Six-dimensional force sensor calibration method and calibration device

Abstract

The application is suitable for the technical field of sensor calibration, and provides a six-dimensional force sensor calibration method and a calibration device.The calibration method comprises a test component installation step, a single-point decoupling step, a calibration weight loading step and a multi-dimensional calibration review step.The six-dimensional force sensor to be detected is installed on a mechanical arm, and a loading piece is connected to the six-dimensional force sensor;the mechanical arm is controlled to move below a single-point calibration assembly to exert a pre-pressure for calibration detection;the mechanical arm adjusts the posture to detect six directions respectively.Subsequently, a calibration weight is connected to the loading piece, the mechanical arm is controlled to drive the calibration weight to move along a preset detection path, and the position is adjusted synchronously, so that the six-dimensional force sensor is subjected to force in multiple postures, and multi-dimensional review decoupling is completed.The embodiment of the application first utilizes single-point directional detection, and then reviews and corrects the decoupling coefficient under multiple postures, realizes full-posture coverage, high-precision automatic calibration, and improves the precision of the detection performance of the six-dimensional force sensor.

Description

Technical Field

[0001] This application relates to the field of sensor calibration technology, and in particular to a calibration method and calibration device for a six-dimensional force sensor. Background Technology

[0002] Six-dimensional force sensors, capable of simultaneously detecting force components (FX, Fy, FZ) and torque components (MX, My, Mz) in three orthogonal directions, are widely used in high-end manufacturing fields such as industrial robots, precision assembly, and automated inspection. Their measurement accuracy directly determines the operational stability of equipment and the product yield rate. Due to the dimensional coupling characteristics of the sensor's internal elastic structure, force signals from different dimensions are prone to mutual interference, leading to a decrease in detection accuracy. Therefore, decoupling calibration of the six-dimensional force sensor is necessary before use to eliminate inter-dimensional coupling errors.

[0003] In related technologies, decoupling methods for six-dimensional force sensors often employ a single-point directional loading approach, applying force or torque only to each dimension individually, and completing decoupling after data acquisition. This calibration method suffers from low decoupling accuracy, incomplete attitude coverage, and lack of verification calibration, and thus cannot improve the sensor's detection accuracy and stability. Summary of the Invention

[0004] In view of this, the present invention provides a calibration method and calibration device for a six-dimensional force sensor to solve the problems of low decoupling accuracy, incomplete attitude coverage, and lack of verification calibration.

[0005] To solve the above problems, the technical solution of this application embodiment is implemented as follows: A calibration method for a six-dimensional force sensor includes: a test component installation step, in which the six-dimensional force sensor to be tested is installed on a robotic arm, and a loading component is connected to the six-dimensional force sensor; a single-point decoupling step, in which the robotic arm is controlled to move to below a single-point calibration component, and the single-point calibration component abuts against the loading component to apply pre-pressure to the six-dimensional force sensor for calibration detection; by adjusting the posture of the robotic arm, the six-dimensional force sensor is subjected to force in the directions Fz, Mx, My, Fy, Fx, and Mz respectively for detection; the host computer obtains the detection values ​​in each direction; and based on the collected values ​​in each direction... The data on the sensor are compared with the values ​​obtained from the standard sensor to obtain the coupling coefficients in each direction. In the calibration weight loading step, the robotic arm is controlled to move the loading component to the weight loading station, and calibration weights are fixedly connected to the loading component via the weight loading assembly. In the multi-dimensional calibration verification step, after the calibration weights are fixedly connected to the loading component, the robotic arm is controlled to move the calibration weights according to a preset detection path, synchronously adjusting their positions so that the six-dimensional force sensor is subjected to forces in multiple postures. The host computer collects data at a preset frequency, compares and corrects the single-point coupling coefficients, and completes the multi-dimensional verification decoupling.

[0006] In some embodiments, in the single-point decoupling step, at least in one detection direction, different magnitudes of pre-pressure are applied to the six-dimensional force sensor multiple times through the single-point calibration component, and different detection data are obtained respectively. Each detection data is compared with the corresponding value obtained on the standard sensor, and the coupling coefficient in the corresponding direction is obtained by calculating the average value; and / or, in the single-point decoupling step, the direction in which the single-point calibration component applies the pre-pressure to the six-dimensional force sensor is the same as or perpendicular to the corresponding detection direction.

[0007] In some embodiments, during the calibration weight loading step, the calibration weight is automatically loaded onto the loading member by the weight loading assembly.

[0008] In some embodiments, during the multidimensional calibration and verification step, the robotic arm drives the calibration weight to move according to a preset detection path, which is a figure-eight shape, a circle, or a rectangle.

[0009] This application embodiment also provides a calibration device for performing the six-dimensional force sensor calibration method described in any of the above embodiments, the calibration device comprising: A single-point calibration component is used to apply the pre-pressure required for calibration detection to the six-dimensional force sensor under test during single-point testing. The single-point calibration component can reciprocate between extending and retracting relative to the six-dimensional force sensor. A calibration weight is used to apply a detection force to the six-dimensional force sensor under test. A robotic arm is used to mount the six-dimensional force sensor and move to adjust the posture of the six-dimensional force sensor for calibration detection. A loading element is connected to the six-dimensional force sensor and is used to detachably connect to the calibration weight to transmit the detection force generated by the calibration weight to the six-dimensional force sensor. A weight loading component is used to automatically load or unload the calibration weight onto the loading element. A host computer is used to collect and analyze test data. Both the single-point calibration component and the six-dimensional force sensor are electrically connected to the host computer. The robotic arm drives the six-dimensional force sensor to perform single-point testing and drives the loading element to connect or separate from the calibration weight.

[0010] In some embodiments, the single-point calibration component includes: a mounting frame with a detection space formed below it; a force-applying element for generating the pre-pressure required for calibration detection, the force-applying element being fixed on the mounting frame with one end located within the detection space; a standard sensor mounted on the force-applying element, the standard sensor being used to output a standard detection value based on the pre-pressure; and a detection head, one end of which is connected to the standard sensor, and the other end of which is used to abut against the six-dimensional force sensor to transmit the pre-pressure; wherein, the robotic arm drives the six-dimensional force sensor into the detection space to abut against the detection head for detection.

[0011] In some embodiments, the weight loading assembly includes: a base having a receiving cavity for accommodating the calibration weight; and a locking member disposed adjacent to the base, the locking member being reciprocatingly extending and retracting to move closer to or further away from the calibration weight; wherein, when the loading member is connected to or separated from the calibration weight, the locking member abuts against the calibration weight to lock the position of the calibration weight.

[0012] In some embodiments, the calibration weight is provided with a first connecting structure, and the loading member is provided with a second connecting structure, wherein the first connecting structure and the second connecting structure are detachably connected; wherein, the weight loading assembly further includes an adjustment component, which is disposed adjacent to the weight loading assembly and can abut against the first connecting structure or the second connecting structure to adjust the mutual locking or unlocking between the first connecting structure and the second connecting structure.

[0013] In some embodiments, the adjusting component includes: a roller, connectable to the first connecting structure to drive the first connecting structure to rotate; a driving member, connected to the roller, the driving member being used to drive the roller to rotate; and an adjusting member, used to drive the roller to reciprocate between approaching and abutting against the first connecting structure and moving away from it, the driving member being mounted on the adjusting member; wherein, a transmission structure is provided on the roller or between the roller and the first connecting structure.

[0014] In some embodiments, the calibration device further includes: a support frame mounted on the weight loading assembly; a first detection component mounted on the support frame, the first detection component being used to detect the position of the first connecting structure in a locked state; and a second detection component mounted on the support frame and located above the first detection component, the second detection component being used to detect the position of the first connecting structure in an unlocked state.

[0015] This application provides a calibration method and apparatus for a six-dimensional force sensor. The calibration method includes a test component installation step, a single-point decoupling step, a calibration weight loading step, and a multi-dimensional calibration verification step. The six-dimensional force sensor to be tested is mounted on a robotic arm, and a loading component is connected to the sensor. The robotic arm is controlled to move below the single-point calibration component, bringing it into contact with the loading component to apply pre-pressure to the six-dimensional force sensor for calibration testing. Simultaneously, the robotic arm adjusts its posture to calibrate the sensor's detection performance in six directions. Furthermore, by connecting calibration weights to the loading component, the robotic arm moves the weights along a preset detection path and adjusts their positions synchronously, subjecting the six-dimensional force sensor to forces in multiple postures, thus completing the multi-dimensional verification and decoupling. This application's embodiments employ a collaborative calibration method with two stages: single-point decoupling and multi-dimensional calibration verification. First, single-point directional loading eliminates fundamental coupling errors in each dimension. Then, calibration weights are applied under multi-dimensional attitudes to verify and correct the decoupling coefficients. This achieves a fully attitude-covered, high-precision automated calibration process, improving the accuracy of the six-dimensional force sensor's detection performance. The calibration device can automatically perform calibration testing on the six-dimensional force sensor, improving the efficiency and repeatability of calibration testing. It can meet the calibration testing needs of different types of six-dimensional force sensors and has a wide range of applications. Attached Figure Description

[0016] Figure 1 This is a flowchart illustrating the six-dimensional force sensor calibration method provided in the embodiments of this application; Figure 2 This is a schematic diagram of the robotic arm in various postures during the multi-dimensional calibration and verification step provided in the embodiments of this application; Figure 3This is a schematic diagram of the calibration device provided in the embodiment of this application in the single-point decoupling step state; Figure 4 This is a schematic diagram of the calibration device provided in the embodiments of this application during the multi-dimensional calibration and verification step; Figure 5 This is a schematic diagram showing the loading element and calibration weight in a separated state according to an embodiment of this application; Figure 6 This is a schematic diagram showing the loading component and calibration weight in a docking state according to an embodiment of this application; Figure 7 This is a schematic diagram of the robotic arm lifting the calibration weight according to an embodiment of this application; Figure 8 This is a cross-sectional schematic diagram of the connecting rod and calibration weight provided in the embodiment of this application in the connected state.

[0017] Explanation of reference numerals in the attached figures: 1. Calibration device; 10. Six-dimensional force sensor; 11. Single-point calibration assembly; 111. Mounting bracket; 112. Force application component; 113. Standard sensor; 114. Detection head; 115. Detection space; 12. Calibration weight; 121. Connector; 1211. Mounting hole; 122. First connection structure; 1221. Connecting hole; 1222. Locking cap; 1223. Pin; 1224. Spring; 13. Robotic arm; 14. 141. Loading component; 1412. Second connecting structure; 1413. Connecting rod; 1414. Locking groove; 15. Weight loading assembly; 151. Base; 1515. Receiving cavity; 1516. Through hole; 152. Locking component; 153. Adjustment assembly; 1531. Roller; 1532. Driving component; 1533. Adjustment component; 1534. Transmission structure; 16. Support frame; 17. First detection assembly; 18. Second detection assembly. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of this application clearer, the application will be further described in detail below with reference to the accompanying drawings. The described embodiments should not be regarded as limitations on this application. All other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0019] The specific technical features described in the specific embodiments can be combined in any suitable manner without contradiction. For example, different combinations of specific technical features can form different embodiments and technical solutions. To avoid unnecessary repetition, the various possible combinations of the specific technical features in this application will not be described separately.

[0020] In the following description, the terms “first, second, third, ……” are used only to distinguish similar objects and do not represent a specific ordering of objects. It is understood that “first, second, third, ……” may be interchanged in a specific order or sequence where permitted, so that the embodiments of this application described herein can be implemented in an order other than that illustrated or described herein.

[0021] It should be understood that the directional descriptions "above", "below", "outside" and "inside" involved in the embodiments of this application refer to the directional descriptions under normal use. The "left" and "right" directions refer to the left and right directions shown in the corresponding schematic diagrams. They can be the left and right directions under normal use or not.

[0022] It should be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element. "A plurality of" means two or more.

[0023] Unless otherwise defined, all technical and scientific terms used in the embodiments of this application have the same meaning as commonly understood by one of ordinary skill in the art. The terminology used in the embodiments of this application is for the purpose of describing the embodiments of this application only and is not intended to limit this application.

[0024] like Figure 1 and Figure 2 As shown in the embodiment of this application, a six-dimensional force sensor calibration method is provided to test the detection performance of a six-dimensional force sensor, thereby improving the accuracy and precision of the detection. This six-dimensional force sensor calibration method achieves full-attitude coverage and high-precision automated calibration through two detection methods: single-point decoupling and multi-dimensional calibration verification, greatly improving the efficiency and reliability of six-dimensional force sensor calibration.

[0025] Specifically, the six-dimensional force sensor calibration method provided in this application embodiment includes a test component installation step, a single-point decoupling step, a calibration weight loading step, and a multi-dimensional calibration verification step. (Refer to...) Figures 3 to 8 This method relies on a robotic arm 13, a single-point calibration component 11, a weight loading component 15, and a host computer control system (not shown in the figure) to automatically perform calibration and detection, which is convenient and efficient.

[0026] In the test component installation step, the six-dimensional force sensor 10 to be tested is mounted on the robotic arm 13, specifically on the end effector of the robotic arm 13, ensuring that the installation posture of the six-dimensional force sensor 10 is accurately calibrated with the coordinate system of the robotic arm 13. Simultaneously, a loading element 14 is connected to the six-dimensional force sensor 10. This loading element 14 serves as an intermediate interface for force transmission and can be used to abut against the single-point calibration component 11 or connect to the calibration weight 12, transmitting external loads to the sensitive part of the six-dimensional force sensor 10, causing strain in the sensitive part and generating a detection signal. The loading element 14 can be fixed to the six-dimensional force sensor 10 using threaded connections, flange connections, or quick-connect couplings. Its structure can be designed with a ball head, a flat surface, or a dedicated mating surface according to calibration requirements, offering good flexibility in its design.

[0027] In the single-point decoupling step, the robotic arm 13 is moved to a position below the single-point calibration assembly 11, and the single-point calibration assembly 11 abuts against the loading member 14 to apply pre-pressure to the six-dimensional force sensor 10 for calibration detection. The single-point calibration assembly 11 typically includes a force-applying component (such as a cylinder or electric cylinder) capable of generating pre-pressure and a standard sensor 113 for providing a standard force with a known value. By adjusting the posture of the robotic arm 13, the six-dimensional force sensor 10 is sequentially subjected to force in six directions: Fz (axial force), Mx (torque about the X-axis), My (torque about the Y-axis), Fy (Y-direction force), Fx (X-direction force), and Mz (torque about the Z-axis) for detection. For example, during Fz direction calibration, the robotic arm 13 adjusts its posture so that the force-bearing surface of the loading element 14 is perpendicular to the force direction of the single-point calibration component 11, applying an axial force of known magnitude. During Mx calibration, the robotic arm 13 tilts the six-dimensional force sensor 10 around the X-axis at a certain angle, causing the contact point between the loading element 14 and the single-point calibration component 11 to deviate from the sensor center, generating a torque around the X-axis. The host computer collects the detection values ​​of the six-dimensional force sensor 10 in each direction and compares them with the standard values ​​obtained from the standard sensor 113. Based on the differences between the two, the coupling coefficient in each direction is calculated, which is the quantitative index of mutual interference between dimensions. This step, through single-point directional loading, specifically eliminates the basic coupling errors in each dimension, providing an initial calibration benchmark for subsequent verification.

[0028] In the calibration weight loading step, the robotic arm 13 is controlled to move, causing the loading component 14 to move to the weight loading station. The calibration weight 12 is then fixedly connected to the loading component 14 via the weight loading assembly 15. The weight loading assembly 15 can employ an automated weight gripping and locking mechanism, such as a pneumatic gripper with a locking cap structure, to reliably fix the calibration weight 12 of known mass onto the loading component 14. Alternatively, it can be manually connected and fixed. The mass of the weight can be selected according to the range of the six-dimensional force sensor 10, typically covering multiple ranges such as 20%, 50%, and 80% of the full scale of the six-dimensional force sensor 10, to verify linearity.

[0029] In the multidimensional calibration and verification step, after the calibration weight 12 is fixedly connected to the loading component 14, refer to Figure 2 ( Figure 2 The diagram above illustrates how, during the detection process, the robotic arm 13 adjusts the six-dimensional force sensor 10 in different postures. The robotic arm 13 moves the calibration weight 12 along a preset detection path, simultaneously adjusting its position to subject the six-dimensional force sensor 10 to forces in multiple postures. The preset detection path can be a series of spatial trajectories, such as conical oscillation, spiral motion, or a rectangular or circular path in three-dimensional space. It can be set according to different detection needs to simulate the force state of the six-dimensional force sensor 10 in actual use scenarios. Under different postures, the calibration weight 12's gravity vector generates three continuously changing force components and three torque components relative to the sensor's coordinate axes. During the movement of the robotic arm 13, the host computer collects the output data of the six-dimensional force sensor 10 at different detection points at a preset frequency (e.g., 100Hz to 1000Hz), and simultaneously records the real-time pose of the robotic arm 13 (obtainable through joint angles or end-effector poses fed back by the robotic arm 13 controller). The host computer calculates the theoretically applicable standard six-dimensional force value to be applied to the six-dimensional force sensor 10 based on the position of the robotic arm 13 and the mass of the weight. Then, it compares the measured value of the six-dimensional force sensor 10 with the theoretical value, uses the initial coupling coefficient obtained from the single-point decoupling step for compensation, and observes whether the residual is within the allowable range. If the residual exceeds the threshold, the coupling coefficient is corrected based on multi-pose data to complete the verification decoupling.

[0030] The calibration method provided in this application employs two different calibration approaches: single-point decoupling and multi-dimensional calibration verification. First, a single-point loading step precisely controls the independent forces in each dimension, specifically identifying and eliminating fundamental coupling errors between dimensions, thus establishing an accurate initial calibration model for the six-dimensional force sensor 10. Then, the multi-dimensional calibration verification step, through continuous multi-pose motion, allows the six-dimensional force sensor 10 to undergo comprehensive verification under simulated actual working conditions, supplementing the composite force attitudes that single-point loading cannot cover, ensuring the effectiveness of the calibration model across the entire workspace. Furthermore, this method achieves automated operation through automatic control of the robotic arm 13's movement, automatic data acquisition, and automatic comparison and correction by the host computer. This collectively solves the problems of low decoupling accuracy, incomplete attitude coverage, and lack of verification calibration in existing methods, while also avoiding the uncertainty and inefficiency caused by manual intervention. It significantly improves the repeatability and reliability of the calibration process, providing a guarantee for the measurement reliability and stability of the six-dimensional force sensor 10 under actual working conditions.

[0031] In one possible embodiment, the six-dimensional force sensor calibration method can be performed according to the following procedure: First, the six-dimensional force sensor 10 to be tested is installed on the end effector of a six-axis industrial robot, and the loading element 14 is fixed to the six-dimensional force sensor 10 via a flange. The robot moves to a single-point calibration station, which is equipped with a high-precision standard force sensor and an automatic lifting loading head. The robot sequentially adjusts to six postures, in which the loading head slowly descends at a speed of 0.5 mm / s, abuts against the loading element 14, and applies an increasing force from 0 to full scale. The host computer synchronously collects the outputs of the six-dimensional force sensor 10 to be tested and the standard sensor 113, and calculates the 6×6 coupling coefficient matrix. Subsequently, the robot moves to the weight loading station, which is equipped with four calibration weights 12 of different masses (1 kg, 2 kg, 5 kg, 10 kg) and their corresponding weight holders. The robot sequentially picks up each calibration weight 12 according to a preset program. After each calibration weight 12 is loaded, the robot moves along a preset spherical spiral trajectory, causing the calibration weight 12 to generate a continuously changing six-dimensional force relative to the six-dimensional force sensor 10. The host computer collects data at a frequency of 200Hz and performs real-time compensation using the coupling coefficient matrix obtained from single-point decoupling. The error after compensation is calculated. If the error exceeds 0.5%FS (FS is an abbreviation for Full Scale), the coupling coefficient matrix is ​​corrected using the least squares method. After all calibration weights 12 have been verified, the final decoupling calibration parameters are output. The entire testing process is automated and requires no manual intervention, greatly improving the efficiency of calibration testing and the reliability of the test results.

[0032] In some embodiments, to improve the accuracy and reliability of coupling coefficient calculation in the single-point decoupling step, in the single-point decoupling step, different magnitudes of pre-pressure are applied multiple times to the six-dimensional force sensor 10 through the single-point calibration component 11 in at least one detection direction, and different detection data are obtained respectively. For example, when calibrating in the Fz direction, five different magnitudes of pre-pressure of 20%, 40%, 60%, 80%, and 100% of the full scale can be applied sequentially, with each magnitude repeated three times, and the output values ​​of each channel of the six-dimensional force sensor 10 are recorded. The detection data of each instance are compared with the corresponding standard values ​​obtained on the standard sensor 113, and the coupling coefficient (such as Fx / Fz, Fy / Fz, Mx / Fz, My / Fz, Mz / Fz) at each loading point is calculated. Then, the final coupling coefficient in that direction is obtained by calculating the average value. By multi-point loading and average calculation, the random errors that may exist in single-point loading (such as small deviations in contact position, material elastic hysteresis, etc.) are eliminated, and the accuracy of the coupling coefficient is improved by averaging calculation. Meanwhile, loading data at different magnitudes can be used to verify whether the coupling coefficient changes with the load size. If nonlinear coupling characteristics are found, a piecewise correction model or a polynomial fitting model can be further established to provide more accurate initial parameters for subsequent multidimensional verification. As a possible implementation scheme, the number of loading points can be flexibly set according to the range and accuracy requirements of the six-dimensional force sensor 10. For example, more than ten loading points can be set for a high-precision six-dimensional force sensor 10. The loading order can be incremental, decremental, or random to eliminate systematic errors. Through this multi-point averaging calculation method, the coupling coefficient obtained in the single-point decoupling step is more accurate and reliable.

[0033] In some embodiments, to improve the reliability of pre-load during single-point decoupling, the direction in which the single-point calibration component 11 applies pre-load to the six-dimensional force sensor 10 is either in the same direction or perpendicular to the direction being detected. Specifically, when detecting Fz (axial force), the force direction of the single-point calibration component 11 is in the same direction as the sensor's Z-axis, generating an axial force, while the output in other directions should be zero. When detecting Mx (torque about the X-axis), the force direction of the single-point calibration component 11 is perpendicular to the sensor's X-axis, i.e., an eccentric load is applied along the Y or Z direction, causing the load line of action to deviate from the sensor center, thereby generating a torque about the X-axis. In this way, by matching the force direction with the detection direction, the singleness of the loading is ensured, minimizing the additional coupling in other directions caused by the deviation in the force direction, so that the measured coupling coefficient can truly reflect the inherent coupling characteristics of the internal elastic body of the six-dimensional force sensor 10, rather than the additional error brought by the single-point calibration component 11. As one possible implementation, the single-point calibration component 11 can employ a universal adjustment mechanism, enabling precise adjustment of the force application direction at any angle in space; alternatively, multiple calibration stations with fixed directions can be set up, with the robotic arm 13 switching between different stations to complete single loading in each dimension. This direction-matching loading method results in a higher signal-to-noise ratio and accuracy in the coupling coefficient obtained from the single-point decoupling step, laying a reliable foundation for subsequent multi-dimensional verification.

[0034] In some embodiments, to improve the automation and operational efficiency of the calibration process and avoid the safety hazards and uncertainties caused by manual loading and unloading of the calibration weights 12, the calibration weights 12 are automatically loaded onto the loading member 14 by the weight loading assembly 15 during the calibration weight loading step. The weight loading assembly 15 typically includes a weight holder, a gripping mechanism, and a locking mechanism. After the gripping mechanism (such as a pneumatic gripper, electromagnetic chuck, or vacuum chuck) grips the calibration weights 12 from the weight holder, it connects them to the loading member 14, and then the locking mechanism (such as a rotary lock cap, elastic buckle, or electromagnetic lock) automatically locks the calibration weights 12 onto the loading member 14, completing the loading. By automating the loading of the calibration weights 12, the risks of slippage and collisions that may occur during manual handling of the weights can be eliminated, protecting the calibration weights 12 and the six-dimensional force sensor 10 from physical damage. Furthermore, the mechanized gripping and locking actions maintain consistency; each time the calibration weights 12 and the loading element 14 are loaded, their relative positions and contact postures are exactly the same, avoiding the problems of deviation in the direction of the loading force or unstable contact caused by differences in manual operation techniques. This improves the reliability of subsequent multi-dimensional calibration verification data. Moreover, the automated loading effect is high, significantly improving the overall calibration efficiency. Of course, it is understandable that the weight loading assembly 15 can also have the function of automatically detaching the calibration weights 12 from the loading element 14, further improving the automation level and operational efficiency of the calibration.

[0035] In one possible implementation, the weight loading assembly 15 can integrate multiple weight holders to hold calibration weights 12 of different mass specifications. The specifications of the calibration weights 12 are confirmed by an automatic identification system (such as an RFID reader or visual recognition, where RFID is radio frequency identification). The robotic arm 13 can sequentially grasp calibration weights 12 of different masses according to a preset program to perform multi-level loading tests, achieving automatic verification across the entire range. The grasping mechanism can also monitor the clamping force in real time during the grasping process to prevent damage to the weight surface due to excessive clamping force or slippage due to insufficient clamping force, further improving the safety and reliability of the loading process. This automatic loading design allows for seamless integration of the calibration weight loading step with the movement of the robotic arm 13 and data acquisition processes, providing support for automated calibration.

[0036] In some embodiments, to achieve comprehensive coverage of the force posture within the entire working space of the six-dimensional force sensor 10 during the multi-dimensional calibration verification step, the robotic arm 13 drives the calibration weight 12 to move according to a preset detection path. This preset detection path can adopt various trajectory forms such as a figure-eight, a circle, or a rectangle. Each trajectory has corresponding motion characteristics and can generate different force / torque coupling modes. During the "circular" trajectory movement, the end of the robotic arm 13 performs uniform circular motion around a fixed point in space, causing the gravity vector of the calibration weight 12 to exhibit continuous sine and cosine changes relative to the sensor coordinate axes. The generated force and torque components are singular, with less interference, facilitating the extraction of dynamic coupling characteristics between dimensions. This is suitable for determining whether the response of the six-dimensional force sensor 10 is consistent under periodic loads. The radius of the circular trajectory can be adjusted according to the range of the six-dimensional force sensor 10. A larger radius generates a larger bending moment component, which can be used to verify the decoupling accuracy of the six-dimensional force sensor 10 under high torque conditions.

[0037] Specifically, the figure-eight trajectory is composed of two tangent circular trajectories. The end effector of the robotic arm 13 moves along the figure-eight path in space, with its acceleration and velocity directions continuously changing. This causes the inertial force component generated by the calibration weight 12 to superimpose with the gravitational component, forming a more complex composite load mode. This trajectory can simulate the non-uniform, variable-direction load scenarios that the six-dimensional force sensor 10 may encounter in practical applications, thus more comprehensively verifying the effectiveness of the decoupling coefficient under dynamic conditions. The two rings of the figure-eight trajectory can use different radii to generate an asymmetric load spectrum, which can be used to verify the decoupling performance of the six-dimensional force sensor 10 under asymmetric force conditions.

[0038] Specifically, during the "rectangular" trajectory movement, the end effector of the robotic arm 13 moves in a straight line along the four sides of the rectangle, generating sudden acceleration changes at the four corners, causing the calibration weight 12 to generate an impact-like inertial force. This trajectory can test the dynamic response characteristics of the six-dimensional force sensor 10 under step loads or impact loads, as well as the real-time compensation capability of the decoupling algorithm under transient conditions. The side length and movement speed of the "rectangular" trajectory can be adjusted according to testing requirements; the side length determines the amplitude of the attitude change, and the speed determines the rate of load change.

[0039] As one possible implementation, "rectangular" and "circular" trajectories can be combined. For example, a "circular" trajectory can be used first for steady-state characteristic testing, followed by a "rectangular" trajectory for dynamic characteristic testing, forming a complete verification test. Regardless of the trajectory used, during the movement of the robotic arm 13, the host computer synchronously acquires the output of the six-dimensional force sensor 10 at different detection points and the end-effector pose of the robotic arm 13 at a preset frequency (e.g., 200Hz to 1000Hz). By comparing the measured values ​​with the theoretical calculated values ​​(based on the mass of the calibration weight 12, gravitational acceleration, and real-time pose), the coupling coefficient obtained in the single-point decoupling step is verified and corrected. This multi-trajectory preset path design allows the multi-dimensional calibration verification step to select the most suitable motion mode according to the expected application scenario of the sensor, or to achieve a more comprehensive performance evaluation through the combination of multiple trajectories, significantly improving the adaptability and reliability of the calibration method.

[0040] like Figures 3 to 8As shown in the figure, this application embodiment also provides a calibration device 1 for performing the six-dimensional force sensor calibration method in any of the above embodiments. The calibration device 1 includes a single-point calibration component 11, a calibration weight 12, a robotic arm 13, a loading component 14, a weight loading component 15, and a host computer (not shown in the figure). The single-point calibration component 11 is used to apply the pre-pressure required for calibration detection to the six-dimensional force sensor 10 to be tested during single-point testing. The single-point calibration component 11 can reciprocate between extending and abutting relative to the six-dimensional force sensor 10 and retracting and separating. The calibration weight 12 is used to apply the detection force to the six-dimensional force sensor 10 to be tested. The robotic arm 13 is used for mounting the six-dimensional force sensor 10 and for adjusting the posture of the six-dimensional force sensor 10 for calibration detection. Furthermore, the robotic arm 13 is also used to drive the six-dimensional force sensor 10 to perform single-point testing, and to drive the loading component 14 to connect or disconnect from the calibration weight 12. The loading component 14 is connected to the six-dimensional force sensor 10 and is detachably connected to the calibration weight 12 to transmit the detection force generated by the calibration weight 12 to the six-dimensional force sensor 10. The weight loading assembly 15 is used to automatically load or unload the calibration weight 12 onto the loading component 14. The host computer is used to collect and analyze the test data, and both the single-point calibration assembly 11 and the six-dimensional force sensor 10 are electrically connected to the host computer. Through this calibration device 1, single-point decoupling of the six-dimensional force sensor 10, multi-dimensional calibration verification, and loading of the calibration weight 12 before multi-dimensional calibration verification can be realized, so that no manual intervention is required during the calibration test process, and the calibration test has good reliability.

[0041] Specifically, the single-point calibration component 11 is used to apply the pre-pressure required for calibration detection to the six-dimensional force sensor 10 during single-point testing, and can reciprocate between extending and retracting relative to the six-dimensional force sensor 10. For example, the single-point calibration component 11 can be installed on a fixed workbench, with a loading head at its end. When single-point calibration is required, the robotic arm 13 moves the six-dimensional force sensor 10 to the extension direction of the loading head, and the drive mechanism of the calibration component (such as a servo cylinder) extends, causing the loading head to abut against the loading element 14 on the six-dimensional force sensor 10, applying a preset force. After the test is completed, the drive mechanism retracts, separating the loading head from the loading element 14, making room for the next action of the robotic arm 13. As one possible implementation, the single-point calibration component 11 uses a servo electric cylinder in conjunction with a high-precision force sensor, achieving force control accuracy within 0.1%FS and steplessly adjusting the loading force from 0 to full scale. A ball joint or flexible connection can be installed between the loading head and the loading element 14 to automatically compensate for alignment deviations, ensuring the direction of the loading force remains consistent with the sensor axis. Alternatively, multiple single-point calibration components 11 with different ranges can be set up to cover testing requirements for small, medium, and large ranges, respectively. The robotic arm 13 can automatically switch to the corresponding station for testing according to a preset program. Specifically, the calibration weight 12 is used to apply the detection force to the six-dimensional force sensor 10 to be tested. Its mass can be selected according to the range required for calibration, such as 0.5kg, 1kg, 2kg, etc. Different specifications can also be used, and different loading forces can be adjusted by stacking multiple standard calibration weights 12. The shape of the calibration weight 12 can be cylindrical, square, or a special shape with a guide structure to facilitate a stable and reliable docking connection with the loading element 14.

[0042] The robotic arm 13 is used to mount the six-dimensional force sensor 10 and can move to adjust the posture of the six-dimensional force sensor 10 during calibration testing. The robotic arm 13 is typically a multi-degree-of-freedom articulated robot, such as a six-axis industrial robot, with a gripper or mounting base at its end for fixing the six-dimensional force sensor 10. The six-dimensional force sensor 10 is fixed on it to achieve position limitation before testing. Through the movement of the robotic arm 13, the six-dimensional force sensor 10 can be accurately moved to the test position of the single-point calibration component 11 for single-point detection calibration in various directions, or moved to the position to dock with the calibration weight 12. After the calibration weight 12 is loaded, the robotic arm 13 can drive the calibration weight 12 to move along a preset trajectory and can adjust the posture of the six-dimensional force sensor 10 to simulate the force state in different directions, achieving comprehensive calibration of the six-dimensional force. Furthermore, after calibration is completed, the robotic arm 13 drives the calibration weight 12 to be reset to the station for storing the calibration weight 12, and with the cooperation of the weight loading assembly 15, the loading component 14 and the calibration weight 12 are disassembled and separated.

[0043] The loading element 14 is fixedly connected to the six-dimensional force sensor 10, for example, through a threaded connection, flange connection, or snap-fit ​​connection. The loading element 14 is detachably connected to the calibration weight 12. The function of the loading element 14 is to transmit the detection force generated by the calibration weight 12 to the six-dimensional force sensor 10 after the calibration weight 12 is connected. The connection method between the loading element 14 and the calibration weight 12 can be mechanical (hook and plug), or it can be magnetic, pneumatic gripper, or electromagnetic chuck, etc., as long as reliable connection and separation are achieved, and the gravity of the calibration weight 12 is accurately transmitted to the six-dimensional force sensor 10 after connection. The mating surfaces of the loading element 14 and the calibration weight 12 can be provided with positioning structures, such as conical guides or pin hole fits, to ensure that the relative positions of the two are precisely consistent each time they are mated, thereby ensuring the accurate direction of the loading force.

[0044] The weight loading assembly 15 is used to automatically load or unload calibration weights 12 onto the loading member 14. In one possible implementation, the weight loading assembly 15 typically includes a weight holder and a locking mechanism. The weight holder has a receiving groove that matches the shape of the calibration weight 12 for storing the calibration weight 12 to be used. After the loading member 14 is docked with the calibration weight 12 in the weight holder, the calibration weight 12 is automatically locked onto the loading member 14 by the locking mechanism (such as a rotary lock cap, elastic buckle, or electromagnetic lock). The weight loading assembly 15 can be set with multiple stations to place calibration weights 12 of different mass specifications, realizing automatic switching of multiple ranges. In another possible implementation, the weight loading assembly 15 integrates a vision positioning system, which uses an industrial camera to identify the positioning marks on the loading member 14 and the calibration weight 12, guiding the robotic arm 13 to drive the loading member 14 to dock and connect, ensuring the success rate of loading and the repeatability of positioning accuracy.

[0045] The host computer is used to collect and analyze the test data. Both the single-point calibration component 11 and the six-dimensional force sensor 10 are electrically connected to the host computer. The host computer acquires the six-channel output values ​​of the six-dimensional force sensor 10 and the readings of the standard force sensor in the single-point calibration component 11 in real time through data acquisition. At the same time, it communicates with the controller of the robotic arm 13 through the data network to obtain the real-time pose information of the robotic arm 13. The host computer runs the calibration software, controls the movement of the robotic arm 13 according to the preset program, controls the loading of the single-point calibration component 11, controls the loading and unloading of the calibration weights 12 by the weight loading component 15, processes the collected data, calculates the coupling coefficient, performs error comparison and correction, and finally generates a calibration report.

[0046] In this embodiment, the calibration device 1 uses a robotic arm 13 as a motion execution unit to drive a six-dimensional force sensor 10 to reciprocate between a single-point calibration component 11 and a weight loading component 15. During single-point decoupling, the robotic arm 13 moves the sensor to the single-point calibration component 11, which extends to apply multi-level pre-pressure in six directions. The host computer collects data and calculates the initial coupling coefficient matrix. During the loading stage of the calibration weight 12, the robotic arm 13 moves the loading component 14 to the weight loading station. The weight loading component 15 automatically loads and locks the corresponding mass of calibration weight 12 onto the loading component 14. During the multi-dimensional calibration verification stage, the robotic arm 13 carries the six-dimensional force sensor 10 loaded with the calibration weight 12 along a preset detection path (such as a figure-eight, circular, or rectangular trajectory). The host computer simultaneously collects the output of the six-dimensional force sensor 10 and the pose of the robotic arm 13, performs real-time compensation using the initial coupling coefficient, compares it with theoretical values, and corrects the coupling coefficient. Throughout the process, the single-point calibration component 11, the weight loading component 15, the robotic arm 13, and the host computer work closely together to achieve fully automated calibration from single-point decoupling to multi-dimensional verification. Through the design of this calibration device 1, in conjunction with the aforementioned calibration method, high-precision, full-attitude, and automated calibration of the six-dimensional force sensor 10 is achieved, significantly improving calibration efficiency and calibration quality.

[0047] In this embodiment, a robotic arm 13 drives a loading member 14 to connect or disconnect from a calibration weight 12. Specifically, during the loading phase, the robotic arm 13 moves the end of the loading member 14, which is equipped with a six-dimensional force sensor 10, to the top of the weight holder, aligning the loading member 14 with the docking structure on the calibration weight 12 and completing the connection. Subsequently, the robotic arm 13 lifts the loading member 14 and the connected calibration weight 12, transferring the gravity of the calibration weight 12 to the six-dimensional force sensor 10 through the loading member 14, thus achieving the loading of the detected force. During this process, the robotic arm 13 can also move according to a preset program, changing the spatial orientation of the calibration weight 12 relative to the six-dimensional force sensor 10, thereby generating forces and torques in different directions to achieve multi-directional calibration of the six-dimensional force. During the unloading phase, the robotic arm 13 moves the loading component 14, carrying the calibration weight 12, back to the station where the calibration weight 12 is stored. The calibration weight 12 is accurately placed down, and then, with the cooperation of the weight loading assembly 15, the loading component 14 separates from the calibration weight 12. The robotic arm 13 can then move away with the six-dimensional force sensor 10, completing one calibration cycle. This automated loading and unloading of the calibration weight 12 avoids the risks of bumps and slippage that may occur during manual handling. The robotic arm 13 docks and separates with precisely controlled speed and a smooth trajectory, without impact or shaking, effectively protecting the six-dimensional force sensor 10 from accidental physical damage and also avoiding the safety hazard of the calibration weight 12 falling and injuring people. Furthermore, the entire testing process is automated, requiring no manual intervention, significantly improving loading and unloading efficiency. The loading and unloading time for a single operation can be reduced from tens of seconds or even minutes of manual operation to just a few seconds, and can be repeated continuously, significantly improving the efficiency of calibration testing. Furthermore, automated operation eliminates the uncertainty caused by human factors, ensuring that the position, angle, and speed of each docking remain consistent, thus improving the repeatability of the calibration process and the reliability of the measurement data.

[0048] In some embodiments, such as Figure 3 and Figure 4As shown, the single-point calibration component 11 includes a mounting frame 111, a force-applying element 112, a standard sensor 113, and a detection head 114. The mounting frame 111 typically adopts a frame structure or a cantilever structure, forming a detection space 115 with multiple openings below it, providing a precise positioning area for the robotic arm 13 to carry the six-dimensional force sensor 10 into this area for contact testing. The force-applying element 112 is used to generate the pre-pressure required for calibration testing. It is fixed on the mounting frame 111, with one end located within the detection space 115. The force-applying element 112 can be a servo electric cylinder, a high-precision cylinder, or a linear motor, capable of outputting precise and controllable thrust or pull force according to instructions from the host computer. A standard sensor 113 is mounted on the force-applying component 112. Its measuring range matches that of the six-dimensional force sensor 10 under test, and it has a detection accuracy no lower than that of the six-dimensional force sensor 10 under test. The standard sensor 113 is used to monitor the actual pre-pressure output by the force-applying component 112 in real time and output a standard detection value as a comparison benchmark for the six-dimensional force sensor 10 under test. One end of the detection head 114 is connected to the standard sensor 113, and the other end is used to abut against the six-dimensional force sensor 10 to transmit the pre-pressure. The end face of the detection head 114 that contacts the loading component 14 can be designed as a spherical or flat surface, and wear-resistant coatings or flexible gaskets can be provided as needed to ensure stable contact and prevent damage to the sensor surface.

[0049] During single-point decoupling testing, the robotic arm 13 moves the six-dimensional force sensor 10 and its loading component 14 accurately into the detection space 115 below the mounting bracket 111, aligning the force-bearing surface of the loading component 14 with the detection head 114 and maintaining an appropriate distance. The host computer issues a command, causing the force-applying component 112 to extend, driving the detection head 114 to move slowly forward. After contacting the loading component 14, a preset pre-pressure is applied. The standard sensor 113 provides real-time feedback of the actual force value, and the six-dimensional force sensor 10 simultaneously outputs the detection values ​​from each channel. The host computer collects and compares the data. After the test, the force-applying component 112 retracts, the detection head 114 separates from the loading component 14, and the robotic arm 13 can then carry the sensor out of the detection space 115 and proceed to the next workstation.

[0050] Specifically, the mounting bracket 111 provides rigid support for the force-applying component 112, effectively preventing equipment swaying or displacement caused by reaction forces during loading, and ensuring the stability and repeatability of the force application direction. The force-applying component 112 can adopt a double guide rod structure, which can maintain good straightness during extension and retraction, ensuring that the pre-pressure direction is strictly consistent with the axis of the detection head 114.

[0051] In some embodiments, such as Figures 3 to 8As shown, the weight loading assembly 15 includes a base 151 and a locking member 152. The base 151 has a receiving cavity 1511 for accommodating the calibration weight 12, and a rubber shock-absorbing pad can be provided within the receiving cavity 1511 to absorb the impact force received by the calibration weight 12. The locking member 152 is disposed adjacent to the base 151 and can reciprocate to move closer to or further away from the calibration weight 12. When the loading member 14 is connected to or separated from the calibration weight 12, the locking member 152 extends and abuts against the calibration weight 12 to lock the position of the calibration weight 12; after the loading and unloading operation is completed, the locking member 152 retracts, releasing the lock on the calibration weight 12. This configuration avoids unexpected changes in the position of the calibration weight 12 and eliminates the thrust or vibration interference that the robotic arm 13 may generate on the calibration weight 12 during the docking process. This ensures that the calibration weight 12 maintains a precise positioning posture during the connection or separation process with the loading component 14, thereby ensuring the success rate of docking between the loading component 14 and the calibration weight 12 and avoiding docking failure or unreliable connection problems caused by the loose position of the calibration weight 12.

[0052] Specifically, the locking element 152 can adopt an electromagnet chuck structure. A magnetic material needs to be placed at the corresponding position of the calibration weight 12. When the electromagnet is energized, the magnetic force generated will attract the calibration weight 12 to the reference surface of the receiving cavity 1511, achieving contactless locking. In another possible implementation, the locking element 152 can adopt an elastic clamping structure, such as a spring-driven floating ball or elastic pressure plate on the side wall of the receiving cavity 1511. When the weight is placed in, the elastic element automatically clamps it to the reference surface, requiring no additional driving source, resulting in a simple and reliable structure. The number and arrangement of the locking elements 152 can also be optimized according to the shape of the weight. For example, one locking element 152 can be placed on each of the two opposite sides of a rectangular weight to form opposing clamping; or three locking elements 152 can be evenly arranged around the circumference of a circular weight to achieve self-centering locking. Through the design of the locking component 152, the weight loading assembly 15 provides a stable and reliable positioning guarantee for the calibration weight 12 during the loading and unloading process, forming an efficient, accurate and safe automated loading and unloading system with the loading component 14.

[0053] In some embodiments, such as Figure 7 and Figure 8As shown, the locking component 152 is a cylinder connected to the base 151. The base 151 has a through hole 1512 communicating with the receiving cavity 1511. The telescopic rod of the cylinder is located in the through hole 1512, and moves closer to or further away from the calibration weight 12 through telescopic movement. By setting the locking component 152 as a cylinder, which uses compressed air as a power source, the cylinder has a fast response speed and sensitive action, which can match the rapid loading and unloading rhythm of the robotic arm 13, avoiding docking waiting caused by locking delay. Moreover, the cylinder drive does not require electricity or lubrication, and will not produce oil or dust pollution, maintaining the cleanliness of the working area. Furthermore, the locking force of the cylinder can be precisely controlled by adjusting the air source pressure, which not only ensures reliable locking of the weight, but also avoids damage to the surface of the weight or positioning deviation caused by excessive locking force. In one possible implementation, the cylinder is a single-acting spring-reset cylinder. Compressed air drives the telescopic rod to extend and lock the weight. When the air supply is interrupted, the spring automatically returns to its original position. This design allows for automatic unlocking in case of an unexpected interruption of the air supply, preventing the calibration weight 12 from becoming stuck. In another possible implementation, the cylinder is a double-acting cylinder. The extension and retraction of the telescopic rod are controlled by a solenoid valve, enabling bidirectional controllable movement and meeting the needs of high-frequency loading and unloading scenarios requiring rapid reciprocating motion. Furthermore, a soft pressure head, such as rubber, can be provided at the end of the telescopic rod to increase the contact friction with the calibration weight 12, while also buffering contact impact and preventing indentations on the weight surface. A dustproof sealing ring can be installed between the through hole 1512 and the telescopic rod to prevent external dust from entering the cylinder and extend its service life.

[0054] In some embodiments, such as Figure 5 and Figure 8 As shown, the calibration weight 12 is provided with a first connecting structure 122, and the loading member 14 is provided with a second connecting structure 141. The first connecting structure 122 and the second connecting structure 141 are detachably connected. Specifically, the first connecting structure 122 can be a lifting ring or a hanging hole, and the second connecting structure 141 can be a hook or a claw; or the first connecting structure 122 can be a magnetic suction cup, and the second connecting structure 141 can be a magnetically conductive contact surface; quick-connect couplings, threaded connections, or snap-fit ​​connections can also be used to ensure that the first connecting structure 122 and the second connecting structure 141 can be quickly assembled and disassembled.

[0055] To further ensure the reliability and controllability of connection and disconnection actions, such as Figures 3 to 5As shown, the weight loading assembly 15 also includes an adjustment assembly 153. The adjustment assembly 153 is disposed adjacent to the weight loading assembly 15 and can abut against the first connecting structure 122 or the second connecting structure 141 to adjust the mutual locking or unlocking between the first connecting structure 122 and the second connecting structure 141. For example, when the first connecting structure 122 is a hook-type structure, the adjustment assembly 153 can be a pneumatic push rod or an electromagnet-driven top block. After the robotic arm 13 drives the second connecting structure 141 to dock with the first connecting structure 122, the adjustment assembly 153 extends and pushes the locking mechanism of the hook to lock it. When separation is required, the adjustment assembly 153 acts again to open the lock and unlock it. As another example, when a magnetic connection is used, the adjustment assembly 153 can be a mechanical separation push rod. When separation is required, it extends to push the calibration weight 12 away from the suction cup to help overcome the magnetic force. By setting the adjustment component 153, the robotic arm 13 can be assisted in loading and unloading the calibration weight 12, making the locking and unlocking actions more precise and controllable. This avoids the problem of incomplete docking or separation that may occur due to relying solely on the movement of the robotic arm 13, especially in connection structures that require a large locking force or have self-locking characteristics.

[0056] In one possible implementation, the adjusting component 153 may also be equipped with a position sensor to detect the docking status of the first connecting structure 122 and the second connecting structure 141, and feed the signal back to the control system to ensure the reliability of each loading and unloading operation. The adjusting component 153 may be driven by a cylinder, electric cylinder, or electromagnet, and a flexible pressure head or roller may be provided on the actuating end to reduce impact and wear when in contact with the connecting structure. By setting the adjusting component 153, the connection and separation actions between the calibration weight 12 and the loading member 14 can be further assisted by control, which, together with the precise positioning of the robotic arm 13 and the stable locking of the locking member 152, constitutes a highly reliable and automated calibration weight loading and unloading system.

[0057] In some embodiments, such as Figures 3 to 8As shown, either the first connecting structure 122 or the second connecting structure 141 includes a connecting rod 1411, and the other of the first connecting structure 122 and the second connecting structure 141 includes a connecting hole 1221 for inserting the connecting rod 1411 and a locking cap 1222 for clamping the connecting rod 1411. For example, the connecting rod 1411 can be provided on the calibration weight 12, and the connecting hole 1221 and the locking cap 1222 can be provided on the loading member 14; or conversely, the connecting rod 1411 can be provided on the loading member 14, and the connecting hole 1221 and the locking cap 1222 can be provided on the calibration weight 12. The locking cap 1222 is usually a rotating part with an internal thread or a cam clamping structure. When the connecting rod 1411 is inserted into the connecting hole 1221, the locking cap 1222 is rotated to form a threaded engagement with the connecting rod 1411 or to press the connecting rod 1411 by the cam surface, thereby achieving mechanical locking. The adjusting component 153 can abut against the locking cap 1222 to drive the locking cap 1222 to rotate and lock or unlock the connecting rod 1411. For example, the adjusting component 153 may include a rotary drive head (such as a friction wheel driven by a pneumatic or electric motor). When the robotic arm 13 inserts the connecting rod 1411 into the connecting hole 1221, the drive head of the adjusting component 153 approaches and presses against the outer periphery of the locking cap 1222, driving the locking cap 1222 to rotate by a predetermined angle through friction or gear engagement, thus locking the connecting rod 1411. When separation is required, the adjusting component 153 rotates the locking cap 1222 in the opposite direction to unlock. The clamping connection of the locking cap 1222 forms a rigid connection, which can reliably transmit the gravity of the calibration weight 12, improving the accuracy of the calibration test. Moreover, the self-locking structure formed by the rotation and locking of the locking cap 1222 ensures that the connection remains even if the air or power supply is unexpectedly interrupted, improving operational safety. Meanwhile, the adjustment component 153 drives the rotation of the lock cap 1222 to achieve fully automatic locking and unlocking without manual intervention, thus improving the efficiency and automation of locking and unlocking operations.

[0058] In one possible implementation, the outer periphery of the locking cap 1222 may be provided with external gears or friction patterns, and the adjusting assembly 153 is provided with a cooperating gear or friction wheel to ensure reliable drive. The locking cap 1222 and the connecting rod 1411 may use trapezoidal or sawtooth threads to ensure both locking force and anti-loosening properties. In another possible implementation, the adjusting assembly 153 may integrate a torque sensor to monitor the locking torque in real time during the locking process. When a preset torque value is reached, rotation automatically stops, ensuring consistency in each connection and preventing loosening due to insufficient locking force or thread damage due to excessive locking force.

[0059] In some embodiments, such as Figure 5 and Figure 8As shown, the first connecting structure 122 is configured to include a connecting hole 1221 and a locking cap 1222. A connector 121 is provided on the calibration weight 12, and the connector 121 has a connecting hole 1221. The locking cap 1222 is fitted onto the outer periphery of the connector 121, and a threaded structure is provided between them. Simultaneously, the second connecting structure 141 includes a connecting rod 1411. One end of the connecting rod 1411 is fixedly connected to the loading member 14, and the other end of the connecting rod 1411 is inserted into the connecting hole 1221. Specifically, the inner wall of the locking cap 1222 is machined with internal threads, and the outer periphery of the connector 121 is machined with matching external threads, allowing the locking cap 1222 to rotate on the connector 121 and achieve axial movement. When the robotic arm 13 moves the loading component 14, causing the connecting rod 1411 to be inserted into the connecting hole 1221 on the calibration weight 12, the adjusting component 153 abuts against the locking cap 1222 and drives the locking cap 1222 to rotate. As the locking cap 1222 is screwed in along the thread, its end face or internal step surface gradually abuts against the positioning surface (such as a step or retaining ring) on ​​the connecting rod 1411, thereby firmly clamping the connecting rod 1411 into the connecting hole 1221 and achieving a locking connection. When separation is required, the adjusting component 153 drives the locking cap 1222 to loosen in the opposite direction, releasing the constraint on the connecting rod 1411, and the connecting rod 1411 can be pulled out from the connecting hole 1221. This locking and unlocking mechanism, achieved through a threaded structure, cleverly utilizes the self-locking property of threads. It is not easily loosened under vibration or accidental impact, ensuring that the calibration weight 12 remains rigidly connected to the loading component 14 throughout the movement of the robotic arm 13. This avoids the risk of unstable loading force or the calibration weight 12 falling due to loose connection. Furthermore, by adjusting the locking cap 1222 and driving it to rotate via the adjusting component 153, fully automatic control of connection and separation is achieved without manual intervention. This works in conjunction with the positioning action of the robotic arm 13, further improving loading and unloading efficiency and operational safety. Moreover, the threaded locking cap 1222 structure has excellent repeatability; after each locking, the relative position of the connecting rod 1411 and the connecting hole 1221 is consistent, ensuring the directional accuracy of the loading force of the calibration weight 12 and the repeatability of measurement data.

[0060] In one possible implementation, the thread between the locking cap 1222 and the connector 121 is a trapezoidal or rectangular thread, which has higher load-bearing capacity and impact resistance, and is suitable for frequent loading and unloading of calibrating weights 12 with large weights. In another possible implementation, the insertion end of the connecting rod 1411 can be provided with a guide cone surface, which allows the robotic arm 13 to be smoothly inserted into the connecting hole 1221 even with slight positioning deviations, thereby improving the docking fault tolerance rate.

[0061] In some embodiments, such as Figure 5 and Figure 8As shown, the first connecting structure 122 further includes a pin 1223, and the second connecting structure 141 further includes a locking groove 1412. Specifically, the connector 121 of the calibration weight 12 has a mounting hole 1211 that communicates with the connecting hole 1221. The pin 1223 is movably inserted into the mounting hole 1211, and one end of the pin 1223 abuts against the inner wall of the locking cap 1222. The connecting rod 1411 is provided with a locking groove 1412 for the pin 1223 to be inserted and locked. When the adjusting component 153 drives the locking cap 1222 to rotate, the locking cap 1222 moves axially along the thread on the connector 121. Since the inner wall of the locking cap 1222 abuts against the end of the pin 1223, the screwing motion of the locking cap 1222 forces the pin 1223 to move radially along the mounting hole 1211 and gradually insert into the locking groove 1412 on the connecting rod 1411, forming a mechanical interlock, thereby firmly locking the connecting rod 1411 in the connecting hole 1221. When the locking cap 1222 rotates in the opposite direction, the pin 1223 automatically exits the locking groove 1412 under the action of the corresponding mechanism (such as a spring), and the lock is released. This secondary locking mechanism, achieved through the pin 1223 and the locking groove 1412, provides double protection in conjunction with the threaded locking cap 1222. The threaded locking cap 1222 provides axial clamping force, while the pin 1223 provides radial shear resistance. Together, they ensure that the connecting rod 1411 will not loosen or rotate relative to each other even when subjected to large tensile and torsional forces. This is particularly suitable for applications involving large-mass calibration weights 12 and situations where the robotic arm 13 generates inertial forces during rapid movement or posture changes, greatly enhancing the reliability and safety of the connection.

[0062] Specifically, to achieve automatic disengagement of the latch 1223 during unlocking, such as Figure 8 As shown, a spring 1224 can be installed inside the mounting hole 1211. The spring 1224 can be sleeved on the end of the pin 1223 that is inserted into the locking groove 1412, generating a driving force to push the pin 1223 away from the locking groove 1412. The end of the pin 1223 that abuts against the lock cap 1222 is spherical, with a portion protruding outside the mounting hole 1211. The protruding length must ensure that the pin 1223 can be inserted into the locking groove 1412 during the locking operation. In this way, when the lock cap 1222 is loosened, the pin 1223 loses the pressure constraint of the inner wall of the lock cap 1222 and automatically exits from the locking groove 1412 under the elastic force of the spring 1224, achieving quick unlocking. This method of reset via the spring 1224 has a simple and reliable structure and good response sensitivity.

[0063] As another possible implementation, the mounting hole 1211 can be configured as a downward-facing inclined hole, with its axis forming a certain angle (e.g., 15 to 30 degrees) with the axis of the connecting hole 1221. Under the action of gravity, the pin 1223 will naturally tend to slide down to the bottom of the mounting hole 1211. When the lock cap 1222 is tightened, the inner wall of the lock cap 1222 overcomes gravity and pushes the pin 1223 into the locking groove 1412; when the lock cap 1222 is loosened, the pin 1223 automatically slides out of the locking groove 1412 under the action of gravity. This gravity-based reset eliminates the need for additional elastic elements, resulting in a simpler structure and avoiding the risk of fatigue failure of the spring 1224.

[0064] Specifically, a tapered mating surface design can be adopted between the connector 121 and the locking cap 1222. This means that an outer tapered surface is provided on the outer periphery of the connector 121, and a matching inner tapered surface is provided on the inner wall of the locking cap 1222. When the locking cap 1222 rotates, the radial component force generated by the tapered mating surface synchronously and evenly pushes the pin 1223 into the locking groove 1412, ensuring that the pin 1223 is inserted into place. Simultaneously, it generates a large clamping force to clamp and lock the connecting rod 1411, further improving the connection's robustness. The tapered mating surface also has a self-centering function, automatically compensating for machining errors and ensuring that the locking cap 1222 and the connector 121 always remain coaxial, preventing the pin 1223 from jamming due to eccentricity.

[0065] Multiple locking slots 1412 can be provided, such as two or more spaced apart along the length of the end into which the connecting rod 1411 is inserted. The pins 1223 can also be symmetrically arranged, for example, one pin 1223 on each radially opposite side of the connector 121, with corresponding mounting holes 1211. This ensures that each locking slot 1412 can be locked by at least two symmetrically arranged pins 1223, meaning two pins 1223 are inserted into the same locking slot 1412 from opposite directions, forming opposing clamping, making the locking of the connecting rod 1411 more stable and the force more balanced. The symmetrical multi-pin 1223 design evenly distributes the radial load of the connecting rod 1411, avoiding potential skewing or stress concentration caused by unilateral force application, and effectively preventing unilateral wobbling of the calibration weight 12 during multi-posture movements of the robotic arm 13. Of course, it is understandable that the locking groove 1412 can also be set as a circle around the circumference of the connecting rod 1411, and the pins 1223 are evenly distributed in the circumference. When locking, multiple pins 1223 are inserted into the same locking groove 1412 to achieve joint locking, which improves the reliability of locking and meets the reliable locking requirements of the heavy calibration weight 12.

[0066] In some embodiments, such as Figures 3 to 7As shown, the adjustment assembly 153 includes a roller 1531, a drive member 1532, and an adjusting member 1533. The roller 1531 is connected to the locking cap 1222 to drive the locking cap 1222 to rotate. The drive member 1532 is connected to the roller 1531 and provides power for the rotation of the roller 1531, for example, using a micro motor or pneumatic motor. The adjusting member 1533 drives the roller 1531 to reciprocate between approaching and moving away from the locking cap 1222. The drive member 1532 is mounted on the adjusting member 1533, allowing the roller 1531 to move and contact the locking cap 1222 when needed, and to retract after driving, avoiding interference with other actions. A transmission structure 1534 is provided on the roller 1531 or between the roller 1531 and the locking cap 1222 to ensure reliable power transmission.

[0067] As one possible implementation, the transmission structure 1534 is a rubber coating layer, such as polyurethane or silicone, disposed on the outer periphery of the roller 1531, which utilizes its high coefficient of friction to form frictional transmission with the outer surface of the lock cap 1222. Moreover, this frictional transmission structure 1534 has a buffering effect, capable of absorbing minor impacts and vibrations, and has a low risk of scratching the surface of the lock cap 1222, reducing the possibility of damage to the lock cap 1222.

[0068] As another possible implementation, the transmission structure 1534 is a cooperating gear structure, that is, a driving gear is set on the outer periphery of the roller 1531, and a driven gear or gear ring is set on the outer periphery of the lock cap 1222, which meshes with it. Precise angular displacement is achieved through gear meshing. Gear transmission has the advantages of precise transmission ratio, no slippage, and the ability to transmit large torque, and is suitable for occasions that require precise control of the rotation angle of the lock cap 1222. The adjusting component 1533 can be driven by a cylinder, electric cylinder or electromagnet. For example, when the lock cap 1222 needs to be driven, the adjusting component 1533 extends so that the roller 1531 abuts against the surface of the lock cap 1222, and the driving component 1532 starts to drive the roller 1531 to rotate, thereby driving the lock cap 1222 to rotate; after locking or unlocking is completed, the adjusting component 1533 retracts to separate the roller 1531 from the lock cap 1222, making room for the movement of the robotic arm 13. This method of using a movable adjustment component 153 to drive the locking cap 1222 not only avoids interference when loading and unloading the calibration weight 12 on the robotic arm 13, but also allows the flexible contact or gear meshing between the roller 1531 and the locking cap 1222 to adapt to certain positional deviations, reducing the positional accuracy requirements of related components, improving the system's fault tolerance, and making it easy to use.

[0069] In some embodiments, such as Figure 3 and Figure 4As shown, to achieve accurate monitoring of the locking and unlocking states of the lock cap 1222 and ensure the controllability and reliability of the loading and unloading process, the calibration device 1 also includes a support frame 16, a first detection component 17, and a second detection component 18. The support frame 16 is mounted on the weight loading component 15 (weight base), providing a stable mounting foundation for the detection components. The first detection component 17 is mounted on the support frame 16 and is used to detect the position of the first connecting structure 122 (lock cap 1222) in the locked state. For example, when the first connecting structure 122 (lock cap 1222) rotates to the locked position, the first detection component 17 is triggered, sending a locked-in signal. The second detection component 18 is mounted on the support frame 16 and located above the first detection component 17. It is used to detect the position of the first connecting structure 122 (lock cap 1222) in the unlocked state. For example, when the first connecting structure 122 (lock cap 1222) rotates in the opposite direction to the fully released position, the second detection component 18 is triggered, sending an unlocked-in signal. In this way, through the combined action of the first detection component 17 and the second detection component 18, precise sensing of the extreme positions at both ends of the rotational stroke of the locking cap 1222 is achieved, providing reliable status feedback for the control system. When the robotic arm 13 completes docking and the adjusting component 153 drives the locking cap 1222 to rotate, the control system monitors the signal of the first detection component 17 in real time to ensure that the locking cap 1222 is locked in place before proceeding to the next loading action. When separation is required, the signal of the second detection component 18 is also monitored to confirm that the unlocking is in place before the robotic arm 13 can be withdrawn. This avoids the risk of the calibration weight 12 falling off during loading due to inadequate locking, or the risk of the robotic arm 13 forcibly withdrawing due to inadequate unlocking, causing damage to the loading component 14 or the calibration weight 12, significantly improving the safety and reliability of the loading and unloading process.

[0070] Specifically, the type of detection component can be selected based on detection accuracy and environmental requirements. As one possible implementation, the first detection component 17 and the second detection component 18 employ proximity switches, such as inductive or capacitive proximity sensors. When the detection bump or sensing plate on the lock cap 1222 moves to the sensor's sensing area, the sensor outputs a switching signal. Proximity switches offer advantages such as non-contact operation, fast response, and long lifespan, making them suitable for automated equipment with frequent operations. As another possible implementation, the detection components can employ photoelectric sensors, detecting the detection marks on the lock cap 1222 through photoelectric transmission or reflection principles, offering advantages such as long detection distance and high accuracy. Of course, it is understandable that the detection components can employ Hall effect sensors, in conjunction with a magnet on the lock cap 1222, to achieve non-contact position detection, which can also reliably detect the movement position of the lock cap 1222.

[0071] The installation height of the detection components can be adjusted according to the stroke of the locking cap 1222 to accommodate the different locking positions of calibration weights 12 of different specifications. The support frame 16 can be integrally formed with the weight loading component 15, or it can be processed separately and fixed with screws. Its structure can be L-shaped or gate-shaped, providing suitable installation positions for the two detection components. Through this dual-detection component position feedback design, and the precise drive of the adjustment component 153, the automatic loading and unloading of the calibration weights 12 is reliably realized.

[0072] In this embodiment, the automatic loading and unloading method for the calibration weight 12, achieved through the coordinated action of the robotic arm 13, the loading component 14, and the weight loading assembly 15, includes a loading step and a disassembly step. The loading step includes: controlling the movement of the robotic arm 13 to precisely align the connecting rod 1411 with the connecting hole 1221 on the connector 121 of the calibration weight 12; then, the robotic arm 13 moves towards the calibration weight 12, inserting the connecting rod 1411 into the connecting hole 1221, and aligning the locking groove 1412 on the connecting rod 1411 with the pin 1223 in the mounting hole 1211. At this time, the locking component 152 is activated, extending and abutting against the calibration weight 12, firmly fixing the calibration weight 12 within the receiving cavity 1511 of the weight holder, preventing displacement of the weight during subsequent locking operations. Next, the adjusting component 1533 is activated, driving the roller 1531 to approach and abut against the locking cap 1222. Then, the driving component 1532 is activated, driving the roller 1531 to rotate and causing the locking cap 1222 to rotate. During rotation, the locking cap 1222 moves axially along the thread of the connector 121, its inner wall gradually pressing against the pin 1223, pushing the pin 1223 into the locking groove 1412 of the connecting rod 1411, thus achieving mechanical interlocking between the connecting rod 1411 and the calibration weight 12. When the locking cap 1222 rotates to the locked position, the first detection component 17 detects the locking cap 1222's arrival signal. The control system then controls the driving component 1532 to stop operating and controls the adjusting component 1533 to move in the opposite direction, separating the roller 1531 from the locking cap 1222. Then, the locking component 152 is activated again, separating it from the calibration weight 12 and releasing the position lock on the calibration weight 12. The robotic arm 13 moves again, removing the calibration weight 12 from the receiving cavity 1511 and moving it along a preset path to enter the calibration and testing process. Throughout the loading process, the locking component 152 fixes the calibration weight 12 before the locking operation and releases it after locking, ensuring the stability of the locking action and the smooth removal of the calibration weight 12.

[0073] During the disassembly process, after the calibration and testing procedures are completed, the robotic arm 13 moves to place the calibration weight 12 back into the receiving cavity 1511 of the weight holder. The locking component 152 is activated, bringing it into contact with the calibration weight 12 and fixing its position again. The adjusting component 1533 is activated, driving the roller 1531 to approach and contact the locking cap 1222. The driving component 1532 is activated, driving the roller 1531 to rotate in the opposite direction, causing the locking cap 1222 to rotate in the opposite direction. During the reverse rotation of the locking cap 1222, the inner wall gradually releases the pressure on the pin 1223. Under the force of the spring 1224 or gravity, the pin 1223 automatically exits from the locking groove 1412, and the connecting rod 1411 is unlocked from the calibration weight 12. When the locking cap 1222 is reversed to the unlocked position, the second detection component 18 detects the unlock signal, the control system controls the drive component 1532 to stop, and the adjusting component 1533 moves in the opposite direction to separate the roller 1531 from the locking cap 1222. Finally, the robotic arm 13 moves again to smoothly pull the connecting rod 1411 out of the connecting hole 1221, completing the disassembly of the calibration weight 12.

[0074] During the loading and unloading of the calibration weight 12, the first detection component 17 and the second detection component 18 accurately sense the extreme positions at both ends of the locking cap 1222, and the locking component 152 provides auxiliary fixation during the loading and unloading process. This method achieves closed-loop control of the entire process from loading to unloading, ensuring the reliability of each docking and locking, the smoothness of unlocking and separation, and the absolute stability of the weight during operation. It completely avoids risks such as bumps, slippage, or incomplete locking caused by improper human operation, and significantly improves the automation level, work efficiency, and safety of the calibration and detection of the six-dimensional force sensor 10.

[0075] In some embodiments, during the loading step, the locking member 152 can be controlled to fix the position of the calibration weight 12 first, and then the robotic arm 13 can be controlled to move so that the connecting rod 1411 is aligned with the connecting hole 1221, and then the subsequent loading steps can be performed. During the disassembly step, after the connecting rod 1411 is pulled out of the connecting hole 1221 and the calibration weight 12 is disassembled, the locking member 152 can be activated to separate the locking member 152 from the calibration weight 12, releasing the position lock on the calibration weight 12, so that subsequent related operations can be performed on the calibration weight 12.

[0076] The above description is merely an embodiment of this application and is not intended to limit the scope of protection of this application. Any modifications, equivalent substitutions, and improvements made within the spirit and scope of this application are included within the scope of protection of this application.

Claims

1. A calibration method for a six-dimensional force sensor, characterized in that, include: The test component installation steps involve installing the six-dimensional force sensor to be tested on the robotic arm, and connecting a loading component to the six-dimensional force sensor; In the single-point decoupling step, the robotic arm is controlled to move to below the single-point calibration component, and the single-point calibration component is brought into contact with the loading component to apply pre-pressure to the six-dimensional force sensor for calibration detection. By adjusting the posture of the robotic arm, the six-dimensional force sensor is subjected to force in the directions Fz, Mx, My, Fy, Fx, and Mz respectively for detection. The host computer obtains the detection values ​​in each direction. Based on the data collected in each direction, the values ​​obtained from the standard sensor are compared to obtain the coupling coefficient in each direction. In the calibration weight loading step, the movement of the robotic arm is controlled to move the loading component to the weight loading station, and the calibration weight is fixedly connected to the loading component through the weight loading assembly. In the multi-dimensional calibration and verification step, after the calibration weight is fixedly connected to the loading component, the robotic arm is controlled to move the calibration weight according to the preset detection path and adjust its position synchronously so that the six-dimensional force sensor is subjected to force in multiple postures. The host computer collects data at a preset frequency, compares and corrects the single-point coupling coefficient, and completes the multi-dimensional verification and decoupling.

2. The calibration method for a six-dimensional force sensor as described in claim 1, characterized in that, In the single-point decoupling step, at least in one detection direction, different magnitudes of pre-pressure are applied multiple times to the six-dimensional force sensor through the single-point calibration component, and different detection data are obtained for each. Each detection data is compared with the corresponding value obtained on the standard sensor, and the coupling coefficient in the corresponding direction is obtained by calculating the average value; and / or, In the single-point decoupling step, the direction in which the single-point calibration component applies pre-pressure to the six-dimensional force sensor is the same as or perpendicular to the direction being detected.

3. The calibration method for a six-dimensional force sensor as described in claim 1, characterized in that, In the calibration weight loading step, the calibration weight is automatically loaded onto the loading component by the weight loading assembly.

4. The calibration method for a six-dimensional force sensor as described in claim 1, characterized in that, In the multidimensional calibration and verification step, the robotic arm drives the calibration weight to move according to a preset detection path, which is a figure-eight shape, a circle, or a rectangle.

5. A calibration device, characterized in that, For performing the six-dimensional force sensor calibration method as described in any one of claims 1 to 4, the calibration device comprises: A single-point calibration component is used to apply the pre-pressure required for calibration detection to the six-dimensional force sensor to be tested during single-point testing. The single-point calibration component can reciprocate between extending and abutting relative to the six-dimensional force sensor and retracting and separating. Calibration weights are used to apply detection force to the six-dimensional force sensor to be tested; A robotic arm is used to mount a six-dimensional force sensor and to move and adjust the posture of the six-dimensional force sensor for calibration and detection. A loading element is connected to the six-dimensional force sensor. The loading element is used to detachably connect to the calibration weight to transmit the detection force generated by the calibration weight to the six-dimensional force sensor. A weight loading assembly is used to automatically load or unload the calibration weights onto the loading member; The host computer is used to collect and analyze detection data. The single-point calibration component and the six-dimensional force sensor are both electrically connected to the host computer. The robotic arm drives the six-dimensional force sensor to perform single-point testing, and also drives the loading component to connect or disconnect from the calibration weight.

6. The calibration device as described in claim 5, characterized in that, The single-point calibration component includes: Mounting bracket, with a detection space formed below the mounting bracket; A force-applying component is used to generate the pre-pressure required during calibration testing. The force-applying component is fixed on the mounting frame, and one end is located within the testing space. A standard sensor is mounted on the force-applying component, and the standard sensor is used to output a standard detection value based on the pre-pressure. A detection head, one end of which is connected to the standard sensor, and the other end of which is used to abut against the six-dimensional force sensor to transmit the pre-pressure; The robotic arm drives the six-dimensional force sensor into the detection space, where it comes into contact with the detection head for detection.

7. The calibration device as described in claim 5, characterized in that, The weight loading assembly includes: The base has a cavity for accommodating the calibration weights; A locking element is provided adjacent to the base body. The locking element can reciprocate to extend and retract to move closer to or further away from the calibration weight. Specifically, when the loading component is connected to or separated from the calibration weight, the locking component abuts against the calibration weight to lock the position of the calibration weight.

8. The calibration device as described in claim 5, characterized in that, The calibration weight is provided with a first connecting structure, and the loading component is provided with a second connecting structure. The first connecting structure and the second connecting structure are detachably connected. The weight loading assembly further includes an adjustment component, which is disposed adjacent to the weight loading assembly and can abut against the first connection structure or the second connection structure to adjust the mutual locking or unlocking between the first connection structure and the second connection structure.

9. The calibration device as described in claim 8, characterized in that, The adjustment component includes: A roller can be connected to the first connecting structure to drive the first connecting structure to rotate; A driving component, connected to the roller, is used to drive the roller to rotate; An adjusting component is used to drive the roller to reciprocate between approaching and abutting with or moving away from the first connecting structure, and the driving component is mounted on the adjusting component; A transmission structure is provided on the roller or between the roller and the first connecting structure.

10. The calibration device as described in claim 8, characterized in that, The calibration device further includes: A support frame is mounted on the weight loading assembly; A first detection component is installed on the support frame, and the first detection component is used to detect the position of the first connecting structure in the locked state; The second detection component is mounted on the support frame and located above the first detection component. The second detection component is used to detect the position of the first connection structure in the unlocked state.

Images