A magnetoelectric flexible tactile sensor

By designing a magnetoelectric flexible tactile sensor with a pyramid array of magnetoelastic materials and combining it with a convolutional neural network to process magnetic field signals, the problems of high production costs and complex processes have been solved. This has enabled the sensor to achieve high sensitivity and real-time multimodal signal analysis, thus promoting its commercial application.

CN119984577BActive Publication Date: 2026-01-09HUAZHONG UNIV OF SCI & TECH +1
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
CN202510077957.4
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2025-01-17
Publication Date
2026-01-09
Estimated Expiration
2045-01-17

AI Technical Summary

Technical Problem

Existing magnetoelectric flexible tactile sensors have high production costs, complex production processes, and are difficult to commercialize on a large scale. Furthermore, these sensors cannot analyze and process multimodal signals in real time.

Method used

Design a structure including a magnetoelastic body, a support, a flexible circuit board, a recess, and a magnetic field sensor. The magnetoelastic body is a pyramid array. Combined with a pre-trained convolutional neural network to process the magnetic field signal, the pressure position and magnitude can be obtained in real time.

Benefits of technology

It reduces production costs, simplifies the production process, and improves the sensitivity and real-time response capability of the sensor, enabling accurate detection of pressure location and magnitude.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119984577B_ABST
    Figure CN119984577B_ABST
Patent Text Reader

Abstract

The application discloses a magnetoelectric flexible tactile sensor, which comprises a magnetoelastic body, a support part, a flexible circuit board, a recess part and a magnetic field sensor, the recess part is located above the support part, the magnetoelastic body is aligned with the recess part, the magnetic field sensor is located on the flexible circuit board and between the support part and the flexible circuit board, the magnetoelastic body is used for generating a corresponding magnetic field distribution according to an external contact pressure, and the magnetic field sensor is used for collecting a magnetic field signal of the magnetoelastic body, wherein a lower part of the magnetoelastic body in contact with the recess part is a pyramid array structure. Through the magnetoelectric flexible tactile sensor, the structure of the magnetoelastic body is pyramid-shaped, the sensitivity of the flexible tactile sensor is improved, and the tactile information on the tactile sensor is directly obtained by pre-training a convolutional neural network to process the magnetic field signal.
Need to check novelty before this filing date? Find Prior Art

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of sensors, and particularly relates to a magneto-electric flexible tactile sensor. BACKGROUND

[0002] Flexible tactile sensors are also known as "electronic skin", and have the characteristics of extensibility and bendability compared with rigid sensors, and can adapt to various shapes of carriers. Due to these characteristics, there are currently extensive research directions for tactile sensors, including resistive, capacitive, piezoelectric, magneto-electric and other sensing principles.

[0003] At present, the application of flexible tactile sensors is still relatively small, and it is still in the laboratory research and development stage, mainly for the following reasons: the production cost of flexible tactile sensors is too high, and it is difficult to commercialize; the production process of flexible tactile sensors is too complex, and there is a lack of simple and large-scale flexible tactile sensor production process; most tactile sensors cannot analyze and process external force and give real-time response of multi-modal signals.

[0004] The tactile of the magneto-electric flexible tactile sensor will change the position of the magnetic source, affect the magnetic field generated by the magnetic source, and the position and size of the tactile are determined by the magnetic field sensor and the signal processing unit. Compared with other flexible tactile sensors of other sensing forms, the sensing part and the circuit part of the magneto-electric flexible tactile sensor are separate parts, so there are the following advantages: the sensing part can be produced separately, and if elastic fatigue occurs after repeated compression, the sensing part can be replaced at any time; by replacing the sensing part, the sensing structure can be changed, so that the elastic body with different pressure ranges and sensitivities can be designed to adapt to different sensing environments; the sensor and the circuit are separated, the preparation process is simple, and the cost is low; the pressure will not affect the circuit part through the sensing part, and the stability of the circuit can be guaranteed.

[0005] However, the existing magneto-electric flexible tactile sensor cannot provide a solution for the commercial large-scale production of tactile sensors. Accordingly, there is an urgent need in the art to make further research and design on this, to design a magneto-electric flexible tactile sensor that can be commercialized, simplified production and easily replaced, so as to better meet the integration and commercialization of humanoid robots or other fields. SUMMARY

[0006] In view of the defects of the prior art and the need for improvement, the present application provides a magneto-electric flexible tactile sensor, comprising: a magneto-elastic body, a support part, a flexible circuit board, a recess part located above the support part, the magneto-elastic body being aligned with the recess part, and a magnetic field sensor located on the flexible circuit board between the support part and the flexible circuit board, the magneto-elastic body being used to generate a corresponding magnetic field distribution according to an external contact pressure, and the magnetic field sensor being used to collect a magnetic field signal of the magneto-elastic body, wherein a lower part of the magneto-elastic body in contact with the recess part is a pyramid array structure.

[0007] Further, the magneto-electric flexible tactile sensor further comprises a signal processing unit containing a trained machine learning model for processing the magnetic field signal and generating tactile information.

[0008] Further, the tactile information includes pressure distribution and size.

[0009] Further, the magnetic field sensor array is arranged on the flexible circuit board.

[0010] Further, the magnetic field sensor is arranged in a 3*3 array, the magnetic field sensor array has a size of 2cm*2cm, and the magnetic field sensor array is divided into 4*4 pressure regions, wherein each pressure region has a size of 0.5cm*0.5cm.

[0011] Further, the pyramid array has a height of 2mm and a base size of 2mm*2mm, and the distance between each pyramid structure is 4mm.

[0012] Further, the magnetic field sensor is a three-dimensional Hall sensor.

[0013] Further, the magneto-elastic body is a polydimethylsiloxane elastomer containing neodymium iron boron magnetic powder.

[0014] Further, the magnetic field sensor outputs the magnetic field signal using an I2C bus.

[0015] Further, when the magneto-elastic body is subjected to an external contact pressure, it is embedded in the recess body.

[0016] Overall, the magneto-electric flexible tactile sensor disclosed in the present application has a pyramid structure of the magneto-elastic body, which improves the sensitivity of the flexible tactile sensor, and the magnetic field signal is processed by a pre-trained convolutional neural network to directly obtain the pressure position and size on the tactile sensor. BRIEF DESCRIPTION OF DRAWINGS

[0017] Figure 1 A structure diagram of a magneto-electric flexible tactile sensor is provided for an embodiment of the present application.

[0018] Figure 2 A magnetic field distribution diagram of a tactile sensor when the magneto-elastic flexible tactile sensor provided by the embodiment of the present application is not subjected to pressure;

[0019] Figure 3 A magnetic field distribution diagram of a tactile sensor when the magneto-elastic flexible tactile sensor provided by the embodiment of the present application is subjected to pressure;

[0020] Figure 4 A comparison diagram of pressure sensitivity of the magneto-elastic flexible tactile sensor provided by the embodiment of the present application and different structured elastomers; and

[0021] Figure 5 A diagram of arrangement of magnetic field sensors and pressure positioning areas in the magneto-elastic flexible tactile sensor provided by the embodiment of the present application.

[0022] Figure 6 A pressure output diagram when different pressures are applied to the magneto-elastic flexible tactile sensor provided by the embodiment of the present application. DETAILED DESCRIPTION

[0023] In order to make the objectives, technical solutions and advantages of the present application clearer, the present application will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and should not be used to limit the present application. In addition, the technical features involved in the various embodiments of the present application described below can be combined with each other as long as they do not conflict with each other.

[0024] In the present application, the terms "first", "second", etc. (if any) in the present application and the accompanying drawings are used to distinguish similar objects, and do not necessarily describe a specific order or sequence.

[0025] Figure 1 A structural schematic diagram of a magneto-elastic flexible tactile sensor provided by the embodiment of the present application. As shown in FIG. 1, the magneto-elastic flexible tactile sensor provided by the embodiment of the present application comprises a flexible substrate 1, a magnetic field sensor 2, a magneto-elastic material 3, and a pressure positioning area 4. Figure 1As shown, the magneto-electric flexible tactile sensor comprises a magneto-elastic body 11, a support part 12, a flexible circuit board 13, a concave part 14, a magnetic field sensor 15, and a signal processing unit 16. The magneto-elastic body 11 is a polydimethylsiloxane elastomer doped with neodymium iron boron magnetic powder, belonging to a magnetic film structure, and is used to generate a corresponding magnetic field distribution according to an external contact pressure. The support part 12 is a support part printed by polylactic acid 3D, and the support part 12 serves as a carrier of the upper magneto-elastic body, and has a concave structure 14 on the surface, which is used for a self-alignment scheme of the magneto-elastic body 11, and can ensure that the position of the magneto-elastic body 11 will not be horizontally deviated. The flexible circuit board 13 is a flexible circuit board of the magnetic field sensor, and the flexible circuit board 13 is distributed with the magnetic field sensor 15, such as a three-dimensional Hall sensor array with a model of MLX90393. The magnetic field sensor 15 is used to collect the magnetic field signal generated by the magneto-elastic body 11. The signal processing unit 16 comprises a machine learning model, is used to process the magnetic field signal collected by the magnetic field sensor 15, and obtains the pressure position and size through the magnetic field signal. The signal processing unit 16 is located above the flexible circuit board 13 or is independent of the flexible circuit board 13. In the magneto-electric flexible tactile sensor disclosed by the present application, the circuit and the sensor are separated and no relative horizontal movement is generated. Further, the part below the magneto-elastic body 11, i.e. the part in contact with the support part 12, is designed as a pyramid array structure. The gap between the pyramids provides a space for the deformation of the magnetic film, and the structure of the pyramids, which is wide at the top and narrow at the bottom, makes it easier to deform when subjected to pressure. Therefore, the tactile sensor can more sensitively and accurately detect the touch or pressure change of different parts, so as to obtain tactile information, including the pressure size and the pressure distribution information.

[0026] Figure 2 The magnetic field distribution diagram of the tactile sensor when the magneto-electric flexible tactile sensor provided by the embodiment of the present application is not subjected to pressure is shown in FIG. 2. Figure 2 As shown, the magneto-electric flexible tactile sensor comprises a magneto-elastic body 21, a support part 22, a flexible circuit board 23, and a magnetic field sensor 24. The magnetic field sensor 24 is an array composed of three-dimensional Hall sensors MLX90393. When the magneto-elastic body 21 is not subjected to pressure, the magnetic field excited by the magneto-elastic body 21 is shown in FIG. 3. Figure 2 As shown, the magnetic field distribution at the center is mainly a component in the vertical direction, and the magnetic field value is small. The magnetic field distribution at the edge is mainly a component in the horizontal direction, and the magnetic field value is large.

[0027] Figure 3The embodiment of the present invention provides a magnetic field distribution diagram of the tactile sensor when subjected to pressure. As shown in the figure, when pressure is applied to the concave part of the magnetoelastic body 21, the deformation energy generated by the magnetoelastic body 21 causes the magnetic field it excites to change. The magnetic field sensor 24 located directly below the pressure will receive different magnetic field change responses due to the difference in the horizontal position of different sensors. Tactile information, namely pressure magnitude and pressure distribution, is calculated based on the magnetic field change.

[0028] Figure 4 A comparison chart of the pressure sensitivity of the magnetoelectric flexible tactile sensor provided in the embodiments of the present invention with that of elastomers with different structures. Figure 4 Combining Figure 1 Describe, such as Figure 4 The diagram shows the pressure sensitivity test results for four different elastomer structures: bottomless, pyramidal, cylindrical, and pyramidal. Specifically, when the same pressure is applied to the surface of the magnetoelastic 11, the deformation and magnetic field changes produced by different structures vary. The curves show that the pyramidal structure has the highest pressure sensitivity, exhibiting higher detection accuracy and measurement range in pressure sensing. Furthermore, different parameters of the pyramid structure also affect the measurement range and sensitivity of the tactile sensor. Larger pyramid structures result in higher sensitivity but a smaller measurement range; greater distances between pyramid structures decrease sensitivity but increase the measurement range. In one embodiment of the invention, the pyramid array has a height of 2mm, a base dimension of 2mm*2mm, and a distance of 4mm between each pyramid structure.

[0029] Figure 5 This diagram illustrates the sensor arrangement and pressure localization area in a magnetoelectric flexible tactile sensor provided in an embodiment of the present invention. The magnetic field sensors 51 are arranged in a 3x3 array, with each sensor spaced 1 cm apart. The 3x3 array is further divided into 4x4 pressure regions 52, and pressure is applied to each of these regions each time pressure is applied. In one embodiment, the magnetic field sensor array is 2 cm x 2 cm in size, arranged in a 3x3 array, with a 1 cm distance between adjacent sensors. Further, to better apply pressure, the 3x3 array is divided into 4x4 regions, each 0.5 cm x 0.5 cm in size. Each time pressure is applied, random pressure is applied to one or more consecutive regions within these 4x4 regions, serving as data for both the test and training sets. Furthermore, the magnetic field transformation signals captured by the magnetic field sensors are input to a pre-trained neural network model to obtain the pressure magnitude and distribution at 16 points within the 4x4 region, thus acquiring the pressure intensity of the tactile sensor at these 16 points. Specifically... Figure 5 CombiningFigure 1 The description is as follows: Figure 5 As shown in the figure, the magnetic field sensors in the magneto-elastic flexible tactile sensor are arranged in a 3*3 array form for measuring the magnetic field changes generated under the action of touch, and each sensor in the array is responsible for capturing the local magnetic field change signal, which reflects the position and size of the touch pressure. The magnetic field sensor 51 transmits the collected magnetic field signal to the signal processing unit 16 through the I2C interface, and after preprocessing such as denoising and normalization, the magnetic field signal is input into the trained neural network model. Among them, the input layer of the network receives three-dimensional data from the magnetic field sensor array, and the feature extraction is performed through the hierarchical structure of the neural network.

[0030] Figure 6 The pressure output diagram of the magneto-elastic flexible tactile sensor provided for the embodiment of the present application under different pressures. It can be seen from Figure 6 that the tactile sensor has strong positioning function under single-point pressure and high recognition accuracy, and the maximum measurement error of the pressure size is 13.5%.

[0031] In an embodiment of the present application, a convolutional neural network architecture is used, and the input magnetic field signal is processed by the weight matrix obtained by training to accurately infer the position and size of the pressure source. Specifically, based on the weight matrix obtained by training the experimental data set, the convolutional neural network can learn the mapping relationship between the magnetic field change and the pressure, and provide high-precision real-time prediction. The neural network not only improves the detection accuracy of the pressure position and size, but also reduces the array size of the magnetic field sensor, so that the pressure sensor can detect the pressure with fewer sensor numbers, and the circuit production cost is reduced.

[0032] The trained matrix can directly convert the input magnetic field signal into the position and size of the pressure through convolution operation and full connection, and in actual use, it can be directly operated with the input magnetic field signal to obtain the actual pressure position and size without further training.

[0033] Through the magneto-elastic flexible tactile sensor disclosed by the present application, the structure of the magneto-elastic body is pyramid-shaped, which improves the sensitivity of the flexible tactile sensor, and the magnetic field signal is processed by the pre-trained convolutional neural network to directly obtain the position and size of the pressure on the tactile sensor.

[0034] Those skilled in the art will readily understand that the above description is only a preferred embodiment of the present application and is not intended to limit the present application, and any modifications, equivalent replacements and improvements made within the spirit and principles of the present application shall be included in the protection scope of the present application.

Claims

1. A magnetoelectric flexible tactile sensor, characterized by, Comprise: A magnetoelastic body, a support, a flexible circuit board, a recessed part, and a magnetic field sensor, the recessed part is above the support, the magnetoelastic body is aligned with the recessed part, the magnetic field sensor is on the flexible circuit board, between the support and the flexible circuit board, the magnetoelastic body is used to generate a corresponding magnetic field distribution according to the external contact pressure, the magnetic field sensor is used to collect the magnetic field signal of the magnetoelastic body, wherein the lower part of the magnetoelastic body in contact with the recessed part is a pyramid array structure.

2. The magnetoelectric flexible tactile sensor according to claim 1, wherein, Also includes a signal processing unit, the signal processing unit contains a trained machine learning model, used to process the magnetic field signal and generate haptic information.

3. The magnetoelectric flexible tactile sensor according to claim 2, wherein, The haptic information includes pressure distribution and size.

4. The magnetoelectric flexible tactile sensor of claim 1, wherein, The magnetic field sensor array is arranged on the flexible circuit board.

5. The magnetoelectric flexible tactile sensor according to claim 4, wherein, The magnetic field sensor is arranged in a 3*3 array, the magnetic field sensor array size is 2cm*2cm, and the magnetic field sensor array is divided into 4*4 pressure areas, wherein each pressure area size is 0.5cm*0.5cm.

6. The magnetoelectric flexible tactile sensor of claim 1, wherein, The height of the pyramid array is 2mm, the base size is 2mm*2mm, and the distance between each pyramid structure is 4mm.

7. The magnetoelectric flexible tactile sensor of claim 1, wherein, The magnetic field sensor is a three-dimensional Hall sensor.

8. The magnetoelectric flexible tactile sensor of claim 1, wherein, The magnetoelastic body is a polydimethylsiloxane elastomer containing neodymium iron boron magnetic powder.

9. The magnetoelectric flexible tactile sensor of claim 1, wherein, The magnetic field sensor outputs the magnetic field signal using an I2C bus.

10. The magnetoelectric flexible tactile sensor of claim 1, wherein, The magnetoelastic body is embedded in the recessed part when it is subjected to external contact pressure.

Citation Information

Patent Citations

  • Pressure detection device and pressure detection method

    CN101520349A

  • Pressure sensitive element based on thermoplastic elastomers and surface load distribution measurement method

    CN103994844A