An automatic loading and unloading control system for semiconductor material plates

Abstract

The present invention relates to the field of semiconductor manufacturing technology, and in particular to an automatic loading and unloading control system for semiconductor material plates, comprising: a visual positioning and shape detection module; a clamping force dynamic control module; a path planning and obstacle avoidance module: dividing the loading and unloading area safety level based on the material plate stacking parameters, and using an improved path search algorithm to generate an obstacle avoidance trajectory; optimizing the acceleration curve of the path according to the load data of the robot arm joint; a closed-loop verification module: comparing the deviation between the actual posture of the end of the robot arm and the target coordinates, and triggering the coordinated reset of the visual positioning and shape detection module to the path planning and obstacle avoidance module if the deviation exceeds the limit; the visual positioning and shape detection module to the closed-loop verification module are executed in sequence, and the output parameters of the previous module are used as the input parameters of the next module. Through the close cooperation between the modules, efficient and accurate automatic loading and unloading control is achieved, which not only improves the automation level, but also ensures the safety and accuracy of operation.

Description

Technical Field

[0001] The present invention relates to the field of semiconductor manufacturing technology, and in particular to an automatic loading and unloading control system for semiconductor material plates. Background Art

[0002] In the semiconductor manufacturing field, the automated loading and unloading of substrates (such as ceramic or metal substrates that support wafers) is a key step in chip packaging and testing processes. As semiconductor devices develop towards higher integration and miniaturization, the thickness of substrates has gradually decreased (<0.3mm) and the requirements for surface flatness have become increasingly stringent (warpage tolerance <0.1mm). This places extremely high demands on the accuracy, stability, and environmental adaptability of the loading and unloading system. However, existing technologies have the following defects, which lead to frequent problems such as wafer breakage, positioning deviation, and low efficiency during the loading and unloading process:

[0003] Existing loading and unloading systems often use a single-wavelength light source and monocular vision positioning solutions, segmenting the edges of the sheet using a fixed threshold. However, semiconductor sheet surfaces often have metal coatings or polishing treatments, which cause strong reflections. Single-wavelength light sources are prone to overexposure of the reflective areas, resulting in loss of edge information.

[0004] Existing loading and unloading systems typically plan robot trajectories based on fixed path templates or static environment assumptions, without incorporating dynamic parameters such as the inclination angle of the sheet stack and the load on the robot joints. For example, increasing the inclination angle of the sheet stack reduces the safe motion space. Traditional A* algorithms don't incorporate the inclination parameter into the path cost function, resulting in conflicts between the planned path and the actual sheet position. Summary of the Invention

[0005] Based on the above objectives, the present invention provides an automatic loading and unloading control system for semiconductor material plates, comprising:

[0006] Visual positioning and shape detection module:

[0007] Multispectral imaging is used to eliminate the reflection interference on the surface of the sheet, and the center coordinates of the sheet are calculated by combining the sub-pixel edge positioning algorithm;

[0008] Generate a 3D point cloud model of the sheet surface based on binocular vision and extract the warpage height and edge inclination parameters;

[0009] Gripping force dynamic control module:

[0010] According to the surface topography parameters and real-time pressure distribution data, the output torque of the gripper servo motor is adjusted through the fuzzy control algorithm;

[0011] When abnormal changes in local pressure are detected, the high-frequency micro-vibration mechanism is triggered to adjust the position of the sheet;

[0012] Path planning and obstacle avoidance module:

[0013] The safety level of the loading and unloading area is divided based on the sheet stacking parameters, and an improved path search algorithm is used to generate obstacle avoidance trajectories.

[0014] Optimize the acceleration curve of the path based on the load data of the robot arm joints;

[0015] Closed-loop verification module:

[0016] Compare the deviation between the actual position of the end of the robot arm and the target coordinates. If the deviation exceeds the limit, the coordinated reset of the visual positioning and shape detection module to the path planning and obstacle avoidance module is triggered;

[0017] The visual positioning and shape detection module to the closed-loop verification module are executed in sequence, and the output parameters of the previous module are used as the input parameters of the subsequent module.

[0018] Preferably, the specific process of the multispectral imaging includes:

[0019] Based on the light reflection characteristics of the sheet material, the reflectivity difference curve in the visible light and infrared bands is calibrated through experiments;

[0020] Selecting a wavelength combination whose reflectivity difference exceeds a preset threshold, and controlling the annular array light source to light up alternately according to the sequence of the wavelength combination;

[0021] Edge contours are extracted from multi-wavelength images captured by the binocular camera, and weight coefficients are dynamically assigned according to the signal-to-noise ratio of edge contrast at each wavelength. Weighted fusion is then performed to generate a synthetic image that is resistant to reflection interference.

[0022] Preferably, the sub-pixel edge location algorithm includes:

[0023] Perform gradient direction detection on the synthetic image and screen candidate edge points whose gradient amplitude exceeds the dynamic threshold;

[0024] Perform bilinear interpolation on candidate edge points to calculate sub-pixel offsets, and generate a 3D point cloud using a binocular disparity model.

[0025] The point cloud plane equation is fitted based on the least squares method, the edge inclination is calculated by the angle between the plane normal vector and the reference plane, and the warping height is calculated by the point cloud height range.

[0026] Preferably, the parameter adjustment logic of the fuzzy control algorithm includes:

[0027] Through experimental calibration of the clamping force-deformation relationship curves of different material plates, a fuzzy rule library with material hardness and friction coefficient as input variables was established;

[0028] The deviation between the real-time pressure distribution data and the target pressure threshold, as well as the deviation change rate, are used as fuzzy input variables to output the servo motor torque correction value;

[0029] The target pressure threshold is dynamically calculated based on the distribution gradient of the surface warpage height to ensure that the clamping force and the deformation trend of the sheet are inversely compensated.

[0030] Preferably, the triggering conditions of the high-frequency micro-vibration mechanism include:

[0031] Real-time monitoring of the pressure gradient change rate in each area of the pressure sensor array. When the pressure gradient change rate in a local area exceeds the safety threshold, it is determined to be a slip risk.

[0032] The safety threshold is dynamically adjusted according to the product of the warping height of the sheet surface and the friction coefficient of the material. The greater the warping height or the lower the friction coefficient, the smaller the safety threshold.

[0033] Preferably, the obstacle avoidance strategy of the improved path search algorithm includes:

[0034] Obtaining 3D contour data of the obstacle using a ToF sensor, and expanding the boundary of the obstacle outward to form an expansion area. The distance of the expansion area is determined by the motion envelope radius of the end effector of the robotic arm.

[0035] During the path search process, the grids in the expansion area are marked as inaccessible nodes, and the dynamic weighted A* algorithm is used to bypass the inaccessible nodes;

[0036] The dynamic weights include path length weight, joint torque demand weight and acceleration smoothness weight, and each weight coefficient is allocated in real time according to the current load inertia of the robotic arm.

[0037] Preferably, the specific process of optimizing the acceleration curve includes:

[0038] Perform quintic polynomial interpolation on the angular displacement sequence of each joint of the robotic arm to generate a continuous and differentiable angular acceleration curve;

[0039] According to the maximum instantaneous torque limit of the joint motor, the derivative extreme value of the interpolation curve is calculated. If the torque limit is exceeded, the interpolation node is reallocated;

[0040] By iteratively adjusting the interpolation coefficient, the actual torque demand is always lower than the preset safety factor of the motor's rated torque.

[0041] Preferably, the deviation comparison logic of the closed-loop verification module includes:

[0042] Set dual judgment conditions of position deviation threshold and pressure deviation threshold. The position deviation threshold is dynamically calculated based on 1 / 2 of the sheet size, and the pressure deviation threshold is linearly scaled according to the material brittleness level.

[0043] When the position deviation exceeds the threshold, the visual positioning module is triggered to recalibrate the material plate coordinates; when the pressure deviation exceeds the threshold, the output torque of the clamping force control module is adjusted first.

[0044] Preferably, the system further includes a data collaboration interface:

[0045] The sheet coordinate parameters output by the visual positioning module are synchronized with the input parameters of the path planning module in real time through shared memory, and the synchronization period is adaptively adjusted according to the movement speed of the robot arm;

[0046] The pressure data of the gripping force control module and the robot arm load data of the path planning module are aligned through a unified timestamp to ensure the timing consistency of the control instructions.

[0047] Preferably, the reset logic of the system includes:

[0048] When the closed-loop verification module detects three consecutive reset commands, it switches to safe mode and starts the manual intervention protocol;

[0049] In safety mode, the robot arm returns to its initial position along a preset safety path, while the power to the gripper servo motor is turned off and the gripping force is released.

[0050] Beneficial effects of the present invention:

[0051] The high-precision position information provided by the visual positioning module provides reliable basic data for the clamping force control and path planning modules, allowing the entire system to be adjusted according to actual conditions. The clamping force dynamic control module can automatically adjust the output according to actual pressure and shape changes, greatly reducing the need for human intervention and reducing the risk of sheet material damage. The path planning and obstacle avoidance module makes the movement of the robot arm smoother through intelligent path generation and acceleration optimization, thereby improving the overall loading and unloading efficiency and the service life of the mechanical equipment. The real-time deviation correction mechanism of the closed-loop verification module can ensure that each step of the action meets the predetermined goal, and respond quickly when an abnormal situation occurs, avoiding system loss of control due to errors. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only for the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0053] Figure 1 It is a structural block diagram of the system of the present invention;

[0054] Figure 2A flowchart of the steps of the sub-pixel edge positioning algorithm in the system of the present invention;

[0055] Figure 3 Flowchart of the steps for optimizing the acceleration curve in the system of the present invention. DETAILED DESCRIPTION

[0056] The present invention is described in detail below with reference to the accompanying drawings and specific embodiments. It is also noted that, to provide a more detailed description, the following embodiments are best and preferred embodiments, and those skilled in the art may employ alternative methods for implementing certain known technologies. Furthermore, the accompanying drawings are intended only to provide a more detailed description of the embodiments and are not intended to limit the present invention.

[0057] See Figure 1-Figure 3 , an embodiment of the present invention provides an automatic loading and unloading control system for semiconductor material plates, which first eliminates the reflective interference on the surface of the material plate through multi-spectral imaging technology. The differences in the reflection characteristics of light of different wavelengths are used to effectively extract edge information, and the sub-pixel edge positioning algorithm calculates the center coordinates of the material plate. The accuracy of this step is the basis for the accuracy of the entire loading and unloading system. Then, based on binocular vision technology, a three-dimensional point cloud model of the material plate surface is generated. Through the processing of point cloud data, the system can extract key parameters such as warping height and edge inclination, which will be passed to subsequent modules as input for clamping force control and path planning.

[0058] The dynamic gripping force control module receives surface topography data and real-time pressure distribution data from the vision positioning and topography detection module, and uses a fuzzy control algorithm to adjust the output torque of the gripper's servo motor in real time. This allows the gripper to dynamically adjust gripping force based on sheet surface deformation, preventing damage to the sheet due to improper gripping. If the system detects abnormal local pressure changes, a micro-vibration mechanism automatically triggers to adjust the sheet's position, ensuring grip stability and further preventing slippage or damage.

[0059] The path planning and obstacle avoidance module uses sheet stacking parameters to classify the loading and unloading area into safety levels and employs an improved path search algorithm to generate obstacle avoidance trajectories. During path planning, the module optimizes the path acceleration curve based on the robot arm's joint load data to ensure the robot does not exceed its load capacity during movement, thereby preventing mechanical failure. By dynamically adjusting the path acceleration curve, impact and vibration are reduced during movement, improving the smoothness of the loading and unloading process.

[0060] The closed-loop verification module compares the actual position of the end-arm against the target coordinates in real time. If the position deviation exceeds a set threshold, the closed-loop verification module triggers a coordinated reset of the aforementioned visual positioning and shape detection module and the path planning and obstacle avoidance module, ensuring the arm's repositioning and replanning of its path, ensuring continued stable system operation. This allows the system to automatically correct its operation when deviations occur, preventing sheet material misoperation caused by position deviations.

[0061] In one possible implementation, the system first calibrates the reflectivity differences between the visible and infrared bands based on the material of the substrate. The differences in reflectivity between substrates of different materials and wavelengths can significantly affect image quality. Therefore, experimentally measuring the reflectivity curve allows for a precise understanding of the reflectivity characteristics of the substrate surface. In this step, the reflectivity of the substrate is measured at different wavelengths and its reflectivity characteristics are recorded. The resulting reflectivity difference curve provides a basis for subsequent light source selection and image processing.

[0062] By analyzing the reflectivity difference curve, the system can select a wavelength combination in which the reflectivity difference between the visible light and infrared bands exceeds a preset threshold. This wavelength combination can maximize the recognition of surface details of the material sheet while avoiding the influence of reflection interference. After selecting the appropriate wavelength combination, the system will control the annular array light source to light up alternately in the order of these wavelengths. The alternating lighting of the light source enables the imaging system to capture the detailed information of the material sheet in different bands, especially in areas with strong surface reflection, and reduce reflection interference through the selection of wavelengths.

[0063] The binocular camera captures multi-wavelength images after alternating illumination with light sources of different wavelengths. During this step, the system extracts edge contours from these images. Because edge contrast and signal-to-noise ratios may vary across different wavelengths, to ensure the quality of the composite image, the system dynamically assigns weighting coefficients based on the edge contrast and signal-to-noise ratio for each wavelength. Specifically, bands with high signal-to-noise ratios are assigned higher weights, while bands with low signal-to-noise ratios are assigned lower weights. These weighted images are then fused to produce a composite image that is resistant to reflection interference.

[0064] In one possible implementation, the system first performs gradient direction detection on the weighted fusion composite image. Gradient direction detection calculates the local gradient values of the image to identify areas of drastic change within the image, which typically correspond to the edges of objects. To further improve detection accuracy, the system filters out candidate edge points whose gradient amplitude exceeds a set dynamic threshold. The dynamic threshold setting takes into account variations in ambient lighting and image quality, thus adapting to image processing requirements in different situations and ensuring accurate edge point screening.

[0065] Once the candidate edge points have been identified, the next task is to precisely determine the edge's actual location. In this step, the system performs bilinear interpolation on the candidate edge points to obtain sub-pixel offsets. This method allows for detailed calculations between image pixels to determine the exact edge position, achieving sub-pixel positioning accuracy. Subsequently, using a binocular parallax model, the system combines the image information acquired from the two viewpoints to calculate the 3D coordinates of each edge point, generating a 3D point cloud. This 3D point cloud provides a detailed description of the sheet's surface structure and serves as the foundation for subsequent plane fitting and warpage height detection.

[0066] Based on the generated 3D point cloud, the system then uses the least squares method to fit a plane to the point cloud. The least squares method effectively fits the point cloud data and finds an optimal plane equation to represent the sheet surface. By fitting the plane equation, the system can calculate the plane normal vector and use the angle between the normal vector and the reference plane to solve for the edge inclination angle, that is, the degree of inclination of the sheet surface relative to the reference plane. At the same time, the system can also calculate the warpage height of the sheet based on the height range of the point cloud, that is, the height variation of the sheet surface caused by manufacturing or other factors.

[0067] In one possible implementation, to accurately control the clamping force of sheets of different materials, the system first calibrates the clamping force-deformation relationship curves for these materials through experiments. This experimental data helps the system establish the deformation characteristics of each material under different clamping forces. These relationship curves serve as an important basis for further developing the fuzzy rule base of the fuzzy control algorithm. The input variables in the fuzzy rule base primarily include material hardness and friction coefficient, which directly affect the relationship between the clamping force and deformation of the sheet.

[0068] Materials with high hardness tend to be less prone to deformation and require less clamping force to prevent damage, while materials with a high coefficient of friction may require greater clamping force to ensure a stable grip. Using a fuzzy control system's rule base, the system can flexibly adjust the clamping force based on these input variables (material hardness, coefficient of friction), ensuring that the clamping process does not cause excessive deformation or damage to the sheet.

[0069] During actual operation, the system collects pressure distribution data in real time and compares it with a preset target pressure threshold to calculate the deviation and rate of change. The target pressure threshold is typically based on a preset clamping force requirement for the sheet surface and is adjusted based on actual conditions. The real-time pressure distribution provides feedback on the current clamping state of the sheet. These deviations and rates of change serve as input variables for the fuzzy control algorithm, which then outputs a torque correction for the servo motor to adjust the clamping force.

[0070] This process ensures the system can dynamically adjust based on real-time feedback, rather than relying on static preset conditions. Through fuzzy control algorithms, the clamping force can be intelligently adjusted based on changes in pressure distribution, avoiding sheet damage or unstable clamping caused by uneven pressure or inappropriate clamping force.

[0071] To better adapt to sheet surface variations, the system dynamically calculates the distribution gradient of the sheet's surface warpage height and adjusts the target pressure threshold. The surface warpage gradient refers to the rate of change or unevenness of the sheet's surface warpage. Because warpage often leads to uneven pressure distribution across different areas of the sheet, the system requires a compensation strategy to ensure uniform clamping force.

[0072] In practice, areas with significant warpage may require less clamping force to prevent further deformation, while areas with less warpage or flatness can be clamped with greater force. By dynamically calculating the distribution gradient of surface warpage height, the system can compensate for this, adjusting the target pressure threshold and automatically modifying the clamping force distribution based on the current warpage state to ensure inverse compensation between sheet deformation trends and clamping force.

[0073] In one possible implementation, an array of pressure sensors deployed in different areas monitors the rate of change of pressure gradients in each area in real time. The pressure gradient refers to the difference in pressure between points on the sheet surface, while the rate of change reflects the rate at which the pressure distribution changes over time or location. By monitoring pressure gradient changes, the system can accurately detect pressure non-uniformity on the sheet surface.

[0074] The pressure sensor array acquires pressure data, converts it into analyzable values, and then calculates the pressure gradient in each area. If the pressure gradient in a certain area changes too quickly, it usually means that the pressure distribution in certain parts of the sheet surface has changed dramatically, which may cause slippage or instability on the sheet surface. In this case, the system will identify the potential slip risk and respond accordingly.

[0075] When the system detects that the rate of change in the pressure gradient in a local area exceeds a set safety threshold, it determines that there is a risk of slippage in that area. This risk means that the pressure distribution on the sheet in that area may be uneven, causing the sheet to shift or deflect during automated loading and unloading, thus affecting loading and unloading accuracy and safety.

[0076] The system's safety threshold isn't a fixed value; it dynamically adjusts based on the sheet's actual surface conditions and material properties. The safety threshold is adjusted based on two primary factors: the sheet's surface warpage height and the material's coefficient of friction. Specifically, a greater warpage height indicates a more uneven sheet surface, and a lower coefficient of friction indicates a greater likelihood of sheet slippage during clamping. Therefore, as the warpage height increases or the coefficient of friction decreases, the system automatically lowers the safety threshold to more sensitively detect slippage risks.

[0077] When the sheet's surface warps, the pressure distribution in certain areas is more likely to fluctuate dramatically. The system needs to lower the safety threshold to detect abnormal pressure gradient changes early. Conversely, when the warp height is small and the friction coefficient is high, the risk of slippage is relatively low, and the safety threshold can be appropriately increased to avoid overreaction.

[0078] When the system detects through pressure sensors that the rate of change of pressure gradient in a local area exceeds a safety threshold and identifies a slip risk, it triggers a micro-vibration response via a high-frequency micro-vibration mechanism. This high-frequency micro-vibration adjusts the sheet's state through tiny local vibrations, redistributing pressure in the local area and helping to eliminate the risk of slip caused by uneven pressure distribution, thereby preventing positioning deviations or unstable clamping during loading and unloading.

[0079] In one possible implementation, the system uses a Time of Flight (ToF) sensor to acquire real-time 3D profile data of obstacles in the environment. By emitting laser light and measuring its reflection time, the ToF sensor accurately captures the distance between the obstacle and the sensor, generating a 3D image of the obstacle. This data provides a detailed description of the obstacle's spatial position and shape, providing fundamental information for path planning algorithms.

[0080] After acquiring the 3D outline data of an obstacle, the system dilates the obstacle's boundaries to form an expansion region. This expansion process simulates the actual range of motion of the robot's end effector by extending the obstacle's boundaries outward by a certain distance. The specific expansion distance is determined by the end effector's motion envelope radius. The motion envelope radius refers to the spatial range occupied by the end effector (such as a gripper or suction cup) during movement. Therefore, the expansion region should cover any space within this area where collisions may occur.

[0081] During the path search process, all grid nodes within the expanded area are marked as impassable, meaning they cannot be used as paths during path planning. This allows the path search algorithm to avoid collisions between the robot arm and obstacles. To achieve obstacle avoidance, the system uses the dynamic weighted A* algorithm, an improved version of the A* algorithm that considers more dynamic factors when calculating the path, finding an optimal path that bypasses impassable nodes.

[0082] The dynamic weighted A* algorithm is an extension of the A algorithm, primarily used for path planning problems, particularly in robotic navigation and autonomous driving. Unlike the standard A algorithm, the dynamic weighted A* algorithm dynamically adjusts the weights of the heuristic function to control the balance during the search process, thereby adapting to different environments or task requirements.

[0083] In the dynamic weight A* algorithm, the system uses dynamic weights to optimize path selection. The dynamic weights include the following aspects:

[0084] Path length weight: The shorter the total length of the path, the better. The system will tend to choose a shorter path to reduce movement time.

[0085] Joint torque requirement weight: Taking into account the dynamic characteristics of the robot arm, paths with smaller joint torque requirements are prioritized during path planning, which helps reduce energy consumption and improve the motion stability of the robot arm.

[0086] Acceleration smoothness weight: The smoother the acceleration change of the path, the better. The system avoids sudden acceleration changes by smoothing the acceleration curve of the path, which helps to extend the service life of the robot arm and reduce unnecessary mechanical wear.

[0087] These weights are assigned in real time based on the robot's current load inertia. When the load inertia is large, the system increases the joint torque demand and smoothness weights to avoid excessive torque burdens and acceleration variations. When the load is light or unloaded, the path length weight is increased, and the system tends to choose shorter routes for improved efficiency.

[0088] In one possible implementation, the system first performs quintic polynomial interpolation on the angular displacement sequence of each joint of the robotic arm. Quintic polynomial interpolation is a commonly used curve fitting technique that generates a smooth, continuous, and differentiable acceleration curve between given discrete angular displacement data points. This interpolation method ensures that the angular velocity and angular acceleration of each joint of the robotic arm change continuously during the transition process, avoiding discontinuities and sudden changes in acceleration, which is crucial for the robotic arm's kinematic stability.

[0089] Because each joint in the robotic arm is driven by a motor, and the motor's maximum instantaneous torque is limited, the generated angular acceleration curve may contain points where the derivative extremes (i.e., the maximum rate of change of acceleration) exceed the motor's maximum instantaneous torque limit. To prevent the motor from overloading at these points, the system calculates the derivative extremes of the interpolated curve and compares them with the joint motor's maximum torque limit. If the torque demand at a particular point exceeds the motor's capacity, the system reallocates the interpolation nodes, adjusts the curve's shape, and reduces the severity of the acceleration change, ensuring that the motor's torque does not exceed its maximum tolerance throughout the entire motion.

[0090] To further optimize the acceleration curve and ensure smooth operation of each joint throughout the entire motion process, the system iteratively adjusts the interpolation coefficients, gradually adjusting the interpolation curve shape to ensure that the actual torque demand always remains within a preset safety factor of the motor's rated torque. This iterative process repeatedly calculates the load on each joint and adjusts the interpolation coefficients in real time to optimize the acceleration curve, ensuring that the system load does not exceed the rated load of the motor.

[0091] In one possible implementation, in order to effectively monitor deviations during system operation, the closed-loop verification module needs to set two key deviation thresholds:

[0092] Position Deviation Threshold: This threshold is dynamically calculated based on the sheet size, specifically 1 / 2 of the sheet size. This is because semiconductor sheet sizes can vary, and dynamically calculating the position deviation threshold ensures the system maintains an appropriate tolerance when handling sheets of varying sizes. Specifically, the maximum allowable position deviation is calculated as half of the sheet's length and width, effectively controlling and correcting position errors during sheet movement.

[0093] Pressure Deviation Threshold: This threshold scales linearly based on the sheet material's brittleness level. Different sheet materials have varying tolerances to pressure, with more brittle materials requiring more precise pressure control. Therefore, by linearly scaling the pressure deviation threshold based on the material's brittleness level, the system can provide real-time control of pressure deviation based on the sheet material's varying requirements, preventing damage to the sheet due to excessive pressure.

[0094] When the system detects a position deviation exceeding a set threshold, the visual positioning module is triggered to recalibrate the sheet's coordinates. The visual positioning module monitors the sheet's position and posture in real time using a camera or other sensor. If the position deviation exceeds a set value, the system recalibrates the position to correct the sheet's coordinates. This process effectively corrects positioning deviations caused by robotic arm operation errors or external interference, ensuring accurate sheet placement and retrieval.

[0095] When the system detects a pressure deviation exceeding a set threshold, it prioritizes adjusting the output torque of the clamping force control module. The clamping force control module regulates the clamp's grip on the sheet. Excessive pressure deviation may indicate inadequate gripping force, posing a risk to the sheet's safety. By adjusting the output torque, the clamp's gripping force is effectively controlled, avoiding excessive pressure on the sheet, thereby preventing deformation or cracking during handling, especially with brittle materials.

[0096] In one possible implementation, the visual positioning module uses image recognition technology to acquire the real-time coordinate information of the sheet material, while the path planning module uses this coordinate information to plan the robot's motion trajectory. To ensure high consistency between the two data types, the system uses shared memory to synchronize the sheet material coordinate parameters output by the visual positioning module with the input parameters of the path planning module in real time.

[0097] Because the robot's movement speed and the size and characteristics of the sheet material (such as weight and material) affect the motion cycle, the system has designed an adaptive synchronization cycle mechanism. Specifically, the synchronization cycle is automatically optimized based on the robot's current movement speed. For example, when the robot's movement speed is fast, the synchronization cycle may be shortened to ensure that the real-time coordinates output by the visual positioning module are transmitted to the path planning module in a timely manner, avoiding trajectory planning errors caused by data delays. When the movement speed is slow, the synchronization cycle can be appropriately extended to improve system efficiency.

[0098] The beneficial effect of this design is that it can ensure a high degree of synchronization between visual information and path planning information within the system, eliminating the problem of inaccurate control caused by information lag, thereby improving the accuracy and real-time performance of the automatic loading and unloading process.

[0099] The gripping force control module is responsible for adjusting the gripper's gripping force, while the path planning module optimizes trajectory planning and control strategies based on the robot's load. To ensure coordination between gripping force and robot load, the system precisely aligns the pressure data output by the gripping force control module with the load data from the path planning module through unified timestamp alignment.

[0100] All data (including pressure and load data) is time-stamped, ensuring that the path planning module receives the correct pressure and load data at the same time. This alignment effectively avoids control timing errors caused by inconsistent data updates, such as excessive or insufficient gripping force or misjudgment of the robot's load, which can lead to improper operation.

[0101] The beneficial effect of this step is that by unifying the timestamp alignment data, it can ensure that each control instruction of the system has consistent timing, thereby improving the system's comprehensive coordination and response speed to load, pressure and clamping force, and optimizing the accuracy and stability of the entire automatic loading and unloading process.

[0102] Through the design of a data collaboration interface, the system ensures real-time information sharing and precise synchronization between different modules, avoiding misoperation caused by data delays, inconsistent timing, or information lags. Specifically, the synchronized adjustment of the visual positioning module and the path planning module enables the robot arm to adjust its motion path in real time to accommodate changes in the sheet's position, while the alignment of the gripping force and load ensures that the robot arm does not cause improper gripping or equipment damage due to pressure mismatch during operation.

[0103] This collaborative mechanism enables the system to maintain high consistency across multiple operational links, reducing errors, improving operational efficiency, and enhancing the precision and reliability of the automated process. It also improves the system's adaptability, automatically adjusting operational strategies to ensure safe and efficient loading and unloading of semiconductor sheets of varying sizes, weights, and materials.

[0104] In one possible implementation, the system uses a closed-loop verification module to monitor device status in real time, specifically the issuance of reset commands. This module continuously receives command information from other system components (such as the robotic arm control module and the gripper control module) and verifies reset commands. If the module detects three consecutive reset commands, it indicates that the system may have encountered an error or fault that cannot be automatically recovered.

[0105] When the system experiences certain faults or abnormal conditions (such as a stuck robot arm, sensor errors, or operational errors), a reset command is issued, and the system attempts to recover. If three reset commands are issued consecutively within a short period of time, the closed-loop verification module will determine that the current state cannot be resolved through a normal reset, and the system should enter safe mode.

[0106] If the closed-loop verification module detects three consecutive reset commands, the system automatically switches to safe mode and initiates the manual intervention protocol. Safe mode is designed to isolate the device from a potentially hazardous state, preventing the robotic arm from continuing to operate and potentially causing further damage to the device or personal injury.

[0107] After entering safe mode, the system prompts the operator to intervene manually to check and correct the system fault. The operator can obtain detailed fault information through the system control interface and take appropriate remedial measures. At this point, the system will not continue to perform automatic operations until the manual intervention is completed.

[0108] In safe mode, the robot arm will return to its initial position along a pre-defined safe path. This pre-defined safe path is a pre-planned path that avoids hazardous areas and typically avoids high-risk or unstable operating areas, such as damaged equipment or areas where people are active. This design ensures that the robot arm will not continue to perform potentially accidental operations while in a faulty state.

[0109] A pre-set safety path ensures the robot arm returns safely and smoothly to its starting position, avoiding collisions with other equipment or personnel. This path is typically carefully designed and validated by system engineers to minimize risk and avoid any unsafe aspects of the robot arm's movement.

[0110] Safety mode also includes a function that shuts down the gripper's servo motor and releases the gripping force. This ensures that the gripper does not continue to apply unnecessary force during its return to its initial position, potentially damaging the sheet or other objects. Furthermore, shutting down the gripper's servo motor prevents continued operation during a fault condition, mitigating electrical risks.

[0111] Releasing the clamping force can effectively prevent the clamping jaws from being unable to release in time when a system failure occurs, thereby avoiding damage to mechanical parts or sheet metal due to excessive clamping force or failure to release the clamping jaws.

[0112] The reset logic design ensures that the system can quickly enter a safe state when encountering an anomaly by monitoring reset instructions, switching to safe mode, returning to a safe path, and shutting down the power to the gripper. It also provides effective protection for manual intervention, thereby greatly improving the safety and reliability of the semiconductor material plate automatic loading and unloading control system.

[0113] The present invention encompasses any alternatives, modifications, equivalents, and solutions that fall within the spirit and scope of the present invention. To provide a thorough understanding of the present invention, specific details are described in detail below in connection with the preferred embodiments of the present invention, but those skilled in the art will be able to fully understand the present invention without these detailed descriptions. Furthermore, to avoid unnecessary confusion regarding the essence of the present invention, well-known methods, processes, procedures, components, and circuits have not been described in detail.

[0114] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications should also be regarded as within the scope of protection of the present invention.

Claims

1. An automatic loading and unloading control system for semiconductor material plates, characterized in that: It includes the following modules connected in sequence: Visual positioning and shape detection module: Multispectral imaging is used to eliminate the reflection interference on the surface of the sheet, and the center coordinates of the sheet are calculated by combining the sub-pixel edge positioning algorithm; Generate a 3D point cloud model of the sheet surface based on binocular vision and extract the warpage height and edge inclination parameters; Dynamic clamping force control module: According to the surface topography parameters and real-time pressure distribution data, the output torque of the gripper servo motor is adjusted through the fuzzy control algorithm; When abnormal changes in local pressure are detected, the high-frequency micro-vibration mechanism is triggered to adjust the position of the sheet; Path planning and obstacle avoidance module: The safety level of the loading and unloading area is divided based on the stacking parameters of the sheet materials, and the obstacle avoidance trajectory is generated using an improved path search algorithm. The stacking parameters include warping height and edge inclination parameters. Optimize the acceleration curve of the path based on the load data of the robot arm joints; The obstacle avoidance strategy of the improved path search algorithm includes: Obtaining 3D contour data of the obstacle using a ToF sensor, and expanding the boundary of the obstacle outward to form an expansion area. The distance of the expansion area is determined by the motion envelope radius of the end effector of the robotic arm. During the path search process, the grids in the expansion area are marked as inaccessible nodes, and the dynamic weighted A* algorithm is used to bypass the inaccessible nodes; The dynamic weights include path length weight, joint torque demand weight, and acceleration smoothness weight, and each weight coefficient is allocated in real time according to the current load inertia of the manipulator; Closed-loop verification module: Compare the deviation between the actual position of the end of the robot arm and the target coordinates. If the deviation exceeds the limit, the coordinated reset of the visual positioning and shape detection module to the path planning and obstacle avoidance module is triggered; The visual positioning and shape detection module to the closed-loop verification module are executed in sequence, and the output parameters of the previous module are used as the input parameters of the next module.

2. The automatic loading and unloading control system for semiconductor material plates according to claim 1, characterized in that: The specific process of the multispectral imaging includes: Based on the light reflection characteristics of the sheet material, the reflectivity difference curve in the visible light and infrared bands is calibrated through experiments; Selecting a wavelength combination whose reflectivity difference exceeds a preset threshold, and controlling the annular array light source to light up alternately according to the sequence of the wavelength combination; Edge contours are extracted from multi-wavelength images captured by the binocular camera, and weight coefficients are dynamically assigned according to the signal-to-noise ratio of edge contrast at each wavelength. Weighted fusion is then performed to generate a synthetic image that is resistant to reflection interference.

3. The automatic loading and unloading control system for semiconductor material plates according to claim 2, characterized in that: The sub-pixel edge positioning algorithm includes: Perform gradient direction detection on the synthetic image and filter candidate edge points whose gradient amplitude exceeds the dynamic threshold; Perform bilinear interpolation on candidate edge points to calculate sub-pixel offsets, and generate a 3D point cloud using a binocular disparity model. The point cloud plane equation is fitted based on the least squares method, the edge inclination is calculated by the angle between the plane normal vector and the reference plane, and the warping height is calculated by the point cloud height range.

4. The automatic loading and unloading control system for semiconductor material plates according to claim 1, characterized in that: The parameter adjustment logic of the fuzzy control algorithm includes: Through experimental calibration of the clamping force-deformation relationship curves of different material plates, a fuzzy rule library with material hardness and friction coefficient as input variables was established; The deviation between the real-time pressure distribution data and the target pressure threshold, as well as the deviation change rate, are used as fuzzy input variables to output the servo motor torque correction value; The target pressure threshold is dynamically calculated based on the distribution gradient of the surface warpage height to ensure that the clamping force and the deformation trend of the sheet are inversely compensated.

5. The automatic loading and unloading control system for semiconductor material plates according to claim 4, characterized in that: The triggering conditions of the high-frequency micro-vibration mechanism include: Real-time monitoring of the pressure gradient change rate in each area of the pressure sensor array. When the pressure gradient change rate in a local area exceeds the safety threshold, it is determined to be a slip risk. The safety threshold is dynamically adjusted according to the product of the warping height of the sheet surface and the friction coefficient of the material. The greater the warping height or the lower the friction coefficient, the smaller the safety threshold.

6. The automatic loading and unloading control system for semiconductor material plates according to claim 1, characterized in that: The specific process of the acceleration curve optimization includes: Perform quintic polynomial interpolation on the angular displacement sequence of each joint of the robotic arm to generate a continuous and differentiable angular acceleration curve; According to the maximum instantaneous torque limit of the joint motor, the derivative extreme value of the interpolation curve is calculated. If the torque limit is exceeded, the interpolation node is reallocated; By iteratively adjusting the interpolation coefficient, the actual torque demand is always lower than the preset safety factor of the motor's rated torque.

7. The automatic loading and unloading control system for semiconductor material plates according to claim 1, characterized in that: The deviation comparison logic of the closed-loop verification module includes: Set dual judgment conditions of position deviation threshold and pressure deviation threshold. The position deviation threshold is dynamically calculated based on 1 / 2 of the sheet size, and the pressure deviation threshold is linearly scaled according to the material brittleness level. When the position deviation exceeds the threshold, the visual positioning module is triggered to recalibrate the material plate coordinates; when the pressure deviation exceeds the threshold, the output torque of the clamping force control module is adjusted first.

8. The automatic loading and unloading control system for semiconductor material plates according to claim 1, characterized in that: The system also includes a data collaboration interface: The sheet coordinate parameters output by the visual positioning module are synchronized with the input parameters of the path planning module in real time through shared memory, and the synchronization period is adaptively adjusted according to the movement speed of the robot arm; The pressure data of the gripping force control module and the robot arm load data of the path planning module are aligned through a unified timestamp to ensure the timing consistency of the control instructions.

9. The automatic loading and unloading control system for semiconductor material plates according to claim 1, characterized in that: The reset logic of the system includes: When the closed-loop verification module detects three consecutive reset commands, it switches to safe mode and starts the manual intervention protocol; In safety mode, the robot arm returns to its initial position along a preset safety path, while the power to the gripper servo motor is turned off and the gripping force is released.

Images