High-precision clamping device for mechanical parts based on flexible force feedback

Abstract

This invention relates to the field of mechanical clamping control, specifically to a high-precision clamping device for mechanical parts based on flexible force feedback. The clamping device includes a drive unit and multiple mechanical claws. A suction cup is mounted at the end of each claw. The drive unit drives the claws and controls the suction cup to clamp the mechanical part. Multiple pressure sensors are evenly installed at multiple locations on the bottom of the suction cup to collect pressure data at each historical moment prior to the current moment. A deployed feedback adjustment module adjusts the output power of the drive unit based on the pressure data at different locations on the bottom of the suction cup, thereby controlling the clamping action of the mechanical part. This invention can prevent mechanical parts from falling during clamping and avoid damage to the parts due to excessive clamping, thus improving clamping accuracy.

Description

Technical Field

[0001] This invention relates to the field of mechanical clamping and control, and specifically to a high-precision clamping device for mechanical parts based on flexible force feedback. Background Technology

[0002] In the field of mechanical manufacturing, high-precision clamping devices are core components for achieving precision machining. They are widely used in various processing and production processes to ensure that parts remain stable during machining and assembly, avoiding movement or vibration, thereby improving processing efficiency and product quality.

[0003] In related technologies, the mechanical jaws of a clamping device typically perform clamping actions on mechanical parts with a fixed output power. However, due to the irregularity of the surface of the mechanical parts and the vibration that easily occurs during the clamping process, the existing clamping devices exert uneven clamping forces on the mechanical parts, which can easily cause the mechanical parts to fall off or be clamped too tightly, damaging the parts and thus reducing clamping accuracy. Summary of the Invention

[0004] To address the technical problems of uneven clamping force in existing clamping devices, which easily leads to parts falling off and excessive clamping damaging parts, thereby reducing clamping accuracy, the present invention aims to provide a high-precision clamping device for mechanical parts based on flexible force feedback. The specific technical solution adopted is as follows:

[0005] This invention proposes a high-precision clamping device for mechanical parts based on flexible force feedback. The clamping device includes a drive unit and multiple mechanical claws. A suction cup is mounted at the end of each mechanical claw. The drive unit drives the mechanical claws and controls the suction cup to clamp the mechanical part. The clamping device also includes a feedback adjustment module. Multiple pressure sensors are evenly installed at multiple positions on the bottom of the suction cup. These pressure sensors acquire pressure data at each historical moment within a preset time period prior to the current moment at their respective positions. The feedback adjustment module adjusts the output power of the drive unit based on the pressure data and controls the clamping action of the mechanical part, including:

[0006] Using any one suction cup as the target suction cup, the stability impact of the part at each position on the bottom of the target suction cup is obtained based on the fluctuation difference of the pressure data between the positions of the pressure sensors at the bottom of the target suction cup; based on the stability impact of the part at each position on the bottom of each suction cup, the real-time probability of the part falling off the clamping device is obtained.

[0007] Based on the real-time probability of the part falling off the clamping device and the real-time pressure data at different positions of the suction cup, the real-time output power of the drive device to the mechanical claw is adjusted to obtain the adjusted output power of the drive device to the mechanical claw at the next moment. The clamping device clamps the mechanical part with the adjusted output power, and predicts the probability of the part falling off at future moments based on the probability of the part falling off the clamping device at different moments, thus obtaining the stopping moment of the clamping action.

[0008] Furthermore, the determination of the component stability influence at each location on the bottom of the target suction cup includes:

[0009] Based on the pressure data fluctuations at each historical moment at the location of each pressure sensor on the bottom of the target suction cup, the degree of pressure fluctuation at each location on the bottom of the target suction cup is obtained;

[0010] The location of any pressure sensor on the bottom of the target suction cup is taken as the target location. The location symmetrical to the target location is taken as the symmetrical location of the target location. All other locations besides the target location and the symmetrical location of the target location are taken as the reference locations of the target location.

[0011] A negative correlation mapping is performed on the difference in pressure fluctuation degree between the target location and the symmetrical location of the target location to obtain the pressure fluctuation similarity of the target location;

[0012] The differences in the degree of pressure fluctuation between the target location and each reference location are accumulated to obtain the pressure fluctuation difference degree at the target location;

[0013] The product of the pressure fluctuation similarity and the pressure fluctuation difference is used as the component stability influence at the target location.

[0014] Furthermore, the degree of pressure fluctuation at each location on the bottom of the target suction cup includes:

[0015] Extract the extreme values ​​from all historical pressure data at the location of each pressure sensor at the bottom of the target suction cup, take two adjacent extreme values ​​as an extreme value group, and take the absolute value of the difference between the two extreme values ​​in each extreme value group as the pressure change amplitude of each extreme value group.

[0016] The average pressure change of all extreme value groups at each position on the bottom of the target suction cup is taken as the pressure fluctuation level at each position on the bottom of the target suction cup.

[0017] Furthermore, the real-time probability of part falling off the clamping device includes:

[0018] At the bottom of the target suction cup, an initial vibration vector is constructed for each pressure sensor location. The direction of the initial vibration vector at each location is the direction from the center of the target suction cup to each location, and the length of the initial vibration vector at each location is the part stability influence at each location on the bottom of the target suction cup.

[0019] The locations of two mutually symmetrical pressure sensors on the bottom of the target suction cup are taken as the symmetrical position group of the target suction cup. The sum of the component stability influence of the two positions in each symmetrical position group is taken as the comprehensive influence of each symmetrical position group. The two positions in the symmetrical position group with the largest comprehensive influence are taken as the key vibration positions on the bottom of the target suction cup.

[0020] Based on the initial vibration vectors of all positions in a preset neighborhood of each key vibration position on the bottom of the target suction cup, two suspected vibration vectors of the target suction cup are obtained.

[0021] Two mutually symmetrical suction cups are taken as a symmetrical suction cup group, and any symmetrical suction cup group is taken as the target symmetrical suction cup group. The projection plane of the target symmetrical suction cup group is constructed. The projection plane is perpendicular to the line connecting the midpoints of the two suction cups in the target symmetrical suction cup group and passes through the center point of the line connecting them.

[0022] The two suspected vibration vectors of each suction cup in the target symmetrical suction cup group are projected onto the projection plane to obtain two projection vectors of each suction cup in the target symmetrical suction cup group.

[0023] In the two suction cups of the target symmetrical suction cup group, a projection vector is selected for each of them. The two selected projection vectors are used as a projection vector group. The angle between the two projection vectors in each projection vector group is used as the reference angle of each projection vector group. The two projection vector groups with the smallest reference angle are used as the reference projection vector groups of the target symmetrical suction cup group.

[0024] Based on the reference angle of each reference projection vector group of the target symmetrical suction cup group, and the projection vector in each reference projection vector group, the part drop coefficient of the target symmetrical suction cup group is obtained.

[0025] The average of the part drop coefficients of all symmetrical suction cup groups is taken as the real-time part drop probability of the clamping device.

[0026] Furthermore, the two suspected vibration vectors of the target suction cup obtained include:

[0027] The sum of the initial vibration vectors of all locations in a preset neighborhood of each key vibration location is used as the suspected vibration vector of the target suction cup.

[0028] Furthermore, the part drop coefficient for obtaining the target symmetrical suction cup assembly includes:

[0029] By performing negative correlation mapping on the average value of the reference angles of the two reference projection vector groups of the target symmetrical suction cup group, the first part drop evaluation value of the target symmetrical suction cup group is obtained.

[0030] The sum of the two projection vectors in each reference projection vector group is used as the composite projection vector of each reference projection vector group, and the sum of the magnitudes of the composite projection vectors of the two reference projection vector groups is used as the second part drop evaluation value of the target symmetrical suction cup group.

[0031] The drop evaluation values ​​of the first and second parts are combined and normalized to obtain the drop coefficient of the target symmetrical suction cup group.

[0032] Furthermore, the adjustment of the mechanical gripper's output power by the drive device at the next moment includes:

[0033] The average real-time pressure data of all pressure sensors at the suction cup at the end of each robotic gripper is taken as the real-time comprehensive pressure value of each robotic gripper; based on the real-time comprehensive pressure value of each robotic gripper and the real-time probability of the part falling off the gripping device, the real-time traction force control weight of each robotic gripper is obtained.

[0034] Based on the calculation formula for adjusting the output power, the adjusted output power of the drive device for each mechanical gripper at the next moment is obtained. The calculation formula for the adjusted output power is as follows:

[0035] P k ′ =w k ×P max ×Δt+P k

[0036] Among them, P k ′ This indicates the adjustment of the output power of the drive unit for the k-th mechanical gripper at the next moment; w k P represents the real-time traction control weight of the k-th robotic gripper; max The maximum output power of the drive unit is represented by Δt; the sampling time interval of the pressure data is represented by P. k This represents the real-time output power of the drive unit to the k-th mechanical gripper.

[0037] Furthermore, obtaining the real-time traction control weights for each robotic gripper includes:

[0038] Based on the calculation formula for traction control weight, the real-time traction control weight of each robotic gripper is obtained. The calculation formula for the traction control weight is as follows:

[0039] w k =(1-norm(F) k ))×U

[0040] Among them, w k F represents the real-time traction control weight of the k-th robotic gripper; k represents the real-time comprehensive pressure value of the k-th mechanical gripper; U represents the real-time probability of the part falling off the gripping device; norm() represents the normalization function.

[0041] Furthermore, the stopping time of the clamping action includes:

[0042] The probability of parts falling off the clamping device at different times is input into the time series prediction algorithm, and the predicted probability of parts falling off at multiple future times is output.

[0043] The stopping time of the clamping action is obtained based on the changes in the predicted probability of the part falling at each future time.

[0044] Furthermore, obtaining the stopping time of the clamping action based on the change in the predicted probability of the part falling at each future time includes:

[0045] Curve fitting is performed on the predicted probability of the part falling at all future moments to obtain the prediction fitting curve;

[0046] On the prediction fitting curve, the slope of each future moment and the sum of the predicted probability of the part falling are used as the judgment parameter for each future moment.

[0047] The future moment with the smallest judgment parameter is taken as the stop moment of the clamping action.

[0048] The present invention has the following beneficial effects:

[0049] This invention addresses the issue of uneven clamping force in existing clamping devices, which can easily cause parts to fall off or damage them due to excessive clamping, thus reducing clamping accuracy. Therefore, multiple pressure sensors are evenly installed at different positions on the bottom of the suction cup, which directly contacts the mechanical part, to collect pressure data between the suction cup and the mechanical part at different locations. Considering that if the pressure varies at different positions on the bottom of the suction cup while clamping the mechanical part, it may create an opening between the suction cup and the part, rendering the vacuum environment ineffective and ultimately causing the mechanical part to fall off. Therefore, this invention first addresses the fluctuation differences in pressure data between the locations of the pressure sensors on the bottom of the target suction cup. The analysis reflects the degree of influence of each position on the bottom of the target suction cup on maintaining the stability of the part when it is clamped. Then, the probability of part falling reflects the likelihood of the part falling during the clamping process. Based on the probability of part falling, the real-time output power of the drive device to the mechanical claw is adjusted. This allows the clamping device to clamp the mechanical part more stably when the output power is adjusted, reducing the impact of vibration on the clamping device and preventing the mechanical part from falling. Finally, at the obtained stop time, the power output and clamping action of the clamping device on the mechanical part are stopped to prevent over-clamping and improve the clamping accuracy of the mechanical part. Attached Figure Description

[0050] To more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0051] Figure 1 This is an overall structural diagram of a high-precision clamping device for mechanical parts based on flexible force feedback, provided in one embodiment of the present invention.

[0052] Figure 2 This is a front view of a high-precision clamping device for mechanical parts based on flexible force feedback, provided in one embodiment of the present invention.

[0053] Figure 3 The image shows a front view of a suction cup for a high-precision clamping device for mechanical parts based on flexible force feedback, provided in one embodiment of the present invention.

[0054] Figure 4 The bottom view of a suction cup of a high-precision clamping device for mechanical parts based on flexible force feedback provided in an embodiment of the present invention;

[0055] Figure 5This is a schematic diagram showing the positional distribution of the pressure sensors at the bottom of the suction cup according to an embodiment of the present invention;

[0056] Figure 6 This is a flowchart illustrating a method for controlling the clamping action of a mechanical part according to an embodiment of the present invention.

[0057] Reference numerals: 1-Motor, 2-Drive connection, 3-Bracket, 4-First link, 5-Second link, 6-Third link, 7-Rubber ring, 8-Connecting hole, 9-Rotating base, 10-Electro-hydraulic rod, 11-Electro-cylinder, 12-Suction cup, 13-Pressure sensor, 14-Cavity. Detailed Implementation

[0058] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of a high-precision clamping device for mechanical parts based on flexible force feedback proposed according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0059] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0060] The following description, in conjunction with the accompanying drawings, details a specific solution for a high-precision clamping device for mechanical parts based on flexible force feedback provided by the present invention.

[0061] Please see Figures 1 to 4The diagram shows an overall structural diagram and a front view of a high-precision clamping device for mechanical parts based on flexible force feedback, according to an embodiment of the present invention. It also shows a front view and a bottom view of the suction cup of the clamping device. The clamping device includes a drive unit and multiple mechanical claws. The drive unit consists of a motor 1, a drive connection part 2, and a rotating base 9, and is fixed to a bracket 3. Each mechanical claw consists of a first connecting rod 4, a second connecting rod 5, a third connecting rod 6, an electro-hydraulic rod 10, and a suction cup 12. The two ends of the second connecting rod 5 are hinged to one end of the first connecting rod 4 and one end of the third connecting rod 6, respectively. Next, the other end of the first connecting rod 4 is hinged to the bracket 3, and the other end of the third connecting rod 6 is fixedly connected to the suction cup 12. One end of the electro-hydraulic rod 10 is fixedly connected to the rotating base 9, and the other end is fixedly connected to the third connecting rod 6. An electro-hydraulic cylinder 11 is installed at the top of the suction cup 12. During the operation of the clamping device, the rotating base 9 controls the electro-hydraulic rod 10 to swing and extend and retract, thereby realizing the clamping action of the mechanical parts. At the same time, when the suction cup contacts the mechanical parts, the electro-hydraulic cylinder 11 is used to extract or release the air in the cavity 14 of the suction cup 12, so that the suction cup 12 can adsorb or release the parts.

[0062] To improve the clamping accuracy of the clamping device for mechanical parts, this embodiment of the invention also installs a rubber ring 7 at the bottom of each suction cup 12, and evenly installs multiple pressure sensors 13 at different positions on the bottom of the suction cup 12. Each pressure sensor 13 is equidistant from the bottom edge of the suction cup 12. The internal sealed space of the rubber ring 7 is filled with flexible particles and air. When the rubber ring 7 comes into contact with the mechanical part, the pressure generated between them compresses the rubber ring 7, causing the rubber ring 7 to trigger the pressure sensor 13. This allows the pressure sensor 13 to collect pressure data for each historical moment within a preset time period prior to the current moment at its location. The preset time period is typically 10 to 20 seconds. In one embodiment of the invention, the preset time period is set to 15 seconds, and the pressure data collection interval is set to 0.1 seconds. The preset time period and the data collection interval can also be set by the implementer according to the specific implementation scenario, and are not limited here. Please refer to [link to relevant documentation]. Figure 5The diagram shows the positional distribution of each pressure sensor at the bottom of the suction cup provided in one embodiment of the present invention. Here, a to h represent the positions of the eight pressure sensors. In one embodiment of the present invention, eight pressure sensors are installed at the bottom of each suction cup 12. It should be noted that the number of pressure sensors 13 installed on each suction cup 12 and the number of suction cups 12 must both be even, so that each pressure sensor 13 or suction cup 12 has a mutually symmetrical pressure sensor 13 or suction cup 12, facilitating subsequent calculation and analysis. Furthermore, the clamping device in this embodiment of the present invention also includes a feedback adjustment module. The feedback adjustment module is used to adjust the output power of the drive device based on the pressure data and control the clamping action of the mechanical parts, thereby achieving precise clamping of the mechanical parts.

[0063] Please see Figure 6 The diagram illustrates a method for controlling the clamping action of a mechanical part according to an embodiment of the present invention, the method comprising:

[0064] Step S1: Take any suction cup as the target suction cup, and obtain the stability impact of the part at each position on the bottom of the target suction cup based on the fluctuation difference of the pressure data between the positions of the pressure sensors at the bottom of the target suction cup; obtain the real-time probability of the part falling off the clamping device based on the stability impact of the part at each position on the bottom of each suction cup.

[0065] Because the maximum suction force of the suction cup cavity on the mechanical part is constant under vacuum, and the part is subjected to the pressure of the suction cup, together they keep the mechanical part in a stationary state relative to the clamping device. However, if the pressure changes at different positions on the bottom of the suction cup while clamping the mechanical part, an opening may appear between the suction cup and the part, rendering the vacuum environment ineffective and ultimately causing the mechanical part to fall off. Therefore, when the clamping device holds the mechanical part and vibrations occur during movement or processing, if the pressure change at a certain position on the bottom of the suction cup differs significantly from other positions, it indicates that the clamping effect at that position is poor, and the part is prone to falling off at that position. Therefore, this embodiment of the invention first... Any suction cup is used as the target suction cup, and the fluctuation differences of pressure data between the locations of the pressure sensors at the bottom of the target suction cup are analyzed. The part stability influence reflects the degree of influence of each position at the bottom of the target suction cup on maintaining the stability of the part when it is clamping the part. The greater the part stability influence, the greater the impact of that position on the clamping stability of the part, which is more likely to cause instability of the part and more likely to cause the part to fall off. Subsequently, based on the part stability influence of different positions at the bottom of each suction cup, the real-time probability of the part falling off can be accurately calculated and analyzed. It should be noted that the positions at the bottom of the suction cup mentioned in the embodiments of the present invention refer to the positions of the pressure sensors at the bottom of the suction cup.

[0066] Preferably, in one embodiment of the present invention, the method for obtaining the stability influence of the component at each position on the bottom of the target suction cup specifically includes:

[0067] Based on the pressure data fluctuations at each historical moment at the location of each pressure sensor on the bottom of the target suction cup, the degree of pressure fluctuation at each location on the bottom of the target suction cup is obtained. The greater the degree of pressure fluctuation, the more obvious the fluctuation or change in the pressure data at each location on the bottom of the target suction cup.

[0068] Preferably, in one embodiment of the present invention, the method for obtaining the pressure fluctuation level at each position on the bottom of the target suction cup specifically includes:

[0069] Extreme values ​​are extracted from the pressure data at all historical moments at the location of each pressure sensor on the bottom of the target suction cup. These extreme values ​​include both maximum and minimum values. Two adjacent extreme values ​​are grouped together, and each extreme value group typically contains one minimum and one maximum value. The absolute value of the difference between the two extreme values ​​in each extreme value group is taken as the pressure change amplitude of each extreme value group. The larger the pressure change amplitude, the greater the difference between the two extreme values ​​in the extreme value group, indicating that the pressure data fluctuation at each location is more obvious. Therefore, the average pressure change amplitude of all extreme value groups at each location on the bottom of the target suction cup can be taken as the pressure fluctuation degree at each location on the bottom of the target suction cup.

[0070] In embodiments of the present invention, extreme values ​​can be selected by curve fitting or local comparison, which is not limited here. The process of selecting extreme values ​​by curve fitting is as follows: First, curve fitting methods such as least squares are used to fit the pressure data of all historical moments at the location of each pressure sensor at the bottom of the target suction cup, and the temperature data corresponding to the point with a slope of 0 on the fitted curve is taken as the extreme value. The process of selecting extreme values ​​by local comparison is as follows: The pressure data of each historical moment is compared with the pressure data of other historical moments in a preset window. If the temperature data of the historical moment is greater than or less than the temperature data of other historical moments in the preset window, then the temperature data of that moment is taken as the extreme value. The length of the preset window can be set to 7, that is, the preset window includes the 6 other historical moments closest to the historical moment and the historical moment itself. The length of the preset window can also be set by the implementer according to the specific implementation scenario, which is not limited here.

[0071] In other embodiments of the present invention, the degree of pressure fluctuation at each position on the bottom of the target suction cup can also be obtained by calculating the variance or standard deviation of the pressure data at all historical moments of each pressure sensor location on the bottom of the target suction cup, which is not limited here.

[0072] Then, the location of any pressure sensor on the bottom of the target suction cup is taken as the target position. The position symmetrical to the target position is taken as the symmetrical position of the target position. All other positions besides the target position and its symmetrical position are taken as reference positions of the target position. Please refer to [link to relevant documentation]. Figure 5 For example, if position a is the target position, then position e is the symmetrical position of position a, and the other positions are reference positions.

[0073] The more similar the pressure fluctuations at the target location and the symmetrical location, the greater the impact of the target location on the stability of the part. Therefore, a negative correlation mapping can be performed on the difference in pressure fluctuation levels between the target location and the symmetrical location to obtain the pressure fluctuation similarity of the target location.

[0074] The greater the difference in pressure fluctuations between the target position and the reference position, the greater the impact of the target position on the stability of the component. The difference in pressure fluctuations between the target position and each reference position is accumulated to obtain the pressure fluctuation difference degree of the target position.

[0075] In the embodiments of the present invention, in the above-described analysis of the difference in pressure fluctuation, the analysis of the difference in pressure fluctuation between the target position and the symmetrical position, as well as the analysis of the difference in pressure fluctuation between the target position and each reference position, can be achieved by calculating the absolute value or the square value of the difference in pressure fluctuation. No limitation is made here.

[0076] The product of pressure fluctuation similarity and pressure fluctuation difference is then used as the component stability influence at the target location.

[0077] As an example, in one embodiment of the present invention, the expression for the stability influence of the component at the target location can be specifically as follows:

[0078]

[0079] Where A represents the stability impact on the target location; V represents the pressure fluctuation at the target location; V ′ Indicates the degree of pressure fluctuation at the symmetrical location of the target position; V i This represents the pressure fluctuation level at the i-th reference position of the target location; I represents the number of reference positions for the target location. Indicates the similarity of pressure fluctuations at the target location; ε1 represents the pressure fluctuation difference at the target location; ε1 represents the preset first adjustment parameter, used to prevent the denominator from being 0. The value range of ε1 is [0.001, 0.01]. In one embodiment of the present invention, ε1 is set to 0.01. The specific value of ε1 can also be set by the implementer according to the specific implementation scenario, and is not limited here.

[0080] It should be noted that negative correlation mapping can also be achieved through other basic mathematical operations in other embodiments of the present invention, which will not be elaborated here.

[0081] Using the same method described above, the stability impact of each pressure sensor at the bottom of the target suction cup and the stability impact of each pressure sensor at the bottom of each suction cup can be obtained. Based on the stability impact of each part at each position at the bottom of each suction cup, the real-time probability of part falling off the clamping device can be obtained. The probability of part falling off reflects the likelihood of part falling off during the clamping process.

[0082] Preferably, in one embodiment of the present invention, the method for obtaining the real-time probability of part falling off the clamping device specifically includes:

[0083] Since vibration has a certain direction, an initial vibration vector is first constructed at the bottom of the target suction cup for each pressure sensor location. The direction of the initial vibration vector at each location is the direction from the center of the target suction cup to each location, and the length of the initial vibration vector at each location is the stability influence of the part at each location on the bottom of the target suction cup.

[0084] The locations of the two symmetrical pressure sensors on the bottom of the target suction cup are defined as the symmetrical position group of the target suction cup. Please refer to [link / reference]. Figure 5 In this context, positions a and e form a symmetrical position group, positions b and f form a symmetrical position group, positions c and g form a symmetrical position group, and positions d and h form a symmetrical position group. The sum of the part stability influence of two positions in each symmetrical position group is then used as the comprehensive influence of each symmetrical position group. The two positions in the symmetrical position group with the largest comprehensive influence are then used as the key vibration positions at the bottom of the target suction cup.

[0085] Then, based on the initial vibration vectors of all positions in the preset neighborhood of each key vibration position on the bottom of the target suction cup, two suspected vibration vectors of the target suction cup are obtained. Since the part usually vibrates in two directions, the directions of the two suspected vibration vectors represent the directions of the part's vibration, and the magnitudes of the two suspected vibration vectors represent the amplitudes of the part's vibration in these two directions.

[0086] Preferably, in one embodiment of the present invention, the method for obtaining the two suspected vibration vectors of the target suction cup specifically includes:

[0087] The sum of the initial vibration vectors of all locations within a preset neighborhood of each critical vibration location is used as the suspected vibration vector of the target suction cup. Since one suspected vibration vector is obtained for each critical vibration location, two suspected vibration vectors can be obtained. All locations within the preset neighborhood include the critical vibration location. The length of the preset neighborhood is set to 5, meaning that the preset neighborhood of a critical vibration location includes the locations of the four nearest other pressure sensors and the critical vibration location itself. (See [link to relevant documentation]). Figure 5 If positions a and e are two key vibration positions, then the preset neighborhood of position a includes positions a, b, c, g, and h, and the preset neighborhood of position e includes positions c, d, e, f, and g. The specific length of the preset neighborhood can also be set by the implementer according to the specific implementation scenario, and is not limited here.

[0088] Using the same method described above, two suspected vibration vectors for each suction cup can be obtained. Then, two mutually symmetrical suction cups are taken as a symmetrical suction cup group, and any symmetrical suction cup group is taken as the target symmetrical suction cup group. The projection plane of the target symmetrical suction cup group is constructed, wherein the projection plane is perpendicular to the line connecting the midpoints of the two suction cups in the target symmetrical suction cup group, and the two suspected vibration vectors of each suction cup in the target symmetrical suction cup group are projected onto the projection plane through the center point of the line connecting the two suction cups, thus obtaining two projection vectors of each suction cup in the target symmetrical suction cup group.

[0089] In the two suction cups of the target symmetrical suction cup group, a projection vector is selected from each of the two suction cups. The two selected projection vectors are used as a projection vector group. The projection vector group contains a projection vector from one suction cup and a projection vector from the other suction cup in the target symmetrical suction cup group. At this time, there are a total of four projection vector groups. Then, the angle between the two projection vectors in each projection vector group is used as the reference angle of each projection vector group. The two projection vector groups with the smallest reference angle are used as the reference projection vector groups of the target symmetrical suction cup group.

[0090] Based on the reference angle of each reference projection vector group of the target symmetrical suction cup group, and the projection vector in each reference projection vector group, the part drop coefficient of the target symmetrical suction cup group is obtained. The larger the part drop coefficient, the greater the possibility of part drop when analyzing the target symmetrical suction cup group.

[0091] Preferably, in one embodiment of the present invention, the method for obtaining the part drop coefficient of the target symmetrical suction cup assembly specifically includes:

[0092] The smaller the reference angle of the two reference projection vector groups of the target symmetrical suction cup group, the more consistent the vibration direction, and the greater the possibility of the part falling off. Therefore, the average value of the reference angle of the two reference projection vector groups of the target symmetrical suction cup group is negatively correlated to obtain the first part falling off evaluation value of the target symmetrical suction cup group.

[0093] The sum of the two projection vectors in each reference projection vector group is used as the composite projection vector of each reference projection vector group. The larger the magnitude of the composite projection vector of the two reference projection vector groups, the greater the vibration amplitude, and the greater the possibility of the part falling off. Therefore, the sum of the magnitudes of the composite projection vectors of the two reference projection vector groups can be used as the second part falling assessment value of the target symmetrical suction cup group.

[0094] Then, the drop evaluation values ​​of the first and second parts are combined and normalized to limit the calculation results to the range of [0,1], thereby obtaining the drop coefficient of the target symmetrical suction cup group.

[0095] In embodiments of the present invention, the sum or product of the first part drop assessment value and the second part drop assessment value can be calculated to achieve a combination of the two, which is not limited here.

[0096] In one embodiment of the present invention, the normalization process can be specifically implemented as maximum-minimum value normalization or by using an activation function or hyperbolic tangent function. Furthermore, the normalization in subsequent steps can all adopt maximum-minimum value normalization. In other embodiments of the present invention, other normalization methods can be selected according to the specific range of values, which will not be elaborated further.

[0097] As an example, in one embodiment of the present invention, the expression for the part drop coefficient of the target symmetrical suction cup assembly can be specifically as follows:

[0098]

[0099] Where G represents the part drop coefficient of the target symmetrical suction cup group; θ represents the average value of the reference angles of the two reference projection vector groups of the target symmetrical suction cup group; L1 and L2 represent the magnitude of the composite projection vector of the two reference projection vector groups, respectively; L1+L2 represent the second part drop evaluation value of the target symmetrical suction cup group; ε2 represents the preset second adjustment parameter, used to prevent the denominator from being 0. The value range of ε2 is [0.001, 0.01]. In one embodiment of the present invention, ε2 is set to 0.01. The specific value of ε2 can also be set by the implementer according to the specific implementation scenario, and is not limited here.

[0100] It should be noted that negative correlation mapping can also be achieved through other basic mathematical operations in other embodiments of the present invention, which will not be elaborated here.

[0101] By using the same method described above, the part drop coefficient of each symmetrical suction cup group can be obtained. Then, the average of the part drop coefficients of all symmetrical suction cup groups can be used as the real-time part drop probability of the clamping device.

[0102] This provides a real-time indication of the likelihood of parts falling off the clamping device.

[0103] Step S2: Based on the real-time probability of the part falling off the clamping device and the real-time pressure data at different positions of the suction cup, the real-time output power of the drive device to the mechanical claw is adjusted to obtain the adjusted output power of the drive device to the mechanical claw at the next moment. The clamping device clamps the mechanical part with the adjusted output power, and predicts the probability of the part falling off at future moments based on the probability of the part falling off the clamping device at different moments, thereby obtaining the stopping moment of the clamping action.

[0104] In addition to the suction cups generating a certain adsorption force on the parts, the clamping device of this invention can also adjust the traction force of the mechanical claws to generate pressure between the suction cups and the parts. This allows it to accommodate heavier parts and reduce vibration. Since different mechanical claws clamp parts of different shapes, the increase in pressure between different suction cups and parts varies when the traction force is increased with the same power. The rubber ring itself functions to protect the surface of the parts while reducing vibration to a certain extent through flexible clamping. When the pressure is too high, the entire clamping device changes from flexible to rigid, resulting in increased vibration amplitude and loss of vibration reduction function. Furthermore, irregular parts can cause different clamping postures of the mechanical claws, resulting in different pressures between different suction cups and parts. Therefore, during the adjustment process, different mechanical claws should be controlled to use different output powers to control different mechanical claws to complete the clamping action with different traction forces, so as to ensure the force balance of the mechanical parts.

[0105] Meanwhile, the greater the likelihood of the part falling off the clamping device in real time, the more likely the mechanical part it is to fall off at the current moment. Therefore, in order to prevent the mechanical part from falling off, this embodiment of the invention adjusts the output power of the drive device to the mechanical claw in real time based on the likelihood of the part falling off the clamping device in real time, so as to obtain the adjusted output power of the drive device to the mechanical claw in the next moment. This allows the clamping device to clamp the mechanical part with the adjusted output power, thereby avoiding the mechanical part from falling off and improving the clamping effect.

[0106] Preferably, in one embodiment of the present invention, the method for obtaining the adjustment output power of the mechanical gripper at the next moment by the driving device specifically includes:

[0107] The average real-time pressure data of all pressure sensors at the end of each robotic gripper is used as the real-time comprehensive pressure value of each robotic gripper. Based on the real-time comprehensive pressure value of each robotic gripper and the real-time probability of parts falling off the gripping device, the real-time traction control weight of each robotic gripper is obtained.

[0108] Preferably, in one embodiment of the present invention, the method for obtaining the real-time traction force control weight of each robotic gripper specifically includes:

[0109] Based on the calculation formula for traction control weight, the real-time traction control weight of each robotic gripper is obtained. The calculation formula for traction control weight is as follows:

[0110] w k =(1-norm(F) k ))×U

[0111] Among them, w k F represents the real-time traction control weight of the k-th robotic gripper; k represents the real-time comprehensive pressure value of the k-th mechanical gripper; U represents the real-time probability of the part falling off the gripping device; norm() represents the normalization function.

[0112] The higher the real-time comprehensive pressure value of a certain mechanical gripper, the greater the pressure between the gripper and the part. In this case, it is necessary to reduce the traction force provided to the gripper to avoid the entire clamping device from changing from flexible to rigid when the pressure is too high, thus losing its vibration damping function. Furthermore, when the real-time probability of the part falling off the clamping device is higher, it is necessary to increase the traction force provided to the gripper to prevent the part from falling off.

[0113] Then, based on the calculation formula for adjusting the output power, the adjusted output power of the drive device for each mechanical gripper at the next moment is obtained. The calculation formula for adjusting the output power is:

[0114] P k ′ =w k ×P max ×Δt+P k

[0115] Among them, P k ′ This indicates the adjustment of the output power of the drive unit for the k-th mechanical gripper at the next moment; w k P represents the real-time traction control weight of the k-th robotic gripper; max The maximum output power of the drive unit is represented by Δt, which is a known value for the drive unit; Δt represents the sampling time interval of the pressure data; P kThis represents the real-time output power of the drive unit to the k-th mechanical gripper, which is also a known value.

[0116] The obtained adjustable output power can then be fed back to the drive device, which controls the mechanical gripper to clamp the part with the corresponding adjustable output power. During clamping, the probability of the part falling off the clamping device at different times can be obtained by the same method described above. In order to prevent the clamping device from over-clamping the part, it is also necessary to predict the probability of the part falling off at future times based on the probability of the part falling off the clamping device at different times, and obtain the stopping time of the clamping action.

[0117] Preferably, in one embodiment of the present invention, the method for obtaining the stopping time of the clamping action specifically includes:

[0118] The probability of the part falling off the clamping device at different times is input into the time series prediction algorithm, and the predicted probability of the part falling off at multiple future times is output. Based on the change of the predicted probability of the part falling off at each future time, the stopping time of the clamping action is obtained. The time series prediction algorithm can be a moving average method, an exponential smoothing method, or an autoregressive moving average model, etc., and is not limited here.

[0119] Preferably, in one embodiment of the present invention, the method for obtaining the stopping time of the clamping action further includes:

[0120] Curve fitting is performed on the predicted probability of parts falling at all future times to obtain the prediction fitting curve. The curve fitting method can be the least squares method or other methods, which are not limited here.

[0121] The smaller the trend of the predicted probability of a part falling at a certain future moment, and the smaller the predicted probability of the part falling, the better the stopping time of the clamping action. Therefore, on the prediction fitting curve, the slope of each future moment and the sum of the predicted probability of the part falling are used as the judgment parameter for each future moment, and the future moment with the smallest judgment parameter is taken as the stopping time of the clamping action.

[0122] Subsequently, when the clamping device clamps and tightens the mechanical parts, it stops further clamping when the stop time is reached to prevent over-clamping from damaging the parts and losing the vibration damping function.

[0123] It should be noted that the order of the above embodiments of the present invention is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0124] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

Claims

1. A high-precision mechanical part clamping device based on flexible force feedback, the clamping device comprising a driving device and a plurality of mechanical claws, the tips of the mechanical claws being provided with suction cups, the driving device being used to drive the mechanical claws and control the suction cup clamping of the mechanical part, characterized in that, The clamping device further comprises a feedback adjustment module, a plurality of pressure sensors are uniformly arranged at a plurality of positions of the bottom of the suction cup, the pressure sensors are used to acquire pressure data of each historical moment within a preset time period before the current moment at the positions of the pressure sensors, the feedback adjustment module is used to perform feedback adjustment on output power of the driving device according to the pressure data, and control the clamping action of the mechanical part, comprising: Taking any one of the suction cups as a target suction cup, obtaining the part stability influence degree of each position of the bottom of the target suction cup according to the fluctuation difference of the pressure data between the positions of the pressure sensors of the bottom of the target suction cup, and obtaining the part falling possibility of the clamping device in real time according to the part stability influence degree of each position of the bottom of each suction cup; According to the part falling possibility of the clamping device in real time and the real-time pressure data of different positions of the suction cup, the output power of the driving device to the mechanical claw in real time is feedback adjusted to obtain the adjustment output power of the driving device to the mechanical claw at the next moment, the clamping device clamps the mechanical part with the adjustment output power, and the part falling possibility at the future moment is predicted according to the part falling possibility of the clamping device at different moments to obtain the stopping moment of the clamping action.

2. The high-precision mechanical part clamping device based on flexible force feedback according to claim 1, characterized in that, The obtaining of the part stability influence degree of each position of the bottom of the target suction cup comprises: According to the fluctuation of the pressure data of each historical moment at the position of each pressure sensor of the bottom of the target suction cup, the pressure fluctuation degree of each position of the bottom of the target suction cup is obtained. Taking any one of the pressure sensors of the bottom of the target suction cup as a target position, taking the positions symmetrical to the target position as the symmetrical positions of the target position, and taking other positions except the target position and the symmetrical positions of the target position as the reference positions of the target position. The difference of the pressure fluctuation degree between the target position and the symmetrical position of the target position is negatively correlated to obtain the pressure fluctuation similarity of the target position. The difference of the pressure fluctuation degree between the target position and each reference position is accumulated to obtain the pressure fluctuation difference degree of the target position. The product value of the pressure fluctuation similarity and the pressure fluctuation difference degree is taken as the part stability influence degree of the target position.

3. The high-precision mechanical part clamping device based on flexible force feedback according to claim 2, characterized in that, The obtaining of the pressure fluctuation degree of each position of the bottom of the target suction cup comprises: From the pressure data of all historical moments at the position of each pressure sensor of the bottom of the target suction cup, extreme values are extracted, two adjacent extreme values are taken as an extreme value group, and the absolute value of the difference between the two extreme values in each extreme value group is taken as the pressure change amplitude of each extreme value group. The average value of the pressure change amplitudes of all extreme value groups of each position of the bottom of the target suction cup is taken as the pressure fluctuation degree of each position of the bottom of the target suction cup.

4. The high-precision mechanical part clamping device based on flexible force feedback according to claim 1, characterized in that, The obtaining of the part falling possibility of the clamping device in real time comprises: An initial vibration vector of each position of each pressure sensor is constructed at the bottom of the target suction cup, wherein the direction of the initial vibration vector of each position is the direction of the center of the target suction cup pointing to each position, and the length of the initial vibration vector of each position is the part stability influence degree of each position of the bottom of the target suction cup. The positions of two pressure sensors on the bottom of the target suction disc are symmetrical to each other, and the positions are taken as a symmetrical position group of the target suction disc; a sum of the part stability influence degrees of the two positions in each symmetrical position group is taken as a comprehensive influence degree of each symmetrical position group; and the two positions in the symmetrical position group with the maximum comprehensive influence degree are taken as key vibration positions of the bottom of the target suction disc; Two suspected vibration vectors of the target suction disc are obtained according to the initial vibration vectors of all positions in the preset neighborhood of each key vibration position of the bottom of the target suction disc; Two suction discs that are symmetrical to each other are taken as a symmetrical suction disc group, any one symmetrical suction disc group is taken as a target symmetrical suction disc group, a projection plane of the target symmetrical suction disc group is constructed, the projection plane is perpendicular to a line connecting the midpoints of the two suction discs in the target symmetrical suction disc group, and the projection plane passes through the center point of the line; Two projection vectors of each suction disc in the target symmetrical suction disc group are obtained by projecting the two suspected vibration vectors of each suction disc in the target symmetrical suction disc group onto the projection plane; In the two suction discs of the target symmetrical suction disc group, one projection vector is selected from each of the two suction discs, the two selected projection vectors are taken as a projection vector group, an included angle between the two projection vectors in each projection vector group is taken as a reference angle of each projection vector group, and the two projection vector groups with the minimum reference angle are taken as reference projection vector groups of the target symmetrical suction disc group; A part falling coefficient of the target symmetrical suction disc group is obtained according to the reference angle of each reference projection vector group of the target symmetrical suction disc group and the projection vectors in each reference projection vector group. An average value of the part falling coefficients of all symmetrical suction disc groups is taken as a part falling possibility of the clamping device in real time.

5. The high-precision mechanical part clamping device based on flexible force feedback according to claim 4, characterized in that, The obtaining of the two suspected vibration vectors of the target suction disc comprises: A sum of the initial vibration vectors of all positions in the preset neighborhood of each key vibration position is taken as a suspected vibration vector of the target suction disc.

6. The high-precision mechanical part clamping device based on flexible force feedback according to claim 4, characterized in that, The obtaining of the part falling coefficient of the target symmetrical suction disc group comprises: A first part falling evaluation value of the target symmetrical suction disc group is obtained by performing negative correlation mapping on an average value of the reference angles of the two reference projection vector groups of the target symmetrical suction disc group; A sum of the reference angles of the two projection vectors in each reference projection vector group is taken as a synthetic projection vector of each reference projection vector group, a sum of the lengths of the synthetic projection vectors of the two reference projection vector groups is taken as a second part falling evaluation value of the target symmetrical suction disc group; The first part falling evaluation value and the second part falling evaluation value are comprehensively processed and normalized to obtain the part falling coefficient of the target symmetrical suction disc group.

7. The high-precision mechanical part clamping device based on flexible force feedback according to claim 1, characterized in that, The obtaining of the adjustment output power of the driving device to the mechanical claw at the next moment comprises: An average value of real-time pressure data of all positions of pressure sensors on the suction disc at the end of each mechanical claw is taken as a comprehensive pressure value of each mechanical claw in real time; and a traction force control weight of each mechanical claw in real time is obtained according to the comprehensive pressure value of each mechanical claw in real time and the part falling possibility of the clamping device in real time. Based on a calculation formula of the adjusting output power, an adjusting output power of the driving device to each mechanical claw at a next time is obtained, and the calculation formula of the adjusting output power is: P k ′ = w k × P max × Δt + P k wherein P k ′ represents the adjustment output power of the driving device to the kth mechanical claw at the next moment; w k represents the real-time traction force control weight of the kth mechanical claw; P max represents the maximum output power of the driving device; Δt represents the sampling time interval of the pressure data; P k represents the real-time output power of the driving device to the kth mechanical claw.

8. The high-precision mechanical part clamping device based on flexible force feedback according to claim 7, characterized in that, The obtaining of the real-time traction force control weight of each mechanical claw comprises: Based on a calculation formula of the traction force control weight, a real-time traction force control weight of each mechanical claw is obtained, and the calculation formula of the traction force control weight is: w k = (1 - norm(F k )) x U wherein w k represents the real-time traction force control weight of the kth mechanical gripper; F k represents the real-time comprehensive pressure value of the kth mechanical gripper; U represents the real-time part falling possibility of the clamping device; norm() represents a normalization function.

9. The high-precision mechanical part clamping device based on flexible force feedback according to claim 1, characterized in that, The obtaining of the stopping time of the clamping action comprises: The part falling probability of the clamping device at different times is input into a time sequence prediction algorithm, and a plurality of future time part falling prediction probabilities are output; According to the change of the part falling prediction probability of each future time, the stopping time of the clamping action is obtained.

10. The high-precision mechanical part clamping device based on flexible force feedback according to claim 9, characterized in that, The obtaining of the stopping time of the clamping action according to the change of the part falling prediction probability of each future time comprises: Curve fitting is performed on the part falling prediction probability of all future times to obtain a prediction fitting curve; On the prediction fitting curve, the sum of the slope of each future time and the part falling prediction probability is taken as a judgment parameter of each future time; The future time with the smallest judgment parameter is taken as the stopping time of the clamping action.

Images