axis feedback components. The intensity of the master-end feedback force signal is dynamically adjusted according to the amplitude change of the corresponding axis force waveform, and the frequency of the master-end feedback force signal changes synchronously with the periodic characteristics of the waveform. The master-end feedback force signal is transmitted to the drive unit of the master-end actuator. The drive unit controls the mechanical structure to generate physical vibrations or force outputs of corresponding intensity, frequency, and direction based on the signal parameters. Operators perceive the multi-dimensional physical feedback generated by the feedback force signal by contacting the master-end actuator. Combining the intensity differences, frequency changes, and directional characteristics of the feedback, operators can distinguish the texture period, spatial geometry, and contact characteristics of the triangular grating surface.

[0056] Specifically, the parameter library of the main actuator is called to read its force output range and dynamic response characteristics. An amplitude mapping algorithm is used to scale the original amplitude of the three-dimensional force waveform to the effective output range of the main actuator. The adapted three-dimensional force waveform is decomposed into three single-axis feedback components along the x, y, and z axes. The x-axis component corresponds to the lateral drive parameters of the main actuator, the y-axis component to the longitudinal drive parameters, and the z-axis component to the vertical drive parameters. Each component contains force data that varies over time. Based on the force changes of each single-axis feedback component, a pulse width modulation (PWM) signal is generated as the main actuator feedback force signal: the larger the force value, the higher the duty cycle of the PWM signal; the smaller the force value, the lower the duty cycle, and the signal frequency is synchronized with the periodic characteristics of the corresponding axis force waveform. The PWM feedback force signals of the x, y, and z axes are transmitted to the corresponding drive units of the main actuator via a data transmission module. After receiving the signals, the drive units control the mechanical structure to generate physical vibrations or force outputs of corresponding intensity, frequency, and direction based on the duty cycle and frequency parameters of the PWM signal. The operator contacts the main actuator and senses the multi-dimensional physical feedback it generates. By observing the differences in intensity, frequency, and direction of the feedback, the operator can distinguish the texture period, spatial geometry, and contact characteristics of the triangular grating surface.

[0057] The method provided in this embodiment, in its first aspect, involves a reconstruction process based on signals optimized in both the time and frequency domains. The time-domain optimized signal, through dynamic gain adjustment and standardization, enhances the distinguishability of amplitude characteristics of gratings of different sizes and eliminates unit differences. The frequency-domain optimized signal filters out noise and retains the dominant frequency and phase information corresponding to the grating texture period, ensuring the accuracy of frequency characteristics and the spatial correspondence of phase. During reconstruction, amplitude references are determined along the x, y, and z axes, and frequencies and phases are extracted. A trigonometric function model is constructed to obtain single-axis waveforms corresponding to transverse friction, longitudinal pressure, and vertical contact pressure, which are then integrated into a three-dimensional force waveform. This split-axis reconstruction based on the optimized signal allows the three-dimensional waveform to accurately carry the physical contact characteristics of the grating in different directions, avoiding dimensional confusion. At the same time, it completely preserves the correspondence between the strength, period, and spatial position of the force, enabling the master feedback to accurately restore the spatial geometric properties of the grating surface, allowing operators to clearly perceive the microstructure of the grating.

[0058] Secondly, during data acquisition, the end-effector obtains distance and current force signals in real time. After calculating the position and force deviation, it converts the force deviation into a position correction amount using virtual mass, damping, and stiffness as adjustment coefficients. Combined with the kinematic model, it calculates the joint adjustment amount and drives the joint movement, enabling the end-effector to adapt to the grating undulations in position and attitude. This alignment operation ensures that the end-effector and the grating surface are always in a stable and precise contact state, avoiding the loss, distortion, or noise of force signals caused by unstable contact or adjustment lag. This allows the acquired real-time force signals to better match the actual physical properties of the grating surface, providing high-quality raw data for subsequent time-domain and frequency-domain optimization and three-dimensional waveform reconstruction. This ensures the accuracy and authenticity of the final remote operation perception from the source, avoiding the disconnect between the master feedback and the actual contact of the slave due to poor raw signal quality.

[0059] Corresponding to the aforementioned embodiment of the triangular grating teleoperation sensing method based on force signal reconstruction, this application also provides an embodiment of a triangular grating teleoperation sensing device based on force signal reconstruction.

[0060] Figure 2 A schematic diagram of the triangular grating teleoperation sensing device based on force signal reconstruction provided in this application. Please refer to... Figure 2 The device provided in this embodiment includes a data acquisition module 210, a processing module 220, a reconstruction module 230, and a sensing module 240.

[0061] The acquisition module 210 is used to acquire the real force signal generated when the slave robotic arm comes into contact with the surface of the triangular grating.

[0062] The processing module 220 is used to perform time-domain optimization processing on the real-time real force signal, and to scale the signal amplitude by using a position gain variable and a unit conversion factor to obtain a time-domain optimized signal; the position gain variable is dynamically adjusted according to the distance between the end effector and the surface of the triangular grating.

[0063] The processing module 220 is also used to perform frequency domain optimization processing on the real-time real force signal, and to extract the dominant frequency component and corresponding phase information in the real-time real force signal by using fast Fourier transform to obtain a frequency domain optimized signal containing frequency features; the dominant frequency component corresponds to the periodic features of the triangular grating surface texture.

[0064] The reconstruction module 230 is used to perform time-domain and frequency-domain hybrid optimization processing on the time-domain optimized signal and the frequency-domain optimized signal, retain the time-domain amplitude characteristics of the time-domain optimized signal, fuse the dominant frequency components and phase information in the frequency-domain optimized signal, and reconstruct the three-dimensional force waveform through the trigonometric function method; the three-dimensional force waveform corresponds to the spatial geometric properties of the triangular grating surface;

[0065] The sensing module 240 is used to generate a main-end feedback force signal based on the three-dimensional force waveform. The main-end feedback force signal drives the main-end actuator to generate physical feedback, and performs remote sensing of the surface properties of the triangular grating based on the physical feedback.

[0066] The apparatus of this embodiment can be used to perform... Figure 1 The steps of the method embodiment shown are similar in principle and process, and will not be repeated here.

[0067] The specific implementation process of the functions and roles of each unit in the above device can be found in the implementation process of the corresponding steps in the above method, and will not be repeated here.

[0068] For the device embodiments, since they basically correspond to the method embodiments, the relevant parts can be referred to in the description of the method embodiments. The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this application according to actual needs. Those skilled in the art can understand and implement this without creative effort.

[0069] The above description is merely a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A method for teleoperation sensing of a triangular grating based on force signal reconstruction, characterized in that, The method includes: Real-time force signals generated when the robotic arm contacts the surface of the triangular grating are collected. The real-time force signal is subjected to time-domain optimization processing. The signal amplitude is scaled by a position gain variable and a unit conversion factor to obtain a time-domain optimized signal. The position gain variable is dynamically adjusted according to the distance between the end effector and the surface of the triangular grating. The real-time real force signal is subjected to frequency domain optimization processing. The dominant frequency component and corresponding phase information in the real-time real force signal are extracted by fast Fourier transform to obtain a frequency domain optimized signal containing frequency features. The dominant frequency component corresponds to the periodic features of the triangular grating surface texture. The time-domain optimized signal and the frequency-domain optimized signal are subjected to time-domain and frequency-domain hybrid optimization processing. The time-domain amplitude characteristics of the time-domain optimized signal are preserved, and the dominant frequency components and phase information in the frequency-domain optimized signal are fused. The three-dimensional force waveform is reconstructed by the trigonometric function method. The three-dimensional force waveform corresponds to the spatial geometric properties of the triangular grating surface. Based on the three-dimensional force waveform, a master-end feedback force signal is generated. The master-end feedback force signal drives the master-end actuator to generate physical feedback. Based on the physical feedback, the remote sensing of the surface properties of the triangular grating is performed.

2. The method according to claim 1, characterized in that, The process of performing time-domain and frequency-domain hybrid optimization on the time-domain optimized signal and the frequency-domain optimized signal, retaining the time-domain amplitude characteristics of the time-domain optimized signal, fusing the dominant frequency components and phase information in the frequency-domain optimized signal, and reconstructing the three-dimensional force waveform using the trigonometric function method includes: For any target axis among the x-axis, y-axis, and z-axis, extract the local force signal peaks where the target axis contacts the surface of the triangular grating from the time-domain optimization signal, calculate the arithmetic mean of the local force signal peaks, and determine the amplitude reference of the force on the target axis. Extract the dominant frequency component and phase information corresponding to the target axis from the frequency domain optimized signal; Based on the amplitude reference, dominant frequency components and phase information, a trigonometric function force model of the target axis is constructed, and a single-axis force waveform is reconstructed; wherein, the x-axis force waveform corresponds to the frictional force variation characteristics of the triangular grating surface sliding laterally, the y-axis force waveform corresponds to the shear force variation characteristics of the triangular grating surface along the longitudinal direction, and the z-axis force waveform corresponds to the pressure variation characteristics of the triangular grating surface in perpendicular contact. Integrate the single-axis force waveforms of the x-axis, y-axis, and z-axis to form a complete three-dimensional force waveform.

3. The method according to claim 2, characterized in that, The construction of the trigonometric function force model of the target axis based on the amplitude reference, dominant frequency component, and phase information includes: The amplitude benchmark is set to the maximum fluctuation range of the trigonometric function; The dominant frequency component is converted into the periodic parameter of a trigonometric function; The phase information is set as the starting position parameter of the trigonometric function; By integrating the maximum fluctuation amplitude, period parameters, and starting position parameters, a sinusoidal continuous force model is constructed; wherein, the fluctuation amplitude of the waveform is determined by the amplitude reference, the repetition frequency of the waveform is determined by the dominant frequency component, and the initial state of the waveform is determined by the phase information.

4. The method according to claim 1, characterized in that, The acquisition of the real-time force signal generated when the robotic arm contacts the surface of the triangular grating includes: The end-effector robotic arm collects the current force signal generated by contact through a force sensor; The difference between the current force signal and the preset contact force threshold is taken as the force deviation; Using preset virtual mass, damping, and stiffness parameters as adjustment coefficients, the force deviation is converted into the target position correction amount of the end effector; wherein, the virtual mass parameter controls the response speed of position adjustment, the damping parameter suppresses vibration during the adjustment process, and the stiffness parameter controls the accuracy of position correction. Based on the target position correction amount of the end effector, and combined with the kinematic model of the slave robot arm, the angle that each joint needs to rotate is calculated as the joint adjustment amount. Drive the robotic arm joints to move according to the joint adjustment amount, control the end effector to adapt to the undulations of the triangular grating surface in position and orientation, and during the dynamic adjustment of the robotic arm, the force sensor synchronously and continuously collects the real force signal generated by the contact.

5. The method according to claim 1, characterized in that, The step of performing time-domain optimization processing on the real-time real force signal, by scaling the signal amplitude using a position gain variable and a unit conversion factor to obtain a time-domain optimized signal, includes: Obtain the raw amplitude value of the real force signal in real time; A position gain variable is introduced, and the position gain variable is multiplied by the original amplitude value to obtain the adjusted amplitude value; the position gain variable is adjusted negatively correlated with the real-time distance between the end effector and the triangular grating surface; Multiply the adjusted amplitude value by the unit conversion factor to convert the signal amplitude to a unified standard unit, thus obtaining the standardized amplitude value. A continuous scaling function is used to perform secondary adjustment on the standardized amplitude value, and the processed amplitude values ​​are integrated to generate a time-domain optimized signal.

6. The method according to claim 5, characterized in that, The process involves using a continuous scaling function to perform secondary adjustment on the standardized amplitude value, integrating the processed amplitude values, and generating a time-domain optimized signal, including: The baseline scaling factor and smoothing adjustment range of the preset continuous scaling function are defined; the continuous scaling function takes the real-time displacement change rate of the end effector as an input parameter, and the output value of the continuous scaling function is within the smoothing adjustment range. When the real-time displacement change rate is positive, the continuous scaling function gradually increases the output value in a linear or curvilinear manner, starting from the reference scaling factor, according to the degree of increase in the real-time displacement change rate. When the real-time displacement change rate is negative, the continuous scaling function gradually decreases the output value in a linear or curvilinear manner, starting from the reference scaling factor, according to the degree of increase in the absolute value of the real-time displacement change rate. When the end effector is stationary or the real-time displacement change rate is close to zero, the output value of the continuous scaling function gradually returns to the reference scaling factor. The output value of the continuous scaling function is multiplied by the standardized amplitude value to obtain the amplitude value after secondary adjustment. The amplitude value after secondary adjustment is integrated in time to remove fluctuations caused by function transition, forming a time-domain optimized signal.

7. The method according to claim 1, characterized in that, The step of performing frequency domain optimization processing on the real-time real force signal, using Fast Fourier Transform to extract the dominant frequency components and corresponding phase information from the real-time real force signal, to obtain a frequency domain optimized signal containing frequency features, includes: The real-time force signal is preprocessed to remove the DC component and noise interference from the signal; The preprocessed real-time real force signal is segmented according to a preset time window. A fast Fourier transform is performed on each segment to convert the time domain signal into a frequency domain signal, thereby obtaining the energy distribution characteristics of each frequency component. Frequency components with energy values ​​exceeding a preset threshold are selected from the frequency domain signal, and the frequency component with the highest energy is determined as the dominant frequency component. Extract the phase information of the dominant frequency component in the frequency domain; The continuity of the dominant frequency components and phase information extracted in each time window is verified, and outliers caused by signal mutations are removed. By integrating the verified dominant frequency components and their corresponding phase information, a frequency-domain optimized signal containing frequency characteristics is generated.

8. The method according to claim 1, characterized in that, The process of generating a master-end feedback force signal based on the three-dimensional force waveform, the master-end feedback force signal driving the master-end actuator to generate physical feedback, and performing remote operation sensing of the surface properties of the triangular grating based on the physical feedback includes: The reconstructed three-dimensional force waveform is adapted to the main end, and the amplitude of the waveform is proportionally adjusted according to the force output range and dynamic response characteristics of the main end actuator. The adapted three-dimensional force waveform is decomposed into single-axis feedback components in three directions: x-axis, y-axis, and z-axis. Each component corresponds to the driving force parameter of the main actuator in the corresponding spatial direction. The main feedback force signal is generated in the form of a pulse width modulation signal based on the single-axis feedback component; the intensity of the main feedback force signal is dynamically adjusted according to the amplitude of the force waveform on the corresponding axis, and the frequency of the main feedback force signal changes synchronously with the periodic characteristics of the waveform. The feedback force signal from the main end is transmitted to the drive unit of the main end actuator. The drive unit controls the mechanical structure to generate physical vibration or force output with corresponding intensity, frequency and direction according to the signal parameters. Operators can perceive the multi-dimensional physical feedback generated by the feedback force signal by contacting the main actuator. By combining the differences in the intensity, frequency changes and directional characteristics of the feedback, operators can distinguish the texture period, spatial geometry and contact characteristics of the triangular grating surface.

9. The method according to claim 1, characterized in that, The unit conversion factor is 0.

001.

10. A triangular grating teleoperation sensing device based on force signal reconstruction, characterized in that, The device includes a data acquisition module, a processing module, a reconstruction module, and a sensing module; The acquisition module is used to acquire real-time force signals generated when the robotic arm contacts the surface of the triangular grating. The processing module is used to perform time-domain optimization processing on the real-time real force signal, and to scale the signal amplitude by using a position gain variable and a unit conversion factor to obtain a time-domain optimized signal; the position gain variable is dynamically adjusted according to the distance between the end effector and the surface of the triangular grating. The processing module is also used to perform frequency domain optimization processing on the real-time real force signal, and to extract the dominant frequency component and corresponding phase information in the real-time real force signal by using fast Fourier transform to obtain a frequency domain optimized signal containing frequency features; the dominant frequency component corresponds to the periodic features of the triangular grating surface texture. The reconstruction module is used to perform time-domain and frequency-domain hybrid optimization processing on the time-domain optimized signal and the frequency-domain optimized signal, retain the time-domain amplitude characteristics of the time-domain optimized signal, fuse the dominant frequency components and phase information in the frequency-domain optimized signal, and reconstruct the three-dimensional force waveform using the trigonometric function method; the three-dimensional force waveform corresponds to the spatial geometric properties of the triangular grating surface; The sensing module is used to generate a main-end feedback force signal based on the three-dimensional force waveform. The main-end feedback force signal drives the main-end actuator to generate physical feedback, and the teleoperation sensing of the surface properties of the triangular grating is performed based on the physical feedback.

Images