A three-point calibration-based automatic calculation method for a gripping center
By using a three-point calibration method to automatically calculate the horizontal coordinates and attitude angle of the robot's gripping center, the calibration problem when the material size changes is solved, achieving efficient and low-cost material adaptation and improving production efficiency and equipment stability.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- 伯朗特机器人股份有限公司
- Filing Date
- 2026-04-30
- Publication Date
- 2026-06-09
AI Technical Summary
In existing technologies, when the size of materials changes during robotic gripping operations, the gripping points need to be frequently recalibrated manually, resulting in long production line downtime, low changeover efficiency, and difficulty in adapting to the needs of multi-variety, small-batch production.
An automatic calculation method for the gripping center based on three-point calibration is adopted. By collecting the planar coordinates of three reference points, constructing and normalizing the direction vector, and combining the material length and width parameters, the horizontal coordinates and attitude angle of the gripping center are automatically calculated. The reference point coordinates are collected and the material size parameters are input only during the initial deployment. When the material is changed, the height parameter is updated to obtain the complete three-dimensional coordinates of the gripping center.
It significantly reduces the calibration workload and manual intervention during production changeovers, improves production efficiency, reduces reliance on operator skills and visual alignment, adapts to the calibration needs of materials of different sizes, and has a simple calculation method, low cost, and is easy to implement on ordinary industrial robot controllers.
Smart Images
Figure CN122165430A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of industrial automation gripping technology, specifically to an automatic calculation method for gripping center based on three-point calibration. Background Technology
[0002] In industrial automated grasping operations, robots need to accurately obtain the grasping center point and posture angle of materials in order to achieve stable and reliable grasping and placement operations. The calibration accuracy of the grasping center directly affects the grasping success rate, production efficiency, and equipment operational stability.
[0003] Currently, the most common robot gripping and positioning method is the teach-and-playback method: the operator manually moves the robot's end effector to the target gripping point on the material using a teach pendant, records the coordinates of that point, and the robot then repeats the process at that point. However, when the material size changes, the original gripping point is no longer applicable, and the operator must re-teach a new gripping point. This means that every time a different size material is used, the X, Y, and Z coordinates need to be manually recalibrated. This results in long production line downtime, low changeover efficiency, and difficulty in adapting to the demands of frequent changes in material size in multi-variety, small-batch production. Summary of the Invention
[0004] To address the problems existing in the prior art, the purpose of this invention is to provide an automatic calculation method for the gripping center based on three-point calibration, which can reduce the calibration workload when the material size changes, thereby improving the calculation efficiency.
[0005] To achieve the above objectives, the technical solution adopted by the present invention is as follows: An automatic calculation method for the grasping center based on three-point calibration, comprising: Collect the planar coordinates of three reference points; In the robot's base coordinate system, the robot teach pendant sequentially collects the two-dimensional coordinates of three non-collinear points on the material; among them, the first reference point P1 is a preset corner point on the material facing the gripping robot, and its coordinates are recorded. The second reference point P2 is any point on the side along the width direction of the material starting from P1, and its coordinates are recorded. The third reference point P3 is any point on the side along the length of the material starting from P1, and its coordinates are recorded. ; Construct the direction vector and normalize it; Based on the collected reference point coordinates, a width direction vector and a length direction vector are constructed, and then normalized. These are then scaled in conjunction with the input material length parameter L and width parameter W to obtain the scaled width direction vector. and the scaled length and direction vector ; Calculate and store the horizontal coordinates of the grab center; the horizontal coordinates of the grab center are... ; When it is necessary to obtain the 3D coordinates of the grab center, the stored horizontal coordinates will be used. Combined with the height parameter of the current material, we obtain When the material size changes, only the stored material height parameter is updated. Stored horizontal coordinates It remains unchanged.
[0006] The specific steps for constructing and normalizing the direction vector include: (1) Calculation of the width direction vector; Using the first reference point P1 as the reference point, a width direction vector is constructed through coordinate difference calculation to represent the spatial placement posture of the material in the width direction. The specific calculation process is as follows: Width vector calculation: Construct the original width vector using the coordinate difference between the second reference point P2 and the first reference point P1. The expression is:
[0007] Vector magnitude calculation: The magnitude of the width vector is calculated using the following expression:
[0008] Normalized unit vector: Converts the original width vector into a unit vector. The expression is:
[0009] Size scaling vector: Based on the material length parameter of the corresponding workstation, the unit vector is scaled proportionally. The expression is:
[0010] In the formula, L is the material length parameter; (2) Calculation of length direction vector; Using the first reference point P1 as the reference point, a length direction vector is constructed through coordinate difference calculation to represent the spatial orientation of the material along its length. The specific calculation process is as follows: Length vector solution: Construct the original length vector using the coordinate difference between the third reference point P3 and the first reference point P1, expressed as:
[0011] Vector magnitude calculation: The magnitude of a length vector is calculated using the following expression:
[0012] Normalized unit vector: Converts the original length vector into a unit vector, expressed as:
[0013] Size scaling vector: Based on the material width parameter of the corresponding workstation, the unit vector is scaled proportionally. The expression is:
[0014] In the formula, W is the material width parameter.
[0015] The method also includes an attitude angle calculation step: The original rotational attitude of the material is calculated based on the width direction vector, and then processed to generate standardized attitude angles. The original vector radians are then converted to... The attitude angle within the standard range is given by the following formula: .
[0016] The preset corner point is the lower left corner or the upper right corner of the material.
[0017] The acquisition of the planar coordinates was accomplished using a robot teach pendant.
[0018] When the method is applied to a dual-station grasping scenario, the acquisition, construction, calculation, storage and combination steps are executed independently for the left and right stations respectively. Each station independently acquires the coordinates of its three reference points, inputs its own material length and width parameters, and independently stores its own grasping center horizontal coordinates and height parameters.
[0019] By adopting the above scheme, this invention only requires collecting the planar coordinates of three reference points and inputting the length and width parameters of the material at the initial deployment to calculate the fixed horizontal coordinates of the gripping center and the posture angle of the material. These horizontal coordinates remain unchanged in subsequent production. When changing to materials of different sizes, there is no need to re-collect the reference points or recalculate the horizontal coordinates; only the height parameters of the material need to be updated to obtain the complete three-dimensional coordinates of the gripping center. This significantly reduces the calibration workload and manual intervention during production changes, greatly improving production efficiency. Furthermore, this invention has low requirements for the accuracy of the reference point acquisition. The second reference point P2 and the third reference point P3 can be any point on the side without precise alignment of the endpoints. This makes the calibration operation simple and easy, reducing reliance on operator skills and visual alignment. Attached Figure Description
[0020] Figure 1 This is a flowchart of the method of the present invention; Figure 2 Schematic diagram of the positions of the three reference points of this invention Figure 1 ; Figure 3 Schematic diagram of the positions of the three reference points of this invention Figure 2 . Detailed Implementation
[0021] like Figure 1 As shown, this invention discloses an automatic calculation method for the grasping center based on three-point calibration, which includes the following steps: Step 1: Collect the planar coordinates of three reference points.
[0022] In the robot's base coordinate system (or a unified world coordinate system), use the robot teach pendant to sequentially collect the two-dimensional coordinates of three non-collinear points on the material: First reference point P1: A preset corner point (lower left corner) on the material facing the gripping robot (e.g., ... Figure 2 (as shown) or the upper right corner (as shown) Figure 3 (As shown). The operator precisely aligns the tip of the robot's end effector with the corner point and records the coordinates. .
[0023] Second reference point P2: Any point on the side along the width direction of the material, starting from the first reference point P1. The endpoint is not required; it only needs to be on that side. Record the coordinates. .
[0024] Third reference point P3: Starting from the first reference point P1, record the coordinates of any point on the side of the material along its length (perpendicular to the width side). .
[0025] Step 2: Construct the direction vector and normalize it.
[0026] Based on the collected reference point coordinates, a direction vector representing the material placement posture is constructed. The reference point spacing error is eliminated through normalization processing, and the vector is scaled in combination with the material size to achieve accurate matching between the vector posture and the actual material placement. This process is divided into two sub-steps: width and length direction vector calculation.
[0027] (1) Calculation of the width direction vector.
[0028] Using the first reference point P1 as the reference point, a width direction vector is constructed through coordinate difference calculation to represent the spatial placement posture of the material in the width direction. The specific calculation process is as follows: Width vector calculation: Construct the original width vector using the coordinate difference between the second reference point P2 and the first reference point P1. The expression is:
[0029] Vector magnitude calculation: Solving for the magnitude of the width vector reflects the physical distance between reference points. The expression is:
[0030] Normalized unit vector: Converts the original width vector into a unit vector, eliminating calculation errors caused by differences in the spacing between reference points. The expression is:
[0031] Size scaling vector: Based on the material length parameter of the corresponding workstation, the unit vector is scaled proportionally to make the vector length match the actual size of the material. The expression is:
[0032] In the formula, L is the material length parameter.
[0033] (2) Calculation of length direction vector.
[0034] Using the first reference point P1 as the reference point, a length direction vector is constructed through coordinate difference calculation to represent the spatial orientation of the material along its length. The specific calculation process is as follows: Length vector solution: Construct the original length vector using the coordinate difference between the third reference point P3 and the first reference point P1, expressed as:
[0035] Vector magnitude calculation: The magnitude of a length vector is calculated using the following expression:
[0036] Normalized unit vector: Converts the original length vector into a unit vector, eliminating interference from the spacing between reference points. The expression is:
[0037] Size scaling vector: Based on the material width parameter of the corresponding workstation, the unit vector is scaled proportionally. The expression is:
[0038] In the formula, W is the material width parameter.
[0039] By combining vector normalization and size scaling, the systematic error caused by the difference in the spacing of the reference points is completely eliminated, and the scaling vector is completely bound to the actual geometric size of the material, ensuring that the vector direction accurately reflects the actual placement posture of the material. This significantly reduces the accuracy requirements of manual measurement and reference point layout, while also better adapting to the differentiated calibration needs of materials with different aspect ratios and specifications.
[0040] Step 3: Calculate and store the horizontal coordinates of the capture center.
[0041] The horizontal coordinates of the center are calculated using a mathematical model of "baseline point coordinates superimposed with scaled vector half-values". This fully utilizes the accuracy of vector superposition and avoids the accumulation of errors in traditional geometric calculations. The specific calculation formula is as follows:
[0042] All coordinate results are retained to three decimal places to further improve positioning accuracy and meet the micron-level positioning requirements of high-speed grasping in industrial equipment.
[0043] Step 4: Calculate and store the attitude angle.
[0044] To adapt to the control logic of industrial equipment and avoid angle exceeding limits and rotation direction errors, the original rotation posture of the material is calculated based on the width direction vector, and the standardized posture angle is generated after processing. At the same time, the parameters are automatically updated and the closed-loop feedback is completed.
[0045] This invention addresses the differences in angle standards between mathematical calculations and industrial control by designing a multi-level normalization process: "radian calculation - angle conversion - clockwise conversion - interval normalization," which converts the original vector radians into... The attitude angles within the standard range, and the final standard angle formula is:
[0046] No processing is required when the material rotation angle is less than or equal to 180°; when the material rotation angle is greater than 180°, it needs to be converted, that is, the material rotation angle is reduced by 360° to obtain the final rotation angle.
[0047] Step 5: Obtain the three-dimensional coordinates and rotational orientation of the grab center.
[0048] The height parameter of the material is obtained from the system storage module and used as the Z-axis coordinate value of the gripping center. Combined with the horizontal coordinate of the gripping center stored in step 3, the three-dimensional coordinates of the gripping center are obtained. ,in, The height of the material. Obtain the attitude angle stored in step 4 and use it as the rotation attitude for grasping.
[0049] When the material dimensions change, the system storage module updates the material's dimension parameters, specifically the length, width, and height. After the update, the updated material height parameter and the horizontal coordinates of the gripping center are retrieved from the system storage module to obtain the three-dimensional coordinates of the gripping center. ,in, This is the updated material height.
[0050] The attitude angle is obtained from the system storage module and used as the rotational attitude for grasping.
[0051] The method of this invention can be applied to both single-station and dual-station grasping scenarios. In a dual-station scenario, the above steps are executed independently for the left and right stations respectively. Each station independently collects the coordinates of its three reference points, inputs its own material length and width parameters, calculates its own fixed grasping center horizontal coordinates and attitude angle, and stores them independently in the system storage module to avoid positioning deviations caused by mixing parameters.
[0052] This invention only requires collecting the planar coordinates of three reference points during initial deployment and inputting the length and width parameters of the material once. It can then calculate the fixed horizontal coordinates of the gripping center and the material's orientation angle. These horizontal coordinates remain unchanged during subsequent production. When changing to materials of different sizes, there is no need to re-collect the reference points or recalculate the horizontal coordinates; only the material's height parameter needs to be updated to obtain the complete three-dimensional coordinates of the gripping center. This significantly reduces the calibration workload and manual intervention during production changes, greatly improving production efficiency. Furthermore, this invention has low requirements for the accuracy of the reference point acquisition. The second reference point P2 and the third reference point P3 can be any point on the side without precise alignment of the endpoints. This simplifies the calibration operation and reduces reliance on operator skills and visual alignment.
[0053] Furthermore, this invention employs a combination of vector normalization and size scaling to eliminate the interference of reference point spacing errors on the calculation results, ensuring the accuracy of the grasping center coordinates. The posture angles undergo multi-level normalization processing, including radian calculation, angle transformation, clockwise conversion, and interval regularization, enabling direct adaptation to industrial robot control logic and avoiding issues such as angle exceeding limits or directional ambiguity. The entire calculation method involves only basic vector operations and trigonometric functions, eliminating the need for complex iterations, matrix decomposition, or expensive measurement equipment (such as laser trackers and depth cameras). It boasts advantages such as high computational efficiency, low hardware cost, and ease of implementation on ordinary industrial robot controllers.
[0054] The above description is merely an embodiment of the present invention and does not constitute any limitation on the technical scope of the present invention. Therefore, any minor modifications, equivalent changes, and alterations made to the above embodiments based on the technical essence of the present invention shall still fall within the scope of the technical solution of the present invention.
Claims
1. A method for automatically calculating the grasping center based on three-point calibration, characterized in that, include: Collect the planar coordinates of three reference points; In the robot's base coordinate system, the robot teach pendant sequentially collects the two-dimensional coordinates of three non-collinear points on the material; among them, the first reference point P1 is a preset corner point on the material facing the gripping robot, and its coordinates are recorded. The second reference point P2 is any point on the side along the width direction of the material starting from P1, and its coordinates are recorded. The third reference point P3 is any point on the side along the length of the material starting from P1, and its coordinates are recorded. ; Construct the direction vector and normalize it; Based on the collected reference point coordinates, a width direction vector and a length direction vector are constructed, and then normalized. These are then scaled in conjunction with the input material length parameter L and width parameter W to obtain the scaled width direction vector. and the scaled length and direction vector ; Calculate and store the horizontal coordinates of the grab center; the horizontal coordinates of the grab center are... ; When it is necessary to obtain the 3D coordinates of the grab center, the stored horizontal coordinates will be used. Combined with the height parameter of the current material, we obtain When the material size changes, only the stored material height parameter is updated. Stored horizontal coordinates It remains unchanged.
2. The automatic calculation method for the grasping center based on three-point calibration according to claim 1, characterized in that, The specific steps for constructing and normalizing the direction vector include: (1) Calculation of the width direction vector; Using the first reference point P1 as the reference point, a width direction vector is constructed through coordinate difference calculation to represent the spatial placement posture of the material in the width direction. The specific calculation process is as follows: Width vector calculation: Construct the original width vector using the coordinate difference between the second reference point P2 and the first reference point P1. The expression is: Vector magnitude calculation: The magnitude of the width vector is calculated using the following expression: Normalized unit vector: Converts the original width vector into a unit vector. The expression is: Size scaling vector: Based on the material length parameter of the corresponding workstation, the unit vector is scaled proportionally. The expression is: In the formula, L is the material length parameter; (2) Calculation of length direction vector; Using the first reference point P1 as the reference point, a length direction vector is constructed through coordinate difference calculation to represent the spatial orientation of the material along its length. The specific calculation process is as follows: Length vector solution: Construct the original length vector using the coordinate difference between the third reference point P3 and the first reference point P1, expressed as: Vector magnitude calculation: The magnitude of a length vector is calculated using the following expression: Normalized unit vector: Converts the original length vector into a unit vector, expressed as: Size scaling vector: Based on the material width parameter of the corresponding workstation, the unit vector is scaled proportionally. The expression is: In the formula, W is the material width parameter.
3. The automatic calculation method for the grasping center based on three-point calibration according to claim 1, characterized in that, The method also includes an attitude angle calculation step: The original rotational attitude of the material is calculated based on the width direction vector, and then processed to generate standardized attitude angles. The original vector radians are then converted to... The attitude angle within the standard range is given by the formula: 。 4. The automatic calculation method for the grasping center based on three-point calibration according to claim 1, characterized in that, The preset corner point is the lower left corner or the upper right corner of the material.
5. The automatic calculation method for the grasping center based on three-point calibration according to claim 1, characterized in that, The acquisition of the planar coordinates was accomplished using a robot teach pendant.
6. The automatic calculation method for the grasping center based on three-point calibration according to claim 1, characterized in that, When the method is applied to a dual-station grasping scenario, the acquisition, construction, calculation, storage and combination steps are executed independently for the left and right stations respectively. Each station independently acquires the coordinates of its three reference points, inputs its own material length and width parameters, and independently stores its own grasping center horizontal coordinates and height parameters.