A MEMS multi-dimensional force sensor with high-resolution positioning function

By designing an array structure and a high-resolution positioning algorithm, a MEMS multidimensional force sensor has been developed, which solves the problems of low positioning resolution and susceptibility to signal interference, and achieves high-precision force perception and detection, applicable to fields such as robotics and medical assistive devices.

CN120084473BActive Publication Date: 2025-11-11HANGZHOU DIANZI UNIV
View PDF 1 Cites 0 Cited by

Patent Information

Application Number
CN202510268564.1
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2025-03-07
Publication Date
2025-11-11
Estimated Expiration
2045-03-07

AI Technical Summary

Technical Problem

Existing MEMS multidimensional force sensors have low positioning resolution and their output signals are easily interfered with, affecting the accuracy and reliability of multidimensional force perception. They are particularly unable to accurately acquire force information in high-precision industrial manufacturing and delicate medical surgery.

Method used

Design a MEMS multidimensional force sensor with a 4×4 array structure, including normal and tangential stress sensing units and capacitive ranging units. Employ a high-resolution positioning algorithm to improve accuracy and stability through array structure and signal processing. Combine the capacitive ranging unit to achieve the fusion of piezoresistive and capacitive sensing. Utilize the least squares optimization method to locate the Gaussian center position.

Benefits of technology

It improves the positioning accuracy and signal stability of the multi-dimensional force sensor, enhances the ability to accurately detect weak forces, and improves the reliability of perception in complex environments.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN120084473B_ABST
    Figure CN120084473B_ABST
Patent Text Reader

Abstract

This invention discloses a MEMS multidimensional force sensor with high-resolution positioning function, comprising an SOI device and an SOI cover plate. The SOI device is designed as a 4×4 array structure, with each array unit including two normal stress sensing units, one X-axis tangential stress sensing unit, one Y-axis tangential stress sensing unit, and one capacitive ranging unit. The capacitive ranging unit is located at the center of the array unit. The two normal stress sensing units are diagonally distributed at the upper right and lower left, while the X-axis and Y-axis tangential stress sensing units are diagonally distributed at the upper left and lower right. The SOI cover plate includes sixty-four evenly distributed tactile protrusions and a silicon encapsulation cover. This invention increases the data volume and reduces signal crosstalk by using the array structure and normal and tangential stress sensing units, thereby improving the accuracy and stability of the multidimensional force sensor.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of multidimensional force sensor technology, and more specifically, to a MEMS multidimensional force sensor with high-resolution positioning function. Background Technology

[0002] Multidimensional force sensors based on Micro-Electro-Mechanical Systems (MEMS) technology can achieve precise detection, localization, and sensing of weak forces. They are typically fabricated using silicon-on-insulator (SOI) wafers on an insulating substrate. MEMS multidimensional force sensors have broad application prospects in fields such as robotic tactile perception and medical assistive devices. For example, when robots interact with their environment, MEMS multidimensional force sensors enable them to perceive the magnitude, direction, and point of application of forces on objects with the same fine detail as humans, thus allowing them to complete various complex tasks more flexibly and intelligently. Specifically, on industrial production lines, robots can use MEMS multidimensional force sensors to precisely grasp and manipulate delicate components, avoiding damage caused by improper force. In service robots, MEMS multidimensional force sensors can help robots interact better with humans, such as gently assisting the elderly or patients, improving the quality and safety of services. In surgical robots, they can provide doctors with force feedback information at the surgical site, enabling them to control surgical instruments more precisely during operations, reducing surgical risks and improving the success rate and safety of surgeries. For rehabilitation training equipment, MEMS multidimensional force sensors can monitor the patient's movement force and joint stress in real time, and based on this data, develop personalized rehabilitation training plans to effectively promote the patient's rehabilitation process. However, currently, MEMS multidimensional force sensors still face problems such as low positioning resolution and weak output signals that are easily interfered with, affecting the accuracy and reliability of multidimensional force perception. This may lead to inaccurate force information acquisition in some applications with stringent positioning accuracy requirements, such as high-precision industrial manufacturing or delicate medical surgical operations, thus affecting the performance and effectiveness of the entire system. Summary of the Invention

[0003] The purpose of this invention is to provide a MEMS multidimensional force sensor with high-resolution positioning function to achieve accurate detection and perception of weak forces, providing a more accurate and reliable solution for weak force perception in various fields.

[0004] The present invention adopts the following technical solution:

[0005] A multi-dimensional force sensor based on MEMS technology includes two main parts: an SOI device and an SOI cover plate, and has high-resolution positioning function.

[0006] The SOI device is designed as a 4×4 array structure. Each array unit includes two normal stress sensing units, one X-axis tangential stress sensing unit, one Y-axis tangential stress sensing unit, and one capacitive ranging unit. The capacitive ranging unit is located at the center of the array unit. With the center of the array unit as the focal point, the array unit is divided into four equal blocks. The two normal stress sensing units are distributed diagonally to the upper right and lower left, and the X-axis tangential stress sensing unit and the Y-axis tangential stress sensing unit are distributed diagonally to the upper left and lower right.

[0007] Furthermore, the normal stress sensing unit, the X-axis tangential stress sensing unit, and the Y-axis tangential stress sensing unit have the same structure, each including a cavity, a support beam, a metal lead wire, and two sensitive beams, to realize the detection of stress in the normal and tangential directions.

[0008] The cavity is a cuboid recessed cavity with a length of 400μm, a width of 300μm, and a height of 500μm, located at the center of each sensing unit, with gaps between the cavities; above each cavity are two evenly distributed support beams with a width of 50μm, which evenly divide the space above the cavity into three rectangular spaces of the same size.

[0009] Furthermore, in the X-axis tangential stress sensing unit and the Y-axis tangential stress sensing unit, the two sensitive beams are both cuboids with a length of 400μm, a width of 50μm, and a thickness of 20μm, located on the upper surface of the support beam and distributed along the long side of the central rectangular space; in the normal stress sensing unit, the two sensitive beams are cuboids located on the upper surface of the support beam, with a width of 50μm and a thickness of 20μm, and lengths of 350μm and 400μm respectively; the metal lead is a metal trace with a width of 10μm and a height of 1μm, located on the upper surfaces of both ends of the two sensitive beams, and in the normal stress sensing unit, a metal trace with a length of 200μm, a width of 50μm, and a height of 1μm is distributed in the middle of the 400μm long sensitive beam for outputting the electrical signal of the sensitive beam.

[0010] Furthermore, in the normal stress sensing unit and the X-axis tangential stress sensing unit, the support beams are vertically distributed, while in the Y-axis tangential stress sensing unit, the support beams are horizontally distributed.

[0011] Furthermore, the capacitive ranging unit is a metal electrode located at the center of the array unit; the metal electrode is a cube with a length of 80μm, a width of 80μm, and a height of 1μm, forming a capacitor plate with the SOI cover plate to realize the detection of the plate spacing.

[0012] Furthermore, the SOI cover plate includes sixty-four evenly distributed tactile protrusions and a silicon encapsulation cover; the silicon encapsulation cover is a cube with a side length of 5000μm and a height of 200μm, and the sixty-four tactile protrusions are all cubes with a side length of 150μm and a height of 300μm, located on the lower surface of the silicon encapsulation cover; the sixty-four tactile protrusions span the two sensitive beams of each cavity and are in direct contact with the upper surface of the sensitive beams to realize stress transmission.

[0013] The implementation process of high-resolution positioning function is as follows:

[0014] Step 1: Collect stress data distribution and coordinates of all stress sensing units and capacitive ranging units in the normal (Z-axis) and tangential (X-axis, Y-axis) directions, and organize them into four sets of data: stress data in the Z-axis / X-axis / Y-axis directions and capacitive ranging data.

[0015] Step 2: Reconstruct the stress data in the Z-axis direction from Step 1 using a bivariate standard Gaussian / normal distribution, and simplify the distribution to a symmetric Gaussian distribution;

[0016] Step 3: Measure the optimal Gaussian distribution of the data, transform the localization problem into a minimization problem, and use the least squares optimization method to find the optimal parameters and obtain the coordinates of the Gaussian center.

[0017] Step 4: Repeat steps 2-3 for the remaining three sets of data to obtain the Gaussian center coordinates of all data sets. Perform a weighted average of the four sets of Gaussian center coordinates to obtain the final force position coordinates, thus achieving high-resolution positioning.

[0018] This invention designs a MEMS multidimensional force sensor with high-resolution positioning function, providing a solution for the design and application of MEMS multidimensional force sensors, and has the following beneficial effects:

[0019] This MEMS multi-dimensional force sensor with high-resolution positioning function increases the amount of data and reduces the impact of signal crosstalk by using an array structure and normal and tangential stress sensing units, thereby improving the accuracy of the multi-dimensional force sensor.

[0020] This MEMS multi-dimensional force sensor with high-resolution positioning function adopts an integrated capacitive ranging unit design to achieve piezoresistive and capacitive fusion sensing, thereby improving the stability of the multi-dimensional force sensor.

[0021] This MEMS multi-dimensional force sensor with high-resolution positioning function achieves full-plane high-resolution positioning through positioning algorithm processing, thereby improving the positioning accuracy of the multi-dimensional force sensor. Attached Figure Description

[0022] Figure 1A schematic diagram of a MEMS multidimensional force sensor with high-resolution positioning function;

[0023] Figure 2 This is a schematic diagram of the SOI device structure for a MEMS multidimensional force sensor.

[0024] Figure 3 This is a schematic diagram of the SOI cover plate structure for a MEMS multidimensional force sensor.

[0025] Figure 4 Fabrication process diagram of a MEMS multidimensional force sensor with high-resolution positioning function;

[0026] Figure 5 A schematic diagram illustrating the principle of triaxial force detection and capacitive ranging in a MEMS multidimensional force sensor.

[0027] Figure 6 Flowchart of the high-resolution positioning algorithm for a MEMS multidimensional force sensor;

[0028] Figure 7 This is a graph showing the high-resolution positioning output data of a MEMS multidimensional force sensor. Detailed Implementation

[0029] The present invention will be further explained below with reference to embodiments and accompanying drawings. The following embodiments are for illustrative purposes only and are not intended to limit the scope of the invention.

[0030] Figure 1 , Figure 2 and Figure 3 This paper presents a MEMS multidimensional force sensor structure with high-resolution positioning function, which consists of SOI device I and SOI cover plate II.

[0031] In this example, the SOI device I is designed as an array structure, comprising 16 array units 1 in a 4×4 configuration. Each array unit 1 includes two normal stress sensing units 2, one X-axis tangential stress sensing unit 3, one Y-axis tangential stress sensing unit 4, and one capacitive ranging unit 5. The X-axis tangential stress sensing unit 3 and the Y-axis tangential stress sensing unit 4 are diagonally distributed, as are the two normal stress sensing units 2, enabling triaxial force detection. The capacitive ranging unit 5 is located at the center of array unit 1, with its metal electrode forming a capacitor plate with the SOI cover plate II above, enabling electrode spacing detection. The cavity 13 of the stress sensing unit is located at the center of the support beam 14, the sensitive beam 12 is located on the upper surface of the cavity 13, and metal leads 11 are located on the upper surfaces of both ends of the sensitive beam 12. The SOI cover plate II is designed as an array structure, including sixty-four tactile protrusions 16 and a silicon encapsulation cover 15. The tactile protrusion 16 is located on the lower surface of the silicon package cover 15 and is in direct contact with the upper surface of the sensitive beam 12 of the stress sensing unit of the SOI device I below, thereby realizing the transmission of stress.

[0032] This application also provides a fabrication process for a MEMS multidimensional force sensor with high-resolution positioning capabilities, see [link to application]. Figure 4 The fabrication process of a MEMS multidimensional force sensor with high-resolution positioning function includes the following steps:

[0033] The top silicon layer is P-type. <100> On a high-resistivity silicon (SOI) wafer with a resistivity >1000 ohm·cm, the surface of the z-axis beam is lightly doped, the sides of the x / y-axis beam are lightly doped, and the surface of the contact metal region is heavily doped.

[0034] A thermally oxidized insulating layer is formed on the surface of the SOI wafer, and the patterned thermally oxidized insulating layer is etched.

[0035] After sputtering a layer of Au metal onto the surface of the SOI wafer, it is patterned using a metal lift-off technology.

[0036] The top silicon layer of the SOI wafer is formed by reactive ion etching (RIE) on the front side to create a beam structure.

[0037] Reactive ion etching (RIE) on the back side of the SOI wafer separates the silicon dioxide film from the underlying silicon, thus releasing the back cavity.

[0038] On the back of another silicon wafer, reactive ion etching is used to form tactile bumps, which are then aligned and bonded to the SOI wafer.

[0039] Thus, the fabrication of a MEMS multidimensional force sensor with high-resolution positioning function has been completed.

[0040] Figure 5This diagram illustrates the principle of triaxial force detection and capacitive ranging in a MEMS multidimensional force sensor. As shown in the left figure, the triaxial force detection principle involves phosphorus ion doping of the sensitive beam of the stress sensing unit to form a P-type silicon varistor. In the tangential stress sensing unit, two varistors, R1 and R2, are doped and located in the middle of the sensitive beam. Under tangential force, the resistance of R1 increases by ΔR1 with beam tension, while the resistance of R2 decreases by ΔR2 with beam compression. In the normal stress sensing unit, varistor R3 is located in the middle of the sensitive beam, and varistor R4 is located at both ends. Under normal stress, the resistance of R3 increases by ΔR3 with beam tension, while the resistance of R4 decreases by ΔR4 with beam compression. The force signal is converted into an electrical signal by the varistors and output through metal leads. When R1 = R2 and R3 = R4, a differential output using a bridge circuit can reduce the influence of capacitive ranging. When the coefficient of friction is known, slippage will not occur if the tangential force is less than the maximum frictional force obtained through the normal force. This helps to avoid the decrease in measurement accuracy caused by slippage and improves the accuracy of force detection. The working principle of capacitive ranging is shown in the right figure. The capacitive ranging unit is located at the center of the array unit. The metal electrode and the SOI cover plate above form capacitor plates. The potential difference between the two plates causes an electrostatic field distribution between them. When subjected to normal stress, the distance between the two plates changes, and the capacitance C changes accordingly. The distance can be detected by the change in capacitance ΔC.

[0041] Figure 6 and Figure 7 This diagram shows the algorithm flowchart and output data of the high-resolution positioning function of a MEMS multi-dimensional force sensor. The Gaussian center coordinates of each data set are obtained by processing, reconstructing, and solving four sets of data (Z-axis stress data, X-axis stress data, Y-axis stress data, and capacitance ranging data). Figure 7 As shown, taking a set of data as an example, the reconstructed data coordinates in the entire plane are (x... i y i The corresponding original data output value is z. i The output of the data after bivariate standard normal distribution is f(x). i y i After simplification to a symmetric Gaussian distribution, the data output is as follows: Where p0,x c0 ,y c0 The four constant terms σ0 are the four Gaussian parameters for the optimal Gaussian distribution, and the Gaussian parameters x c0 and y c0 Combination (x) c0 ,y c0 Let ( ) be the coordinates of the center of the Gaussian distribution. The optimal Gaussian distribution is found using the least squares optimization method. The specific process is as follows: Define the loss function for the data output distribution as... N represents the number of data points in this dataset. The loss function has four Gaussian parameters p0, x... c0 ,y c0 Find the partial derivatives of σ0 and make them zero, then substitute them into z. i f(x) i y i A system of nonlinear equations is obtained. Solving the system of nonlinear equations yields the four Gaussian parameters p0, x for the optimal Gaussian distribution. c0 ,y c0 σ0 and the coordinates of the Gaussian center (x c0 ,y c0 This minimizes the loss function. Finally, the four sets of Gaussian center position coordinates (x, y) are obtained. c0 ,y c0 ), (x c1 ,y c1 ), (x c2 ,y c2 ), (x c3 ,y c3 The final force location coordinates (x, y) are obtained by performing a weighted average. c ,y c This enables high-resolution positioning.

[0042] Finally, it should be noted that the above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A MEMS multidimensional force sensor with high-resolution positioning function, characterized in that, It consists of two parts: SOI device and SOI cover plate; The SOI device is designed as a 4×4 array structure. Each array unit (1) includes two normal stress sensing units (2), one X-axis tangential stress sensing unit (3), one Y-axis tangential stress sensing unit (4), and one capacitive ranging unit (5). The capacitive ranging unit (5) is located at the center of the array unit (1). With the center of the array unit (1) as the focus, the array unit (1) is divided into four blocks of the same size. The two normal stress sensing units (2) are distributed diagonally to the upper right and lower left, and the X-axis tangential stress sensing unit (3) and the Y-axis tangential stress sensing unit (4) are distributed diagonally to the upper left and lower right.

2. The MEMS multidimensional force sensor with high-resolution positioning function according to claim 1, characterized in that, The normal stress sensing unit (2), the X-axis tangential stress sensing unit (3), and the Y-axis tangential stress sensing unit (4) have the same structure, each including a cavity (13), a support beam (14), a metal lead wire (11), and two sensitive beams (12) to realize the detection of normal and tangential stress. The cavity (13) is a rectangular recessed cavity located at the center of each sensing unit, and there is a gap between the cavities (13); above each cavity (13) there are two evenly distributed support beams (14) with a width of 50μm, which divide the space above the cavity (13) into three rectangular spaces of the same size.

3. The MEMS multidimensional force sensor with high-resolution positioning function according to claim 2, characterized in that, The two sensitive beams (12) in the X-axis tangential stress sensing unit (3) and the Y-axis tangential stress sensing unit (4) are both cuboids, located on the upper surface of the support beam (14), and distributed along the long side of the middle rectangular space; the two sensitive beams (12) in the normal stress sensing unit (2) are cuboids located on the upper surface of the support beam (14), with the same width as the support beam (14) and lengths of 350μm and 400μm respectively.

4. The MEMS multidimensional force sensor with high-resolution positioning function according to claim 3, characterized in that, The metal lead (11) is located on the upper surface of both ends of the two sensitive beams, and a metal lead with a length of 200 μm is distributed in the middle of the sensitive beam with a length of 400 μm in the normal stress sensing unit (2), with the same width as the sensitive beam, for outputting the electrical signal of the sensitive beam.

5. The MEMS multidimensional force sensor with high-resolution positioning function according to claim 4, characterized in that, In the normal stress sensing unit (2) and the X-axis tangential stress sensing unit (3), the support beam (14) is vertically distributed, and in the Y-axis tangential stress sensing unit (4), the support beam (14) is horizontally distributed.

6. The MEMS multidimensional force sensor with high-resolution positioning function according to claim 5, characterized in that, The capacitance ranging unit (5) is a metal electrode located at the center of the array unit (1); the metal electrode is a cube and forms a capacitor plate with the SOI cover plate to realize the detection of the plate spacing.

7. The MEMS multidimensional force sensor with high-resolution positioning function according to claim 6, characterized in that, The SOI cover plate includes sixty-four evenly distributed tactile protrusions (16) and a silicon encapsulation cover (15); the silicon encapsulation cover is a cube, and the sixty-four tactile protrusions (16) are all cubes with a side length of 150μm and a height of 300μm, located on the lower surface of the silicon encapsulation cover (15); the sixty-four tactile protrusions span the two sensitive beams (12) of each cavity (13) and are in direct contact with the upper surface of the sensitive beams to realize stress transmission.

Citation Information

Patent Citations

  • TSV switching-based tactile sensor

    CN120136021A