Pre-alignment method based on high-precision multi-axis displacement calculation

Abstract

The invention provides a pre-alignment method based on high-precision multi-axis displacement resolving, a marked graph on a to-be-transmitted object has first position information, and the pre-alignment method comprises the following steps: S1, detecting the marked graph to obtain second position information of a detected graph; s2, judging whether the detection graph is complete or not according to the position information of the detection graph; if not, multi-axis displacement calculation is carried out according to the first position information and the second position information, the translation amount of the mechanical wrist axis is obtained, and the step S3 is executed; if yes, the deflection angle is calculated according to the first position information and the second position information, and the step S4 is executed; s3, the manipulator moves the to-be-conveyed object according to the translation amount of the wrist shaft, and the step S1 is executed again; s4, whether the deflection angle is smaller than a set threshold value or not is judged; if not, the manipulator rotates according to the deflection angle and returns to the step S1; and if yes, pre-alignment of the to-be-transmitted object is completed. According to the pre-alignment method disclosed by the invention, the deflection angle can be controlled within an extremely small angle, and the pre-alignment precision is improved.

Description

Technical Field

[0001] This disclosure relates to the field of optical inspection technology, specifically to a pre-alignment method based on high-precision multi-axis displacement calculation. Background Technology

[0002] The wafer and mask transport system provides a fast, stable, reliable, collision-free, and clean transport solution for the entire system. The system includes transport equipment such as: a wafer load port, a pre-alignment device, a robotic arm, a front-opening universal cassette (FOUP), an Epin module, and a standard mechanical interface (SMIF).

[0003] In the transfer process, after the Load Port loads the FOUP box, the robot arm removes the wafer from the FOUP box and places it into the pre-alignment device for calibration. This process precisely calibrates the wafer's angle. After calibration, the robot arm retrieves the wafer again and places it onto the Epin module (i.e., the wafer loading position) on the workpiece stage. Therefore, the wafer loading repeatability positioning accuracy is closely related to the robot arm's repeatability positioning accuracy at the wafer loading position, and the same applies to the mask transfer process. How to improve the wafer loading repeatability positioning accuracy and complete pre-alignment is a topic that R&D personnel continuously focus on and research.

[0004] The information disclosed in this section is only for understanding the background of the inventive concept of this disclosure, and therefore may include information that does not constitute prior art. Summary of the Invention

[0005] In view of this, this disclosure provides a method for pre-aligning an object to be transferred based on high-precision multi-axis displacement calculation. A robotic arm is used to transfer the object to be transferred. The object to be transferred has a marked pattern, and the marked pattern has first position information at the target position on the object to be transferred, including:

[0006] S1, detect the marked graphic on the object to be transmitted to obtain the second position information of the detected graphic;

[0007] S2, determine whether the detection pattern is complete based on the detection pattern; if not, perform multi-axis displacement calculation based on the first position information and the second position information to obtain the translation amount of the mechanical wrist axis, and proceed to step S3; if yes, calculate the deflection angle based on the first position information and the second position information, and proceed to step S4.

[0008] S3, the robotic arm moves the object to be transferred according to the translation amount of the wrist axis and returns to step S1;

[0009] S4, determine whether the deflection angle is less than the set threshold; if not, the robot arm rotates according to the deflection angle and returns to step S1; if yes, the pre-alignment of the object to be transferred is completed.

[0010] According to embodiments of this disclosure, S2, the multi-axis displacement calculation based on the first position information and the second position information, includes:

[0011] S21, establish a first coordinate system with the rotation center of the wrist axis of the robotic arm as the origin and establish a second coordinate system with the center of the object to be transferred as the origin;

[0012] S22, in the first coordinate system, calculate the displacement of the marked graphic and the detected graphic, as well as the displacement before and after the wrist axis rotation, based on the first position information and the second position information;

[0013] S23. Based on the transformation relationship between the first coordinate system and the second coordinate system, the translation amount of the wrist axis in the first coordinate system is obtained.

[0014] According to an embodiment of this disclosure, the rotation of the robotic arm in S4 according to the deflection angle includes:

[0015] S41, determine the magnitude of the deviation of the deflection angle from the preset value; if the deviation of the deflection angle from the preset value is large, the robot arm rotates by a first fixed amount; if the deviation of the deflection angle from the preset value is small, the robot arm rotates by a second fixed amount; the first fixed amount is greater than the second fixed amount.

[0016] According to an embodiment of this disclosure, the rotation of the robotic arm in S4 according to the deflection angle includes:

[0017] S411, divide the deviation of the deflection angle from the preset value into at least two levels, each level corresponding to a preset fixed amount;

[0018] S412, sequentially determine whether the value of the deflection angle deviating from the preset value is in the current gear; if yes, the robot arm rotates by the preset fixed amount corresponding to the current gear and returns to step S1; if no, determine whether the value of the deflection angle deviating from the preset value is in the next gear, until all gears are determined.

[0019] According to an embodiment of this disclosure, the value in each gear position in S411 is greater than a preset fixed amount corresponding to the gear position.

[0020] According to embodiments of this disclosure, S1, detecting the marked pattern on the object to be transmitted includes:

[0021] S11, use a camera to capture images of the marked graphics on the object to be transmitted to obtain image information of the detected graphics; wherein, the marked graphics include at least two sets;

[0022] S12 performs filtering, binarization, image segmentation, and target detection on the image information to obtain the second position information of the detected graphic; the second position information includes the center coordinates of the detected graphic.

[0023] According to embodiments of this disclosure, calculating the deflection angle based on the first position information and the second position information in step S2 includes:

[0024] The deflection angle is calculated using the arctangent function based on the center coordinates of the detected graphic and the center coordinates of the marked graphic.

[0025] According to embodiments of this disclosure, step S2, before determining whether the detected image is complete, further includes:

[0026] S20, determine whether a detection pattern has been detected; if yes, proceed to step S2; if no, re-perform mechanical pre-alignment.

[0027] According to embodiments of this disclosure, determining whether the detected image is complete in step S2 further includes:

[0028] S24, determine whether the orientation of the markers in the detected graphic is consistent; if yes, calculate the deflection angle based on the first position information and the second position information, and proceed to step S4; if no, the robot arm performs orientation compensation based on the first position information and the second position information.

[0029] According to an embodiment of this disclosure, S24, calculating the deflection angle based on the first position information and the second position information, further includes:

[0030] The sign of the deflection angle is determined based on the first and second position information.

[0031] This disclosed pre-alignment method, based on high-precision multi-axis displacement calculation, utilizes the results of multi-axis displacement calculation to guide the precise movement of the multi-axis end effector of a robot, ensuring that the marked pattern can return to the detection range while maintaining a substantially constant rotation angle. Simultaneously, this method introduces an adaptive small-amount algorithm, employing corresponding fixed small-amount rotation adjustments for different deflection angles, thereby effectively controlling cumulative errors. The combination of these techniques significantly improves the accuracy and reliability of pre-alignment. Attached Figure Description

[0032] The above and other objects, features and advantages of this disclosure will become clearer from the following description of embodiments with reference to the accompanying drawings, in which:

[0033] Figure 1 A schematic diagram of the structure of a wafer and mask transfer system according to an embodiment of the present disclosure is shown.

[0034] Figure 2 The schematic diagram illustrates the structure of an optical pre-alignment device according to an embodiment of the present disclosure;

[0035] Figure 3 A schematic diagram illustrating the calculation of the deflection angle according to an embodiment of the present disclosure is shown.

[0036] Figure 4 The illustration shows a schematic diagram of the target image position marker and detection pattern displayed within the field of view of the CCD according to an embodiment of the present disclosure;

[0037] Figure 5 A flowchart illustrating a pre-alignment method based on high-precision multi-axis displacement calculation according to an embodiment of the present disclosure is shown schematically.

[0038] Figure 6 The illustration shows a schematic diagram of the wrist axis movement direction and CCD position according to an embodiment of the present disclosure;

[0039] Figure 7 The illustration schematically shows a model of the motion control structure of the optical pre-alignment device according to an embodiment of the present disclosure;

[0040] Figure 8 A flowchart illustrating a pre-alignment method according to an embodiment of the present disclosure is shown schematically;

[0041] Figure 9 Another flowchart of the pre-alignment method according to an embodiment of the present disclosure is illustrated schematically;

[0042] Figure 10 A schematic diagram of a wafer and mask transfer system Rz detection method according to an embodiment of the present disclosure is shown. Detailed Implementation

[0043] To make the objectives, technical solutions, and advantages of this disclosure clearer, the following detailed description is provided in conjunction with specific embodiments and the accompanying drawings.

[0044] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit this disclosure. The terms “comprising,” “including,” etc., as used herein indicate the presence of features, steps, operations, and / or components, but do not exclude the presence or addition of one or more other features, steps, operations, or components.

[0045] All terms used herein (including technical and scientific terms) have the meanings commonly understood by those skilled in the art, unless otherwise defined. It should be noted that the terms used herein are to be interpreted in a manner consistent with the context of this specification, and not in an idealized or overly rigid way.

[0046] The accompanying drawings illustrate several block diagrams and / or flowcharts. It should be understood that some blocks, or combinations thereof, in the block diagrams and / or flowcharts can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that, when executed by the processor, these instructions can create means for implementing the functions / operations described in these block diagrams and / or flowcharts. The technology of this disclosure can be implemented in hardware and / or software (including firmware, microcode, etc.). Alternatively, the technology of this disclosure can take the form of a computer program product stored on a computer-readable storage medium, which is available for use by or in conjunction with an instruction execution system.

[0047] The use of ordinal numbers such as "first," "second," "third," etc., in the specification and claims to modify the corresponding elements does not in itself imply or represent any ordinal number of the element, nor does it represent the order of one element with another element, or the order of manufacturing methods. The use of these ordinal numbers is only to enable a clearly distinguishable element with a certain name from another element with the same name.

[0048] Figure 1 A schematic diagram of the structure of a wafer and mask transfer system according to an embodiment of the present disclosure is shown. Figure 2 The schematic diagram illustrates the structure of an optical pre-alignment device according to an embodiment of the present disclosure.

[0049] like Figure 1 As shown, taking a wafer transport system as an example, this system includes a wafer load port, pre-alignment devices (including mechanical and optical pre-alignment devices), a robotic arm, a front-opening universal cassette (FOUP), and a workpiece stage. A high-precision position deviation detection device, namely the optical pre-alignment device, is equipped at the wafer loading position. Figure 1 As shown, the function of the optical pre-alignment device is to detect the deviation of the object to be transmitted in the position on the upper film in real time.

[0050] In practical applications, the high-precision micro-motion stage in the workpiece stage has a small travel range; for example, its rotary axis travel is only ±0.5 mRad. This requires the loading repeatability positioning accuracy to reach ≤±50 μRad. However, the accuracy of existing transfer systems is approximately ±1.7 mRad, which cannot meet the requirements for loading repeatability positioning accuracy. In the transfer system, the minimum step size and repeatability positioning accuracy of the robot are the key factors determining the final loading repeatability positioning accuracy.

[0051] The object to be transferred is positioned inside the machine. Due to space limitations in height and width, the optical pre-alignment device needs to be both compact and highly accurate. Typically, high-precision detection devices have a small detection range. When the robotic arm's end effector carries the object to be transferred and makes minor adjustments to its deflection angle, it may exceed the detection range. In this situation, it's impossible to determine whether the object has rotated to the target position. To maintain the rotation angle of the object as constant as possible, the robotic arm's end effector needs to perform multi-axis displacement motion to bring the object back into the detection range.

[0052] When the robotic arm of the transfer system transports the object to be transferred (wafer or mask) to the designated position, there is a transfer positioning error of approximately ±418 μRad, which needs to be reduced to within ±50 μRad. Theoretically, when the robotic arm transports the object to be transferred to the wafer mounting position, fine-tuning the minimum step deflection angle can improve the overall repeatability and positioning accuracy of the transfer system. However, in reality, the detection of the minimum step deflection angle is limited by the measurement accuracy of the robotic arm's own encoder or sensor, and can only measure about 17 μRad (thousandths). Therefore, in the process of gradually moving to the target wafer mounting position using the minimum step rotation, inaccurate positioning may occur (overshooting or undershooting), and the robotic arm's own measurement limits may misjudge that the movement has been completed, resulting in an increasing cumulative error.

[0053] To address the aforementioned issues, this disclosure provides a pre-alignment method based on high-precision multi-axis displacement calculation.

[0054] Due to limitations in the height and width of the internal space, the optical pre-alignment device must be designed to be both compact and highly accurate. Therefore, two CCD cameras, coupled with telecentric lenses and right-angle adapters, are used. This design effectively meets the requirement of a thin structure for the inspection device. A schematic diagram of the optical pre-alignment device is shown below. Figure 2 As shown, two sets of CCD cameras are used to acquire images of two sets of marked graphics on the object to be transmitted.

[0055] Based on the required detection accuracy, two sets of marker patterns are designed. When the CCD camera can acquire a complete marker pattern, the center coordinates of the detected marker pattern and the center coordinates of the marker pattern on the target image can be calculated. The deflection angle is obtained based on these two center coordinates. The robotic arm then fine-tunes the deflection angle until it is less than a threshold, completing the pre-alignment. A schematic diagram for calculating the deflection angle is shown below. Figure 3 As shown. The CCD camera performs pre-alignment by acquiring images of the marked patterns. and It is the center of the two sets of marked graphics on the target image. This center refers to the outline center of the marked graphics on the object to be transmitted on the target image. and L1 is the detection contour center of the image captured by the CCD camera when the object to be transferred is actually transferred to the upper film position by the robotic arm. and The vertical distance between them, L2 is and The vertical distance between the two marks. According to the arctangent function, formula (1) shows that the greater the distance between the two marks (the larger R is), the smaller the angle θ that can be calculated, and the higher the detection accuracy.

[0056] (1)

[0057] When the CCD camera cannot capture a complete detection pattern, the incomplete detection pattern leads to inaccurate center coordinates, making it impossible to determine whether the object to be transferred has moved to the target image position. In this case, while keeping the rotation angle of the object to be transferred as constant as possible, the end effector of the robot arm needs to perform multi-axis displacement motion to bring the actual marking pattern back into the field of view of the CCD camera. Specifically, the vertical distances L1 and L2 are the shorter sides of the CCD camera's field of view. Assuming that the marking pattern at the target image position is in the middle of the CCD's field of view when the object to be transferred is pre-aligned, then when the object to be transferred rotates more than 697 μRad, the actual captured detection pattern will be outside the CCD camera's field of view, meaning the detection pattern in the image will be incomplete or unmarked. Figure 4 As shown, the blue box represents the marker graphic at the target image position, the blue circle represents the center of the marker graphic at the target image position, the red box represents the position of the real-time detected graphic, and the red circle represents the center of the detected graphic at the actual image position. Because the detected graphic is incomplete, the detected center is inaccurate, making it impossible to determine whether the object to be transferred has rotated to the target image position. Therefore, while keeping the rotation angle of the object to be transferred as constant as possible, the robotic arm's end effector needs to perform multi-axis displacement motion to bring the detected graphic of the actual object to be transferred back into the field of view of the CCD camera.

[0058] Figure 5 A flowchart illustrating a pre-alignment method based on high-precision multi-axis displacement calculation according to an embodiment of the present disclosure is shown.

[0059] like Figure 5 As shown, the pre-alignment method may include operations S1 to S4. A robotic arm is used to transfer an object to be transferred. The object to be transferred has a marked pattern, and the marked pattern has first position information at the target position on the object to be transferred.

[0060] In operation S1, the marked graphic of the object to be transmitted is detected to obtain the second position information of the detected graphic.

[0061] In operation S2, the detection pattern is judged to be complete based on the detection pattern; if not, multi-axis displacement calculation is performed based on the first position information and the second position information to obtain the translation amount of the mechanical wrist axis, and then proceed to step S3; if yes, the deflection angle is calculated based on the first position information and the second position information, and then proceed to step S4.

[0062] In operation S3, the robotic arm moves the object to be transferred according to the translation amount of the wrist axis and returns to step S1.

[0063] In operation S4, it is determined whether the deflection angle is less than the set threshold; if not, the robot arm rotates according to the deflection angle and returns to step S1; if yes, the pre-alignment of the object to be transferred is completed.

[0064] Embodiments of this disclosure employ two sets of CCD cameras to acquire images of two sets of marker patterns on an object to be transmitted, such as a wafer or mask. By performing a series of image processing operations on the acquired detection patterns, including filtering, binarization, image segmentation, and target detection, the center coordinates of the detection patterns can be obtained. Subsequently, the distance between the center coordinates of the detected patterns and the center coordinates of the marker pattern's target wafer position is calculated, thereby obtaining the deviation in the X / Y direction between the actual wafer position of the object to be transmitted and the target wafer position. Furthermore, based on the distance between the centers of the detection patterns and the marker patterns at the target wafer position, the arctangent function can be used to calculate the deflection angle θ of the wafer position. The robotic arm performs fine adjustments based on the calculated deflection angle θ until the deflection angle θ is less than a set threshold, thereby completing the pre-alignment process.

[0065] The process of multi-axis displacement calculation is described in detail below.

[0066] Based on the above embodiments, S2, which performs multi-axis displacement calculation based on the first position information and the second position information, includes: S21, establishing a first coordinate system with the rotation center of the robot's wrist axis as the origin and establishing a second coordinate system with the center of the object to be transmitted as the origin; S22, calculating the displacement of the marked graphic and the detected graphic at the target's upper position, as well as the displacement before and after the wrist axis rotation, based on the first position information and the second position information in the first coordinate system; S23, obtaining the translation of the wrist axis in the first coordinate system based on the transformation relationship between the first coordinate system and the second coordinate system.

[0067] Taking a wafer as an example, after the robotic arm transfers the wafer to the loading position, the direction of its wrist axis movement and the position of the CCD camera are illustrated as follows: Figure 6As shown, the wrist axis of the robotic arm is the joint structure connecting the end effector to the robotic arm. The wrist axis motion includes the A+ axis, X+ axis, and R+ axis. The wrist axis rotates the wafer with point O as the origin, and also translates the wafer along the A+ and X+ axes with point O as the origin. The angle between the A+ and X+ axes is θ1.

[0068] Using a rectangle as an example, the motion control structure of the optical pre-alignment device is modeled as follows: Figure 7 As shown, the position, size, and center coordinates A″ and B″ of the target wafer mounting position box are recorded. After the robotic arm transfers the wafer to the mounting position, the rotation center of the robotic wrist axis is point O. A Cartesian coordinate system O-XY is established with point O as the origin. A Cartesian coordinate system O'-X'Y' is established with the wafer center O' as the origin, where the line OX∥O'X', and O'X' coincides with AB. The angle formed by OO' and OX is fixed at θ1. The distance between point O' and point A' is r, and A' and B' are centrally symmetric about point O'.

[0069] The angle between the line formed by the centers of the two square markers on the wafer (in this case, the detection pattern) at coordinates A' and B' and the line formed by the centers of the square markers AB is θ2. Since the robotic arm firmly holds the wafer, there is no rotation between the wafer and the robotic arm; therefore, the angle between OO' and A'O' is... It does not change as the robotic arm rotates around point O.

[0070] When the robotic arm's wrist axis rotates (i.e., the center O' moves around the origin O with radius R) to A"B"∥AB, it is considered that the target wafer mounting position has been reached, and the pre-alignment is completed. At this time, a Cartesian coordinate system O"-X"Y" is established with the wafer center O" as the origin, where the straight line OX∥O"X", and A"B" coincides with O"X".

[0071] ∵OX∥O"X"∥A"B"

[0072] ∴ =

[0073] Also, since OX∥O'X'∥AB

[0074] ∴

[0075] ∵

[0076] Therefore, proof is obtained.

[0077] Let point O be the center of rotation of the mechanical wrist axis. Establish a Cartesian coordinate system O-XY with point O as the origin, where the coordinates of each point are (x, y, 1). This coordinate system is the first coordinate system. Establish a Cartesian coordinate system O'-X'Y' with the wafer center O' as the origin, where the coordinates of each point are (x', y', 1). This coordinate system is the second coordinate system. The transformation formula is:

[0078] (2)

[0079] in, .

[0080] The coordinates of each point in a Cartesian coordinate system with the wafer center O' as the origin are shown in Table 1. According to the transformation formula, the coordinates of each point in a Cartesian coordinate system O-XY with point O as the origin are shown in Table 2.

[0081] Table 1. Coordinates of points in the second coordinate system

[0082]

[0083] Table 2 Coordinates of points in the first coordinate system

[0084]

[0085] According to Table 2, the formulas for the displacement in the X / Y directions before and after wrist rotation are shown in formulas (3) and (4). The calculation results are marked with "±", where "+" indicates a positive movement in the coordinate axis and "-" indicates a negative movement in the coordinate axis. The target's position (A / B) and the detection position (A / B) are known. The displacement formulas between points are shown in formulas (5) and (6), and the calculation results are marked with "±". During image processing, to prevent the acquired detection image from exceeding the field of view of the CCD camera, point O needs to be translated in the X and Y directions to bring the detection image back into the CCD field of view. The amount of translation... and The calculation formula is shown in formula (7), and the calculation result is marked with "±".

[0086] (3)

[0087] (4)

[0088] (5)

[0089] (6)

[0090] (7)

[0091] At this point, the wrist axis needs to be translated along the X+ and A+ axes. Since the X+ axis of the wrist axis is parallel to OY, and the angle formed by the A+ axis and OX is... The translation formulas of the X+ axis and A+ axis of the wrist axis are calculated based on the translation of the origin O in the X and Y directions, as shown in equations (8) and (9).

[0092] (8)

[0093] The transformation yields:

[0094] (9)

[0095] Based on the solution and The robot arm performs multi-axis end effector movement, bringing the detected pattern back into the detection range while keeping the rotation angle as constant as possible.

[0096] The following details the process of implementing the adaptive small-quantity algorithm.

[0097] Based on the above embodiments, the rotation of the robot arm according to the deflection angle in S4 includes: S41, determining the magnitude of the deviation of the deflection angle from the preset value; if the magnitude of the deviation of the deflection angle from the preset value is greater than the preset threshold, the robot arm rotates by a first fixed amount; if the magnitude of the deviation of the deflection angle from the preset value is less than the preset threshold, the robot arm rotates by a second fixed amount; the first fixed amount is greater than the second fixed amount.

[0098] As described in the multi-axis displacement calculation process above, θ2 represents the rotation of the robotic wrist axis. However, the measurement accuracy of the robotic arm's encoder or sensors limits the feedback of the minimum deflection angle to only 17 μRad. This means that during the process of gradually moving to the target position using the minimum step size, over-movement or failure to reach the target position may occur. Due to the measurement limitations of the device itself, the system may misjudge that the movement has been completed, leading to a gradual increase in cumulative error.

[0099] To address this issue, an adaptive small-amount algorithm adjustment strategy is employed in the embodiments of this disclosure. Different preset thresholds are set based on different deflection angles, and different fixed amounts are selected for rotation adjustment. Specifically, when the detected deflection angle θ2 (equivalent to the attached...)... Figure 3 When θ is of different sizes, a suitable fixed amount will be selected for rotation according to the actual situation, thereby effectively controlling the cumulative error and ensuring alignment accuracy.

[0100] Figure 8 A flowchart illustrating a pre-alignment method according to an embodiment of the present disclosure is shown schematically.

[0101] Based on the above embodiment, the rotation of the robotic arm according to the deflection angle in S4 includes: S411, dividing the value of the deflection angle deviating from the preset value into at least two levels, each level corresponding to a preset fixed amount; S412, sequentially determining whether the value of the deflection angle deviating from the preset value is within the current level; if yes, the robotic arm rotates by the preset fixed amount corresponding to the current level and returns to step S1; if no, determining whether the value of the deflection angle deviating from the preset value is within the next level, until all levels have been determined. In S411, the value of the deflection angle deviating from the preset value in each level is greater than the preset fixed amount corresponding to the level.

[0102] like Figure 8 As shown, the adaptive small-scale algorithm follows the steps after calculating the deflection angle. When the deflection angle θ2 is large, the robotic wrist axis rotates by a fixed, large angle to minimize the marker image from exceeding the CCD's field of view. When the deflection angle θ2 is small, the robotic wrist axis rotates by a fixed, small angle to control some of the accumulated error. Each rotation by a fixed, small angle ensures that even if the motion feedback from the robotic wrist is inaccurate, it will not produce a significant error. Before rotating again, the marker image is inspected, and the rotation error is estimated before proceeding with the next rotation. Controlling accumulated error in this way simplifies the automated alignment process and improves its efficiency.

[0103] For example, the deviation of the deflection angle from the preset value is divided into four levels: |deflection angle - preset value| > 0.02, |deflection angle - preset value| > 0.01, |deflection angle - preset value| > 0.005, and |deflection angle - preset value| > 0.001. The judgment is made sequentially from the level with the largest value to the level with the smallest value. If the judgment result is yes, the current level is rotated by a fixed amount, and the process returns to S1 to re-perform the mark pattern detection.

[0104] Based on the above embodiments, S1 includes detecting the marked graphics on the object to be transmitted, which includes: S11, taking pictures of the marked graphics on the object to be transmitted using a camera to obtain image information of the detected graphics; wherein the marked graphics include at least two sets; S12, filtering, binarizing, segmenting the image information and detecting the target to obtain the second position information of the detected graphics; the second position information includes the center coordinates of the detected graphics.

[0105] The robotic arm moves the object to be transferred to a position below the CCD camera. Light emitted from the light source is transmitted through a lens group to the marked area. The reflected light carrying the marked graphic information then passes through the lens group again to the CCD camera, where the image acquired by the CCD camera is processed. The image is filtered to remove noise; binarization sets the grayscale value of each pixel in the image to 0 or 255 to distinguish the image from blank areas; the grayscale image is segmented, and the unmarked areas are filled with white; finally, target detection is performed to obtain the second position information of the detected graphic, which is represented by the center coordinates of the detected graphic.

[0106] Based on the above embodiment, before determining whether the detected pattern is complete in S2, the method further includes: S20, determining whether a marker pattern is detected; if yes, proceed to step S2; if no, perform mechanical pre-alignment again.

[0107] Before determining whether the detection pattern is complete, it is also necessary to determine whether the marker pattern has been detected. If so, proceed to the subsequent pre-alignment process; otherwise, mechanical pre-alignment must be performed again.

[0108] Based on the above embodiments, S2 further includes determining whether the detection pattern is complete: S24, determining whether the marking orientation of the detection pattern is consistent; if yes, then calculate the deflection angle based on the first position information and the second position information, and proceed to step S4; if no, the robot arm performs orientation compensation based on the first position information and the second position information.

[0109] After determining that the detected image is complete, it is possible to further determine whether the orientation of the detected image is consistent with the position marker image on the target image, such as... Figure 4 As shown, if the centers of the two red boxes are respectively located to the left, upper left, above, upper right, right, lower right, lower left, and lower left of the centers of the two blue boxes, then the orientation of the detected image and the target upper image position mark are considered to be consistent. If so, orientation compensation is performed after determining the robot arm's movement direction, including compensating for X and Y displacement errors. If not, the aforementioned multi-axis displacement calculation method is used to obtain the translation amount of the robot arm axis and perform deflection angle compensation.

[0110] Based on the above embodiments, S24, calculating the deflection angle according to the first position information and the second position information, further includes: confirming the sign of the deflection angle according to the first position information and the second position information.

[0111] After calculating the deflection angle, confirming the sign of the deflection angle is to determine the direction of movement of the robot arm. The sign indicates the direction of movement: "+" indicates movement in a clockwise direction along the coordinate axis, and "-" indicates movement in a counterclockwise direction along the coordinate axis. For example... Figure 6 The arrows in the image point in the positive direction.

[0112] The pre-alignment method disclosed herein utilizes the results of multi-axis displacement calculations to guide the precise movement of the multi-axis end effector of the robot, ensuring that the marked pattern can return to the detection range while maintaining a substantially constant rotation angle. Simultaneously, this method introduces an adaptive small-amount algorithm, employing corresponding fixed small-amount rotation adjustments for different deflection angles, thereby effectively controlling cumulative errors.

[0113] The present disclosure will be further described below through specific embodiments. The method for detecting the above-mentioned step signal will be specifically described in the following embodiments. However, the following embodiments are merely illustrative of the present disclosure, and the scope of the present disclosure is not limited thereto.

[0114] Figure 9 Another flowchart of the pre-alignment method according to an embodiment of the present disclosure is illustrated schematically.

[0115] First, based on the marked image of the target's upper position and the real-time acquired detection image (a box mark in this embodiment), it is determined whether a box mark has been detected and whether the box mark is complete. If no box mark is detected, a prompt is made to re-perform mechanical pre-alignment. The mechanical pre-alignment steps can refer to general mechanical pre-alignment steps. If the detected box mark is incomplete, multi-axis displacement calculation is performed based on the first and second position information to obtain the translation amount of the robotic wrist axis, and the translation is fine-tuned using the robotic wrist axis. Then, the deflection angle θ2 is calculated, and it is determined whether the deflection angle θ2 is less than a set threshold. If it is, the pre-alignment process ends; if not, the robotic arm first rotates θ2, and then calculates... , , and Next, determine whether the box marker exceeds the CCD's field of view; if so, calculate the movement of the origin O. and and solve and Then the wrist axis translates. and If not, the robotic arm only rotates θ2. Repeat the above steps until θ2 is less than the threshold, at which point it is considered that A"B"∥AB.

[0116] The robotic arm sequentially picks up wafers from the wafer loading / unloading machine or standard mechanical interface. After passing through a mechanical pre-alignment device, the object to be transferred is transferred to the Epin module or chuck. The alignment measurement system calculates the alignment position deviation in the X direction between two consecutive alignment operations. This step is repeated n times, recording the position deviation each time and the cumulative position deviation. Finally, the average and standard deviation of the position deviations are calculated to estimate the accuracy range of the transfer system, ensuring its high precision and reliability. A schematic diagram of the Rz detection method for the transfer system is shown below. Figure 10 As shown.

[0117] In transmission systems, the method for detecting Rz in related technologies is to add the absolute values ​​of the errors of the aligned moiré fringes on both sides, i.e. ( Given that the spacing between the moiré fringes on both sides is R, the arctangent function is used to obtain the Rz data, as shown in formula (10).

[0118] (10)

[0119] Repeated tests were conducted, and the difference between the position coordinates of two consecutive tests was taken as a set of alignment position deviations. A total of n sets of alignment position deviations were obtained. ,in:

[0120] (11)

[0121] The average and standard deviation of the n sets of alignment position deviations are calculated as shown in formulas (12) and (13):

[0122] (12)

[0123] (13)

[0124] In embodiments of this disclosure, the repeatability positioning accuracy test steps for the transmission system include:

[0125] (1) The robot arm picks up the wafer from the wafer loading and unloading machine or standard mechanical interface, and transfers it to the E-pin or chuck through the optical pre-alignment device. The alignment measurement system is used to detect the alignment operation between the objects to be transferred. After the alignment is completed, the workpiece stage position coordinates are recorded. ,calculate ;

[0126] (2) The robotic arm retrieves the wafer or mask and puts it back into the front-opening universal box or standard mechanical interface;

[0127] (3) Repeat steps (1) and (2) 11 times. The difference between the position coordinates of the two consecutive times is a set of position deviations. A total of 10 sets of position deviations (such as the X direction) are obtained. As shown in Formula 14;

[0128] (14)

[0129] (4) Calculate the average and standard deviation of the 10 sets of positional deviations, as shown in formulas (15) and (16), where n=10, to obtain the X-direction accuracy of the wafer transmission system. The calculation method for the Y direction is the same as that for the X direction;

[0130] (15)

[0131] (16)

[0132] (5) The accuracy of the transmission system can be estimated according to formula (14). .

[0133] A transmission mechanism was built on a key technology experimental platform to verify the key technology of optical pre-alignment of the transmission system. During the experiment, the same object to be transmitted was repeatedly picked up and placed from a front-opening universal box or standard mechanical interface, placed in the optical pre-alignment device for detection, and then handed over to the E-pin or suction cup. The alignment measurement system calculated the difference in alignment position in the X / Y directions between two consecutive alignment operations, repeating this step n times, recording the position deviation and cumulative position deviation each time. Finally, the average value and standard deviation of the position deviation were calculated to estimate the repeatability positioning accuracy of the transmission system. The statistical results of the alignment position deviation are shown in Table 3. The transmission accuracy in the X direction is ±24.869µm, the transmission accuracy in the Y direction is ±17.778µm, and the transmission accuracy in the Rz direction is ±12.93µRad. Therefore, the embodiments of this disclosure can control the Rz deflection angle of the upper piece within 50µRad, significantly improving the pre-alignment accuracy.

[0134] Table 3 Transmission Error Statistics

[0135]

[0136] Those skilled in the art will understand that the features described in the various embodiments and / or claims of this disclosure can be combined or combined in various ways, even if such combinations or combinations are not explicitly described in this disclosure. In particular, the features described in the various embodiments and / or claims of this disclosure can be combined or combined in various ways without departing from the spirit and teachings of this disclosure. All such combinations and / or combinations fall within the scope of this disclosure.

[0137] Although this disclosure has been shown and described with reference to specific exemplary embodiments thereof, those skilled in the art will understand that various changes in form and detail may be made to this disclosure without departing from the spirit and scope of the disclosure as defined by the appended claims and their equivalents. Therefore, the scope of this disclosure should not be limited to the above embodiments, but should be defined not only by the appended claims, but also by their equivalents.

Claims

1. A pre-alignment method based on high-precision multi-axis displacement calculation, wherein a robot arm is used to transfer an object to be transferred, the object to be transferred having a marked pattern, the marked pattern having first position information at a target position on the object to be transferred, characterized in that, include: S1, detect the marked pattern on the object to be transmitted to obtain the second position information of the detected pattern; S2, determine whether the detection pattern is complete based on the detection pattern; if not, perform multi-axis displacement calculation based on the first position information and the second position information to obtain the translation amount of the mechanical wrist axis, and proceed to step S3; If so, calculate the deflection angle based on the first position information and the second position information, and proceed to step S4; S3, the robotic arm moves the object to be transferred according to the translation amount of the wrist axis, and returns to step S1; S4, determine whether the deflection angle is less than a set threshold; If not, the robotic arm rotates according to the deflection angle and returns to step S1; If so, then the pre-alignment of the object to be transmitted is completed.

2. The pre-alignment method based on high-precision multi-axis displacement calculation according to claim 1, characterized in that, The multi-axis displacement calculation based on the first position information and the second position information described in S2 includes: S21, establish a first coordinate system with the rotation center of the wrist axis of the robotic arm as the origin and establish a second coordinate system with the center of the object to be transferred as the origin; S22, In the first coordinate system, calculate the displacement of the marker graphic relative to the target position of the object to be transmitted and the displacement of the detection graphic, as well as the displacement before and after wrist rotation, based on the first position information and the second position information. S23, based on the transformation relationship between the first coordinate system and the second coordinate system, obtain the translation amount of the wrist axis in the first coordinate system.

3. The pre-alignment method based on high-precision multi-axis displacement calculation according to claim 1, characterized in that, In S4, the rotation of the robotic arm according to the deflection angle includes: S41, determine the magnitude of the deviation of the deflection angle from the preset value; if the deviation of the deflection angle from the preset value is greater than the preset threshold, the robot arm rotates by a first fixed amount; if the deviation of the deflection angle from the preset value is less than the preset threshold, the robot arm rotates by a second fixed amount; the first fixed amount is greater than the second fixed amount.

4. The pre-alignment method based on high-precision multi-axis displacement calculation according to claim 3, characterized in that, In S4, the rotation of the robotic arm according to the deflection angle includes: S411, the deviation of the deflection angle from the preset value is divided into at least two levels, each level corresponding to a preset fixed amount; S412, sequentially determine whether the value of the deviation of the deflection angle from the preset value is in the current gear; if yes, the robot arm rotates by the preset fixed amount corresponding to the current gear and returns to step S1; if no, determine whether the value of the deviation of the deflection angle from the preset value is in the next gear, until all gears are determined.

5. The pre-alignment method based on high-precision multi-axis displacement calculation according to claim 4, characterized in that, In S411, the deviation of the deflection angle of each gear from the preset value is greater than the preset fixed amount corresponding to the gear.

6. The pre-alignment method based on high-precision multi-axis displacement calculation according to claim 1, characterized in that, S1 includes detecting the marked pattern on the object to be transmitted, including: S11, use a camera to capture images of the marked patterns on the object to be transmitted to obtain image information of the detected patterns; wherein, the marked patterns include at least two sets; S12, the image information is filtered, binarized, segmented, and target detected to obtain the second position information of the detected graphic; the second position information includes the center coordinates of the detected graphic.

7. The pre-alignment method based on high-precision multi-axis displacement calculation according to claim 6, characterized in that, S2 calculates the deflection angle based on the first position information and the second position information, including: The deflection angle is calculated using the arctangent function based on the center coordinates of the detected graphic and the center coordinates of the marked graphic at the target position.

8. The pre-alignment method based on high-precision multi-axis displacement calculation according to claim 1, characterized in that, Before determining whether the detected image is complete in S2, the following steps are also included: S20, determine whether the detection pattern has been detected; if yes, proceed to step S2; if no, re-perform mechanical pre-alignment.

9. The pre-alignment method based on high-precision multi-axis displacement calculation according to claim 1, characterized in that, S2 further includes determining whether the detected image is complete: S24, determine whether the orientation of the markers in the detected pattern is consistent; if yes, calculate the deflection angle based on the first position information and the second position information, and proceed to step S4; if no, the robot arm performs orientation compensation based on the first position information and the second position information.

10. The pre-alignment method based on high-precision multi-axis displacement calculation according to claim 9, characterized in that, S24 further includes calculating the deflection angle based on the first position information and the second position information: The sign of the deflection angle is confirmed based on the first position information and the second position information.

Images