A method for precise feeding of a mechanical arm based on coordinate transformation

Abstract

The application provides a kind of based on coordinate transformation realizes accurate feeding method of mechanical arm, this method is first assembled calibration needle and completes system self-checking, three-point method is used to establish user coordinate system of material taking area and feeding area respectively;Visual planning feeding and unloading point, generate no compensation initial coordinates;Through single feeding fine adjustment, collect compensation value, and in 0°, clockwise 90°, counterclockwise 90° three attitudes, batch collection multiple samples;Subsequently, the compensation data is substituted into the coordinate transformation algorithm, and the basic error of the origin offset and axial deflection of the two area coordinate systems is solved;Finally, combined with the actual rotation angle of the workpiece and the rotation transformation formula, the dynamic compensation parameters are calculated in real time and superimposed to the initial coordinates, and the mechanical arm is driven to complete the no demonstration accurate feeding.The application does not need artificial demonstration and three-dimensional model, and the calibration cost is low, the adaptability is strong, the full attitude adaptive compensation can be realized, the positioning accuracy and operation efficiency of mechanical arm feeding and unloading are effectively improved, and it is suitable for precision automatic production line.

Description

Technical Field

[0001] This invention relates to the field of industrial automation technology, specifically to a method for achieving precise material feeding by a robotic arm based on coordinate transformation. Background Technology

[0002] In industrial automation production, robotic arms are widely used in various loading and unloading operations. Their operational accuracy and efficiency directly affect the automation level of the production line and the product qualification rate. As the manufacturing industry transforms towards flexible and precision production, the precision machining process places higher demands on the accuracy of robotic arm loading and unloading operations. Traditional teachable robotic arm loading and unloading technology can no longer meet the core production needs of no-teachability and high precision. In current industrial scenarios, robotic arm loading and unloading operations mainly rely on two modes: online teaching and offline teaching. Both have significant technical drawbacks. Online teaching is inefficient, and its positioning accuracy is highly dependent on the operator's experience. In addition, on-site teaching operations pose safety hazards. Offline teaching does not require on-site manual operation, but it must rely on a complete 3D model of the workpiece and has extremely high requirements for equipment installation accuracy. In actual production processes, it is difficult to completely eliminate positioning errors, and the cost of scenario adaptation is high, which cannot meet the needs of precision production. Coordinate transformation technology can realize spatial position transformation between multiple coordinate systems, which can provide support for the precise positioning of robotic arms. However, the application of existing coordinate transformation technology still has the following drawbacks. 1. Limited to simple position conversion, it has not formed a complete technical system adapted to non-teachable operation scenarios, and cannot compensate for errors such as calibration deviation and rotation offset, making it difficult to meet the requirements of precision feeding; 2. Coordinate transformation is not deeply integrated with motion control, and there is no complete solution that adapts to multiple postures and teaching scenarios, making it impossible to balance low-cost calibration and high-precision operation. Summary of the Invention

[0003] The purpose of this invention is to address the shortcomings of existing technologies by proposing a method for precise material loading by a robotic arm based on coordinate transformation.

[0004] To achieve the above objectives, the present invention adopts the following technical solution: a method for precise material feeding by a robotic arm based on coordinate transformation, the specific steps of which are as follows: S1: Tooling calibration and assembly and preliminary preparation: The calibration needle is coaxially fixed to the end of the robotic arm to complete system initialization, axis zero point reset and communication self-test; Select a standard calibration pin that is compatible with the end flange of the robotic arm, and install the calibration pin coaxially on the end face of the flange by bolting to ensure that the relative position of the calibration pin and the end effector of the robotic arm is fixed. After the tooling assembly is completed, start the robotic arm control system, complete system initialization, axis zero point reset and communication self-test, and ensure that the equipment is in standby calibration state.

[0005] S2: Standard calibration of user coordinate system in material handling area: The calibration needle is manipulated to touch the reference point of the material picking area, and the user coordinate system of the material picking area is established by fitting the three-point method. The robotic arm is controlled to move the end-effector calibration needle to touch the preset coordinate reference points in the material picking area in sequence, and collect the spatial position signals of the reference points. The control system receives the point data and automatically fits and generates the user coordinate system of the material picking area. This step adopts the conventional on-site calibration process, without introducing high-precision calibration equipment and complex calibration algorithms, and quickly completes the construction of the coordinate reference of the material picking area.

[0006] Three non-collinear, rigidly fixed, and clearly positioned reference points are pre-arranged on the working plane of the material reclaiming area, and are denoted as the origin reference points. X-axis reference point Y-axis reference point The three points are non-collinear and approximately perpendicularly distributed; simultaneously, the distance between the three points is ≥50mm to ensure computational stability; the robotic arm is manipulated to drive the end-effector calibration needle to perpendicularly touch the three points respectively, and the sampling is repeated 3 times and the average is taken to obtain the coordinates in the base coordinate system. , , ; The robotic arm is manipulated to move the end-effector calibration needle vertically in a low-speed inching mode. Record the coordinates of the reference point center in the base coordinate system: ; Touch in the same way Benchmark point: ; Touch in the same way Benchmark point: ; When touching, keep the Z-axis vertically downward each time to avoid lateral force causing displacement, and repeat the sampling 3 times and take the average value; Calculate the origin of the coordinate system of the material picking area, with Origin of the coordinate system ; Origin of user coordinate system in material picking area Directly using the benchmark point Base coordinates: ; Depend on point to Construct the X-axis vector and normalize it; ; Normalize the X-axis vector: ; in, ; Depend on point to Construct a temporary Y-axis vector. ; Schmidt orthogonalization ensures perpendicularity to the X-axis and normalization; ; Normalize the Y-axis vector: ; Calculate the Z-axis unit vector using the right-hand rule; ; Finally, the user coordinate system of the material picking area is generated. and homogeneous transformation matrix The system automatically checks the included angle and collinearity, and completes the system establishment after determining that the calibration is valid.

[0007] ; in, , , for The three components of the unit vector along the axis; , , for The three components of the unit vector along the axis; , , for The three components of the unit vector along the axis.

[0008] Pre-set the origin reference point on the rigid platform in the material handling area. X-axis reference point Y-axis reference point The three points are not collinear and are approximately perpendicularly distributed. The robotic arm is manipulated to move the end-effector calibration needle to perpendicularly touch the three points respectively. This sampling is repeated three times, and the average is taken to obtain the coordinates in the base coordinate system. , , .

[0009] by Origin of the coordinate system ,Depend on point to Construct the X-axis vector and normalize it; by point to A temporary Y-axis vector is constructed, orthogonalized using a Schmitt trigger to ensure its perpendicularity to the X-axis, and then normalized. The Z-axis unit vector is calculated using the right-hand rule. Finally, the user coordinate system for the material handling area is generated. and homogeneous transformation matrix The system automatically checks the included angle and collinearity, and completes the system establishment after determining that the calibration is valid.

[0010] S3: Standard calibration of user coordinate system in the loading area: Using the calibration method from step S2, independently establish the user coordinate system for the material loading area; Following the calibration method in step S2, the calibration pin of the robotic arm is manipulated to touch the preset coordinate reference point of the loading area to complete the establishment of the user coordinate system of the loading area; the coordinate systems of the picking area and the loading area are calibrated independently, both adopting low-precision conventional calibration mode, simplifying the on-site implementation process and reducing the calibration threshold and hardware cost. On the rigid worktable / mold positioning surface of the loading area, three non-collinear reference points are pre-machined and fixed, which serve as the origin reference points of the coordinate system of the loading area. Reference point in the X-axis direction of the loading area Y-axis reference point of the loading area The three points are non-collinear and approximately perpendicularly distributed; simultaneously, the distance between the three points is ≥50mm to ensure computational stability; the robotic arm is manipulated to drive the end-effector calibration needle to perpendicularly touch the three points respectively, and the sampling is repeated 3 times and the average is taken to obtain the coordinates in the base coordinate system. , , ; The robotic arm is manipulated to drive the end-effector calibration needle, which touches three reference points sequentially in a low-speed, vertically downward motion. This process is repeated three times, and the average value is taken to eliminate random errors. Touching the origin of the coordinate system of the feeding area Record the coordinates in the base coordinate system: ; Touch the reference point in the X-axis direction of the loading area in the same way. : ; Touch the reference point in the Y-axis direction of the loading area in the same way. : ; Calculate the origin of the coordinate system of the feeding area, with Origin of the coordinate system ; Origin of user coordinate system in material picking area Directly using the benchmark point Base coordinates: ; Depend on point to Construct the X-axis vector and normalize it; ; Normalize the X-axis vector: ; in, ; Depend on point to Construct a temporary Y-axis vector. ; Schmidt orthogonalization ensures perpendicularity to the X-axis and normalization; ; Normalize the Y-axis vector: ; Calculate the Z-axis unit vector using the right-hand rule; ; Finally, the user coordinate system of the material picking area is generated. and homogeneous transformation matrix 0 The system automatically checks the included angle and collinearity, and completes the system establishment after determining that the calibration is valid.

[0011] ; in, , , for The three components of the unit vector along the axis; , , for The three components of the unit vector along the axis; , , for The three components of the unit vector along the axis.

[0012] Pre-set the origin reference point of the coordinate system of the loading area on the rigid worktable of the loading area. Reference point in the X-axis direction of the loading area Y-axis reference point of the loading area The three points are not collinear and are approximately perpendicularly distributed; the calibration needle at the end of the robotic arm touches the three points perpendicularly, and the average is obtained by repeated sampling to obtain the coordinates in the base coordinate system. , , ;by Origin of the coordinate system ,Depend on point to Construct and normalize the X-axis vector, orthogonalize the Y-axis vector with a Schmitt trigger to ensure perpendicularity, generate the Z-axis unit vector according to the right-hand rule, and finally establish the user coordinate system for the loading area. And generate the homogeneous transformation matrix. The system automatically verifies the included angle and repeatability, and completes the calibration of the coordinate system of the loading area after determining its validity.

[0013] S4: Visual planning of loading and unloading reference points: In the control system, plan the material pick-up point, loading point and target posture, and generate uncompensated initial loading and unloading coordinates; Enter the loading and unloading planning interface of the control system software, manually select the target gripping point of the sheet in the picking area, the target placement point and posture of the workpiece in the loading area; the system automatically calculates the movement path of the robotic arm based on the calibrated user coordinate system of the two areas, and generates the initial picking coordinate point and the initial loading coordinate point in the uncompensated state. User coordinate system in the material picking area Within the selection area, the gripping center on the workpiece (such as the geometric center of the sheet metal or the center of the positioning hole) is recorded as the target gripping point in the material handling area. : ; in, These are the position coordinates; To capture the attitude angle; User coordinate system in the material loading area Within the selected processing station / mold target placement center, record it as the target placement point of the loading area. : ; in, These are the position coordinates; To set the attitude angle.

[0014] Initial material handling coordinates generation without compensation: Map the user coordinates in the material handling area to the robot arm's base coordinate system to obtain the uncompensated initial material handling coordinates. ; ; The grasping posture is directly and equivalently transmitted in the user coordinate system without rotation compensation. .

[0015] Generate initial loading coordinates without compensation: Map the user coordinates in the loading area to the robot arm's base coordinate system to obtain the initial uncompensated loading coordinates. ; ; The placement posture is directly transmitted according to the set angle without error compensation. .

[0016] In the control system's visualization interface, select the target grabbing point within the user coordinate system of the material grabbing area. Select the target placement point within the user coordinate system of the material loading area. The system transforms the material handling area matrix respectively. Transformation matrix of feeding area The system maps the grab and placement points in the user's coordinate system to the robot arm's base coordinate system, directly generating uncompensated initial material handling coordinates without any error correction. Initial loading coordinates without compensation And complete the initial motion path calculation.

[0017] S5: Single-time attitude loading and compensation value acquisition: Perform a single loading operation, manually fine-tune to the required accuracy, and record the X and Y two-dimensional compensation values; The control system drives the robotic arm to perform the loading operation according to the initial position. Due to the inherent deviation of the conventional calibration of the coordinate system, the initial position cannot achieve accurate loading of the workpiece. The robotic arm is controlled by manual fine-tuning mode to correct the end position along the X and Y axes until the workpiece reaches the loading accuracy requirements. The X and Y two-dimensional compensation values ​​at this time are recorded to complete the single set of data acquisition. The control system uses the uncompensated initial material handling coordinates generated by S4. Initial loading coordinates without compensation Drive the robotic arm to perform the complete loading action: The robotic arm moves to the safe picking point and descends to... Complete workpiece gripping → lift to safe height; Spatial translation to the loading safety point → vertical descent along the Z-axis, bringing the workpiece closer. ; The workpiece stops at the target position due to calibration error and is in a state of deviation. Switch the control system to user coordinate system manual incremental fine-tuning mode and lock the following parameters: Lock the Z-axis height (to prevent the workpiece from colliding or rubbing against the fixture); Lock the attitude angles (α, β, γ remain constant); Only X-axis and Y-axis translational fine-tuning permissions are enabled.

[0018] Manually and slowly fine-tune the end of the robotic arm along the X and Y axes until the loading accuracy judgment conditions are met: the reference edge of the workpiece is completely in contact with the positioning edge of the mold; there is no interference between the positioning hole of the workpiece and the mold pin, and the gap is uniform; the position deviation of the measuring tool is ≤0.02mm.

[0019] The precise point location is recorded as the actual target coordinates in the base coordinate system: ; Initial loading coordinates without compensation: ; The single attitude compensation value is calculated using the following formula: ; ; That is: Compensation value = Actual precise coordinates - Uncompensated initial coordinates; The system automatically reads the current actual coordinates of the robotic arm, subtracts them from the initial coordinates, and obtains the compensation value for this operation: ( , ); The control system drives the robotic arm to perform a loading action according to the uncompensated initial coordinates, and stops after the workpiece reaches the loading position; the control system switches to the user coordinate system manual fine-tuning mode, locks the Z-axis height and attitude angle, and only allows the X and Y-axis translational adjustment; manual fine-tuning is performed until the workpiece meets the preset loading accuracy requirements, and the difference between the current actual coordinates of the robotic arm and the uncompensated initial coordinates is recorded to obtain the two-dimensional compensation values ​​of the X and Y axes. , ), to complete the data collection for a single compensation.

[0020] S6: Batch acquisition of multi-pose compensation values: Three sets of independent compensation samples were collected under three working conditions: the workpiece's original 0° posture, 90° clockwise rotation, and 90° counterclockwise rotation. Repeat the compensation acquisition process in step S5, and switch between three loading conditions to carry out the operation: the first is to load the workpiece directly in its original posture, the second is to load the workpiece in a posture with the workpiece rotated 90° clockwise, and the third is to load the workpiece in a posture with the workpiece rotated 90° counterclockwise; perform position fine-tuning and precise loading for each condition, and record three sets of independent (X, Y) compensation data to obtain multi-posture compensation samples. Posture 1: 0° original posture; The workpiece gripping direction and placement direction are completely consistent with the X1-Y1 axis of the user coordinate system in the material handling area, without any rotation.

[0021] Posture 2: Clockwise rotation 90° (+90° / CW90°) The workpiece rotates 90° clockwise around the Z-axis in the horizontal plane, with the geometric center remaining unchanged, only the orientation changes.

[0022] Posture 3: Counterclockwise rotation 90° posture (-90° / CCW90°) The workpiece rotates 90° counterclockwise around the Z-axis in the horizontal plane, with the geometric center remaining unchanged, only the orientation changes.

[0023] The system automatically executes the following each time a new posture is changed: Keep the gripping point, loading point, and Z-axis height constant; Based on the current attitude angle, the uncompensated initial loading coordinates generated in step S4 are... 0 Perform a pure rotation transformation to obtain the new initial coordinates in this pose: ; ; ; , These are the initial coordinates before rotation (uncompensated initial point); , The coordinates are after rotation; in, =0°, +90°, -90°; Only attitude rotation transformation is performed, without any error compensation; Strictly following the same process as S5, the loading, fine-tuning, and compensation value calculation are completed independently for each of the three postures, without interference: Group 1: 0° original posture: Execute loading → Lock Z and attitude → Manually fine-tune X and Y to position. Calculate the compensation value: ; ; Storage: Attitude = 0°, Compensation = ( , ); Group 2: 90° clockwise posture Attitude switch → Initial coordinate recalculation → Execute loading → Fine-tuning in place Calculate the compensation value: ; ; Storage: Attitude = +90°, Compensation = ( , ); Group 3: 90° counterclockwise posture Attitude switch → Initial coordinate recalculation → Execute loading → Fine-tuning in place Calculate the compensation value: ; ; Storage: Attitude = -90°, Compensation = ( , ).

[0024] The system sequentially switches the workpiece placement angle according to three standardized postures: 0° original posture, 90° clockwise rotation, and 90° counterclockwise rotation. Based on the posture angle, the system performs a pure rotation transformation on the uncompensated initial coordinates, repeating the loading, alignment, fine-tuning, and compensation value calculation process of S5, and independently collects three sets of data. , Compensation samples: The system uniformly stores the correspondence between attitude and compensation, and completes the data validity verification to form a multi-attitude compensation sample set for subsequent basic error calculation.

[0025] S7: Calibration baseline error calculation: Substitute the three sets of compensation samples collected in step S6 into the coordinate transformation algorithm to solve the origin offset and axial deflection basic error of the coordinate system between the material taking area and the material feeding area. The three sets of multi-attitude compensation values ​​are filled into the corresponding compensation calculation module of the system. The system internally calls the corresponding algorithm to analyze the offset law between the calibration coordinate system of the material picking area and the actual ideal coordinate system, calculates the position deviation of the calibration origin of the two coordinate systems relative to the actual origin, and extracts the constant basic error value. The positioning error of loading and unloading consists of two types of constant basic errors: Offset of the origin of the coordinate system in the material handling area: , ; Offset of the origin of the coordinate system in the loading area: , ; Coordinate system axial deflection error: (X / Y axes are not perpendicular, axial deflection); The comprehensive compensation model uses coordinate rotation transformation plus origin offset superposition: ; ; In the formula: , for; For the workpiece rotation angle (0°, +90°, −90°); , The origin of the material taking area is kept at a constant offset; , The origin of the feeding area is kept at a constant offset; This is the axial deflection coefficient of the coordinate system; , These are the initial coordinates; Substituting the three sets of attitude data (α=0°, 90°, −90°) collected in step S6 into the equations, we obtain: Posture 1: =0° (cos0=1, sin0=0) = + ; = + ; Posture 2: = +90° (cos90 = 0, sin90 = 1) = - ; = + ; Posture 3: =−90° (cos(−90)=0, sin(−90)=−1) = + ; = - ; By solving the above three sets of six equations simultaneously, we can directly solve for the four fundamental error parameters in a closed-loop manner: Material picking area origin offset: ; ; Offset of origin in loading area: ; ; Axial deflection error coefficient .

[0026] The three sets of multi-pose compensation values ​​(0°, 90° clockwise, and 90° counterclockwise) collected by S6 were substituted into the coordinate transformation error model to establish an overdetermined set of equations including the offset of the origin of the material picking area, the offset of the origin of the material loading area, and the axial deflection of the coordinate system. The offset of the origin of the coordinate system in the material picking area was obtained through closed-loop analytical solution. , Offset of the origin of the coordinate system in the material loading area , and axial deflection coefficient The system performs residual verification on the solution results. If the result is deemed valid, the basic error is stored and fixed for subsequent dynamic compensation calculations.

[0027] S8: Dynamic Compensation Parameter Generation and Execution: Based on the actual rotation angle of the workpiece, combined with the basic error and rotation transformation formula, dynamic compensation parameters are calculated in real time and superimposed on the initial coordinates to drive the robotic arm to complete the teachingless precision feeding. An error compensation model is established by combining the workpiece loading rotation angle, basic error value and coordinate transformation algorithm; for the workpiece posture and rotation angle parameters of each loading and unloading operation, the corresponding precise compensation parameters are calculated in real time through the model, and the control system superimposes the compensation parameters onto the initial position to drive the robotic arm to perform teachingless precision loading and unloading operations. Based on the constant fundamental error calculated in step S7, combined with the actual rotation angle of the current workpiece. By combining coordinate rotation transformation with origin offset, the unique dynamic compensation value required for this material loading is calculated in real time. , This achieves "one-time calibration, full attitude adaptive compensation": Attitude rotation transformation (workpiece rotation) horn) ; = - ; = + ; , These are the initial coordinates after rotation; The current feeding posture angle (any angle from 0° to 360°); , The initial coordinates without compensation in step S4; Calculation of comprehensive dynamic compensation value: ; ; In the formula: , The offset of the origin of the material picking area coordinate system calculated in step S7; This is the offset of the origin of the coordinate system of the loading area calculated in step S7; This represents the current workpiece rotation angle; The compensation value is superimposed on the uncompensated initial coordinates to obtain the true and accurate coordinates in the base coordinate system: ; ; The Z-axis coordinate and attitude angle remain unchanged at the set values. ; Finally, the accurate material loading coordinates after compensation are ( , , ).

[0028] The system reads the origin offset of the material picking area calculated in step S7. , The origin of the loading area is offset. , Based on the current actual rotation angle of the workpiece By using coordinate rotation transformation and error superposition formula, dynamic compensation values ​​are calculated in real time. , The compensation value is superimposed on the uncompensated initial coordinates generated by S4 to obtain the final accurate target coordinates. , , The control system drives the robotic arm to perform actions according to the corrected coordinates, completing the untaught, full-attitude adaptive precision feeding.

[0029] Preferably, the present invention also provides a robotic arm control system, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, it executes all the steps of S1 to S8 mentioned above. Through coordinate system calibration, error calculation and dynamic coordinate compensation, the robotic arm is driven to complete teachingless, high-precision loading and unloading operations to meet the precision loading requirements of flexible automated production lines.

[0030] Preferably, the present invention also provides a computer-readable storage medium storing a computer program thereon. When the program is executed by a processor, it implements all the steps of the aforementioned method for achieving precise material feeding of a robotic arm based on coordinate transformation, enabling the robotic arm control system to complete the calibration of the material feeding coordinate system, multi-posture compensation acquisition, basic error calculation and dynamic compensation execution, thereby achieving precise material feeding without teaching.

[0031] Compared with the prior art, the beneficial effects of the present invention are as follows: Compared with the existing robotic arm loading and unloading technology, the present invention abandons the traditional online teaching and offline teaching modes, does not rely on human experience, three-dimensional models and high-precision calibration equipment, and adopts the three-point method to quickly and independently establish the user coordinate system for picking up and loading materials, which significantly reduces calibration costs, operation difficulty and on-site safety risks, solves the defects of conventional coordinate transformation that cannot compensate for calibration deviation and rotation offset, and forms a complete technical solution adapted to teaching-free operation.

[0032] This invention uses multi-pose compensation sample acquisition and coordinate transformation algorithms to accurately calculate the basic errors of coordinate system origin offset and axial deflection. Combined with rotation transformation, it achieves dynamic compensation for all poses. One calibration can adapt to the workpiece rotation angle for adaptive and precise loading. The positioning accuracy meets the requirements of precision production. It effectively balances low-cost calibration and high-precision operation, and greatly improves the loading and unloading efficiency and product qualification rate of automated production lines. Attached Figure Description

[0033] Figure 1This is a schematic diagram illustrating the specific process of the present invention. Detailed Implementation

[0034] To provide a further understanding of the purpose, structure, features, and functions of the present invention, detailed descriptions are provided below with reference to specific embodiments.

[0035] Example 1: Detailed implementation of a robotic arm for precise material feeding without teaching based on coordinate transformation.

[0036] This embodiment discloses a method for precise material loading of a robotic arm based on coordinate transformation. It is applicable to the loading and unloading station of flat workpieces in a precision machining automated production line. The robotic arm is a six-axis industrial robotic arm with a repeatability accuracy of ±0.01mm. The control system supports user coordinate system calibration, coordinate transformation, and motion closed-loop control. The following is in conjunction with the appendix... Figure 1 Detailed instructions with steps: S1: Tooling calibration and assembly and preliminary preparation: Select a standard calibration needle that matches the end flange of the robotic arm, and use bolts to coaxially fix the calibration needle to the center of the end effector of the robotic arm, ensuring that the axis of the calibration needle coincides with the central axis of the end of the robotic arm.

[0037] Start the robotic arm control system, perform system initialization, complete axis zero point reset, servo enable, and communication link self-test; confirm that the robotic arm has no alarms, coordinate data is normal, and I / O signals are stable, and the equipment enters standby calibration state.

[0038] S2: Routine calibration of the user coordinate system in the material handling area: Three non-collinear reference points are pre-machined on the rigid platform in the material handling area: the origin reference point. X-axis reference point Y-axis reference point The three points are approximately perpendicularly distributed and the distance between them is ≥50mm.

[0039] The robotic arm is controlled in a low-speed inching mode to drive the calibration needle to vertically touch the three points mentioned above. Each point is touched three times, and the average coordinates are taken to obtain the coordinates in the base coordinate system. , , .

[0040] by Origin of the coordinate system ,Depend on point to Construct the X-axis vector and normalize it; by point to Construct a temporary Y-axis vector, and then perform Schmitt orthogonalization to ensure it is perpendicular to the X-axis and normalize it; calculate the Z-axis unit vector according to the right-hand rule.

[0041] Finally, the user coordinate system of the material picking area is generated. and homogeneous transformation matrix The system automatically checks the included angle and collinearity, and completes the system establishment after determining that the calibration is valid.

[0042] S3: Standard calibration of user coordinate system in the loading area: Following the three-point calibration process of S2, a reference point is preset on the rigid mold positioning surface in the loading area. X-axis reference point Y-axis reference point .

[0043] The calibration probe is manipulated to vertically touch three points, and the samples are repeated three times and averaged to obtain the coordinates in the base coordinate system. , , .

[0044] by Origin Construct and normalize the X-axis vector; perform Schmitt orthogonalization and normalization on the Y-axis vector; generate the Z-axis unit vector according to the right-hand rule.

[0045] Finally, establish the user coordinate system for the material loading area. and homogeneous transformation matrix After the system verifies the validity, the calibration of the loading area is completed.

[0046] The coordinate systems of the material handling area and the material loading area are calibrated independently, using conventional low-cost calibration methods without the use of high-precision calibration equipment.

[0047] S4: Visual planning of loading and unloading reference points: Enter the control system visualization planning interface, select the workpiece geometric center as the target gripping point G in the user coordinate system of the material picking area; select the mold positioning center as the target placement point H in the user coordinate system of the loading area, and set the gripping and placement posture angles.

[0048] The system transforms the material handling area matrix. Transformation matrix of feeding area The user coordinates (G and H) are mapped to the robot arm's base coordinate system without any error correction, directly generating uncompensated initial material handling coordinates. Uncompensated initial loading coordinates And complete the initial motion path calculation.

[0049] S5: Single-time attitude loading and compensation value acquisition: The control system drives the robotic arm to perform the following loading operation according to the initial coordinates: material pick-up → translation → descent of the loading position.

[0050] Due to inherent calibration deviations, the workpiece was not precisely positioned. Switch the system to manual incremental mode in the user coordinate system, lock the Z-axis height and attitude angle, and only grant fine-tuning permissions for the X and Y axes.

[0051] Manually adjust the end of the robotic arm at low speed until the workpiece reference edge and the mold positioning edge are completely aligned, the gap between the positioning hole and the pin is uniform, and the positional deviation is ≤0.02mm.

[0052] Record the difference between the current actual coordinates and the initial coordinates to obtain the X and Y two-dimensional compensation values. , Complete the data collection for a single compensation operation.

[0053] S6: Batch acquisition of multi-pose compensation values: Keeping the gripping point, loading point, and Z-height constant, switch between the three workpiece postures and repeat the S5 process to independently collect three sets of compensation samples: Attitude 1 (0°): Workpiece and material handling area coordinate system - Same direction, collection compensation ( , ); Posture 2 (90° clockwise): The workpiece rotates 90° clockwise around the Z-axis, and compensation data is collected. , ); Attitude 3 (90° counterclockwise): The workpiece rotates 90° counterclockwise around the Z-axis, and the data is collected and compensated ( , ).

[0054] The system performs a pure rotational transformation on the initial coordinates without adding error compensation, and the three sets of data are stored independently without interfering with each other.

[0055] S7: Calibration baseline error calculation: Substituting the three sets of compensation samples into the coordinate transformation error model, a model was established that includes the offset of the origin of the material picking area, the offset of the origin of the material feeding area, and the axial deflection. The system of equations.

[0056] The overdetermined system of equations was solved using the least squares method, and four constant fundamental errors were calculated using a closed-loop solution: Material pick-up area origin offset , ; Offset of origin of loading area , ; Coordinate system axial deflection coefficient .

[0057] The system performs residual verification on the solution results, and stores the basic error after determining that it is valid.

[0058] S8: Dynamic Compensation Parameter Generation and Execution: The system reads the calculated basic error and, based on the actual rotation angle of the workpiece being loaded,... The dynamic compensation value is calculated using the two-dimensional rotation transformation formula: ; ; The compensation value is superimposed on the uncompensated initial coordinates to obtain the final accurate target coordinates: ; ; The control system drives the robotic arm to perform loading and unloading according to the corrected coordinates, achieving full-attitude adaptive and precise loading without the need for further teaching.

[0059] The present invention has been described in the above-described embodiments; however, these embodiments are merely examples for implementing the present invention. It must be noted that the disclosed embodiments do not limit the scope of the present invention. Conversely, any modifications and refinements made without departing from the spirit and scope of the present invention are within the scope of patent protection of the present invention.

Claims

1. A method for achieving precise material feeding by a robotic arm based on coordinate transformation, characterized in that: The specific steps of the method are as follows: S1: Tooling calibration and assembly and preliminary preparation: The calibration needle is coaxially fixed to the end of the robotic arm to complete system initialization, axis zero point reset and communication self-test; S2: Standard calibration of user coordinate system in material handling area: The calibration needle is manipulated to touch the reference point of the material picking area, and the user coordinate system of the material picking area is established by fitting the three-point method. S3: Standard calibration of user coordinate system in the loading area: Using the calibration method from step S2, independently establish the user coordinate system for the material loading area; S4: Visual planning of loading and unloading reference points: In the control system, plan the material pick-up point, material loading point, and target posture, and generate uncompensated initial material loading coordinates. , ); S5: Single-time attitude loading and compensation value acquisition: Perform a single loading operation, manually fine-tune to the required accuracy, and record the X and Y two-dimensional compensation values; S6: Batch acquisition of multi-pose compensation values: Three sets of independent compensation samples were collected under three working conditions: the workpiece's original 0° posture, 90° clockwise rotation, and 90° counterclockwise rotation. S7: Calibration baseline error calculation: Substitute the three sets of compensation samples collected in step S6 into the coordinate transformation algorithm to solve the origin offset and axial deflection basic error of the coordinate system between the material taking area and the material feeding area. S8: Dynamic Compensation Parameter Generation and Execution: Based on the actual rotation angle of the workpiece, combined with the basic error and rotation transformation formula, dynamic compensation parameters are calculated in real time and superimposed on the initial coordinates to drive the robotic arm to complete the unteachable precision feeding.

2. The method for precise material feeding of a robotic arm based on coordinate transformation as described in claim 1, characterized in that: In step S2, a reference point is preset on the rigid platform of the material handling area. X-axis reference point Y-axis reference point The three points are not collinear and are approximately perpendicularly distributed; the robotic arm is manipulated to drive the end-effector calibration needle to touch the three points perpendicularly, and the samples are repeated three times and averaged to obtain the coordinates in the base coordinate system. , , ; by Origin of the coordinate system ,Depend on point to Construct the X-axis vector and normalize it; by point to Construct a temporary Y-axis vector, ensure its perpendicularity to the X-axis by Schmitt orthogonalization, and normalize it; calculate the Z-axis unit vector according to the right-hand rule. Finally, the user coordinate system of the material picking area is generated. and homogeneous transformation matrix The system automatically checks the included angle and collinearity, and completes the system establishment after determining that the calibration is valid.

3. The method for precise material feeding of a robotic arm based on coordinate transformation as described in claim 2, characterized in that: In step S3, the calibration method of step S2 is used to establish the reference point of the coordinate system origin of the loading area on the rigid worktable in the loading area. Reference point in the X-axis direction of the loading area Y-axis reference point of the loading area The three points are not collinear and are approximately perpendicularly distributed; the robotic arm is manipulated to drive the end-effector calibration needle to touch the three points perpendicularly, and the samples are repeated three times and averaged to obtain the coordinates in the base coordinate system. , , ; by Origin of the coordinate system ,Depend on point to Construct the X-axis vector and normalize it; by point to Construct a temporary Y-axis vector, ensure its perpendicularity to the X-axis by Schmitt orthogonalization, and normalize it; calculate the Z-axis unit vector according to the right-hand rule. Finally, the user coordinate system of the material picking area is generated. and homogeneous transformation matrix The system automatically checks the included angle and collinearity, and completes the system establishment after determining that the calibration is valid.

4. The method for precise material feeding of a robotic arm based on coordinate transformation as described in claim 3, characterized in that: In step S4, on the control system visualization interface, the target grabbing point G is selected in the user coordinate system of the material grabbing area, and the target placement point H is selected in the user coordinate system of the loading area; the system then uses the material grabbing area transformation matrix to... Transformation matrix of feeding area The system maps the grab and placement points in the user's coordinate system to the robot arm's base coordinate system, directly generating uncompensated initial material handling coordinates without any error correction. Initial loading coordinates without compensation And complete the initial motion path calculation.

5. The method for precise material feeding of a robotic arm based on coordinate transformation as described in claim 1, characterized in that: In step S6, three sets of attitude compensation samples are used to separate the calibration origin error from the attitude rotation error.

6. The method for precise material feeding of a robotic arm based on coordinate transformation as described in claim 1, characterized in that: In step S7, the basic error is solved by using the least squares method to solve the overdetermined system of equations to obtain constant compensation parameters.

7. The method for precise material feeding of a robotic arm based on coordinate transformation as described in claim 1, characterized in that: Step S8 uses a two-dimensional rotation transformation formula: ； , These are the initial coordinates after rotation; The current feeding posture angle (any angle from 0° to 360°); , The coordinates are the initial uncompensated loading coordinates in step S4.

8. The method for precise material feeding of a robotic arm based on coordinate transformation as described in claim 1, characterized in that: The comprehensive compensation model in step S8 is as follows: ； ； ； ； In the formula: , The initial loading coordinates are without compensation; , Offset of the origin of the material picking area; , This refers to the offset of the origin point of the loading area in step S4; The workpiece rotation angle; The precise coordinates of the true X-axis in the base coordinate system; The accurate coordinates of the true X-axis in the base coordinate system.

9. A robotic arm control system, characterized in that, The control system executes the method described in any one of claims 1-8 to achieve precise feeding without teaching.

10. A computer-readable storage medium storing a computer program, characterized in that, When the program is executed by a processor, it implements the method according to any one of claims 1-8.

Images