Attitude angle resolving method of square section bending sensor
By calculating the average curvature and resistance change of the sensing unit, combining the sensitivity coefficient and the total length after stretching, the problem of large posture angle error of the bending sensor is solved, and the accurate description of joint motion and attitude angle solution is achieved.
Patent Information
- Application Number
- CN202410956961.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Priority Date
- 2023-07-21
- Filing Date
- 2024-07-17
- Publication Date
- 2024-11-08
AI Technical Summary
When bending sensors describe joint motion postures, the posture angle error is large, making it difficult to accurately describe the movement postures at the joint.
By obtaining the deformation of the sensing unit, calculating its average curvature and resistance change, combining the sensitivity coefficient and the total length after stretching, the attitude angle is calculated to ensure an accurate description of joint movement.
Accurate description of joint motion is achieved, attitude angle error is reduced, and attitude angle solution accuracy of the sensing unit is improved.
Smart Images

Figure CN118913077A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of attitude angle calculation of a sensor unit, and in particular to an attitude angle calculation method of a square cross-section bending sensor. Background Art
[0002] Flexible stretchable bending sensors with elastic integrated structures have gradually become the research focus in the field of kinematic measurement in flexible environments in recent years. They not only provide good interface compatibility, but also show good measurement accuracy. They can be applied to sports training, health monitoring, disease rehabilitation, soft robots, wearable devices, human-computer interaction and many other fields.
[0003] There are two main requirements for bending sensors in flexible multi-directional motion perception. First, in large-scale applications, since it is impossible to debug and calibrate the sensors one by one for each joint of each specific user, the device should be adaptive to different joints. Second, in three-dimensional space, the joint state in open-chain motion (such as human motion) should be described by a rotation matrix, so the sensor output should provide information for the rotation matrix to achieve joint bending description, such as using Euler angles or quaternions. However, the posture angle error of the bending sensor obtained in this way is large, and it is difficult to accurately describe the motion posture at the joint. Summary of the invention
[0004] The main technical problem solved by the present invention is to provide a method for calculating the posture angle of a square cross-section bending sensor, which solves the problem that the posture angle error of the bending sensor is large and it is difficult to accurately describe the motion posture at the joint, comprising the steps of:
[0005] Step S1: obtaining a deformation difference of the sensing unit according to the deformation of the symmetrically arranged sensing units, and obtaining an average curvature of the sensing unit from the deformation difference;
[0006] Step S2: obtaining the resistance change of the sensor unit according to the average curvature, the total length of the sensor unit after the sensor unit is stretched, and the angle change of the sensor unit;
[0007] Step S3: acquiring the attitude angle of the sensor unit according to the resistance change, the sensitivity coefficient of the sensor unit, the total length of the sensor unit after the sensor unit is stretched, and the angle change of the sensor unit.
[0008] The beneficial effects of the present invention are as follows: the present invention targets a symmetrically arranged sensor unit, obtains the deformation difference of the sensor unit according to the deformation of the sensor unit, and obtains the average curvature of the sensor unit from the deformation difference; then obtains the resistance change of the sensor unit from the average curvature, the total length of the sensor unit after the sensor unit is stretched, and the angle change of the sensor unit; finally, obtains the attitude angle of the sensor unit according to the resistance change, the sensitivity coefficient of the sensor unit, the total length of the sensor unit after the sensor unit is stretched, and the angle change of the sensor unit. In this way, the performance parameters of the sensor unit itself can be fully considered, and a relatively accurate attitude angle of the sensor unit can be obtained, thereby ensuring the accuracy of the description of the joint movement. BRIEF DESCRIPTION OF THE DRAWINGS
[0009] Figure 1 is a schematic structural diagram of a bending sensor according to an embodiment of the present invention;
[0010] Figure 2 is a schematic diagram of an exploded structure of a bending sensor according to an embodiment of the present invention;
[0011] Figure 3 is a schematic structural diagram of one end of a bending sensor according to an embodiment of the present invention;
[0012] Figure 4 is a schematic structural diagram of the other end of the bending sensor according to an embodiment of the present invention;
[0013] Figure 5 is a schematic structural diagram of a flexible substrate of a bending sensor according to an embodiment of the present invention;
[0014] Figure 6 is a schematic structural diagram of one end of a flexible substrate according to an embodiment of the present invention;
[0015] Figure 7 is a schematic structural diagram of the other end of the flexible substrate according to an embodiment of the present invention;
[0016] Figure 8 is a schematic diagram of a sensor under the pitch motion rotation axis according to an embodiment of the present invention;
[0017] Fig. 9 is a schematic diagram of a sensor under a yaw motion rotation axis according to an embodiment of the present invention;
[0018] Fig.10 is a schematic diagram of a signal modification circuit according to an embodiment of the present invention;
[0019] Fig.11 is a schematic diagram of signal quality comparison according to an embodiment of the present invention;
[0020] Fig.12 is a schematic diagram of a circuit structure connected according to an embodiment of the present invention;
[0021] Fig.13 is a schematic flow chart of a method for calculating an attitude angle of a bending sensor according to an embodiment of the present invention;
[0022] Fig.14 is a schematic diagram of bending at a joint according to an embodiment of the present invention;
[0023] Fig.15 is a schematic diagram of a process for manufacturing a bending sensor according to an embodiment of the present invention;
[0024] Fig.16 is a schematic structural diagram of a rigid mold according to an embodiment of the present invention;
[0025] Fig.17 is a schematic cross-sectional structural diagram of a rigid mold according to an embodiment of the present invention;
[0026] Fig.18 is a schematic cross-sectional structural diagram of one end of a rigid mold according to an embodiment of the present invention;
[0027] Fig.19 is a schematic cross-sectional view of the other end of the rigid mold according to an embodiment of the present invention;
[0028] Fig. 20 is a schematic diagram of an anti-interference test according to an embodiment of the present invention;
[0029] Fig.21 is a schematic diagram of a resistance-strain response curve of a sensor according to an embodiment of the present invention;
[0030] Fig. 22 is a schematic diagram of a sensor stability test according to an embodiment of the present invention;
[0031] Fig.23 is a schematic diagram of a sensor resolution test according to an embodiment of the present invention;
[0032] Fig.24 is a data comparison diagram of a fractured wrist joint according to an embodiment of the present invention;
[0033] Fig.25 is a graph of spinal status monitoring data according to an embodiment of the present invention;
[0034] Fig.26 This is a lumbar vertebrae status monitoring data diagram according to an embodiment of the present invention. DETAILED DESCRIPTION
[0035] In order to facilitate the understanding of the present invention, the present invention is described in more detail below in conjunction with the accompanying drawings and specific embodiments. Preferred embodiments of the present invention are provided in the accompanying drawings. However, the present invention can be implemented in many different forms and is not limited to the embodiments described in this specification. On the contrary, the purpose of providing these embodiments is to make the understanding of the disclosure of the present invention more thorough and comprehensive.
[0036] It should be noted that, unless otherwise defined, all technical and scientific terms used in this specification have the same meaning as those commonly understood by those skilled in the art of the present invention. The terms used in the specification of the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention. The term "and / or" used in this specification includes any and all combinations of one or more related listed items.
[0037] Figure 1-Figure 7 An embodiment of the flexible multi-directional bending sensor of the present invention is shown, comprising: a flexible substrate 1, and a plurality of pairs of sensing units 2 symmetrically and evenly distributed on the surface of the flexible substrate 1, wherein the flexible substrate 1 is arranged at a position to be measured and deforms with the movement of the position to be measured, and the sensing unit 2 is used to detect the deformation of the flexible substrate 1 in different directions; the sensing unit 2 comprises a first electrode 21, a sensitive element 22 and a second electrode 23, the sensitive element 22 is connected between the first electrode 21 and the second electrode 23, and the first electrode 21 and the second electrode 23 are located at both ends of the flexible substrate 1 in the length direction.
[0038] In the present application, the flexible substrate 1 can adapt to the complex mechanical environment of the soft surface, and can adaptively match the size and shape of the tested part (elbow joint, knee joint, wrist joint and other moving joints). When the tested part moves, the flexible substrate 1 produces corresponding deformation along with the movement of the tested part. Multiple groups of two-to-two corresponding sensing units 2 can detect the deformation of the flexible substrate 1 in multiple different directions, and then obtain the movement information of the tested part. The deformation of the tested part can be accurately obtained with high accuracy.
[0039] Preferably, Figure 3 As shown, the sensing unit 2 in a sensor includes a vertical sensing unit 200 and a lateral sensing unit 210. The vertical sensing unit 200 is arranged on the upper and lower sides of the flexible substrate 1 to detect the vertical deformation of the flexible substrate 1. The lateral sensing unit 210 is arranged on the left and right sides of the flexible substrate 1 to detect the lateral deformation of the flexible substrate 1.
[0040] When the sensor unit 2 is measuring, the posture angles related to bending include pitch angle and yaw angle, which rotate around the x-axis and y-axis of the mechanism section respectively. Since the angles of rotation around the x-axis and y-axis are vectors, the sensor array topology must contain at least four outputs to ensure the existence of a unique solution to the non-homogeneous linear equations. A sensor array consisting of at least four sensor units is required to characterize the multi-directional movement of the joint.
[0041] Preferably, the vertical sensing unit 200 and the horizontal sensing unit 210 are each a pair.
[0042] Preferably, Figure 5 As shown, a first groove 11 adapted to the first electrode 21 is provided on the side surface of one end of the flexible substrate 1, and the first electrode 21 is arranged in the first groove 11. A second groove 12 adapted to the second electrode 23 is provided on the side surface of the other end of the flexible substrate 1. The second electrode 23 is arranged in the second groove 12. A sensitive groove 13 is provided on the side surface of the flexible substrate 1, and the sensitive element 22 is arranged in the sensitive groove 13. The first groove 11, the second groove 12 and the sensitive groove 13 facilitate fixing the first electrode 21, the second electrode 23 and the sensitive element 22.
[0043] Preferably, the outer surface of the first electrode 21 and / or the second electrode 23 is flush with the corresponding outer surface of the flexible substrate 1. This ensures the accuracy of the sensing unit 2 in detecting the deformation of the flexible substrate 1.
[0044] The first electrodes 21 and the second electrodes 23 may correspond one to one, or a plurality of first electrodes 21 may correspond to one second electrode 23 .
[0045] Preferably, Figure 3 and Figure 4 As shown, a plurality of first electrodes 21 in the sensing unit 2 are connected to a second electrode 23 .
[0046] Preferably, Figure 4 As shown, a connector 24 is provided on the end surface of the flexible substrate 1 at one end of the second electrode 23, and the connector 24 has a plurality of connection ends 241, one connection end 241 is connected to the second electrode 23, and the other connection ends are respectively connected to the corresponding sensitive components 22. In this way, a plurality of first electrodes 21 and sensitive components 22 can share a second electrode 23. Four sensing units 2 are connected in parallel, and the four sensing units 2 are grounded and only led out through one second electrode 23, thereby reducing the heterogeneous interface in the sensing unit 2 and improving the overall stability of the sensor.
[0047] Preferably, Figure 7As shown, a second groove 12 is provided on the side surface of the flexible substrate 1 where one end of the second electrode 23 is provided, and a connecting groove 14 adapted to the connecting member 24 is provided on the end surface of the flexible substrate 1 corresponding to the connecting member 24, and the outer end surface of the connecting member 24 is flush with the corresponding end surface of the flexible substrate 1.
[0048] Preferably, the sensor unit 2 is provided with four connecting members 24, and the shape of the connecting groove 14 is a cross.
[0049] Preferably, the flexible substrate 1 is in a rod-shaped structure.
[0050] The cross-sectional shape of the sensor formed by the flexible substrate 1 , the first electrode 21 and the second electrode 23 is rectangular, circular or square.
[0051] The definition of section resistance moment is: the moment of inertia of the cross section of the bent rod about its neutral axis divided by the distance from the neutral axis to the outermost edge of the section. Section resistance moment characterizes the magnitude of the normal stress generated in the section after the bending moment is loaded on the section. The larger the section resistance moment, the greater the resistance of the rod when it is bent, and the easier it is to produce surface deformation, reducing the accuracy of the measurement results. That is, the larger the section resistance moment, the less accurate the measurement results.
[0052] Due to the extremely high cross-sectional resistance moment, the flexible substrate will produce unpredictable out-of-plane wrinkle deformation when bending sideways, which will lead to high system errors. Therefore, when performing multi-degree-of-freedom measurements, it is necessary to select a suitable cross-sectional shape of the sensor.
[0053] When the cross-sectional shape of the sensor is a rectangle, the cross-sectional resistance moment W of the sensor is rect for:
[0054]
[0055] Among them, W rect is the section resistance moment, b is the section length in the direction perpendicular to the bending moment, and h is the section length in the direction of the bending moment.
[0056] like Figure 8 and Fig. 9 As shown, when pitching motion is performed, that is, when the sensor is rotated with the central axis in the width direction at the center of the sensor as the axis, the width direction is b and the height direction is h. When yaw motion is performed, that is, when the sensor is rotated in the vertical direction perpendicular to the axis at the center of the sensor, the width direction is h and the height direction is b.
[0057] During pitch motion, b is much larger than h, and the cross-sectional resistance moment is within an acceptable range. During yaw motion, the positions of b and h are interchanged, and the cross-sectional resistance moment increases by almost h / b times. According to the principle of minimum energy, a huge cross-sectional resistance moment will cause structural instability, resulting in uncontrollable surface deformation. In this way, if the resistance moments of the rectangular cross section in two directions are expected to be similar and as small as possible, a miniaturized symmetrical cross section is required, that is, b=h.
[0058] In order to make b=h, the cross-sectional shape of the sensor is made circular. When the cross-sectional shape of the sensor is circular, its cross-sectional resistance moment W cir It is expressed as:
[0059]
[0060] Among them, W cir is the cross-sectional resistance moment of a circular cross section; d is the diameter of the circular cross section. Corresponding to the rectangular case, the parameter situation at this time is b=h=d. Since π / 32<1 / 6, the cross-sectional resistance moment of the circular cross section is also very low.
[0061] However, although the circular cross-section has a low section resistance moment, which prevents the sensor from undergoing surface deformation, the circular cross-section is prone to unexpected torsion under the action of friction, which makes it impossible to accurately follow the deformation of the joint skin and introduces new system errors.
[0062] Therefore, in order to solve the problem that the rectangular cross-section is prone to uncontrollable surface deformation and the circular cross-section cannot accurately follow the deformation of the joint skin, new system errors are introduced.
[0063] Preferably, the cross-section of the sensor is a square, including the cross-section of the flexible substrate is a square, and the cross-section of the sensing unit is a square or a rectangle.
[0064] When the cross-sectional shape of the sensor is a square, it is symmetrical in the vertical, horizontal and diagonal directions, that is, b=h, which reduces the cross-sectional resistance moment and is not easy to twist. This avoids the problem that the rectangular cross-section is prone to uncontrollable surface deformation, and the circular cross-section cannot accurately follow the deformation of the joint skin, introducing new system errors.
[0065] To accurately calculate the omnidirectional bending angle, obtaining a high-quality signal is the most important factor. A 1kΩ resistor is used as a voltage divider to extract the resistance change of the liquid metal, and the Fig.10The signal modification circuit shown modifies the signal. The signal modification circuit includes a first resistor R1, a first amplifier OP1, a third resistor R3, a fourth resistor R4, a fifth resistor R5 and a first filter OP2 connected in series in sequence. One end of the first resistor R1 is grounded, and the other end is connected to the negative electrode of the first amplifier OP1. The positive electrode of the first amplifier OP1 is used as an input end. The output end of the first amplifier OP1 is electrically connected to the third resistor R3, and the first amplifier OP1 is connected in parallel with a second resistor R2. A first capacitor C1 is electrically connected between the third resistor R3 and the fourth resistor R4 and then grounded. A first end of a third capacitor C3 is electrically connected between the fourth resistor R4 and the fifth resistor R5, and the second end of the third capacitor C3 is connected to the negative electrode of the first filter OP2. One end of a second capacitor C2 is electrically connected between the fifth resistor R5 and the first filter OP2, and the other end of the second capacitor C2 is grounded, and is electrically connected to a sixth resistor R6 and then connected to the second end of the third capacitor C3. The fifth resistor R5 is electrically connected to the positive electrode of the first filter OP2.
[0066] The reasons for not using impedance matching technology are as follows:
[0067] In terms of signal quality, the proposed circuit design can significantly improve the signal-to-noise ratio. Fig.11 Four cases are compared, where A is impedance matching and a high voltage divider; B is no amplification; C is amplification but no low-pass filter; and D is amplification and low-pass filter. The resistance ranges from 5Ω to 9Ω, the power supply is 3.3V DC, and the noise is 100mV 50Hz.
[0068] The results show that in the unmatched case, the signal-to-noise ratio (SNR) of the original signal is only 1.1dB, while in the matched case it can reach 24.67dB, close to the requirements of normal analog signals. After careful signal modification, the SNR soared to 43.96dB, which can accurately capture the resistance change of the sensor.
[0069] In addition, the power consumption of impedance matching is unacceptable in this system. Considering the continuity of wearing, we only use a 3.3V lithium battery for power supply. The resistance of liquid metal is about 5Ω, so if the standard impedance matching procedure is used, the total resistance of one branch is only 10Ω:
[0070]
[0071] Among them, R b1 To R b4is the resistance of each branch. Therefore, the system input resistance will be less than 3Ω and the current will exceed 1A, which is higher than the current value that the lithium battery can provide. On the other hand, with a resistance of 1kΩ, the shunt circuit only needs less than 15mA. Therefore, the proposed method (unbalanced impedance and signal conditioning) is better than direct impedance matching in terms of both noise and power consumption.
[0072] like Fig.12 As shown, in this application, the sensitive part can be made of liquid metal. When connecting the signal conditioning circuit, the liquid metal is connected to the voltage divider circuit, amplifier, filter circuit, ADC, and microcontroller in sequence. The microcontroller is connected to the battery and WIFI module. The battery is connected to the power management module. The WIFI module can transmit the circuit detected by the sensor to the host computer. ADS1256 is used as the ADC chip, and it is controlled by the STM32F10C8T63 chip. The digitized signal is sent to the host computer through ATK-ESP-01.
[0073] In the present application, the cross-sectional shape of the sensor is a square, and four sensing units are symmetrically arranged on the flexible substrate. They are the first sensing unit, the second sensing unit, the third sensing unit and the fourth sensing unit. The first sensing unit and the second sensing unit form a group, forming a vertical sensing unit, which detects the vertical deformation of the flexible substrate. The third sensing unit and the fourth sensing unit are lateral sensing units that detect the lateral deformation of the flexible substrate. The meaning of using the numbers 1 / 2 / 3 / 4 to replace i in the formula below will not be repeated.
[0074] In the following, the posture angle calculation method of the bending sensor is determined. Fig.13 As shown, the steps include:
[0075] Step S1: obtaining a deformation difference of the sensing unit according to the deformation of the symmetrically arranged sensing units, and obtaining an average curvature of the sensing unit from the deformation difference;
[0076] Step S2: obtaining the resistance change of the sensor unit according to the average curvature, the total length of the sensor unit after the sensor unit is stretched, and the angle change of the sensor unit;
[0077] Step S3: acquiring the attitude angle of the sensor unit according to the resistance change, the sensitivity coefficient of the sensor unit, the total length of the sensor unit after the sensor unit is stretched, and the angle change of the sensor unit.
[0078] The deformation ε of the sensing unit i It is expressed as:
[0079] ε i =ε xi +ε yi
[0080]
[0081] Wherein, i is the number of the sensor unit, i is represented by 1 / 2 / 3 / 4, which respectively correspond to the first sensor unit, the second sensor unit, the third sensor unit and the fourth sensor unit in the present application, and the meaning of using the numbers 1 / 2 / 3 / 4 to replace i in the formula below will not be repeated. xi is the deformation of the i-th sensor unit when it bends around the x-axis, ε yi is the deformation of the i-th sensor unit when it bends around the y-axis. xi_strain , ε yi_strain are the tensile deformations of the i-th sensor unit around the x-axis and y-axis, ε xi_bending , ε yi_bending are the bending deformations of the i-th sensing unit around the x-axis and y-axis respectively.
[0082] The projection positions of the first sensor unit and the second sensor unit on the vertical plane are the same, so when rotating around the y-axis, ε x1_bending =ε y2_bending Similarly, the projection positions of the third sensor unit and the fourth sensor unit on the horizontal plane are the same, so when rotating around the x-axis, ε x3_bending =ε x4_bending At the same time, considering that the four sensor units are located at the center of the surface, when rotating around the y-axis, ε x1_bending =ε y2_bending =0; when rotating around the x-axis, ε x3_bending =ε x4_bending =0.
[0083] Since the tensile / compressive deformation experienced by each sensor unit is the same after removing the bending deformation, the deformation of the mutually symmetrical sensor units is subtracted and the effect of stretching is ignored. After considering the above relationship, the deformation difference of the sensor unit is expressed as:
[0084]
[0085] Furthermore, the average curvatures around the x-axis and around the y-axis are obtained by deformation difference and are κ x , κ y :
[0086]
[0087] Among them, t x is the thickness of the first sensing unit and the second sensing unit, t y is the thickness of the third sensing unit and the fourth sensing unit.
[0088] The above calculations show that when acquiring the curvature, whether the deployment position of the sensor unit is on the central axis of the joint has no effect.
[0089] Furthermore, the tensile deformation ε of the sensing unit is expressed by the sum of the mutually symmetrical sensing units in any group of sensing units:
[0090] ε1+ε2=2(ε x_strain +ε y_strain )=2ε
[0091] Considering the existence of errors, the present application averages the sum of the two groups of sensor units to reduce the influence of the errors. In this way, the tensile deformation ε of the sensor unit is expressed as:
[0092]
[0093] Since the sensor unit is to be deployed on a flexible interface, its extended shape cannot be a standard arc, such as Fig.14 However, this irregular deformation does not affect the result of the angle solution.
[0094] For x-axis rotation, in any small area ds between AB, we have:
[0095]
[0096] Wherein: k1 and k2 are the sensitivity coefficients of the first sensing unit and the second sensing unit respectively, △R1 and △R2 are the resistance and resistance change of the first sensing unit and the second sensing unit respectively, and dR is the differential of resistance.
[0097] Considering the relationship dR / ds=R / L, dθ / ds=1 / r, substituting it into the integral, we can get:
[0098]
[0099] Where: L is the total length of the sensor unit after stretching, r is the average curvature radius of the sensor unit, and dθ is the angular change of the sensor unit.
[0100] Since the resistance change of the sensing unit is:
[0101] ΔR i =k i ε i
[0102] k i is the sensitivity coefficient. The initial angle is obtained by rotating the angle θ A -θ B :
[0103] θ A -θ B =κL
[0104] Where: θ A is the angle of the tangent line when the rotation starts; θB is the angle of the tangent after the rotation. So here is the angle of rotation. κ is the inverse of the radius of curvature.
[0105] The rotation angle of the two axes of the sensor unit relative to the initial state can be represented by the resistance change of the four surfaces:
[0106]
[0107] Among them, θ x With θ y are the rotation angles of the sensor unit around the x-axis and the y-axis respectively, l is the original length of the sensor unit, and t represents the thickness of the sensor unit in two axial directions.
[0108] When the flexible substrate is stretched, its transverse interface deformation is not uniform everywhere, and the Poisson's ratio cannot be used to describe the thickness state. Therefore, the concept of "average thickness" is introduced when calculating the average curvature. Since the structure of the sensor unit is symmetrical on all four sides, the mechanical properties of the sensor unit are isotropic. It is believed that the transverse interface deformation of the sensor unit is not uniform everywhere but isotropic. Therefore, the average thickness t av It is expressed as:
[0109]
[0110] Where: V represents the volume of the sensor unit.
[0111] Then the determining formula of the two attitude angles is expressed as:
[0112]
[0113] Among them, t origin is the initial thickness of the sensing unit.
[0114] Therefore, it is determined that the angle algorithm in the present application is not affected by the bending state.
[0115] In application, since the signal needs to flow through the signal conditioning circuit first, the voltage change is used for solution. Since the resistance is proportional to its divided voltage, the amplifier is also in the linear range, so the above proof still holds.
[0116] For the symmetrically arranged sensor units, the deformation difference of the sensor units is obtained according to the deformation of the sensor units, and the average curvature of the sensor units is obtained from the deformation difference; the resistance change of the sensor units is then obtained from the average curvature, the total length of the sensor units after the sensor units are stretched, and the angle change of the sensor units; finally, the attitude angle of the sensor units is obtained according to the resistance change, the sensitivity coefficient of the sensor units, the total length of the sensor units after the sensor units are stretched, and the angle change of the sensor units. In this way, the performance parameters of the sensor units themselves can be fully considered, and a more accurate attitude angle of the sensor units can be obtained, ensuring the accuracy of the description of the joint movement.
[0117] In the above, the resistance-strain parameters are all defaulted to linear, and pre-stretching is performed during measurement, so materials with high linearity and low hysteresis are required to meet the needs of accurate sensing. Due to the physical and chemical properties of carbon materials themselves, their derivative materials generally have multiple linear intervals and a certain hysteresis. This characteristic will make the algorithm unable to distinguish the current deformation when measuring multiple degrees of freedom, reducing accuracy.
[0118] In order to avoid the above problems, the material of the sensitive component is liquid metal that is liquid at room temperature, such as indium, gallium, indium gallium alloy, etc. Liquid metal as a sensitive material has ultra-high linearity and ultra-low hysteresis, and can solve the problem of low sensitivity of carbon materials through signal conditioning in the electronic circuit design part.
[0119] Preferably, the material of the sensitive part is indium gallium alloy. When exposed to air, gallium will generate gallium oxide and wrap the entire object to prevent further oxidation inside. The presence of this oxide film changes the mechanical properties of the alloy - in an oxygen-free environment, the surface tension of a pure indium gallium alloy droplet is only 75% of that of an indium gallium alloy with an oxide film, which makes the latter approximate an elastic solid under low surface stress. Moreover, the oxide layer also gives the alloy stronger adhesion, which is also the basis of the physical properties of indium gallium alloy coated on a flexible substrate. Sensors made of indium gallium alloy have excellent linearity, high resolution and low error.
[0120] The mass ratio of indium to gallium in the indium gallium alloy is: Ga:In=60-80:20-40.
[0121] Preferably, the mass ratio of indium to gallium in the indium-gallium alloy is: Ga:In=75.5:24.5, expressed as GaIn24.5.
[0122] The preparation method of the indium gallium alloy is as follows: 75.5 g of gallium metal and 24.5 g of indium metal are weighed and continuously heated at 100°C until both are melted into liquid. Thereafter, a magnetic stirrer is used to stir for 5 minutes to evenly mix them, thereby obtaining high-purity GaIn24.5.
[0123] Furthermore, the material was oxidized at 70°C for 1 hour to produce a liquid metal ink with high viscosity. Gallium oxide will form an oxide film on the surface of the material to prevent further oxidation. Therefore, a magnetic stirrer is required to stir continuously during oxidation. This improves the adhesion of the indium gallium alloy.
[0124] Preferably, the material of the flexible substrate is elastic rubber, silicone rubber or the like.
[0125] Preferably, the material of the flexible substrate is platinum-catalyzed silicone rubber.
[0126] If the elastic modulus of the flexible substrate and the package in the sensitive part are inconsistent, the two will produce different modes of strain under the same external tension, so that the upper and lower surfaces of the liquid metal are subjected to strong shear force, which interferes with its linearity and cannot reflect the real deformation. Moreover, during long-term use, this shear force will also make the sensor prone to breakage and premature failure. Therefore, platinum-catalyzed silicone rubber is used as the material for the flexible substrate and package.
[0127] It can be seen from the derivation of the above algorithm that the accuracy of joint angle calculation depends largely on the relative position of the sensor. Taking the first sensor and the second sensor as an example, if the position deviation between the two is too large, the deformation produced when the y-axis rotates will not be consistent, so the influence of the y-axis rotation cannot be eliminated after the difference is made. This influence will be further amplified with the calculation of the decoupling matrix, thereby bringing about a non-negligible system error, which seriously affects the calculation accuracy. Therefore, during manufacturing, it is necessary to try to ensure the position accuracy of the alignment sensor.
[0128] Traditional rigid processing technology uses slotting, laser alignment and other methods to complete the positioning of materials on different surfaces, but these methods are not feasible when facing soft materials. Soft materials with low elastic modulus can no longer be "fixed" after leaving the mold and are prone to deformation, so it is difficult to maintain accuracy during multi-layer processing, and it is impossible to use rigid mechanical positioning to process the sensor in this application. Therefore, this application adopts the following preparation method to prepare a flexible multi-directional bending sensor.
[0129] The present application provides a method for preparing a flexible multi-directional bending sensor, such as Fig.15 As shown, the steps include:
[0130] Step S1: preparing a rigid mold that matches the shape of the flexible substrate of the bending sensor;
[0131] Step S2: melting the flexible material for preparing the flexible substrate and pouring it into the rigid mold; cooling the flexible material to form the flexible substrate, and pulling the flexible substrate out of the rigid mold;
[0132] Step S3: placing the liquid metal for preparing the sensing unit of the bending sensor at the corresponding position of the flexible substrate to form the sensing unit;
[0133] Step S4: encapsulating the side surface of the flexible substrate with the flexible material; sealing the sensing unit so that the outer end surfaces of the first electrode and the second electrode of the sensing unit are exposed to form a bending sensor.
[0134] The present invention sets a corresponding rigid mold according to the shape of the designed flexible substrate, and then melts the flexible material for preparing the flexible substrate and pours it into the rigid mold to make a flexible substrate, so that the flexible substrate can adapt to different joints. Then, the liquid metal is set on the flexible substrate, and the first electrode and the second electrode stroke sensing unit are formed, and finally encapsulated by the flexible material, thereby ensuring the stability of the bending sensor, and finally making the bending sensor have better performance.
[0135] Preferably, a rigid mold with a protrusion on the inner wall is manufactured by 3D printing, and the shape of the protrusion corresponds to the first groove, the sensitive groove, the second groove and the connecting groove on the flexible substrate. The rigid mold is printed with PLA (polylactic acid). The protrusion of the sensitive part is reserved in the rigid mold to form the sensitive groove on the flexible substrate, which ensures the accuracy of the position of the sensitive part and ensures the success rate of the above algorithm solution.
[0136] like Figure 16-Figure 19 As shown, the rigid mold 30 is a hollow rod-shaped structure, and the hollow cross section corresponds to the cross section of the flexible substrate. The bottom surface of one end of the rigid mold is provided with a connecting protrusion adapted to the connecting groove, the side surface of one end of the rigid mold is provided with a second protrusion 320 adapted to the second groove, and the side surface of the other end of the rigid mold is provided with a first protrusion 310 adapted to the first groove, and a positioning protrusion 330 adapted to the sensitive groove is provided between the second protrusion 320 and the first protrusion 310.
[0137] The liquid metal can be set to the corresponding position of the flexible substrate by coating or printing.
[0138] The first electrode and the second electrode are 0.5 mm*2 mm copper foils, which are attached to the surface-treated silicone substrate through SU-8 photoresist.
[0139] Silicone base: size 100mm×3mm×3mm
[0140] Liquid metal channel: The size is 70mm×0.5mm, evenly distributed on the median line around the silicone base.
[0141] After the preparation is completed, the first electrode and the second electrode are connected to the signal conditioning circuit by welding.
[0142] Hereinafter, the performance of the sensor prepared by the above method was verified.
[0143] The liquid metals tested in this application are all printed on sensors with a size of 100 mm × 3 mm × 3 mm, of which the effective sensitive part has a size of 70 mm × 0.5 mm.
[0144] The tensile equipment used in the verification is a universal mechanical electronic test platform, and the resistance is obtained through an Agilent source meter. The anti-interference test, resistance-strain test, stability test and resolution test of the sensitive components were carried out in turn.
[0145] The purpose of anti-interference test is to verify whether it has the function to work in the complex environment of the human body surface. Extremely strong environmental interactions will cause permanent failure of the sensor, and this failure can only be avoided by the user during use. But at the same time, during the use of the sensor, some small interferences such as bumps, slips, and scratches are inevitable. These interferences will not cause the sensor to fail but will affect its measurement accuracy. This application studies the impact of these weak environmental effects on the sensor by applying pressure to the sensor. The experiment used an Edelberg pressure gauge to apply 2N and 4N pressure to the sensor respectively. The results show that the R / R0 of the sensor only changed by about 0.5%, such as Fig. 20 The main reason for the resistance change here is that the pressure causes the device to undergo a certain elastic deformation, which causes the sensor to produce a corresponding response. From the results, these responses can be ignored.
[0146] The electrical properties of the sensor were obtained through resistance-strain test. The sensor was stretched to 50% and then released back to its original state. The results showed that the sensitivity coefficient of the sensor was 0.74. This value is sufficient for a variety of human motion applications. At the same time, the linear correlation coefficient of the sensor's resistance-strain curve reached 0.9989, which is basically completely linear. Fig.21 In addition, the hysteresis rate of the sensor is only 0.92%, which will only introduce very small errors in measurement. The experiments at this stage show that the resistance-strain response characteristics of the liquid metal-based sensor are excellent, and it can accurately reflect the strain of the device.
[0147] Stability testing is also an important part of sensor performance testing. This application stretched the sensor from 0%-50% 900 times and obtained a very stable response curve, such as Fig. 22As shown. It can be seen that after stabilization, the output of the sensor hardly changes, indicating that it has excellent stability. In addition, there will be some responses within 5% overshoot at the beginning of the stability experiment, because the sensor has not yet entered a steady state. After stretching several times, the various materials inside the sensor will cooperate perfectly. This phenomenon occurs in all sensing materials and does not mean that the characteristics of this device are poor.
[0148] The resolution of the sensing material was tested by stretching the device by 0.5%, 1%, 2%, 4%, 8%, 16%, and 32%. Fig.23 As shown in the figure, the sensitive material still has a significant response to the 350um stretch, indicating that its resolution can collect all human motion signals. Moreover, the maximum response of each stretching group is exactly half of the maximum response of the next group, proving the accuracy of the sensor measurement. The stretching linearity and resolution of this sensitive material determine its position in high-precision joint bending testing.
[0149] In the following, the performance of the sensor in actual application is further verified.
[0150] The sensor data acquisition system consists of a front-end preamplifier, filter, analog switch, ADC, microcontroller and host computer. This application calibrates two modes, single-degree-of-freedom bending and multi-degree-of-freedom bending, and conducts human body verification.
[0151] The wrist is an essential part of the human body for fine movements, but it is also one of the most vulnerable parts of the human body. When the wrist is injured (such as a fracture), rehabilitation training is needed to restore its normal function. At present, the rehabilitation assessment of wrist joint function mainly relies on professional doctors. With the improvement of human lifestyle, it has become a trend to conduct domestic training and assessment without exposing privacy.
[0152] We recruited patients with left wrist fractures and evaluated their wrist range of motion before and after rehabilitation. Data on the fractured wrist were measured using a dynamic goniometer (rotator).
[0153] Fig.24 The results show that in the figure, S1 / S2 / S3 / S4 represent the detection data of the first sensor unit, the second sensor unit, the third sensor unit and the fourth sensor unit respectively, Pitch represents the pitch angle, and Yaw represents the yaw angle. The vertical axis time represents time, the horizontal axis Angke represents the angle, and the horizontal axis Strain represents strain. Part A represents the data before rehabilitation training, and Part B represents the data after rehabilitation training.
[0154] Due to ligament stiffness, the flexion data before training was only 41° (GT=47°), which is far below the normal range of motion. After rehabilitation training, the flexion data can reach 76° (GT=78°), which proves the effectiveness of rehabilitation training. Similarly, before rehabilitation training, the subject's ulnar deviation was only 17° (GT=15°), which is also less than the normal range of motion. The value after training rose to 35° (GT=39°), which also proves the feasibility of training. Accurate data allows doctors to track the rehabilitation process online and improve treatment efficiency during the pandemic.
[0155] Cervical spondylosis is one of the most common diseases of the cervical spine (C spine), which can seriously reduce the quality of life and cause dizziness and memory loss. Currently, the incidence of cervical spondylosis has increased significantly due to frequent use of mobile phones. In order to test whether the sensor can continuously and accurately monitor the status of the C spine, we adopted two typical bad postures when using mobile phones, including lying on the left side and bowing the head when walking. Participants were asked to hold the phone with their right hand and read the text on the phone within ten seconds. Here, the sensor is attached to the neck to continuously monitor the status of the spine.
[0156] Fig.25 The data showed that in the ascending state, the subject bent his neck 11° to the left for support. In the stable phase, the subject leaned forward 14° to read small-font text, and then gradually retracted his head to 0° yaw for comfort. After reading, he quickly returned his head to the main state of -16° pitch. In the head-down experiment, the subject's neck initially bent rapidly to 43°, almost reaching the limit of forward flexion, causing pressure on the cervical spine. In simple one-way activities, the cervical spine was almost static, and smoothly entered the descending phase after reading. These angle measurements can remind users of their cervical spine conditions in real time, help correct bad postures, and reduce the incidence of cervical spondylosis.
[0157] Diseases related to the lumbar joints, such as scoliosis and lumbar disc herniation, are often caused by incorrect sitting posture. To further test its performance in preventing diseases, we installed the developed sensor on the lumbar spine to detect its posture in real time. The participants were asked to lie on the table and imitate two typical bad postures, right and forward. Both actions were controlled by the participants themselves. The experimental results are shown in Figure 2. Fig.26 shown.
[0158] In the lateral bending test, the subject first bent forward 12° and then moved to the right -31°. In the stabilization phase, the subject adjusted the body to 16° pitch and -9° yaw according to the body shape, and finally returned to the initial state after hearing the end command. In the next test, the subject was asked to imitate different sitting postures and bend forward ten times. Since the bending was controlled by the subject himself, the reaction time and bending angle were different. In the ascending phase, the subject's bending angles were 8°, 3°, 6°, 2°, 4°, 4°, 5°, 2°, 7°, and 6°, and finally reached 47°. These precisely measured data can continuously update the individual's lumbar spine condition, thereby reducing the number of patients with scoliosis and lumbar disc herniation.
[0159] The results of three experiments showed that the proposed sensor can accomplish various tasks related to omnidirectional joint flexion regardless of joint shape and flexion mode. This outstanding performance suggests that the sensor can be used in different application areas such as continuous monitoring for joint disease prevention and home rehabilitation assessment.
[0160] The above are only embodiments of the present invention, and are not intended to limit the patent scope of the present invention. Any equivalent structural transformations made using the contents of the present invention's specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.
Claims
1. A method for calculating the attitude angle of a square cross-section bending sensor, characterized in that: Includes steps: Step S1: obtaining a deformation difference of the sensing unit according to the deformation of the symmetrically arranged sensing units, and obtaining an average curvature of the sensing unit from the deformation difference; Step S2: obtaining the resistance change of the sensor unit according to the average curvature, the total length of the sensor unit after the sensor unit is stretched, and the angle change of the sensor unit; Step S3: acquiring the attitude angle of the sensor unit according to the resistance change, the sensitivity coefficient of the sensor unit, the total length of the sensor unit after the sensor unit is stretched, and the angle change of the sensor unit.
2. The method for calculating the attitude angle of a square cross-section bending sensor according to claim 1, characterized in that: The deformation ε of the sensing unit i It is expressed as: e i =e xi +e yi Where i is the sensor unit number, ε xi is the deformation of the i-th sensor unit when it bends around the x-axis, ε yi is the deformation of the i-th sensor unit when it bends around the y-axis; ε xi_strain , ε yi_strain are the tensile deformations of the i-th sensor unit around the x-axis and y-axis, ε xi_bending , ε yi_bending are the bending deformations of the i-th sensing unit around the x-axis and y-axis respectively.
3. The method for calculating the attitude angle of a square cross-section bending sensor according to claim 2, characterized in that: The deformation of the mutually symmetrical sensing units is subtracted, and the effect of stretching is ignored; the deformation difference of the sensing unit is expressed as:
4. The method for calculating the attitude angle of a square cross-section bending sensor according to claim 3, characterized in that: The average curvatures around the x-axis and around the y-axis obtained by the deformation difference are κ x , κ y : Among them, t x is the thickness of the first sensing unit and the second sensing unit, t y is the thickness of the third sensing unit and the fourth sensing unit.
5. The method for calculating the attitude angle of a square cross-section bending sensor according to claim 4, characterized in that: For x-axis rotation, in any small area ds between AB, we have: Wherein: k1 and k2 are the sensitivity coefficients of the first sensing unit and the second sensing unit respectively, △R1 and △R2 are the resistance and resistance change of the first sensing unit and the second sensing unit respectively; dR is the differential value of the resistance; Assuming dR / ds=R / L, dθ / ds=1 / r, we can get: Wherein: L is the total length of the sensor unit after stretching, r is the average curvature radius of the sensor unit; dθ is the angular change of the sensor unit.
6. The method for calculating the attitude angle of a square cross-section bending sensor according to claim 5, characterized in that: The resistance change of the sensing unit is: ΔR i =k i e i k i is the sensitivity coefficient; the initial angle is obtained by rotating the angle θ A -θ B : i A -θ B =κL Where: θ A is the angle of the tangent line when the rotation starts; θ B is the angle of the tangent after the rotation is completed, is the angle rotated; κ is the inverse of the radius of curvature.
7. The method for calculating the attitude angle of a square cross-section bending sensor according to claim 6, characterized in that: The rotation angle of the two axis directions of the sensor unit relative to the initial state is represented by the resistance change of the four surfaces: Among them, θ x With θ y are the rotation angles of the sensor unit around the x-axis and the y-axis respectively; l is the original length of the sensor unit; and t represents the thickness of the sensor unit in two axial directions.
8. The method for calculating the attitude angle of a square cross-section bending sensor according to claim 7, characterized in that: The average thickness of the sensing unit is t av It is expressed as: Where: V represents the volume of the sensor unit; The two attitude angles are expressed as: Among them, t origin is the initial thickness of the sensing unit.
9. The method for calculating the attitude angle of a square cross-section bending sensor according to claim 1, characterized in that: The cross section of the sensing unit is square.
