A control method and system for a glass insulator processing robot

Abstract

This invention relates to the field of robotic arm control technology, specifically to a control method and system for a robotic arm used in glass insulator processing. The method includes the following steps: obtaining the insulator's radius thickness, gripper finger width, and closed stroke; constructing a radial coordinate sequence to calculate the thickness difference; filtering gripping coordinates based on abrupt change limits and axis distance; calculating a rotation correction angle based on the porcelain cap end face, skirt coordinates, end angle, and fixture reference; generating a gripping posture; and determining the transport trajectory based on the release time, acceptance permission, suspension duration, and deceleration distance. In this invention, avoidance zones and stable force zones are identified through thickness changes, allowing the gripper to avoid abrupt weak points. The end zero position is calibrated through coordinate deviation to ensure the workpiece posture aligns with the fixture reference. The movement path is constrained by cross-station time differences, ensuring that waiting, deceleration, and holding actions match the processing cycle, thus improving gripping stability, placement accuracy, and transport continuity, while reducing the risk of localized crush damage and suspension waiting.

Description

Technical Field

[0001] This invention relates to the field of robotic arm control technology, and in particular to a control method and system for a robotic arm used in glass insulator processing. Background Technology

[0002] The field of robotic arm control technology belongs to the technical fields related to industrial automation and electromechanical collaborative control. Its core aspects include robotic arm body motion control, end effector gripping posture control, joint displacement and speed setting, servo motor drive, pneumatic or electric gripper opening and closing control, workpiece position identification basis, processing station sequence connection, and safety limit constraints. This technical field usually revolves around the cooperation relationship between the robotic arm base, rotating joints, lifting mechanism, telescopic mechanism, end effector gripper, servo driver, position sensor, proximity switch, limit switch, and station positioning fixture. It controls the movement path, action sequence, start and stop timing, and gripping force of the robotic arm during material picking, handling, positioning, clamping, rotation, placement, and resetting processes.

[0003] The traditional control method for glass insulator processing robots refers to the robot motion control method for glass insulators in processing stages such as cutting, grinding, drilling, cleaning, inspection, or transfer. The technical issues it addresses mainly include the stable gripping, accurate placement, posture adjustment, and process connection of glass insulator blanks or finished products between the processing table, conveyor line, clamping fixture, and receiving position. The traditional method usually relies on the conveyor belt positioning switch, workpiece positioning block, gripper closure limit switch, rotary angle encoder, upper and lower limit signals of the lifting cylinder, and start and stop signals of the processing equipment to control the robot to descend to the picking height, close the gripper to hold the glass insulator, lift the robot arm, rotate horizontally or move it linearly to the processing station, adjust the workpiece orientation at the set angle by the end fixture, release the gripper after placing it in the positioning fixture, and pick up the workpiece again after processing and transfer it to the next station or receiving area.

[0004] Traditional control methods rely on position switches, positioning blocks, limit switches, angle encoders, and limit signals to operate in a fixed sequence. The clamping position is mostly determined by the preset height and fixture reference, which makes it difficult to reflect the continuous change in thickness from the outer edge to the root. This can easily lead to concentrated compression in areas of abrupt thickness changes, resulting in uneven force on the workpiece surface. During transport, the end posture deviation is difficult to correct with the relative offset of the ceramic cap and the umbrella skirt. When the cycle time is mismatched across workstations, the workpiece is still transported along the predetermined path, which can lead to decreased clamping stability, accumulation of positioning errors, and increased suspension waiting time. Summary of the Invention

[0005] To address the technical problems existing in the prior art, embodiments of the present invention provide a control method for a glass insulator processing robot, comprising the following steps: To achieve the above objectives, the present invention adopts the following technical solution: a control method for a glass insulator processing robot, comprising the following steps: S1: Obtain insulator outer edge radius data, insulator root radius data, insulator edge thickness data, insulator root thickness data, gripper finger width data, and closed target stroke data. Construct an intermediate radial coordinate sequence based on the insulator outer edge radius data and insulator root radius data, and generate adjacent coordinate thickness differences through a cubic spline interpolation function. S2: Obtain the thickness mutation limit, compare the thickness difference between adjacent coordinates with the thickness mutation limit, filter the avoidance zone coordinate data and candidate stress zone coordinate data, combine the insulator axis coordinate data with the Euclidean distance function to generate candidate zone distance values, and construct the manipulator gripping coordinates; S3: Obtain the coordinates of the ceramic cap end face, the coordinates of the glass umbrella skirt, the current end rotation angle, the fixture positioning reference angle, and the gripper centerline angle; generate the lateral deviation and longitudinal deviation of the ceramic cap; compare them with the fixture positioning reference angle; and construct the end rotation correction angle. S4: Adjust the current end rotation angle based on the end rotation correction angle, generate end rotation zero position data, combine with the robot gripping coordinates to construct robot gripping posture data, obtain fixture release time, permission receiving time, deceleration stop distance parameters, suspension allowable time and waiting coordinate data, and calculate cross-station time difference; S5: Based on the cross-station time difference, the target movement trajectory parameters are filtered by combining the lateral movement trajectory parameters, deceleration waiting coordinates and in-situ holding coordinates. The robot arm transfer control data is generated by combining the deceleration stopping distance parameters and waiting coordinate data.

[0006] As a further aspect of the present invention, the specific steps of S1 are as follows: S101: Obtain insulator outer edge radius data, insulator root radius data, insulator edge thickness data, insulator root thickness data, gripper finger width data, and closed target stroke data. Perform subtraction calculation based on the insulator outer edge radius data and insulator root radius data to obtain the radial span. Set the offset length according to the radial span. Divide the interval where the insulator root radius data is located into equal intervals according to the offset length to establish an intermediate radial coordinate sequence. S102: Call the intermediate radial coordinate sequence, insulator edge thickness data and insulator root thickness data, perform pairing with the nodes in the intermediate radial coordinate sequence and the insulator edge thickness data and insulator root thickness data to obtain the coordinate correlation matrix, input the coordinate correlation matrix into the cubic spline interpolation function to perform fitting operation to calculate the thickness value, and summarize the thickness values ​​to obtain the radial thickness distribution set; S103: Call the gripper finger width data, closed target stroke data and radial thickness distribution set, extract the displacement based on the gripper finger width data, select adjacent items in the radial thickness distribution set according to the displacement and perform subtraction to obtain the initial fluctuation difference, set boundary conditions for the closed target stroke data, filter out out-of-bounds values ​​based on the boundary conditions and generate adjacent coordinate thickness difference values.

[0007] As a further aspect of the present invention, the specific steps of S2 are as follows: S201: Obtain the preset thickness change limit, collect the thickness difference between adjacent coordinates, read the value of each thickness difference between adjacent coordinates, perform a size comparison judgment between the read value and the thickness change limit, peel off the corresponding node when the read value is greater than the thickness change limit, extract the corresponding node when the read value is not greater than the thickness change limit, aggregate the corresponding nodes, and establish a candidate force coordinate set. S202: Obtain the insulator axis coordinate data, extract the internal node coordinate components for the candidate force coordinate set, perform subtraction operation on the node coordinate components and the corresponding components of the insulator axis coordinate data to obtain the coordinate deviation, perform square operation on the coordinate deviation and perform summation, perform square root operation on the summation value to obtain the radial distance, and generate a candidate area distance sequence. S203: Call the candidate force coordinate set, perform ascending sorting processing on the distributed values ​​in the distance sequence of the candidate area to obtain the distance order parameter, extract the first and second minimum values ​​according to the distance order parameter, perform reverse index matching in the candidate force coordinate set according to the minimum value, locate the node information associated with the minimum value, extract the horizontal and vertical parameters corresponding to the node information as the action execution point, and establish the gripping coordinates of the robot arm.

[0008] As a further aspect of the present invention, the method for obtaining the thickness mutation limit is as follows: obtain the root thickness parameter and edge thickness parameter of the insulator standard sample, perform a difference operation on the two to calculate the global thickness difference, simultaneously obtain the outer edge radius parameter and root radius parameter of the insulator standard sample, perform a difference operation on the two to calculate the effective radial span value; divide the global thickness difference by the effective radial span value to calculate the base thickness change rate, obtain the deformation tolerance coefficient corresponding to the flexible pad material of the gripper, perform a multiplication operation on the base thickness change rate and the deformation tolerance coefficient to perform numerical correction, and establish the corrected output result as the thickness mutation limit.

[0009] As a further aspect of the present invention, the specific steps of S3 are as follows: S301: Collect the coordinates of the porcelain cap end face and the glass umbrella skirt. Extract the horizontal and vertical axis components of the porcelain cap end face coordinates. Extract the horizontal and vertical axis components of the glass umbrella skirt coordinates. Perform subtraction calculation on the horizontal axis components of the end face and the horizontal axis components of the umbrella skirt to obtain the horizontal axis distance. Perform subtraction calculation on the vertical axis components of the end face and the vertical axis components of the umbrella skirt to obtain the vertical axis distance. Associate the horizontal axis and vertical axis distances to generate a two-dimensional deviation set. S302: Obtain the end rotation angle and the gripper centerline angle; extract the horizontal axis distance and vertical axis distance from the two-dimensional deviation set; perform sign determination on the horizontal axis distance and vertical axis distance to obtain quadrant assignment; combine quadrant assignment with the arctangent operation of the distance ratio to obtain the original offset angle; perform summation calculation on the original offset angle, the end rotation angle, and the gripper centerline angle to generate the actual workpiece offset angle. S303: Obtain the preset fixture positioning reference angle, subtract the actual offset angle of the workpiece from the fixture positioning reference angle to obtain the angle compensation difference, set a rotation limit range for the angle compensation difference, perform a limit comparison between the angle compensation difference and the rotation limit range, perform a direction reversal operation on the angle compensation difference based on the comparison result, and establish the end rotation correction angle.

[0010] As a further aspect of the present invention, the method for obtaining the fixture positioning reference angle is as follows: before the robot arm performs the transfer task, a vision camera fixed above the workstation is driven to acquire a top view image of the unloaded fixture. Edge contour detection is performed on the image to extract the geometric feature edge lines of the mechanical positioning groove or mechanical stop on the surface of the fixture. Based on the geometric feature edge lines, a central symmetry axis vector of the fixture positioning groove is generated. The central symmetry axis vector is mapped to the global reference coordinate system of the robot arm base. The angle between the central symmetry axis vector and the horizontal reference axis of the global coordinate system is calculated. The obtained angle parameter is recorded as the fixture positioning reference angle.

[0011] As a further aspect of the present invention, the specific steps of S4 are as follows: S401: Call the end-effector rotation correction angle, obtain the current end-effector rotation angle, extract the origin rotation component within the current end-effector rotation angle, convert the end-effector rotation correction angle into a numerical compensation amount, perform an addition operation on the origin rotation component and the numerical compensation amount to obtain the attitude sum, perform inverse coordinate calculation on the attitude sum to obtain the joint parameters, and establish end-effector rotation zero position data; S402: Call the end-effector rotation zero-position data and the robot gripping coordinates, extract the three-dimensional translation component in the robot gripping coordinates, extract the rotation component in the end-effector rotation zero-position data, map the three-dimensional translation component to the translation area of ​​the transformation matrix, map the rotation component to the rotation area of ​​the transformation matrix, and perform splicing on the translation area and the rotation area to generate robot gripping posture data; S403: Based on the gripping posture data of the robotic arm, obtain the jig release time, the receiving permission time, the deceleration and stop distance parameters, the suspension allowable duration and the waiting coordinate data. Perform a subtraction operation on the receiving permission time and the jig release time to obtain the initial time difference. Perform a multiplication conversion on the initial time difference and the clock frequency to obtain the span duration. Perform sign verification on the span duration to remove negative values ​​and generate the cross-station time difference.

[0012] As a further aspect of the present invention, the specific steps of S5 are as follows: S501: Obtain the preset allowable suspension time of the equipment, perform subtraction calculation on the cross-workstation time difference and the allowable suspension time to obtain the time surplus, perform positive and negative sign extraction and judgment on the time surplus value, match the preset logical identifier according to the extracted state parameters, and establish the time difference comparison result. S502: Call the time difference comparison result, collect the lateral trajectory parameters, deceleration waiting coordinates and in-situ holding coordinates issued by the motion unit, perform branch route selection based on the built-in logical identifier of the time difference comparison result, extract the lateral trajectory parameters and deceleration waiting coordinates and perform spatial vector concatenation when the identifier represents a positive direction, extract the in-situ holding coordinates and perform coordinate node copying when the identifier represents a negative direction, map the concatenated or copied node values ​​to the motion sequence matrix, and establish the target movement trajectory parameters; S503: Call the target movement trajectory parameters, obtain the deceleration and stopping distance parameters and waiting coordinate data, extract the terminal node coordinates inside the target movement trajectory parameters, extract the approach direction vector inside the waiting coordinate data, perform a quantity product operation on the direction vector and the deceleration and stopping distance parameters to obtain the spatial offset compensation vector, perform coordinate addition processing on the terminal node coordinates and the spatial offset compensation vector to obtain the corrected pose coordinates, replace the original terminal node according to the corrected pose coordinates, and generate the robot arm transfer control data.

[0013] As a further aspect of the present invention, the method for setting the allowable suspension time is as follows: The initial clamping force data of the clamping mechanism in the grasping state and the force attenuation coefficient of the mechanism in the continuous pressure-holding state are obtained; the critical safe force threshold for the insulator to overcome the frictional force of the gripper finger surfaces and generate relative slippage is read; a subtraction operation is performed on the initial clamping force data and the critical safe force threshold to obtain the clamping redundancy; a division operation is performed on the clamping redundancy and the force attenuation coefficient to obtain the theoretical limit holding time; the theoretical limit holding time is multiplied by a preset safe loss ratio; and the allowable suspension time is determined based on the calculated time span.

[0014] A control system for a glass insulator processing robot, the system comprising: The contour parameter reconstruction module acquires insulator outer edge radius data, insulator root radius data, insulator edge thickness data, insulator root thickness data, gripper finger width data, and closed target stroke data. Based on the insulator outer edge radius data and insulator root radius data, it constructs an intermediate radial coordinate sequence and generates the thickness difference between adjacent coordinates through a cubic spline interpolation function. The clamping area determination module obtains the thickness change limit, compares the thickness difference between adjacent coordinates with the thickness change limit, filters the avoidance area coordinate data and candidate stress area coordinate data, and generates candidate area distance values ​​by combining the insulator axis coordinate data with the Euclidean distance function to construct the manipulator clamping coordinates. The end-point attitude calibration module obtains the coordinates of the ceramic cap end face, the coordinates of the glass umbrella skirt, the current end-point rotation angle, the fixture positioning reference angle, and the gripper centerline angle. It generates the lateral deviation and longitudinal deviation of the ceramic cap and compares them with the fixture positioning reference angle to construct the end-point rotation correction angle. The grasping posture generation module adjusts the current end rotation angle based on the end rotation correction angle, generates end rotation zero position data, and constructs robot gripping posture data by combining the robot gripping coordinates. It also obtains jig release time, permission receiving time, deceleration stop distance parameters, suspension allowable time and waiting coordinate data, and calculates cross-station time difference. The transfer cycle control module, based on the cross-workstation time difference, combines the lateral trajectory parameters, deceleration waiting coordinates, and in-situ holding coordinates to filter the target movement trajectory parameters, and combines the deceleration stopping distance parameters and waiting coordinate data to generate robot transfer control data.

[0015] Compared with the prior art, the advantages and positive effects of the present invention are as follows: In this invention, a radial thickness variation relationship is constructed based on the outer radius, root radius, edge thickness, root thickness, gripper finger width, and closed target stroke. The thickness difference is linked with the abrupt change limit to screen the avoidance zone and the stress zone. The clamping coordinate is formed by constraining the center distance, so that the gripper avoids the weak position of the abrupt change and fits the stress area. The rotation correction angle is derived by the coordinate deviation between the ceramic cap end face and the glass umbrella skirt, so that the end zero position is consistent with the fixture reference. The lateral trajectory is constrained by the release time, the receiving permission time, the allowable suspension time, and the deceleration stop distance. This makes the transfer process match the rhythm between waiting, deceleration, and in-situ holding, reducing crush damage and posture deviation. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying 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.

[0017] Figure 1 This is a schematic diagram of the steps of the present invention; Figure 2 This is a detailed schematic diagram of S1 of the present invention; Figure 3 This is a detailed schematic diagram of S2 of the present invention; Figure 4 This is a detailed schematic diagram of S3 of the present invention; Figure 5 This is a detailed schematic diagram of S4 of the present invention; Figure 6 This is a detailed schematic diagram of S5 of the present invention; Figure 7 This is a system module diagram of the present invention. Detailed Implementation

[0018] The technical solution of the present invention will now be described with reference to the accompanying drawings.

[0019] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0020] Please see Figure 1 This invention provides a method for controlling a robot arm in glass insulator processing, comprising the following steps: S1: Obtain insulator outer edge radius data, insulator root radius data, insulator edge thickness data, insulator root thickness data, gripper finger width data, and closed target stroke data. Construct an intermediate radial coordinate sequence based on the insulator outer edge radius data and insulator root radius data. Input the intermediate radial coordinate sequence into a cubic spline interpolation function to calculate and generate the thickness difference between adjacent coordinates. S2: Obtain the thickness change limit, compare the thickness difference between adjacent coordinates with the thickness change limit to filter the avoidance zone coordinate data and candidate stress zone coordinate data, obtain the insulator axis coordinate data, input the candidate stress zone coordinate data and the insulator axis coordinate data into the Euclidean distance function to calculate and generate candidate zone distance values, and filter the candidate stress zone coordinate data based on the candidate zone distance values ​​to construct the robot gripping coordinates; S3: Obtain the coordinates of the ceramic cap end face, the coordinates of the glass umbrella skirt, the current end rotation angle, the fixture positioning reference angle, and the gripper centerline angle. Calculate the ceramic cap end face coordinates and the glass umbrella skirt coordinates to generate the ceramic cap lateral deviation and the ceramic cap longitudinal deviation. Compare the ceramic cap lateral deviation and the ceramic cap longitudinal deviation with the fixture positioning reference angle to construct the end rotation correction angle. S4: Adjust the current end-effector rotation angle based on the end-effector rotation correction angle to generate end-effector rotation zero-position data. Construct robot gripping posture data based on robot gripping coordinates and end-effector rotation zero-position data. Obtain fixture release time, permission reception time, deceleration stop distance parameters, suspension allowable time and waiting coordinate data. Calculate the time difference between fixture release time and permission reception time to generate cross-station time difference. S5: Obtain the lateral trajectory parameters, deceleration waiting coordinates, and in-situ holding coordinates. Compare the cross-station time difference with the allowable suspension time to generate a time difference comparison result. Based on the time difference comparison result, filter the lateral trajectory parameters, deceleration waiting coordinates, and in-situ holding coordinates to construct the target movement trajectory parameters. Based on the deceleration stop distance parameters and waiting coordinate data, adjust the target movement trajectory parameters to generate robot transfer control data. Please see Figure 2 The specific steps of S1 are as follows: S101: Obtain insulator outer edge radius data, insulator root radius data, insulator edge thickness data, insulator root thickness data, gripper finger width data, and closed target stroke data. Perform subtraction calculation based on the insulator outer edge radius data and insulator root radius data to obtain the radial span. Set the offset length according to the radial span. Divide the interval where the insulator root radius data is located into equal intervals according to the offset length to establish an intermediate radial coordinate sequence. Three-dimensional point cloud data of the insulator surface is acquired using an industrial 3D vision camera. Statistical filtering is performed on the 3D point cloud data to remove spatially discrete noise. Edge feature point fitting is then performed on the filtered data to directly obtain the insulator's outer edge radius as 150 mm, the insulator's root radius as 60 mm, the insulator's edge thickness as 15 mm, and the insulator's root thickness as 30 mm. The gripper finger width data from the device driver library is 12 mm, and the closed target stroke data from the device control process is 80 mm. Based on these parameters, a subtraction calculation is performed between the insulator's outer edge radius (150 mm) and root radius (60 mm). The calculation is expressed as 150 - 60 = 90, yielding a radial span of 90 mm. Based on the obtained radial span of 90, a multiplication calculation is performed using a preset sampling resolution factor of 0.15. This calculation process is expressed by the formula 90*0.15=13.5, with the step size set at 13.5 mm. According to the set step size of 13.5, the outward extension interval containing the insulator root radius data 60 is equally spaced. Using the insulator root radius data 60 as the starting point, numerical accumulation is performed sequentially to generate segmented coordinate values ​​73.5, 87, 100.5, 114, 127.5, and 141, establishing an intermediate radial coordinate sequence containing all the above nodes. The advantage of this operational logic is that by combining the sampling resolution with the multiplication of the radial span, adaptive mesh generation for workpieces of different sizes is achieved.

[0021] Table 1. Insulator Foundation Parameter Characteristics Table Original measurement Insulator outer radius 150 millimeters Original measurement Insulator root radius 60 millimeters Original measurement Insulator edge thickness 15 millimeters Original measurement Insulator root thickness 30 millimeters Original measurement gripper finger width 12 millimeters Secondary operation Radial span value 90 millimeters Secondary operation Divide and walk length 13.5 millimeters Table 1 lists the dimensional parameters and step sizes obtained through point cloud measurement and preliminary calculations.

[0022] S102: Call the intermediate radial coordinate sequence, insulator edge thickness data and insulator root thickness data, perform pairing with the nodes in the intermediate radial coordinate sequence and the insulator edge thickness data and insulator root thickness data to obtain the coordinate correlation matrix, input the coordinate correlation matrix into the cubic spline interpolation function to perform fitting operation to calculate the thickness value, and summarize the thickness values ​​to obtain the radial thickness distribution set; The established intermediate radial coordinate sequence is invoked, along with the insulator edge thickness data 15 and insulator root thickness data 30 obtained from the aforementioned measurements. For each node within the intermediate radial coordinate sequence, the surface scattered elevation information retained during the actual measurement process is extracted and paired with the insulator edge thickness data 15 and insulator root thickness data 30 to obtain a coordinate correlation matrix containing the initial mapping relationship between radial coordinates and thickness. This coordinate correlation matrix is ​​then input into a cubic spline interpolation function to perform a fitting operation, constructing an independent cubic polynomial coefficient matrix equation system for each interval. A constraint of equal first and second derivatives is imposed at the boundaries of adjacent intervals, and a natural boundary condition is introduced to set the second derivatives at both ends to 0. The tridiagonal linear equation system composed of node spacing and thickness difference is solved using a chasing method to obtain the constant terms of the equations for each sub-interval. Node 73.5 is extracted and substituted into the corresponding cubic polynomial. Based on the solved coefficients 0.01, -0.2, 0.8, and 30, the calculation process is expressed as 0.01*73.5*73.5*73.5-0.2*73.5*73.5+0.8*73.5+30=27.2, yielding a thickness of 27.2 mm at this location. This equation is repeated for all nodes within the intermediate radial coordinate sequence, and the calculated thickness values ​​of 30, 27.2, 24.5, 21.9, 19.4, 17.1, and 15 are summarized to obtain a complete radial thickness distribution set mapping the surface undulations. The advantage of this operational logic is that the Runge phenomenon error at the interpolation edges is eliminated through the boundary second derivative zeroing constraint.

[0023] S103: Call the gripper finger width data, closed target stroke data and radial thickness distribution set, extract the displacement based on the gripper finger width data, select adjacent items in the radial thickness distribution set according to the displacement and perform subtraction to obtain the initial fluctuation difference, set boundary conditions for the closed target stroke data, filter out out-of-bounds values ​​based on the boundary conditions and generate adjacent coordinate thickness difference; The process retrieves the read gripper finger width data 12, the closed target stroke data 80, and the generated radial thickness distribution set. Based on the gripper finger width data 12, the displacement compensation is extracted, and the pre-calibrated finger edge gap coefficient 0.2 is extracted. A multiplication operation is performed between 12 and 0.2, expressed as 12*0.2=2.4, yielding a displacement of 2.4 mm. Based on the displacement of 2.4, adjacent items in the radial thickness distribution set are selected according to the one-way traversal rule, and subtraction operations are performed. Subtraction is performed between the thickness value 27.2 and the initial value 30, expressed as 27.2-30=-2.8. Subtraction is performed between the thickness values ​​24.5 and 27.2, expressed as 24.5-27.2=-2.7. This process is repeated for all adjacent nodes to obtain multiple sets of initial fluctuation differences. Boundary conditions are set for the closed target travel data 80, and the travel tolerance percentage parameter 0.04 is extracted. A multiplication operation is performed between 80 and 0.04, expressed as 80 * 0.04 = 3.2. ±3.2 is established as the absolute boundary condition. Based on this boundary condition, a limiting filter operation is performed on the full initial fluctuation difference. The absolute value of each fluctuation difference is compared with 3.2. If an absolute value of any initial fluctuation difference is found to be greater than 3.2, this out-of-bounds value is discarded, retaining valid differences such as -2.8 and -2.7, which are within the legal range, generating cleaned adjacent coordinate thickness differences. The advantage of this operation logic is that it locks in abnormal abrupt changes in thickness based on the actual travel tolerance.

[0024] Please see Figure 3 The specific steps of S2 are as follows: S201: Obtain the preset thickness change limit, collect the thickness difference between adjacent coordinates, read the value of each thickness difference between adjacent coordinates, perform a size comparison judgment between the read value and the thickness change limit, peel off the corresponding node when the read value is greater than the thickness change limit, extract the corresponding node when the read value is not greater than the thickness change limit, aggregate the corresponding nodes, and establish a candidate force coordinate set. The thickness mutation limit is obtained by acquiring the root thickness parameter and edge thickness parameter of the insulator standard sample, performing a difference operation on the two to calculate the global thickness difference, simultaneously acquiring the outer edge radius parameter and root radius parameter of the insulator standard sample, performing a difference operation on the two to calculate the effective radial span value; dividing the global thickness difference by the effective radial span value to calculate the base thickness change rate, acquiring the deformation tolerance coefficient corresponding to the flexible pad material of the gripper, performing a multiplication operation on the base thickness change rate and the deformation tolerance coefficient for numerical correction, and establishing the corrected output result as the thickness mutation limit; The process involves acquiring a preset thickness mutation limit from the storage medium, collecting the thickness difference between adjacent coordinates generated in the previous stage, and reading the values ​​for each adjacent coordinate thickness difference. For each read difference, the absolute value is compared with the thickness mutation limit. If the absolute value of the current read value is greater than the thickness mutation limit, the corresponding coordinate node is stripped. If the absolute value is not greater than the thickness mutation limit, the corresponding node is extracted and retained. All nodes that meet the "not greater than" condition are aggregated and assembled to establish a candidate set of force coordinates capable of withstanding clamping forces. The thickness mutation limit is acquired using the following procedure: the root thickness parameter 32 and edge thickness parameter 14 of the insulator standard sample are obtained from the factory quality control standard. A difference calculation is performed between the two, expressed as 32-14=18, resulting in a global thickness difference of 18 mm. Simultaneously, the outer edge radius parameter 152 and root radius parameter 58 of the insulator standard sample are acquired, and a difference calculation is performed between them, expressed as 152-58=94, resulting in an effective radial span of 94 mm. The global thickness difference of 18 is divided by the effective radial span value of 94. This calculation process is expressed as 18 / 94=0.19, yielding a base thickness change rate of 0.19. The deformation tolerance coefficient 15 corresponding to the flexible padding material of the gripper is obtained from the material property library. The base thickness change rate of 0.19 and the deformation tolerance coefficient 15 are multiplied for numerical amplification correction. This calculation process is expressed as 0.19*15=2.85. The corrected output result of 2.85 mm is established as the thickness mutation limit under the current working condition. The advantage of this calculation logic is that it fully integrates the deformation capacity of the padding material with the inherent geometric curvature of the standard product, obtaining a constraint threshold that highly matches the physical clamping characteristics.

[0025] S202: Obtain the insulator axis coordinate data, extract the internal node coordinate components for the candidate force coordinate set, perform subtraction operation on the corresponding components of the node coordinate components and the insulator axis coordinate data to obtain the coordinate deviation, perform square operation on the coordinate deviation and perform summation, perform square root operation on the summation value to obtain the radial distance, and generate the candidate area distance sequence. The insulator axis coordinate data sent via the visual positioning interface includes planar horizontal and vertical parameters of 100 and 200. For the established candidate force coordinate set, the coordinate components of all internal nodes are extracted, including horizontal and vertical axis position data. Subtraction is performed one by one between the horizontal and vertical axis components of each node and the corresponding planar components of the insulator axis coordinate data. The extracted node's horizontal coordinate is set to 140 and its vertical coordinate to 230. 140 is subtracted from the axis horizontal parameter 100; this calculation is expressed as 140 - 100 = 40. 230 is subtracted from the axis vertical parameter 200; this calculation is expressed as 230 - 200 = 30. This yields a horizontal axis coordinate deviation of 40 and a vertical axis coordinate deviation of 30. The two extracted coordinate deviations are squared separately, as expressed by the formulas 40*40=1600 and 30*30=900. These two deviations are then summed, resulting in the formula 1600+900=2500. The summed value of 2500 is then squared to obtain the square root as the linear spatial span, expressed by the formula 2500^0.5=50. This gives the radial distance of the current node from the axis center as 50 mm. This process of subtraction, squaring, summing, and square root operations is repeated for all nodes within the candidate force coordinate set. All radial distance values ​​are then summed to generate a structured sequence of candidate region distances.

[0026] S203: Call the candidate force coordinate set, perform ascending sorting of the distributed values ​​in the distance sequence of the candidate area to obtain the distance order parameter, extract the first and second minimum values ​​based on the distance order parameter, perform reverse index matching in the candidate force coordinate set based on the minimum value, locate the node information associated with the minimum value, extract the horizontal and vertical parameters corresponding to the node information as the action execution point, and establish the gripping coordinates of the robot arm; The system retrieves a candidate force coordinate set containing the coordinates and parameters of each node. It then performs a bubble sort process on the distributed values ​​within the generated candidate distance sequence, placing the smallest value at the beginning and proceeding sequentially to the largest, thus obtaining the distance order parameter reflecting the proximity relationship. Based on the sorted distance order parameter, it directly extracts the value 50, which is the first minimum value in the sorted sequence. Using this minimum value 50, it performs a numerical equivalence check and reverse index matching within the candidate force coordinate set to locate the node information uniquely corresponding to the minimum value at distance 50. Based on this node information, it extracts the corresponding horizontal parameter 140 and vertical parameter 230, establishing these two-dimensional coordinate parameters as the action execution points of the device endpoints. These coordinates are then converted into a spatial vector format to establish the robotic arm gripping coordinates that guide the device control program.

[0027] Please see Figure 4 The specific steps of S3 are as follows: S301: Collect the coordinates of the porcelain cap end face and the glass umbrella skirt. Extract the horizontal and vertical axis components of the porcelain cap end face coordinates. Extract the horizontal and vertical axis components of the glass umbrella skirt coordinates. Perform subtraction calculation on the horizontal axis components of the end face and the horizontal axis components of the umbrella skirt to obtain the horizontal axis distance. Perform subtraction calculation on the vertical axis components of the end face and the vertical axis components of the umbrella skirt to obtain the vertical axis distance. Associate the horizontal axis and vertical axis distances to generate a two-dimensional deviation set. The end-point camera is controlled to activate the exposure action, acquiring the coordinate values ​​300, 400, and 150 of the upper ceramic cap end face, and simultaneously identifying and acquiring the coordinate values ​​302, 395, and 100 of the lower glass umbrella skirt. The horizontal axis component 300 and the vertical axis component 400 of the ceramic cap end face coordinates are extracted, and the horizontal axis component 302 and the vertical axis component 395 of the glass umbrella skirt coordinates are extracted. A subtraction calculation is performed between the horizontal axis component 300 and the horizontal axis component 302 of the umbrella skirt, expressed as 300 - 302 = -2, yielding a horizontal distance of -2 mm. A subtraction calculation is performed between the vertical axis component 400 and the vertical axis component 395 of the umbrella skirt, expressed as 400 - 395 = 5, yielding a vertical distance of 5 mm. These two values, -2 and 5, are extracted and correlated as geometric offset state characterization parameters for the planar projection, generating a two-dimensional deviation set.

[0028] S302: Obtain the end rotation angle and the gripper centerline angle, extract the horizontal axis distance and vertical axis distance from the two-dimensional deviation set, perform sign determination on the horizontal axis distance and vertical axis distance to obtain quadrant assignment, combine quadrant assignment with the arctangent operation of the distance ratio to obtain the original offset angle, perform summation calculation on the original offset angle, end rotation angle and gripper centerline angle to generate the actual workpiece offset angle; By reading the encoder register value of the servo motor at the current moment, the end-effector rotation angle is obtained as 10 degrees. Simultaneously, the factory-preset gripper centerline angle of the gripper mechanical component is read as 5 degrees. The horizontal axis distance of -2 and the vertical axis distance of 5 are extracted from the established two-dimensional deviation set. A sign determination is performed on the horizontal axis distance of -2 and the vertical axis distance of 5. Since the horizontal axis distance is negative and the vertical axis distance is positive, the plane quadrant is determined to belong to the second quadrant. Combining the second quadrant assignment characteristics and the absolute ratio of the vertical axis distance of 5 to the horizontal axis distance of -2, an arctangent angle calculation is performed. Since the calculated acute angle is 68 degrees, it is converted to the second quadrant. This calculation process is expressed by the formula 180-68=112, obtaining the original offset angle as 112 degrees. An arithmetic summation is performed on the original offset angle 112, the end rotation angle 10, and the gripper centerline angle 5. This calculation is expressed by the formula 112 + 10 + 5 = 127. By combining various attitude deviations, the actual workpiece offset angle of 127 degrees, reflecting the current actual attitude, is generated. The advantage of this calculation logic is that it accurately maps spatial displacement deviations to the required angular deflection of the mechanical actuator axis.

[0029] S303: Obtain the preset fixture positioning reference angle, subtract the actual workpiece offset angle from the fixture positioning reference angle to obtain the angle compensation difference, set the rotation limit range for the angle compensation difference, perform a limit comparison between the angle compensation difference and the rotation limit range, perform a direction reversal operation on the angle compensation difference based on the comparison result, and establish the end rotation correction angle. The method for obtaining the fixture positioning reference angle is as follows: before the robot arm performs the transfer task, the vision camera fixed above the workstation is driven to acquire a top view image of the unloaded fixture. Edge contour detection operation is performed on the image to extract the geometric feature edge lines of the mechanical positioning groove or mechanical stop on the surface of the fixture. Based on the geometric feature edge lines, the central symmetry axis vector of the fixture positioning groove is generated by fitting. The central symmetry axis vector is mapped to the global reference coordinate system of the robot arm base. The angle between the central symmetry axis vector and the horizontal reference axis of the global coordinate system is calculated. The obtained angle parameter is recorded as the fixture positioning reference angle. The database is used to retrieve a preset fixture positioning reference angle of 90 degrees. The actual workpiece offset angle of 127 generated in the previous operation is directly subtracted from the fixture positioning reference angle of 90 degrees. This calculation is expressed as 127 - 90 = 37, yielding an angle compensation difference of 37 degrees for positive errors. The physical rotation limit range of the motor is set to -180 degrees to +180 degrees for this angle compensation difference of 37. A limit comparison is performed between the angle compensation difference of 37 and this rotation limit range. If 37 falls within the valid range, a numerical sign is directly assigned based on the comparison result. If it exceeds 180 degrees, a direction reversal operation is performed by subtracting 360. If it does not exceed the limit, the original value is maintained, establishing an end-rotation correction angle of 37 degrees. The method for obtaining the fixture positioning reference angle is as follows: Before the robot arm performs the transfer task, a vision camera fixed above the workstation is driven to acquire top-view image data of the fixture in an unloaded state. The Sobel edge contour detection operator is performed on this image to obtain gradient extrema points, and the histogram geometric feature edge line pixels of the mechanical positioning grooves or mechanical stops on the fixture surface are extracted. Based on the geometric feature edge line pixels, a least-squares linear fitting operation is performed to generate a central symmetry axis vector representing the direction of the fixture positioning groove. This central symmetry axis vector is mapped to the global reference coordinate system of the robot arm base through coordinate system translation and rotation operations. The plane angle between the central symmetry axis vector and the horizontal reference X-axis of the global coordinate system is calculated, and this angle parameter is persistently recorded as the fixture positioning reference angle of 90 degrees.

[0030] Please see Figure 5 The specific steps of S4 are as follows: S401: Call the end-effector rotation correction angle, obtain the current end-effector rotation angle, extract the origin rotation component within the current end-effector rotation angle, convert the end-effector rotation correction angle into a numerical compensation value, perform an addition operation on the origin rotation component and the numerical compensation value to obtain the attitude sum, perform inverse coordinate calculation on the attitude sum to obtain the joint parameters, and establish end-effector rotation zero position data; The established end-effector rotation correction angle of 37 degrees is invoked. The end-effector rotation angle of 10 degrees within the current cycle is obtained via the communication bus, and the origin rotation component of 10 degrees within the current end-effector rotation angle is extracted. The end-effector rotation correction angle of 37 is directly converted into a numerical compensation amount of equal amplitude 37. A pure algebraic addition operation is performed on the origin rotation component 10 and the numerical compensation amount 37. This calculation process is expressed by the formula 10 + 37 = 47, obtaining the attitude sum value reflecting the overall absolute attitude as 47 degrees. The structural length parameter matrix of the robot arm's forward kinematics model is extracted. A coordinate inverse solution operation based on the Jacobian matrix is ​​performed on the attitude sum value 47 to obtain the absolute pulse commands and joint parameters required for each drive joint. Based on the solution results, end-effector rotation zero-position data guiding the end-effector back to a safe initial state is established. The advantage of this operational logic is that it quickly decomposes the spatial tilt angle deviation into independent feed control commands for each rotary joint.

[0031] S402: Call the end-effector rotation zero-position data and the robot gripping coordinates, extract the three-dimensional translation component within the robot gripping coordinates, extract the rotation component within the end-effector rotation zero-position data, map the three-dimensional translation component to the translation area of ​​the transformation matrix, map the rotation component to the rotation area of ​​the transformation matrix, and perform splicing on the translation area and the rotation area to generate robot gripping posture data. The process involves retrieving the acquired end-effector rotation zero-position data and referencing the previously established manipulator gripping coordinates (x-coordinate 140, y-coordinate 230). Three-dimensional translation components within the manipulator gripping coordinates are extracted, and a Z-axis constant of 150 in the height direction is added to the two-dimensional data, resulting in translation components 140, 230, and 150. Three rotational components (0, 0, and 47) containing yaw, pitch, and roll are extracted from the end-effector rotation zero-position data. The three-dimensional translational components 140, 230, and 150 are directly mapped to the first three rows of the fourth column of a standard fourth-order transformation matrix. The rotational components 0, 0, and 47 are converted into cosine and sine numerical matrices and then mapped to the upper left 3x3 rotational region of the same transformation matrix. Data concatenation is performed on the translation and rotational matrix elements, filling the fourth row of the matrix with general constants 0, 0, 0, and 1, generating complete manipulator gripping posture data to guide the six-DOF manipulator to the specified pose.

[0032] Table 2. Robotic Arm Posture Integration Data Table Translation component region X-axis translation offset 140 millimeters Translation component region Y-axis translation offset 230 millimeters Translation component region Z-axis translation offset 150 millimeters Rotational component region X-axis roll angle 0 Spend Rotational component region Y-axis pitch angle 0 Spend Rotational component region Z-axis yaw angle 47 Spend As shown in Table 2, the three-dimensional positions and Euler angle parameters of each region filled into the transformation matrix are summarized in detail.

[0033] S403: Based on the gripping posture data of the robotic arm, obtain the jig release time, the receiving permission time, the deceleration and stop distance parameters, the suspension allowable duration and the waiting coordinate data. Perform a subtraction operation on the receiving permission time and the jig release time to obtain the initial time difference. Perform a multiplication conversion on the initial time difference and the clock frequency to obtain the span duration. Perform sign verification on the span duration to remove negative values ​​and generate the cross-station time difference. Based on the constructed robotic arm gripping posture data, the simulation determines the action. It acquires the fixture release time (500 milliseconds) reported by the lower-level machine in the current process flow, the receive permission time (800 milliseconds) issued by the main control module, the deceleration stop distance parameter (30 mm) set by the driver, the built-in suspension allowable duration parameter (400 milliseconds), and the waiting coordinate data (300, 300, and 200) of the buffer area. A subtraction operation is performed between the receive permission time (800) and the fixture release time (500), expressed as 800 - 500 = 300, yielding an initial time difference of 300. The clock frequency conversion factor 1 is extracted, and a multiplication operation is performed between the initial time difference (300) and the clock frequency 1, expressed as 300 * 1 = 300, yielding the actual waiting period spanning 300 milliseconds. Perform a sign verification operation on the time span of 300, determine that 300 is a positive number to eliminate possible negative interference items, and generate a cross-station time difference of 300 milliseconds to control the cycle time of the process.

[0034] Please see Figure 6 The specific steps of S5 are as follows: S501: Obtain the preset allowable suspension time of the equipment, perform subtraction calculation between the cross-station time difference and the allowable suspension time to obtain the time surplus, perform positive and negative sign extraction and judgment on the time surplus value, match the preset logical identifier according to the extracted status parameters, and establish the time difference comparison result. The specific method for setting the allowable suspension time is as follows: obtain the initial clamping force data of the clamping mechanism in the grasping state and the force attenuation coefficient of the mechanism in the continuous pressure holding state; read the critical safe force threshold at which the insulator overcomes the friction of the gripper finger surface to generate relative slippage; perform a subtraction operation on the initial clamping force data and the critical safe force threshold to obtain the clamping redundancy; perform a division operation on the clamping redundancy and the force attenuation coefficient to obtain the theoretical limit holding time; perform a multiplication conversion on the theoretical limit holding time and the preset safety loss ratio; and determine the allowable suspension time based on the converted time span. The device control cabinet's preset suspension allowance of 400 milliseconds is obtained. A subtraction calculation is performed between the obtained cross-station time difference of 300 and the suspension allowance of 400, expressed as 300 - 400 = -100, yielding a time surplus of -100 milliseconds. A sign extraction and determination operation is performed on the time surplus value of -100, extracting its state parameter as negative polarity. Based on the machine intelligence state machine scheduling rules, the extracted negative polarity state parameter is matched with the corresponding preset logical identifier B in the configuration table. If the polarity is positive, logical identifier A is matched. Thus, the time difference comparison result for the current cycle scheduling is identified by identifier B. The aforementioned setting of the suspension allowance of 400 milliseconds is specifically executed as follows: the initial clamping force data of 80 Newtons in the gripping state is obtained through a pressure sensor. The force attenuation coefficient of 0.1 Newtons per millisecond in the continuous pressure holding state is extracted through force gauge attenuation curve fitting. The critical safe force threshold of 30 Newtons for the insulator to overcome the frictional force of the gripper fingers and achieve relative slippage is read from the material testing document. The initial clamping force data 80 is subtracted from the critical safe force threshold 30, expressed as 80-30=50, yielding a usable clamping redundancy of 50 Newtons. The clamping redundancy of 50 is then divided by the force attenuation coefficient 0.1, expressed as 50 / 0.1=500, resulting in a theoretical limit holding time of 500 milliseconds under this condition. A preset safety attenuation ratio of 0.8 is extracted, and the theoretical limit holding time of 500 is multiplied by the preset safety attenuation ratio 0.8, expressed as 500*0.8=400, establishing a safe holding allowance of 400 milliseconds based on the converted value. The advantage of this calculation logic is that it transforms pure time waiting into a calculation of the safety margin in the dynamic mechanics of clamping, reducing the risk of the workpiece falling midway.

[0035] S502: Call the time difference comparison result, collect the lateral trajectory parameters, deceleration waiting coordinates and in-situ holding coordinates issued by the motion unit, and perform branch route selection based on the built-in logical identifier of the time difference comparison result. When the identifier indicates a positive direction, extract the lateral trajectory parameters and deceleration waiting coordinates and perform spatial vector concatenation. When the identifier indicates a negative direction, extract the in-situ holding coordinates and perform coordinate node copying. Map the concatenated or copied node values ​​to the motion sequence matrix to establish the target movement trajectory parameters. The system invokes the established time difference comparison result identifier B, and collects the lateral trajectory parameter sequence set issued by the motion unit through the communication port. It collects the deceleration and waiting coordinates 300, 300, and 200 for transition, and the current stationary coordinates 140, 230, and 150 for the robot arm. Based on the logical identifier B built into the time difference comparison result, a branch path selection operation is performed. If identifier B indicates a negative attribute, the stationary coordinates 140, 230, and 150 are extracted and a coordinate node copying operation is performed to generate a series of equal stationary coordinate points. When the identifier indicates a positive attribute, the lateral trajectory parameters and deceleration and waiting coordinates are extracted and concatenated into a three-dimensional spatial vector. The copied and generated equal stationary coordinate node values ​​are directly assigned and mapped into the motion sequence matrix of the underlying servo drive, thereby establishing the target movement trajectory parameters to avoid premature movement.

[0036] S503: Call the target movement trajectory parameters, obtain the deceleration and stopping distance parameters and waiting coordinate data, extract the terminal node coordinates inside the target movement trajectory parameters, extract the approach direction vector inside the waiting coordinate data, perform a quantity product operation on the direction vector and the deceleration and stopping distance parameters to obtain the spatial offset compensation vector, perform coordinate addition processing on the terminal node coordinates and the spatial offset compensation vector to obtain the corrected pose coordinates, replace the original terminal node according to the corrected pose coordinates, and generate the robot arm transfer control data; The generated target trajectory parameter set is invoked to obtain the deceleration and stopping distance parameter 30 from the preceding process, and to obtain the waiting coordinate data containing the transition direction vector. The coordinates of the last terminal node (500, 500, and 200) within the target trajectory parameters are extracted, and the unit values ​​of the approach direction vector pointing to the placement platform (0, 1, and 0) within the waiting coordinate data are extracted. A three-dimensional multiplication operation is performed on each element of the approach direction vector unit values ​​0, 1, and 0 with the deceleration and stopping distance parameter 30. This calculation process is expressed by the formulas 0*30=0, 1*30=30, and 0*30=0, obtaining the spatial offset compensation vectors 0, 30, and 0. The terminal node coordinates 500, 500, and 200 are added to the generated spatial offset compensation vectors 0, 30, and 0 using component-dimensional coordinate addition. This calculation process is expressed by the formulas 500+0=500, 500+30=530, and 200+0=200, obtaining the compensated corrected pose coordinates 500, 530, and 200. Based on the generated corrected pose coordinates 500, 530, and 200, the terminal nodes 500, 500, and 200 at the end of the original trajectory sequence are replaced, and the actions are sent out to generate robot transport control data that guides the robot to achieve smooth transfer and placement interactions. The advantage of this operation logic is that it retreats a certain safe deceleration distance along the normal direction before the target endpoint, reducing the impact of rigid collisions.

[0037] Please see Figure 7 A control system for a glass insulator processing robot includes: The contour parameter reconstruction module acquires insulator outer edge radius data, insulator root radius data, insulator edge thickness data, insulator root thickness data, gripper finger width data, and closed target stroke data. Based on the insulator outer edge radius data and insulator root radius data, it constructs an intermediate radial coordinate sequence and generates the thickness difference between adjacent coordinates through a cubic spline interpolation function. The clamping area determination module obtains the thickness change limit, compares the thickness difference between adjacent coordinates with the thickness change limit, filters the avoidance area coordinate data and candidate stress area coordinate data, and combines the insulator axis coordinate data to generate candidate area distance values ​​through the Euclidean distance function to construct the manipulator clamping coordinates. The end-point attitude calibration module obtains the coordinates of the ceramic cap end face, the coordinates of the glass umbrella skirt, the current end-point rotation angle, the fixture positioning reference angle, and the gripper centerline angle. It generates the lateral deviation and longitudinal deviation of the ceramic cap and compares them with the fixture positioning reference angle to construct the end-point rotation correction angle. The grasping posture generation module adjusts the current end-effector rotation angle based on the end-effector rotation correction angle, generates end-effector rotation zero-position data, and constructs robot gripping posture data by combining it with robot gripping coordinates. It also obtains jig release time, permission reception time, deceleration stop distance parameters, suspension allowable duration and waiting coordinate data, and calculates cross-station time difference. The transfer cycle control module, based on the cross-workstation time difference, combines the lateral trajectory parameters, deceleration waiting coordinates, and in-situ holding coordinates to filter the target movement trajectory parameters, and combines the deceleration stopping distance parameters and waiting coordinate data to generate the robot transfer control data.

[0038] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of protection of the described technical solutions.

Claims

1. A control method for a robotic arm used in glass insulator processing, characterized in that, Includes the following steps: S1: Obtain insulator outer edge radius data, insulator root radius data, insulator edge thickness data, insulator root thickness data, gripper finger width data, and closed target stroke data. Construct an intermediate radial coordinate sequence based on the insulator outer edge radius data and insulator root radius data, and generate adjacent coordinate thickness differences through a cubic spline interpolation function. S2: Obtain the thickness mutation limit, compare the thickness difference between adjacent coordinates with the thickness mutation limit, filter the avoidance zone coordinate data and candidate stress zone coordinate data, combine the insulator axis coordinate data with the Euclidean distance function to generate candidate zone distance values, and construct the manipulator gripping coordinates; S3: Obtain the coordinates of the ceramic cap end face, the coordinates of the glass umbrella skirt, the current end rotation angle, the fixture positioning reference angle, and the gripper centerline angle; generate the lateral deviation and longitudinal deviation of the ceramic cap; compare them with the fixture positioning reference angle; and construct the end rotation correction angle. S4: Adjust the current end rotation angle based on the end rotation correction angle, generate end rotation zero position data, combine with the robot gripping coordinates to construct robot gripping posture data, obtain fixture release time, permission receiving time, deceleration stop distance parameters, suspension allowable time and waiting coordinate data, and calculate cross-station time difference; S5: Based on the cross-station time difference, the target movement trajectory parameters are filtered by combining the lateral movement trajectory parameters, deceleration waiting coordinates and in-situ holding coordinates. The robot arm transfer control data is generated by combining the deceleration stopping distance parameters and waiting coordinate data.

2. The control method for a glass insulator processing robot according to claim 1, characterized in that, The specific steps of S1 are as follows: S101: Obtain insulator outer edge radius data, insulator root radius data, insulator edge thickness data, insulator root thickness data, gripper finger width data, and closed target stroke data. Perform subtraction calculation based on the insulator outer edge radius data and insulator root radius data to obtain the radial span. Set the offset length according to the radial span. Divide the interval where the insulator root radius data is located into equal intervals according to the offset length to establish an intermediate radial coordinate sequence. S102: Call the intermediate radial coordinate sequence, insulator edge thickness data and insulator root thickness data, perform pairing with the nodes in the intermediate radial coordinate sequence and the insulator edge thickness data and insulator root thickness data to obtain the coordinate correlation matrix, input the coordinate correlation matrix into the cubic spline interpolation function to perform fitting operation to calculate the thickness value, and summarize the thickness values ​​to obtain the radial thickness distribution set; S103: Call the gripper finger width data, closed target stroke data and radial thickness distribution set, extract the displacement based on the gripper finger width data, select adjacent items in the radial thickness distribution set according to the displacement and perform subtraction to obtain the initial fluctuation difference, set boundary conditions for the closed target stroke data, filter out out-of-bounds values ​​based on the boundary conditions and generate adjacent coordinate thickness difference values.

3. The control method for a glass insulator processing robot according to claim 1, characterized in that, The specific steps of S2 are as follows: S201: Obtain the preset thickness change limit, collect the thickness difference between adjacent coordinates, read the value of each thickness difference between adjacent coordinates, perform a size comparison judgment between the read value and the thickness change limit, peel off the corresponding node when the read value is greater than the thickness change limit, extract the corresponding node when the read value is not greater than the thickness change limit, aggregate the corresponding nodes, and establish a candidate force coordinate set. S202: Obtain the insulator axis coordinate data, extract the internal node coordinate components for the candidate force coordinate set, perform subtraction operation on the node coordinate components and the corresponding components of the insulator axis coordinate data to obtain the coordinate deviation, perform square operation on the coordinate deviation and perform summation, perform square root operation on the summation value to obtain the radial distance, and generate a candidate area distance sequence. S203: Call the candidate force coordinate set, perform ascending sorting processing on the distributed values ​​in the distance sequence of the candidate area to obtain the distance order parameter, extract the first and second minimum values ​​according to the distance order parameter, perform reverse index matching in the candidate force coordinate set according to the minimum value, locate the node information associated with the minimum value, extract the horizontal and vertical parameters corresponding to the node information as the action execution point, and establish the gripping coordinates of the robot arm.

4. The control method for a glass insulator processing robot according to claim 3, characterized in that, The thickness abrupt change limit is obtained by acquiring the root thickness parameter and edge thickness parameter of the insulator standard sample, performing a difference operation on the two to calculate the global thickness difference, and simultaneously acquiring the outer edge radius parameter and root radius parameter of the insulator standard sample, performing a difference operation on the two to calculate the effective radial span value. Divide the global thickness difference by the effective radial span value to calculate the base thickness change rate, obtain the deformation tolerance coefficient corresponding to the flexible pad material of the gripper, perform a multiplication operation on the base thickness change rate and the deformation tolerance coefficient to perform numerical correction, and establish the corrected output as the thickness mutation limit.

5. The control method for a glass insulator processing robot according to claim 1, characterized in that, The specific steps for S3 are as follows: S301: Collect the coordinates of the porcelain cap end face and the glass umbrella skirt. Extract the horizontal and vertical axis components of the porcelain cap end face coordinates. Extract the horizontal and vertical axis components of the glass umbrella skirt coordinates. Perform subtraction calculation on the horizontal axis components of the end face and the horizontal axis components of the umbrella skirt to obtain the horizontal axis distance. Perform subtraction calculation on the vertical axis components of the end face and the vertical axis components of the umbrella skirt to obtain the vertical axis distance. Associate the horizontal axis and vertical axis distances to generate a two-dimensional deviation set. S302: Obtain the end rotation angle and the gripper centerline angle; extract the horizontal axis distance and vertical axis distance from the two-dimensional deviation set; perform sign determination on the horizontal axis distance and vertical axis distance to obtain quadrant assignment; combine quadrant assignment with the arctangent operation of the distance ratio to obtain the original offset angle; perform summation calculation on the original offset angle, the end rotation angle, and the gripper centerline angle to generate the actual workpiece offset angle. S303: Obtain the preset fixture positioning reference angle, subtract the actual offset angle of the workpiece from the fixture positioning reference angle to obtain the angle compensation difference, set a rotation limit range for the angle compensation difference, perform a limit comparison between the angle compensation difference and the rotation limit range, perform a direction reversal operation on the angle compensation difference based on the comparison result, and establish the end rotation correction angle.

6. The control method for a glass insulator processing robot according to claim 5, characterized in that, The method for obtaining the fixture positioning reference angle is as follows: before the robot arm performs the transfer task, a vision camera fixed above the workstation is driven to acquire a top view image of the unloaded fixture. Edge contour detection is performed on the image to extract the geometric feature edge lines of the mechanical positioning groove or mechanical stop on the surface of the fixture. Based on the geometric feature edge lines, a central symmetry axis vector of the fixture positioning groove is generated. The central symmetry axis vector is mapped to the global reference coordinate system of the robot arm base. The angle between the central symmetry axis vector and the horizontal reference axis of the global coordinate system is calculated. The obtained angle parameter is recorded as the fixture positioning reference angle.

7. The control method for a glass insulator processing robot according to claim 1, characterized in that, The specific steps of S4 are as follows: S401: Call the end-effector rotation correction angle, obtain the current end-effector rotation angle, extract the origin rotation component within the current end-effector rotation angle, convert the end-effector rotation correction angle into a numerical compensation amount, perform an addition operation on the origin rotation component and the numerical compensation amount to obtain the attitude sum, perform inverse coordinate calculation on the attitude sum to obtain the joint parameters, and establish end-effector rotation zero position data; S402: Call the end-effector rotation zero-position data and the robot gripping coordinates, extract the three-dimensional translation component in the robot gripping coordinates, extract the rotation component in the end-effector rotation zero-position data, map the three-dimensional translation component to the translation area of ​​the transformation matrix, map the rotation component to the rotation area of ​​the transformation matrix, and perform splicing on the translation area and the rotation area to generate robot gripping posture data; S403: Based on the gripping posture data of the robotic arm, obtain the jig release time, the receiving permission time, the deceleration and stop distance parameters, the suspension allowable duration and the waiting coordinate data. Perform a subtraction operation on the receiving permission time and the jig release time to obtain the initial time difference. Perform a multiplication conversion on the initial time difference and the clock frequency to obtain the span duration. Perform sign verification on the span duration to remove negative values ​​and generate the cross-station time difference.

8. The control method for a glass insulator processing robot according to claim 1, characterized in that, The specific steps of S5 are as follows: S501: Obtain the preset allowable suspension time of the equipment, perform subtraction calculation on the cross-workstation time difference and the allowable suspension time to obtain the time surplus, perform positive and negative sign extraction and judgment on the time surplus value, match the preset logical identifier according to the extracted state parameters, and establish the time difference comparison result. S502: Call the time difference comparison result, collect the lateral trajectory parameters, deceleration waiting coordinates and in-situ holding coordinates issued by the motion unit, perform branch route selection based on the built-in logical identifier of the time difference comparison result, extract the lateral trajectory parameters and deceleration waiting coordinates and perform spatial vector concatenation when the identifier represents a positive direction, extract the in-situ holding coordinates and perform coordinate node copying when the identifier represents a negative direction, map the concatenated or copied node values ​​to the motion sequence matrix, and establish the target movement trajectory parameters; S503: Call the target movement trajectory parameters, obtain the deceleration and stopping distance parameters and waiting coordinate data, extract the terminal node coordinates inside the target movement trajectory parameters, extract the approach direction vector inside the waiting coordinate data, perform a quantity product operation on the direction vector and the deceleration and stopping distance parameters to obtain the spatial offset compensation vector, perform coordinate addition processing on the terminal node coordinates and the spatial offset compensation vector to obtain the corrected pose coordinates, replace the original terminal node according to the corrected pose coordinates, and generate the robot arm transfer control data.

9. The control method for a glass insulator processing robot according to claim 8, characterized in that, The specific method for setting the allowable suspension time is as follows: obtain the initial clamping force data of the clamping mechanism in the grasping state and the force attenuation coefficient of the mechanism in the continuous pressure holding state; read the critical safe force threshold at which the insulator overcomes the friction of the gripper finger surface to generate relative slippage; perform a subtraction operation on the initial clamping force data and the critical safe force threshold to obtain the clamping redundancy; perform a division operation on the clamping redundancy and the force attenuation coefficient to obtain the theoretical limit holding time; perform a multiplication conversion on the theoretical limit holding time and the preset safety loss ratio; and determine the allowable suspension time based on the converted time span.

10. A control system for a glass insulator processing robot, characterized in that, The system is used to implement the control method for a glass insulator processing robot according to any one of claims 1-9, the system comprising: The contour parameter reconstruction module acquires insulator outer edge radius data, insulator root radius data, insulator edge thickness data, insulator root thickness data, gripper finger width data, and closed target stroke data. Based on the insulator outer edge radius data and insulator root radius data, it constructs an intermediate radial coordinate sequence and generates the thickness difference between adjacent coordinates through a cubic spline interpolation function. The clamping area determination module obtains the thickness change limit, compares the thickness difference between adjacent coordinates with the thickness change limit, filters the avoidance area coordinate data and candidate stress area coordinate data, and generates candidate area distance values ​​by combining the insulator axis coordinate data with the Euclidean distance function to construct the manipulator clamping coordinates. The end-point attitude calibration module obtains the coordinates of the ceramic cap end face, the coordinates of the glass umbrella skirt, the current end-point rotation angle, the fixture positioning reference angle, and the gripper centerline angle. It generates the lateral deviation and longitudinal deviation of the ceramic cap and compares them with the fixture positioning reference angle to construct the end-point rotation correction angle. The grasping posture generation module adjusts the current end rotation angle based on the end rotation correction angle, generates end rotation zero position data, and constructs robot gripping posture data by combining the robot gripping coordinates. It also obtains jig release time, permission receiving time, deceleration stop distance parameters, suspension allowable time and waiting coordinate data, and calculates cross-station time difference. The transfer cycle control module, based on the cross-workstation time difference, combines the lateral trajectory parameters, deceleration waiting coordinates, and in-situ holding coordinates to filter the target movement trajectory parameters, and combines the deceleration stopping distance parameters and waiting coordinate data to generate robot transfer control data.

Images