Semiconductor component cutting control method and system supporting automatic deviation correction

Abstract

The invention provides a semiconductor component cutting control method and system supporting automatic deviation rectification, and relates to the technical field of deviation rectification control, and the method comprises the steps: executing the state recognition of a semiconductor component and the initialization of cutting equipment, and carrying out the positioning of a cutting stage, and obtaining a cutting period; establishing a variation predictor, and performing deviation rectification plan deployment under the cutting period on the intrinsic state information based on the variation predictor; and real-time micro-correction control is executed based on the correction plan, the predicted variation is compressed back to the allowable tolerance range, and a correction processing result is obtained. By means of the cutting control method and device, the technical problem that in the prior art, due to the interweaving effect of various factors of components, displacement deviation or structural deformation or contour errors are prone to occurring in the cutting process, and the cutting precision and the product quality are further affected can be solved, accurate cutting control based on real-time data and dynamic adjustment is achieved, and the product quality is improved. The technical effect of improving the intelligence, the self-adaptability and the accuracy of the cutting process is achieved.

Description

Technical Field

[0001] The present application relates to the field of deviation correction control technology, and in particular to a semiconductor component cutting control method and system supporting autonomous deviation correction. Background Art

[0002] In the current cutting and manufacturing process of semiconductor components, high-precision cutting equipment combined with CNC systems is commonly used to accurately process component geometry in order to meet the micron- or even nanometer-level dimensional control requirements of precision devices. However, in actual applications, due to the interweaving of multiple factors such as the component's material properties, thermal effects during the manufacturing process, differences in equipment operating stability, and external vibrations, the cutting process is prone to problems such as displacement deviation, structural deformation, or contour errors. These errors are often nonlinear, discontinuous, strongly coupled, and difficult to detect in advance, making traditional control methods that rely on fixed paths, static compensation, or preset tool compensation values ​​difficult to cope with dynamic changes.

[0003] In summary, the existing technology has technical problems such as displacement deviation, structural deformation or contour error that are prone to occur during the cutting process due to the interweaving of multiple factors such as component material properties, thermal effects, differences in equipment operation stability and external vibrations, further affecting the cutting accuracy and product quality. Summary of the Invention

[0004] The purpose of this application is to provide a semiconductor component cutting control method and system that supports autonomous correction, so as to solve the technical problem in the prior art that due to the interweaving of multiple factors such as component material properties, thermal effects, differences in equipment operation stability and external vibrations, displacement deviations, structural deformations or contour errors are easily caused during the cutting process, further affecting the cutting accuracy and product quality.

[0005] In view of the above problems, the present application provides a semiconductor component cutting control method and system that supports autonomous correction.

[0006] In the first aspect, the present application provides a semiconductor component cutting control method that supports autonomous correction, which is implemented through a semiconductor component cutting control system that supports autonomous correction, including: executing semiconductor component state identification and cutting equipment initialization, and performing cutting stage positioning to obtain a cutting cycle; establishing a variation predictor, and deploying a correction plan under the cutting cycle for the intrinsic state information based on the variation predictor; executing real-time micro-correction control based on the correction plan, and compressing the predicted variation back to the allowable tolerance range to obtain a correction processing result.

[0007] Preferably, the semiconductor component cutting control method supporting autonomous deviation correction also includes: obtaining the intrinsic state information through material property collection, geometric dimension measurement, surface stress detection and structural stability judgment of the semiconductor component; importing target cutting indicators and target cutting indicator values ​​into the cutting equipment to obtain cutting task requirements; locating the cutting stage based on the intrinsic state information and the cutting task requirements to obtain the cutting cycle.

[0008] Preferably, the semiconductor component cutting control method supporting autonomous correction also includes: generating a control cycle with the intrinsic state information as the control starting point and the cutting task requirement as the control end point; dividing the control cycle by the spatial dimension to obtain a spatially divided control cycle; dividing the control cycle by equal periods by the time dimension to obtain a temporally divided control cycle; performing weighted calculation on the spatially divided control cycle and the temporally divided control cycle to obtain the cutting cycle, wherein the cutting cycle includes a state monitoring threshold obtained based on the intrinsic state information and the cutting task requirement.

[0009] Preferably, the semiconductor component cutting control method that supports autonomous correction also includes: extracting historical mutation positions and historical deviation vectors from historical cutting data to obtain deviation features; performing supervised training based on the deviation features and historical cutting control data in the historical cutting data to establish the mutation predictor, wherein the mutation predictor has a decision linkage with the cutting control interface, and is visually displayed on the cutting control interface based on the decision linkage.

[0010] Preferably, the semiconductor component cutting control method supporting autonomous correction also includes: performing variation prediction on the intrinsic state information based on the variation predictor to obtain the predicted variation; positioning in the cutting cycle according to the predicted variation position of the predicted variation to obtain the predicted variation stage and the corresponding variation stage state monitoring threshold; performing variation correction simulation within the variation stage state monitoring threshold based on the predicted deviation vector to generate the correction plan.

[0011] Preferably, the semiconductor component cutting control method supporting autonomous correction also includes: if the number of correction plans exceeds the preset number of plans, setting a conditional trigger mechanism to trigger the correction plan: static threshold triggering based on the real-time acquisition status of the semiconductor component; dynamic trend triggering when the predicted anomaly grows continuously; multi-condition joint triggering based on the auxiliary correction index and the anomaly stage status monitoring threshold.

[0012] Preferably, the semiconductor component cutting control method supporting autonomous correction also includes: adjusting the predicted cutting control data corresponding to the predicted variation through the correction plan to generate an initial correction processing result; obtaining a real-time acquisition state according to the initial correction processing result; comparing the real-time acquisition state with the pre-correction state of the correction plan to obtain a correction state to be supplemented; presetting an allowable tolerance range for the correction plan, and obtaining a corresponding allowable tolerance correction execution result based on the allowable tolerance range; if the allowable tolerance correction state of the allowable tolerance correction execution result does not meet the correction state to be supplemented, performing supplementary correction in the next stage of the predicted variation stage until the correction state to be supplemented is met, and obtaining the correction processing result.

[0013] Preferably, the semiconductor component cutting control method that supports autonomous correction also includes: the micro-correction control includes automatic correction and human-computer interaction correction, the automatic correction is to trigger a correction plan based on the real-time acquisition status, and the human-computer interaction correction is to select fine-tuning parameters based on the cutting control interface to obtain the correction processing result.

[0014] Preferably, the semiconductor component cutting control method supporting autonomous deviation correction further includes: retraining parameters of the variation predictor, wherein the retraining process performs incremental learning based on a dynamic sampling window, and determines whether to update the variation predictor based on a preset stability threshold.

[0015] In the second aspect, the present application also provides a semiconductor component cutting control system that supports autonomous correction, which is used to execute a semiconductor component cutting control method that supports autonomous correction as described in the first aspect, including: a cutting stage positioning module, which is used to perform semiconductor component state identification and cutting equipment initialization, and perform cutting stage positioning to obtain a cutting cycle; a correction plan deployment module, which is used to establish a variation predictor, and deploy a correction plan under the cutting cycle for the intrinsic state information based on the variation predictor; a correction processing result acquisition module, which is used to perform real-time micro-correction control based on the correction plan, and compress the predicted variation back to the allowable tolerance range to obtain a correction processing result.

[0016] The technical solution provided in this application has at least the following technical effects or advantages: by realizing precise cutting control based on real-time data and dynamic adjustment, the technical effect of improving the intelligence, adaptability and accuracy of the cutting process is achieved.

[0017] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, which can be implemented in accordance with the contents of the description, and to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are specifically listed below. It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present application, nor is it intended to limit the scope of the present application. Other features of the present application will become easy to understand through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions in this application 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 merely exemplary, and a person of ordinary skill in the art can obtain other drawings based on the provided drawings without creative work.

[0019] Figure 1 This is a flow chart of a semiconductor component cutting control method supporting autonomous deviation correction according to the present application.

[0020] Figure 2 This is a structural diagram of a semiconductor component cutting control system that supports autonomous correction in this application.

[0021] Description of the accompanying drawings: cutting stage positioning module 11, correction plan deployment module 12, correction processing result acquisition module 13. DETAILED DESCRIPTION

[0022] This application provides a semiconductor component cutting control method and system that supports autonomous deviation correction. This solves the existing technical problem that displacement deviation, structural deformation, or contour error are easily caused during the cutting process due to the intertwined effects of multiple factors such as component material properties, thermal effects, differences in equipment operating stability, and external vibration, further affecting cutting accuracy and product quality. This method achieves precise cutting control based on real-time data and dynamic adjustment, achieving the technical effect of improving the intelligence, adaptability, and accuracy of the cutting process.

[0023] Below, the technical solutions in this application will be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of this application, rather than all the embodiments of this application. It should be understood that this application is not limited to the example embodiments described herein. Based on the embodiments of this application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application. It should also be noted that, for the convenience of description, only the parts related to this application, rather than all of them, are shown in the accompanying drawings.

[0024] For example 1, please refer to the attached Figure 1 The present application provides a semiconductor component cutting control method supporting autonomous deviation correction, which is applied to a semiconductor component cutting control system supporting autonomous deviation correction, and specifically includes the following steps: S1: Execute semiconductor component status recognition and cutting equipment initialization, and perform cutting stage positioning to obtain the cutting cycle.

[0025] Specifically, before cutting begins, the semiconductor components to be processed are comprehensively inspected and feature extracted, including material properties such as crystal type and conductivity, geometric dimensions such as length, width and thickness, surface stresses such as thermal stress or residual stress distribution, and structural stability such as the degree of deformation of the component under small vibrations. Cutting equipment initialization refers to starting and setting parameters for the cutting device, such as calibrating the initial position of the tool, setting the laser focal length or track speed, and ensuring normal data communication between the sensor and the execution system. Cutting stage positioning refers to determining the various stages in the entire cutting process based on the above-mentioned identification information and equipment parameters. It is divided into pre-acceleration stage, stable cutting stage and finishing stage. The duration and process parameters of each stage are directly related to the status of the component. The cutting cycle is the total control time window or spatial path interval formed by the combination of stages, which is used to constrain subsequent path planning, error assessment and correction plan execution.

[0026] S2: establishing a variation predictor, and deploying a correction plan for the cutting cycle based on the intrinsic state information.

[0027] Specifically, a variation predictor is established to identify possible deviations or abnormalities in advance by analyzing the real-time status information of semiconductor components. By combining historical cutting data, real-time monitoring information and physical models, the variation predictor can identify deviations that may be caused by factors such as thermal expansion, material non-uniformity, and equipment errors. For example, the predictor can predict the geometric deformation that may be caused by thermal expansion based on temperature sensor data and component thickness data. Based on the variation predictor, the correction plan is deployed under the cutting cycle for the intrinsic state information. During the entire cutting cycle, the cutting parameters are adjusted in time according to the output of the predictor to ensure that the component is always maintained within the set tolerance range during the processing process. The cutting cycle refers to the time window from the start to the completion of cutting.

[0028] S3: Execute real-time micro-correction control based on the deviation correction plan, and compress the predicted variation back to the allowable tolerance range to obtain a deviation correction processing result.

[0029] Specifically, during the cutting process, the cutting parameters are adjusted in real time according to the deployed correction plan to ensure that each operation link is carried out within a controllable range. For example, the correction plan may include fine-tuning the tool angle, adjusting the feed speed, or changing the laser focus position to deal with predicted potential deviations. Real-time micro-correction control responds quickly at every moment to deal with sudden deviation changes during the dynamic cutting process to ensure that accuracy is always effectively maintained. After executing micro-correction control, it is ensured that the cutting error will not exceed the preset allowable error range. If the predicted anomaly causes the deviation to exceed the tolerance range, the error is automatically corrected through micro-correction. After executing micro-correction control, the corrected result is output and compared with the target cutting size to ensure that the accuracy of the final product meets the design requirements.

[0030] Furthermore, the present application also includes: obtaining the intrinsic state information by collecting the material properties of the semiconductor component, measuring the geometric dimensions, detecting the surface stress and judging the structural stability; importing the target cutting index and the target cutting index value into the cutting equipment to obtain the cutting task requirements; locating the cutting stage based on the intrinsic state information and the cutting task requirements to obtain the cutting cycle.

[0031] Specifically, intrinsic state information is obtained through the collection of material properties, geometric dimension measurement, surface stress detection and structural stability judgment of semiconductor components. Material property collection refers to the detection of basic properties such as the composition, hardness, thermal expansion coefficient, and conductivity of semiconductor components to help understand the performance of materials in different environments. Geometric dimension measurement is the precise measurement of the shape and size of semiconductor components to ensure that they meet design requirements. Surface stress detection measures the stress state on the surface of the component to understand its possible deformation tendency during the processing process. Structural stability judgment analyzes the stability of semiconductor components during processing and predicts whether they may accidentally deform or break during the cutting process.

[0032] Next, the specific requirements of the cutting task are input into the cutting equipment, such as the target size, accuracy, tolerance range, and required cutting speed. The target cutting parameters include the thickness of the material being cut, edge smoothness, and required accuracy. The target cutting index value is a specific numerical definition of the parameters, such as requiring a cutting error of no more than ten microns.

[0033] The targeted cutting phase refers to identifying specific stages of the cutting process that require special attention, taking into account the inherent state of the semiconductor component and the cutting task requirements. For example, if the material is hard, the cutting speed or temperature may need to be adjusted to avoid damage. The cutting cycle refers to the time period from the start to the completion of the cutting process, and this time period is adjusted accordingly based on the performance of the cutting equipment, the material characteristics, and the task requirements.

[0034] Furthermore, the present application also includes: generating a control cycle with the intrinsic state information as the control starting point and the cutting task requirement as the control end point; dividing the control cycle by the spatial dimension to obtain a spatially divided control cycle; dividing the control cycle by equal periods by the time dimension to obtain a temporally divided control cycle; performing weighted calculation on the spatially divided control cycle and the temporally divided control cycle to obtain the cutting cycle, wherein the cutting cycle includes a state monitoring threshold obtained based on the intrinsic state information and the cutting task requirement.

[0035] Specifically, a control cycle is generated, with intrinsic state information as the starting point for control and the cutting task requirements as the control endpoint. During the cutting process, basic material information (such as hardness, stress, and size) needs to be monitored and controlled as the initial reference point, while the cutting task requirements (such as target size and tolerance requirements) determine the ultimate goal of the cutting process. The control cycle refers to the time period from the start to the completion of cutting. The generation of the entire cycle is based on the differences and connections between the two key points. The control cycle may vary depending on the material properties and task requirements. For example, harder materials require longer to cut.

[0036] The control cycle is divided by spatial dimension to obtain spatial partitioning control cycles. Spatial partitioning refers to segmenting the cutting process based on the spatial characteristics of the component (such as shape, cutting path, geometric features, etc.). Each spatial partitioning control cycle corresponds to a part of the component, which may be a specific area or a section of the cutting path.

[0037] The time-divided control cycle is obtained by dividing the control cycle into equal periods using the time dimension. Time dimension division divides the entire cutting process into several equal time periods based on time intervals. Within each time period, the cutting process is controlled according to the set conditions such as rate and temperature to ensure that the operation at each time point meets the accuracy requirements.

[0038] A weighted calculation is performed on the spatial and temporal control cycles to produce a cutting cycle. This cutting cycle includes a state monitoring threshold based on the intrinsic state information and the cutting task requirements. Weighted calculation combines the spatial and temporal division results, performing calculations at a specific ratio to produce a comprehensive cutting cycle. Weights are assigned based on the impact of spatial division (such as the cutting characteristics of different areas) and temporal division (such as cutting speed and temperature control). Ultimately, the cutting cycle includes not only temporal and spatial arrangements but also a state monitoring threshold for real-time monitoring of whether the cutting process meets requirements.

[0039] Furthermore, the present application also includes: extracting historical mutation positions and historical deviation vectors from historical cutting data to obtain deviation features; performing supervised training based on the deviation features and historical cutting control data in the historical cutting data to establish the mutation predictor, wherein the mutation predictor has a decision linkage with the cutting control interface, and is visually displayed on the cutting control interface based on the decision linkage.

[0040] Specifically, the historical deviation locations and deviation vectors are extracted from historical cutting data to generate deviation features. During the completed semiconductor component cutting process, the locations where errors occurred are identified from the historical records, known as the deviation locations. The direction and magnitude of the errors, known as the deviation vectors, are then analyzed. The deviation locations refer to the specific areas where the actual cutting process deviated from the ideal trajectory, while the deviation vectors represent the numerical direction of the error, such as a 0.02 mm offset to the left or a 0.05 mm offset upward.

[0041] Next, supervised training is performed based on the deviation features and historical cutting control data from the historical cutting data to establish an anomaly predictor. Supervised training is a machine learning method that uses training samples consisting of input features and known results to teach the model the corresponding relationship between features and results. The historical cutting control data provides the input parameters of the control system at the time, such as tool speed, feed rate, and laser power, while the deviation features provide the representation of the output results. Through supervised training, an anomaly predictor can be established. This can identify the conditions under which errors are likely to occur in future cutting processes and identify possible anomaly trends, thereby enabling early warning and proactive intervention.

[0042] The anomaly predictor and the cutting control interface have a decision-making linkage. This linkage means that the predictor's judgment results can directly influence the operating strategy of the cutting control system. When the predictor determines that a deviation may occur at a certain stage, it issues adjustment suggestions to the control system, such as reducing speed, changing trajectory, or initiating a correction program. The cutting control interface is a human-computer interaction platform used to manage the entire cutting process, displaying system operating status, control parameters, alarm information, and other information.

[0043] Finally, the decision-making linkage is visualized on the cutting control interface, displaying real-time graphics or data on the control interface, making it easier for operators to understand the current system status and potential risks. For example, the control interface may display a high-risk area marked as a mutation zone, or a red warning bar may indicate the direction and degree of possible deviation from the current trajectory, thereby improving the operator's perception and response efficiency.

[0044] Furthermore, the present application also includes: performing mutation prediction on the intrinsic state information based on the mutation predictor to obtain the predicted mutation; positioning the predicted mutation position in the cutting cycle according to the predicted mutation to obtain the predicted mutation stage and the corresponding mutation stage state monitoring threshold; performing mutation correction simulation within the mutation stage state monitoring threshold based on the predicted deviation vector to generate the correction plan.

[0045] Specifically, the anomaly predictor predicts anomalies in intrinsic state information, analyzing the intrinsic state information of the semiconductor component currently being cut to predict anomalies such as positional offset, angular error, or material deformation that may occur during the actual cutting process. Intrinsic state information includes characteristics such as the component's material composition, geometry, surface stress, and structural integrity. Predicting anomalies involves the model inferring potential problems that may arise during future cutting, such as predicting that a certain area will shift to the right by 0.04 mm after heating.

[0046] The predicted mutation location is located during the cutting cycle to determine the predicted mutation stage and the corresponding mutation state monitoring threshold. The cutting cycle refers to the entire cutting process from the start to the end, which is subdivided into several stages. The predicted mutation location is the component location where an error may occur, as determined by the predictor. By mapping this location to the cutting cycle in time, the cutting stage is determined. The state monitoring threshold is the error tolerance range set for this stage and is used to determine whether the error in the actual cutting exceeds the acceptable range.

[0047] Based on the predicted deviation vector, an abnormal correction simulation is performed within the state monitoring threshold of the abnormal stage to generate a correction plan. The predicted deviation vector is the model's specific numerical expression of the direction and size of the abnormal behavior, such as indicating that the offset direction is the positive direction of the X-axis and the size is 0.05 mm. Under the premise of knowing the predicted deviation and the state monitoring threshold, simulation calculations can be performed to simulate whether different adjustment strategies can compress the predicted error to within the threshold as an abnormal correction simulation, and then find the optimal or feasible correction plan. The final generated correction plan is a set of preset control instructions or operating procedures, such as reducing the feed speed at a certain point in time, fine-tuning the laser focus, adjusting the trajectory curvature, etc., which are used to actively control the deviation during the cutting process.

[0048] Furthermore, the present application also includes: if the number of plans for the correction plan exceeds the preset number of plans, a conditional trigger mechanism is set to trigger the correction plan: static threshold triggering is performed based on the real-time acquisition status of the semiconductor component; dynamic trend triggering is performed when the predicted anomaly is continuously growing; based on the auxiliary correction indicator, multi-condition joint triggering is performed in combination with the anomaly stage status monitoring threshold.

[0049] Specifically, when the number of corrective actions generated by the predictive model exceeds a pre-set maximum, an automated mechanism is activated to determine whether further action is necessary. This maximum number of actions is set as a threshold, and exceeding this threshold indicates that there may be issues in the cutting process that cannot be resolved through conventional means, necessitating the activation of a trigger mechanism to ensure a timely response.

[0050] Static threshold triggering is performed based on the real-time data collected from semiconductor components. Static threshold triggering involves comparing real-time data (such as temperature, stress, and speed) with pre-set fixed thresholds. Once the real-time data exceeds the set threshold, a correction plan is triggered. For example, if the temperature exceeds 50 degrees Celsius, a plan may be activated to adjust the cutting speed to prevent the high temperature from affecting the semiconductor component.

[0051] When the predicted anomaly shows a continuous increase, a dynamic trend trigger is executed. Dynamic trend triggering determines the continuous growth trend of the anomaly based on the prediction model. When the prediction results indicate that a certain deviation is gradually increasing and the rate of change exceeds the expected range, a plan is triggered to address this trend. For example, if the deviation has gradually increased from 0.01 mm to 0.05 mm over the past few cutting processes and has not been effectively curbed, the trend can be used to determine whether more urgent corrective measures are needed. This can better cope with the impact of long-term changes, especially when the deviation gradually increases, allowing for early identification of the problem and automatic adjustments.

[0052] Based on the auxiliary correction indicators, combined with the abnormal stage state monitoring threshold, multiple conditions are jointly triggered. Auxiliary correction indicators are additional parameters related to cutting quality, such as cutting force, vibration frequency, tool wear, etc., which can provide more information and assist in determining whether correction is needed. The abnormal stage state monitoring threshold is the maximum value of the deviation allowed within a specific cutting stage. A more comprehensive judgment is made by combining different factors. For example, if the cutting force at a certain stage exceeds the set threshold, and the temperature and speed also exceed the limit value, multiple indicators are combined to trigger and execute the correction plan.

[0053] Furthermore, the present application also includes: adjusting the predicted cutting control data corresponding to the predicted variation through the correction plan to generate an initial correction processing result; obtaining a real-time acquisition state according to the initial correction processing result; obtaining a correction state to be supplemented based on the real-time acquisition state and the pre-correction state of the correction plan; presetting an allowable tolerance range for the correction plan, and obtaining a corresponding allowable tolerance correction execution result based on the allowable tolerance range; if the allowable tolerance correction state of the allowable tolerance correction execution result does not meet the correction state to be supplemented, performing the next stage of supplementary correction in the predicted variation stage until the correction state to be supplemented is met, and obtaining the correction processing result.

[0054] Specifically, the correction plan adjusts the predicted cutting control data corresponding to the predicted variation to generate an initial correction result. After predicting possible variations in the semiconductor component, the control parameters of the cutting process are adjusted according to the pre-prepared correction plan. This control data may include the tool's trajectory, cutting speed, laser focus position, and other factors. By adjusting these parameters, it is expected that the predicted variation can be reduced or eliminated.

[0055] The real-time acquisition status is obtained based on the initial correction process results. After the initial correction process results are executed, the current operating status, including temperature, stress, speed, vibration and other indicators, is monitored in real time and fed back to the control system to verify whether the initial correction measures have achieved the expected results.

[0056] The real-time data collection status is compared with the pre-correction status in the correction plan to determine the status to be supplemented. By comparing the real-time data collection status with the preset ideal state (i.e., the pre-correction state), it is determined whether the current cutting is within the acceptable error range. If there is a deviation between the real-time data collection status and the preset correction state, further adjustment is required.

[0057] Preset the tolerance range for the correction plan and obtain the corresponding tolerance correction results based on the tolerance range. The tolerance range refers to the allowable error range, which specifies the maximum acceptable deviation value, and then calculates the correction results that meet the tolerance.

[0058] If the tolerance correction result does not meet the required supplementary correction status, the next stage of supplementary correction in the predicted abnormality phase is performed until the required supplementary correction status is met, and the correction result is obtained. Additional correction measures are gradually implemented until the correction effect meets the preset target. For example, if the error after the initial supplementation still does not reach the tolerance value, further compensation is performed by adjusting the parameters in the next control cycle until the error is within the allowable range.

[0059] Furthermore, the present application also includes: the micro-correction control includes automatic correction and human-computer interaction correction, the automatic correction is to trigger the correction plan based on the real-time acquisition status, and the human-computer interaction correction is to select fine-tuning parameters based on the cutting control interface to obtain the correction processing result.

[0060] Specifically, micro-correction control includes automatic correction and human-computer interaction correction, which means that the system has set up two micro-correction methods to ensure cutting accuracy and control stability during the cutting process of semiconductor components. Micro-correction control refers to small-scale, continuous control parameter adjustments without interrupting the cutting process. Automatic correction relies on the adaptive control capability of the system, while human-computer interaction correction requires the operator to intervene based on interface feedback. Human-computer interaction correction means that when automatic correction cannot fully meet the accuracy or respond to complex emergencies, the operator can intervene through the human-computer interface. The cutting control interface is a graphical control terminal used to display key data of the cutting process and allow operations. The operator can select appropriate fine-tuning parameters based on the displayed information.

[0061] Furthermore, the present application also includes: retraining parameters of the mutation predictor, wherein the retraining process performs incremental learning based on a dynamic sampling window, and determines whether to update the mutation predictor according to a preset stability threshold.

[0062] Specifically, the variation predictor undergoes parameter retraining. Building on its initial learning capabilities, the model is continuously updated with new cutting data to optimize and adjust its parameters, making it more accurate to the actual state of the equipment and components. The variation predictor is an intelligent model used to predict potential deviations or anomalies in the semiconductor cutting process. Constructed using a neural network, decision tree, or ensemble learning model, it identifies potential errors in advance, thereby assisting in decision-making adjustments.

[0063] The retraining process uses incremental learning based on a dynamic sampling window. This process continuously collects recent cropped data through a sliding or variable-length data window and continuously feeds it into the model as training samples. The dynamic sampling window automatically adjusts in size based on the rate of data change. For example, during stable device operation, the window width may be 100 seconds, expanding to 300 seconds during periods of frequent anomalies to ensure representativeness and timeliness of the sampling. Incremental learning emphasizes that the model continuously absorbs new data while preserving existing training results, gradually refining its ability to predict future states.

[0064] The decision to update the mutation predictor is based on a preset stability threshold. The new model's reliability is determined by evaluating metrics such as error reduction, output volatility, or prediction accuracy. The stability threshold is a set of boundary conditions used to measure prediction performance. Otherwise, the new model is discarded and the old model is used to avoid error amplification or unstable behavior. Table 1 shows a partial record of the most recent mutation predictor retraining.

[0065] Table 1: Partial records of the latest mutation predictor retraining

[0066] In summary, the semiconductor component cutting control method supporting autonomous correction provided by this application has the following technical effects: by realizing precise cutting control based on real-time data and dynamic adjustment, the technical effect of improving the intelligence, adaptability and accuracy of the cutting process is achieved.

[0067] In the second embodiment, based on the same inventive concept as the semiconductor component cutting control method supporting autonomous correction in the above embodiment, the present application also provides a semiconductor component cutting control system supporting autonomous correction, please refer to the attached Figure 2 , including: a cutting stage positioning module 11, used to perform semiconductor component state recognition and cutting equipment initialization, and perform cutting stage positioning to obtain a cutting cycle; a correction plan deployment module 12, used to establish a variation predictor, and deploy a correction plan under the cutting cycle based on the intrinsic state information of the variation predictor; a correction processing result acquisition module 13, used to perform real-time micro-correction control based on the correction plan, and compress the predicted variation back to the allowable tolerance range to obtain a correction processing result.

[0068] Furthermore, the semiconductor component cutting control system that supports autonomous correction is also used to: obtain the intrinsic state information through material property collection, geometric dimension measurement, surface stress detection and structural stability judgment of the semiconductor component; import target cutting indicators and target cutting indicator values ​​into the cutting equipment to obtain cutting task requirements; locate the cutting stage based on the intrinsic state information and the cutting task requirements to obtain the cutting cycle.

[0069] Furthermore, the semiconductor component cutting control system that supports autonomous correction is also used to: generate a control cycle with the intrinsic state information as the control starting point and the cutting task requirement as the control end point; divide the control cycle by the spatial dimension to obtain a spatially divided control cycle; divide the control cycle by equal periods by the time dimension to obtain a temporally divided control cycle; perform weighted calculation on the spatially divided control cycle and the temporally divided control cycle to obtain the cutting cycle, wherein the cutting cycle includes a state monitoring threshold obtained based on the intrinsic state information and the cutting task requirement.

[0070] Furthermore, the semiconductor component cutting control system that supports autonomous correction is also used to: extract historical mutation positions and historical deviation vectors from historical cutting data to obtain deviation features; perform supervised training based on the deviation features and historical cutting control data in the historical cutting data to establish the mutation predictor, wherein the mutation predictor has a decision linkage with the cutting control interface, and is visually displayed on the cutting control interface based on the decision linkage.

[0071] Furthermore, the semiconductor component cutting control system that supports autonomous correction is also used to: perform mutation prediction on the intrinsic state information based on the mutation predictor to obtain the predicted mutation; locate the predicted mutation position in the cutting cycle according to the predicted mutation to obtain the predicted mutation stage and the corresponding mutation stage state monitoring threshold; perform mutation correction simulation within the mutation stage state monitoring threshold based on the predicted deviation vector to generate the correction plan.

[0072] Furthermore, the semiconductor component cutting control system that supports autonomous correction is also used to: if the number of correction plans exceeds the preset number of plans, set a conditional trigger mechanism to trigger the correction plan: static threshold triggering based on the real-time acquisition status of the semiconductor component; dynamic trend triggering when the predicted anomaly grows continuously; multi-condition joint triggering based on the auxiliary correction index and combined with the anomaly stage status monitoring threshold.

[0073] Furthermore, the semiconductor component cutting control system that supports autonomous correction is also used to: adjust the predicted cutting control data corresponding to the predicted variation through the correction plan to generate an initial correction processing result; obtain a real-time acquisition status based on the initial correction processing result; compare the real-time acquisition status with the pre-correction status of the correction plan to obtain a correction status to be supplemented; preset an allowable tolerance range for the correction plan, and obtain the corresponding allowable tolerance correction execution result based on the allowable tolerance range; if the allowable tolerance correction status of the allowable tolerance correction execution result does not meet the correction status to be supplemented, perform supplementary correction in the next stage of the predicted variation stage until the correction status to be supplemented is met to obtain the correction processing result.

[0074] Furthermore, the semiconductor component cutting control system that supports autonomous correction is also used for: the micro-correction control includes automatic correction and human-computer interaction correction, the automatic correction is to trigger a correction plan based on the real-time acquisition status, and the human-computer interaction correction is to select fine-tuning parameters based on the cutting control interface to obtain the correction processing result.

[0075] Furthermore, the semiconductor component cutting control system supporting autonomous deviation correction is also used to: retrain parameters of the variation predictor, wherein the retraining process performs incremental learning based on a dynamic sampling window, and determines whether to update the variation predictor based on a preset stability threshold.

[0076] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The semiconductor component cutting control method supporting autonomous correction and the specific examples in the aforementioned embodiment one are also applicable to the semiconductor component cutting control system supporting autonomous correction in this embodiment. Through the aforementioned detailed description of the semiconductor component cutting control method supporting autonomous correction, those skilled in the art can clearly understand the semiconductor component cutting control system supporting autonomous correction in this embodiment, so for the sake of brevity of the specification, it will not be described in detail here.

[0077] The above description of the disclosed embodiments is intended to enable one skilled in the art to implement or use the present application. Various modifications to these embodiments will be readily apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application is not limited to the embodiments shown herein, but is intended to conform to the widest scope consistent with the principles and novel features disclosed herein.

[0078] Obviously, those skilled in the art may make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalents, the present application is intended to include these modifications and variations.

Claims

1. A semiconductor component cutting control method supporting autonomous deviation correction, characterized in that: include: Perform semiconductor component status recognition and cutting equipment initialization, and perform cutting stage positioning to obtain the cutting cycle; Establishing a variation predictor, and deploying a correction plan under the cutting cycle based on the intrinsic state information by the variation predictor; Based on the deviation correction plan, real-time micro-correction control is performed, and the predicted variation is compressed back to the allowable tolerance range to obtain a deviation correction processing result.

2. A semiconductor component cutting control method supporting autonomous deviation correction according to claim 1, characterized in that: Perform semiconductor component status recognition and cutting equipment initialization, and perform cutting stage positioning to obtain the cutting cycle, including: Acquiring the intrinsic state information by collecting material properties, measuring geometric dimensions, detecting surface stress, and judging structural stability of the semiconductor component; Importing a target cutting index and a target cutting index value into the cutting device to obtain a cutting task requirement; The cutting phase is located based on the intrinsic state information and the cutting task requirement, and the cutting cycle is obtained.

3. A semiconductor component cutting control method supporting autonomous deviation correction according to claim 2, characterized in that: Locating the cutting stage based on the intrinsic state information and the cutting task requirements to obtain the cutting cycle includes: Generate a control cycle with the intrinsic state information as the control starting point and the cutting task requirement as the control end point; Dividing the control period by a spatial dimension to obtain a spatially divided control period; Dividing the control period into equal periods by the time dimension to obtain a time-divided control period; A weighted calculation is performed on the spatial division control period and the temporal division control period to obtain the cutting period, wherein the cutting period includes a state monitoring threshold obtained based on the intrinsic state information and the cutting task requirement.

4. The semiconductor component cutting control method supporting autonomous deviation correction according to claim 1, wherein: Build a mutation predictor, including: Extract historical anomaly positions and historical deviation vectors from historical cropping data to obtain deviation features; Based on the deviation features and the historical cutting control data in the historical cutting data, supervised training is performed to establish the mutation predictor, wherein the mutation predictor has a decision linkage with the cutting control interface, and is visualized on the cutting control interface based on the decision linkage.

5. The semiconductor component cutting control method supporting autonomous deviation correction according to claim 1, wherein: Deploying a correction plan for the cutting cycle based on the intrinsic state information by the variation predictor includes: performing mutation prediction on the intrinsic state information based on the mutation predictor to obtain the predicted mutation; Positioning the predicted mutation position during the cutting cycle according to the predicted mutation, to obtain a predicted mutation stage and a corresponding mutation stage state monitoring threshold; Based on the predicted deviation vector, an abnormality correction simulation is performed within the abnormality stage state monitoring threshold to generate the correction plan.

6. The semiconductor component cutting control method supporting autonomous deviation correction according to claim 5, characterized in that: If the number of the correction plan exceeds the preset number of plans, a conditional trigger mechanism is set to trigger the correction plan: Performing static threshold triggering based on the real-time acquisition status of the semiconductor component; When the predicted abnormality shows continuous growth, dynamic trend triggering is performed; Based on the auxiliary correction index, combined with the abnormal phase state monitoring threshold, multi-condition joint triggering is performed.

7. The semiconductor component cutting control method supporting autonomous deviation correction according to claim 1, wherein: Based on the correction plan, real-time micro-correction control is performed, and the predicted variation is compressed back to the allowable tolerance range to obtain the correction processing result, including: Adjust the predicted cutting control data corresponding to the predicted variation through the correction plan to generate an initial correction processing result; Obtaining a real-time acquisition status according to the initial correction processing result; Comparing the real-time acquisition status with the pre-correction status of the correction plan to obtain a state to be supplemented with correction; Presetting an allowable tolerance range for the deviation correction plan, and obtaining a corresponding allowable tolerance deviation correction execution result based on the allowable tolerance range; If the allowable tolerance correction state of the allowable tolerance correction execution result does not meet the state to be supplemented, the next stage supplementary correction of the predicted abnormality stage is performed until the state to be supplemented is met to obtain the correction processing result.

8. The semiconductor component cutting control method supporting autonomous deviation correction according to claim 7, characterized in that: The micro-correction control includes automatic correction and human-computer interaction correction. The automatic correction is to trigger a correction plan based on the real-time acquisition status, and the human-computer interaction correction is to select fine-tuning parameters based on the cutting control interface to obtain the correction processing result.

9. The semiconductor component cutting control method supporting autonomous deviation correction according to claim 1, wherein: Parameters of the mutation predictor are retrained, wherein the retraining process performs incremental learning based on a dynamic sampling window, and determines whether to update the mutation predictor according to a preset stability threshold.

10. A semiconductor component cutting control system supporting autonomous deviation correction, characterized in that: The steps for implementing the semiconductor component cutting control method supporting autonomous deviation correction as described in any one of claims 1 to 9 include: The cutting stage positioning module is used to perform semiconductor component status recognition and cutting equipment initialization, and to perform cutting stage positioning to obtain the cutting cycle; A correction plan deployment module is used to establish a variation predictor and deploy a correction plan under the cutting cycle based on the intrinsic state information of the variation predictor; The deviation correction processing result obtaining module is used to perform real-time micro-correction control based on the deviation correction plan, and compress the predicted variation back to the allowable tolerance range to obtain the deviation correction processing result.

Images