Double-contact flexible tactile sensing system based on ultrasonic waves

Abstract

The invention provides a double-contact flexible tactile sensing system based on ultrasonic waves, and the system comprises a flexible silicone tube which is internally provided with a sealed cavity, and piezoelectric ceramics are fixed at the two ends of the flexible silicone tube respectively; the glycerol is filled in the cavity of the silicone tube and is used as an ultrasonic conduction medium; the cover is packaged at the two ends of the silicone tube to form a sealing structure; the signal processing module comprises a data acquisition module, a preprocessing module and a signal decoupling module; the piezoelectric ceramic on one side of the flexible silicone tube serves as an excitation end, high-frequency ultrasonic waves are generated under driving of control signals, and the high-frequency ultrasonic waves are transmitted to the piezoelectric ceramic on the other side of the flexible silicone tube through glycerin to serve as a receiving end and are converted into electric signals. And the signal processing module decouples to obtain the pressing force and position information of the double contacts based on the amplitude, propagation time and phase change of the electric signals.

Description

Technical Field

[0001] The present invention relates to the field of tactile sensing technology, and in particular to an ultrasonic-based dual-contact flexible tactile sensing system. Background Art

[0002] In the field of tactile sensing, with the rapid development of robotics, medical equipment, and human-computer interaction systems, the demand for high-precision and high-sensitivity tactile sensors is increasing.

[0003] Existing tactile sensing technologies are mainly based on principles such as resistance, capacitance and magnetic induction. For example, patent application number CN119573929A discloses a tactile sensing device and method, which includes multiple neodymium iron boron magnetic structures, a measurement circuit and a processor. Its working principle is to determine the magnitude and position of the force by measuring the change in the resistance value of the tactile sensor under external force, combined with the initial three-dimensional morphology of the sensor. However, this technical solution has the following shortcomings: First, the introduction of multiple magnetic structures and complex measurement circuits increases the manufacturing difficulty and cost of the sensor, which is not conducive to large-scale production and application. Secondly, due to the use of rigid magnetic materials, the sensor may have limitations when adapting to flexible surfaces or non-flat contact surfaces, limiting its application in flexible electronics and wearable devices. In addition, magnetic materials are easily interfered by external magnetic fields, which may affect the measurement accuracy and stability. In particular, the reliability of the sensor may decrease in complex electromagnetic environments. Summary of the Invention

[0004] In view of the defects in the prior art, the object of the present invention is to provide an ultrasonic-based dual-contact flexible tactile sensing system.

[0005] The ultrasonic dual-contact flexible tactile sensing system provided by the present invention includes:

[0006] A flexible silicone tube with a sealed cavity inside and piezoelectric ceramics fixed at both ends;

[0007] Glycerol, filled in the cavity of the silicone tube, serving as an ultrasonic wave conducting medium;

[0008] A cover is sealed at both ends of the silicone tube to form a sealing structure;

[0009] Signal processing module, including data acquisition module, preprocessing module and signal decoupling module;

[0010] The piezoelectric ceramic on one side of the flexible silicone tube acts as the excitation end, generating high-frequency ultrasonic waves driven by a control signal. The high-frequency ultrasonic waves are then transmitted through the glycerol to the piezoelectric ceramic on the other side, acting as the receiving end, and converted into electrical signals.

[0011] The signal processing module decouples and obtains the pressing force and position information of the dual contacts based on the amplitude, propagation time and phase change of the electrical signal.

[0012] Preferably, the piezoelectric ceramic is installed as follows:

[0013] Two piezoelectric ceramics are symmetrically embedded in the inner walls of the silicone tube at both ends, and their polarization direction is consistent with the propagation direction of the ultrasonic wave;

[0014] The resonant frequency of the piezoelectric ceramic is 1 MHz to 10 MHz, and an ultrasonic pulse is generated by a high-frequency excitation signal, with a pulse width of 5 μs to 20 μs.

[0015] Preferably, the filling of the glycerol satisfies the following conditions:

[0016] The volume of glycerol accounts for 90% to 95% of the volume of the silicone tube cavity, and the remainder is reserved for deformation space;

[0017] The difference between the acoustic impedance of glycerin and the acoustic impedance of the silicone tube is less than 10%, so as to reduce interface reflection loss.

[0018] Preferably, the signal processing flow of the pre-processing module includes:

[0019] Perform bandpass filtering on the received signal to remove noise outside the frequency range. The passband frequency is ±20% of the excitation frequency.

[0020] The filtered signal is normalized, and the amplitude normalization formula is:

[0021]

[0022] Among them, V raw is the original signal amplitude, V min and V max are the minimum and maximum values ​​of the signal dynamic range;

[0023] Extract time domain features, including signal envelope slope, energy integral, and zero-crossing rate.

[0024] Preferably, the signal decoupling module adopts a support vector regression SVR model, specifically comprising:

[0025] Training phase: Sample signals with known pressing positions and forces are collected, and time-domain and frequency-domain features are extracted to construct a training set. The frequency-domain features are decomposed into 6 layers using wavelet packets, and the energy proportions of the 3rd to 5th frequency bands are extracted.

[0026] Prediction stage: The feature vector of the real-time signal is input into the trained SVR model, and the touch point coordinates (x, y) and the pressing force F are output. The calculation formula is:

[0027]

[0028] Where Δt is the propagation time delay, is the phase offset, k and b are the coefficients obtained from model training.

[0029] Preferably, the system supports multi-point expansion, specifically:

[0030] An array of multiple silicone tubes connected in parallel is used, with each silicone tube corresponding to an independent signal processing channel;

[0031] The piezoelectric ceramic excitation timing of each channel is staggered to avoid signal crosstalk.

[0032] Preferably, the relationship between the deformation and acoustic impedance of the silicone tube satisfies:

[0033] Local pressure causes the lumen cross-sectional area to decrease by ΔS, and the acoustic impedance change ΔZ is inversely proportional to ΔS, that is:

[0034]

[0035] Where Z0 and S0 are the initial acoustic impedance and cross-sectional area, respectively.

[0036] Preferably, the packaging structure of the cover is:

[0037] The cover is made of the same flexible silicone as the silicone tube, with an annular groove on the inner wall, which fits with the end of the silicone tube in an interference fit.

[0038] The outer surface of the lid is coated with anti-slip texture to enhance the stability of the pressing operation.

[0039] Preferably, the frequency domain feature extraction adopts a wavelet packet method, specifically:

[0040] Decomposing the signal to the 6th layer, we get 64 frequency bands;

[0041] Calculate the energy proportion of each frequency band Among them, W i,j is the coefficient of the jth node in the i-th frequency band, M is the total number of nodes, and k is the frequency band index.

[0042] Preferably, the workflow of the system includes:

[0043] The piezoelectric ceramic at the excitation end transmits ultrasonic pulses, and the piezoelectric ceramic at the receiving end collects the transmitted and reflected signals;

[0044] The data acquisition module records signals at a sampling rate of no less than 20MS / s;

[0045] The preprocessing module eliminates high-frequency noise through sliding average filtering, with a window length of 10 sampling points;

[0046] The signal decoupling module combines the characteristic vectors of the two signals and outputs the weighted average of the dual contact positions and forces.

[0047] Compared with the prior art, the present invention has the following beneficial effects:

[0048] (1) By adopting the structure of "piezoelectric ceramics + glycerin + flexible silicone tube", the problem that traditional rigid tactile sensors are not suitable for flexible environments is solved;

[0049] (2) By using liquid media to conduct ultrasonic signals, the problems of unstable solid wave propagation and signal distortion in flexible materials are avoided;

[0050] (3) By using a dual piezoelectric ceramic structure in conjunction with high-frequency ultrasound, the problems of low precision and poor spatial resolution of traditional sensing methods are solved;

[0051] (4) By introducing signal processing modules and intelligent algorithms, the system's multi-point recognition capability and response sensitivity are improved, making it suitable for a variety of complex interactive applications. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Other features, objects and advantages of the present invention will become more apparent upon reading the detailed description of non-limiting embodiments with reference to the following drawings:

[0053] Figure 1 Schematic diagram of the structure of the ultrasonic-based dual-contact flexible tactile sensing system;

[0054] Figure 2 Schematic diagram of the structure of the ultrasonic-based dual-contact flexible tactile sensing system;

[0055] In the figure, cover 1, piezoelectric ceramic 2, silicone tube 3, glycerin 4. DETAILED DESCRIPTION

[0056] The present invention will be described in detail below with reference to specific embodiments. The following examples will help those skilled in the art to further understand the present invention, but are not intended to limit the present invention in any form. It should be noted that, for those skilled in the art, several changes and improvements can be made without departing from the scope of the present invention. These all fall within the scope of protection of the present invention.

[0057] Example 1

[0058] like Figure 1 and Figure 2 The flexible tactile sensing system of the present invention includes: a cover 1, a piezoelectric ceramic 2, a silicone tube 3, a glycerin 4 and a signal processing module.

[0059] The silicone tube 3 is a flexible cavity structure, with piezoelectric ceramics 2 fixed at both ends of the interior. Glycerol 4 is injected into the cavity of the silicone tube 3 as an ultrasonic wave conducting medium, and the entire silicone tube 3 is encapsulated and sealed by a cover 1.

[0060] During operation, the piezoelectric ceramic on one side acts as an excitation end and generates a high-frequency ultrasonic signal driven by a control signal. The signal is transmitted to the piezoelectric ceramic at the receiving end on the opposite side through the glycerol 4 and converted into an electrical signal.

[0061] When the user presses on the silicone tube 3, a local area deforms, squeezing the tube lumen and reducing its cross-sectional area. This structural change increases the local equivalent acoustic impedance, altering the propagation characteristics of ultrasound in that area. Specifically, the increased acoustic impedance causes ultrasound, which would otherwise be transmitted smoothly, to be more strongly reflected in that area, resulting in an enhanced echo signal at the receiving end and a significant weakening of the transmitted signal.

[0062] Furthermore, due to changes in the thickness and propagation path of the deformed region, the speed and distance of the sound wave propagating through the region also change, causing propagation time delay and phase shift at the receiving end. Amplitude attenuation can also be caused by energy scattering or enhanced absorption due to localized compression.

[0063] Therefore, the geometric changes and acoustic changes caused by pressing the silicone tube are coupled together, affecting the propagation path and characteristics of the ultrasonic signal. This allows the receiver to detect a series of signal changes related to the pressure intensity and location. These changes provide a reliable physical basis for identifying the user's specific actions. By extracting and analyzing the features of these changes through the signal processing module, tactile perception in flexible environments can be achieved.

[0064] The workflow of software and hardware working together:

[0065] The system includes a hardware structure (a cover 1, piezoelectric ceramics 2, a silicone tube 3, and glycerin 4), a data acquisition module, a preprocessing module, and a signal decoupling module.

[0066] The piezoelectric ceramic 2 and the glycerin 4 are encapsulated using a cover 1 and a silicone tube 3 .

[0067] When piezoelectric ceramics are excited, they generate ultrasonic waves;

[0068] The ultrasonic wave is transmitted through the glycerol 4 to the piezoelectric ceramic at the receiving end, which converts it into an electrical signal;

[0069] The data acquisition module collects the electrical signals and transmits them to the preprocessing module;

[0070] The preprocessing module first performs basic processing on the received ultrasonic signal, including denoising, filtering, and normalization, to improve the stability and accuracy of subsequent feature extraction. In terms of time-domain feature extraction, the module extracts common statistical parameters such as the signal's maximum amplitude, mean, envelope variation, rise time, zero-crossing rate, and energy. In terms of frequency-domain feature extraction, the module uses wavelet packet analysis to decompose the signal into multiple frequency bands, extracting metrics such as the energy distribution, spectral entropy, and energy fraction of each band to capture subtle frequency changes caused by compression.

[0071] The signal decoupling module uses the multi-dimensional features extracted above as input and utilizes the trained machine learning model to recognize and decode the pressure information. The model input is the feature vectors corresponding to the two ultrasonic receiving paths, and the output is the pressure position and applied force of the two contacts. Specifically, during the training phase, a data set is constructed by collecting signal samples under known pressure positions and force conditions, and the extracted features are used to train the support vector regression model. During the prediction phase, the system inputs the features extracted in real time into the trained model, and outputs the current corresponding pressure position and force estimation values.

[0072] The system can achieve multi-point expansion based on a multi-channel structure to adapt to different flexible tactile application requirements.

[0073] Example 2

[0074] The present invention provides an ultrasonic-based dual-contact flexible tactile sensing system, comprising: a flexible silicone tube, the interior of which is a sealed cavity, with piezoelectric ceramics fixed at both ends; glycerin, filled in the cavity of the silicone tube and serving as an ultrasonic wave conduction medium; a cover, encapsulated at both ends of the silicone tube to form a sealed structure; a signal processing module, comprising a data acquisition module, a preprocessing module and a signal decoupling module; wherein the piezoelectric ceramic on one side of the flexible silicone tube serves as an excitation end, and generates high-frequency ultrasonic waves under the drive of a control signal, and the high-frequency ultrasonic waves are transmitted through the glycerin to the piezoelectric ceramic on the other side serving as a receiving end and converted into electrical signals; the signal processing module decouples the pressing force and position information of the dual contacts based on the amplitude, propagation time and phase change of the electrical signals.

[0075] The piezoelectric ceramics are installed in the following manner: two piezoelectric ceramics are symmetrically embedded in the inner walls at both ends of the silicone tube, and their polarization direction is consistent with the direction of ultrasonic propagation; the resonant frequency of the piezoelectric ceramics is 1MHz~10MHz, and ultrasonic pulses are generated by high-frequency excitation signals, with a pulse width of 5μs~20μs.

[0076] The filling of the glycerol meets the following conditions: the volume of the glycerol accounts for 90% to 95% of the volume of the silicone tube cavity, and the remainder is reserved for deformation space; the difference in acoustic impedance between the glycerol and the silicone tube is less than 10% to reduce interface reflection loss.

[0077] The signal processing flow of the pre-processing module includes: performing band-pass filtering on the received signal to filter out noise outside the frequency range, where the band-pass frequency is ±20% of the excitation frequency; and performing normalization on the filtered signal. The amplitude normalization formula is: Among them, V raw is the original signal amplitude, V min and V max is the minimum and maximum value of the signal dynamic range; extract time domain features, including signal envelope slope, energy integral, and zero-crossing rate.

[0078] The signal decoupling module uses a support vector regression (SVR) model, specifically including the following: a training phase: collecting sample signals with known pressing positions and forces, extracting time domain and frequency domain features to construct a training set, decomposing the frequency domain features into 6 layers using wavelet packets, and extracting the energy proportion of the 3rd to 5th layer frequency bands; a prediction phase: inputting the feature vector of the real-time signal into the trained SVR model, outputting the contact point coordinates (x, y) and the pressing force F, calculated using the following formula: Where Δt is the propagation time delay, is the phase offset, k and b are the coefficients obtained from model training.

[0079] The system supports multi-point expansion, specifically: multiple silicone tubes connected in parallel are used to form an array, each silicone tube corresponds to an independent signal processing channel; the piezoelectric ceramic excitation timing of each channel is staggered to avoid signal crosstalk.

[0080] The relationship between the deformation and acoustic impedance of the silicone tube satisfies the following: local pressure causes the cross-sectional area of ​​the tube cavity to decrease by ΔS, and the acoustic impedance change ΔZ is inversely proportional to ΔS, that is: Where Z0 and S0 are the initial acoustic impedance and cross-sectional area, respectively.

[0081] The packaging structure of the lid is as follows: the lid is made of flexible silicone of the same material as the silicone tube, and the inner wall is provided with an annular groove, which is interference fit with the end of the silicone tube; the outer surface of the lid is coated with anti-slip texture to enhance the stability of the pressing operation.

[0082] The frequency domain feature extraction adopts the wavelet packet method, specifically: decomposing the signal to the 6th layer to obtain 64 frequency bands; calculating the energy proportion of each frequency band Among them, W i,j is the coefficient of the jth node in the i-th frequency band, M is the total number of nodes, and k is the frequency band index.

[0083] The system's workflow includes: the piezoelectric ceramic at the excitation end emits ultrasonic pulses, and the piezoelectric ceramic at the receiving end collects the transmitted and reflected signals; the data acquisition module records the signals at a sampling rate of no less than 20MS / s; the preprocessing module eliminates high-frequency noise through a sliding average filter with a window length of 10 sampling points; and the signal decoupling module combines the eigenvectors of the two signals to output a weighted average of the dual contact positions and forces.

[0084] In the description of this application, it should be understood that the terms "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", etc., indicating the orientation or position relationship, are based on the orientation or position relationship shown in the accompanying drawings, and are only for the convenience of describing this application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on this application.

[0085] Those skilled in the art will appreciate that, in addition to implementing the system, device, and various modules provided by the present invention in purely computer-readable program code, it is entirely possible to implement the same program in the form of logic gates, switches, application-specific integrated circuits, programmable logic controllers, embedded microcontrollers, and the like by logically programming the method steps. Therefore, the system, device, and various modules provided by the present invention can be considered a hardware component, and the modules included therein for implementing various programs can also be considered structures within the hardware component; the modules for implementing various functions can also be considered both software programs for implementing the method and structures within the hardware component.

[0086] The above describes specific embodiments of the present invention. It should be understood that the present invention is not limited to the specific embodiments described above, and those skilled in the art may make various changes or modifications within the scope of the claims, which do not affect the essence of the present invention. The embodiments of this application and the features in the embodiments may be combined with each other in any manner unless there is a conflict.

Claims

1. An ultrasonic-based dual-contact flexible tactile sensing system, characterized in that: include: A flexible silicone tube with a sealed cavity inside and piezoelectric ceramics fixed at both ends; Glycerol, filled in the cavity of the silicone tube, serving as an ultrasonic wave conducting medium; A cover is sealed at both ends of the silicone tube to form a sealing structure; Signal processing module, including data acquisition module, preprocessing module and signal decoupling module; The piezoelectric ceramic on one side of the flexible silicone tube acts as the excitation end, generating high-frequency ultrasonic waves driven by a control signal. The high-frequency ultrasonic waves are then transmitted through the glycerol to the piezoelectric ceramic on the other side, acting as the receiving end, and converted into electrical signals. The signal processing module decouples and obtains the pressing force and position information of the dual contacts based on the amplitude, propagation time and phase change of the electrical signal.

2. The ultrasonic dual-contact flexible tactile sensing system according to claim 1, characterized in that: The piezoelectric ceramics are installed as follows: Two piezoelectric ceramics are symmetrically embedded in the inner walls of the silicone tube at both ends, and their polarization direction is consistent with the propagation direction of the ultrasonic wave; The resonant frequency of the piezoelectric ceramic is 1 MHz to 10 MHz, and an ultrasonic pulse is generated by a high-frequency excitation signal, with a pulse width of 5 μs to 20 μs.

3. The ultrasonic dual-contact flexible tactile sensing system according to claim 1, characterized in that: The filling of the glycerol meets the following conditions: The volume of glycerol accounts for 90% to 95% of the volume of the silicone tube cavity, and the remainder is reserved for deformation space; The difference between the acoustic impedance of glycerin and the acoustic impedance of the silicone tube is less than 10%, so as to reduce interface reflection loss.

4. The ultrasonic dual-contact flexible tactile sensing system according to claim 1, characterized in that: The signal processing flow of the pre-processing module includes: Perform bandpass filtering on the received signal to remove noise outside the frequency range. The passband frequency is ±20% of the excitation frequency. The filtered signal is normalized, and the amplitude normalization formula is: Among them, V raw is the original signal amplitude, V min and V max are the minimum and maximum values ​​of the signal dynamic range; Extract time domain features, including signal envelope slope, energy integral, and zero-crossing rate.

5. The ultrasonic dual-contact flexible tactile sensing system according to claim 1, characterized in that: The signal decoupling module adopts the support vector regression SVR model, which specifically includes: Training phase: Sample signals with known pressing positions and forces are collected, and time-domain and frequency-domain features are extracted to construct a training set. The frequency-domain features are decomposed into 6 layers using wavelet packets, and the energy proportions of the 3rd to 5th frequency bands are extracted. Prediction stage: The feature vector of the real-time signal is input into the trained SVR model, and the touch point coordinates (x, y) and the pressing force F are output. The calculation formula is: Where Δt is the propagation time delay, is the phase offset, k and b are the coefficients obtained from model training.

6. The ultrasonic dual-contact flexible tactile sensing system according to claim 1, characterized in that: The system supports multi-point expansion, specifically: An array of multiple silicone tubes connected in parallel is used, with each silicone tube corresponding to an independent signal processing channel; The piezoelectric ceramic excitation timing of each channel is staggered to avoid signal crosstalk.

7. The ultrasonic dual-contact flexible tactile sensing system according to claim 1, characterized in that: The relationship between the deformation and acoustic impedance of the silicone tube satisfies: Local pressure causes the lumen cross-sectional area to decrease by ΔS, and the acoustic impedance change ΔZ is inversely proportional to ΔS, that is: Where Z0 and S0 are the initial acoustic impedance and cross-sectional area, respectively.

8. The ultrasonic dual-contact flexible tactile sensing system according to claim 1, characterized in that: The packaging structure of the lid is: The cover is made of the same flexible silicone as the silicone tube, with an annular groove on the inner wall, which fits with the end of the silicone tube in an interference fit. The outer surface of the lid is coated with anti-slip texture to enhance the stability of the pressing operation.

9. The ultrasonic dual-contact flexible tactile sensing system according to claim 1, characterized in that: The frequency domain feature extraction adopts the wavelet packet method, specifically: Decomposing the signal to the 6th layer, we get 64 frequency bands; Calculate the energy proportion of each frequency band Among them, W i,j is the coefficient of the jth node in the i-th frequency band, M is the total number of nodes, and k is the frequency band index.

10. The ultrasonic dual-contact flexible tactile sensing system according to claim 1, characterized in that: The workflow of the system includes: The piezoelectric ceramic at the excitation end transmits ultrasonic pulses, and the piezoelectric ceramic at the receiving end collects the transmitted and reflected signals; The data acquisition module records signals at a sampling rate of no less than 20MS / s; The preprocessing module eliminates high-frequency noise through sliding average filtering, with a window length of 10 sampling points; The signal decoupling module combines the characteristic vectors of the two signals and outputs the weighted average of the dual contact positions and forces.

Images