Crystal transition and crucible transition coordinated regulation and control system based on oxygen and carbon content feedback

Abstract

The invention relates to the technical field, in particular to a crystal transition and crucible transition coordinated regulation and control system based on oxygen and carbon content feedback, and belongs to the technical field of single crystal growth. Aiming at the defects of lack of multi-parameter linkage, regulation and control non-stage pertinence and no effect verification in the prior art, the system comprises five modules, namely an oxygen-carbon time sequence data acquisition module, an oxygen-carbon feature extraction module, a steering coordinated regulation and control module, a rotating speed parameter correction module and a regulation and control effect verification module. The system integrates oxygen and carbon content, growth stage identification and in-furnace temperature and pressure parameters, extracts oxygen and carbon dynamic characteristics in stages, cooperatively regulates and controls the rotation direction, time sequence and rotation speed difference of crystal rotation and crucible rotation, and verifies the regulation and control effect through uniformity comparison. According to the invention, deep synergy of oxygen-carbon feedback and mechanism regulation and control is realized, the regulation and control pertinence and accuracy are improved, the single crystal growth environment is stabilized, the cultivation of a single crystal material with high lattice regularity is assisted, and the requirements of the high-tech field on a core base material are met.

Description

Technical Field

[0001] This invention relates to the field of single crystal growth technology, and in particular to a synergistic control system for crystal rotation and crucible rotation based on oxygen and carbon content feedback. Background Technology

[0002] The field of single crystal growth technology includes technologies related to the cultivation and preparation of crystal materials. Its core content is to promote the formation of single crystal materials with regular lattice structures by precisely controlling the growth environment and process parameters. This field systematically covers key aspects such as equipment design, process optimization, process monitoring, and parameter control for single crystal growth. It is widely used in many high-tech fields such as semiconductor optoelectronic materials and aerospace materials, providing core basic materials for the manufacture of various high-performance devices.

[0003] Among them, the crystal rotation and crucible rotation coordinated control system based on oxygen and carbon content feedback refers to the technical solution for parameter control during single crystal growth. The technical issues it addresses include real-time acquisition of oxygen and carbon content and coordinated control of the movement states of crystal rotation and crucible rotation. Specifically, it collects oxygen and carbon content data in the single crystal growth environment through sensors, uses this data as the basis for control, and achieves coordinated operation of crystal rotation and crucible rotation by adjusting the rotation speed of the crystal rotation mechanism and the rotation speed of the crucible rotation mechanism.

[0004] Existing technologies rely solely on oxygen and carbon content data to regulate the rotation speed of the crystal transfer crucible, without considering growth stage, temperature, pressure, or other operating parameters. This lack of stage-specific control and multi-parameter linkage makes it impossible to capture the dynamic changes and spatial distribution differences of oxygen and carbon content. Focusing only on rotation speed adjustment ignores the coordination of turning direction and timing, and there is no verification process for the control effect. In actual single crystal growth, control lag or deviation is likely to occur, leading to fluctuations in the growth environment, affecting the regularity of the single crystal lattice, reducing the yield of high-performance single crystal preparation, and making it difficult to meet the high standards required for materials in fields such as semiconductors. Summary of the Invention

[0005] The purpose of this invention is to overcome the shortcomings of the prior art and propose a synergistic control system for crystal rotation and crucible rotation based on oxygen and carbon content feedback.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: a crystal rotation and crucible rotation coordinated control system based on oxygen and carbon content feedback, the system comprising:

[0007] Oxygen and carbon time series data acquisition module: Collects real-time data on oxygen and carbon content in the single crystal growth furnace, synchronously collects growth stage identification parameters, associates furnace temperature and pressure operating parameters, integrates them according to time nodes, and generates an oxygen and carbon operating condition time series dataset.

[0008] Oxygen and carbon feature extraction module: Based on the oxygen and carbon operating condition time series dataset, determine the single crystal growth stage, call the corresponding time window threshold, calculate the change amplitude and rate of adjacent oxygen and carbon data, compare them with the stage threshold respectively, filter out the feature points that exceed the standard, retain the clusters that meet the standard after clustering operation, and generate a stage-specific oxygen and carbon feature set.

[0009] Steering Coordination and Control Module: Calls the stage-specific oxygen and carbon feature set, calculates the rate of change of oxygen and carbon content and compares it with the critical value, determines the trend of change, and controls the crystal rotation and crucible rotation steering and start-stop sequence to obtain the crystal rotation and crucible rotation steering parameter set.

[0010] Speed ​​parameter correction module: Based on the stage-specific oxygen and carbon feature set, calculate the radial and axial gradient values ​​of oxygen and carbon and sum them with weights. After comparing with the gradient threshold, correct the speed difference benchmark value and allocate it to the corresponding parameter to obtain the dynamic speed difference parameter.

[0011] Control effect verification module: Collect oxygen and carbon data corresponding to the crystal rotating crucible rotation parameter group and dynamic speed difference parameter, calculate the uniformity and compare it with the threshold, and generate oxygen and carbon control adaptation coefficient.

[0012] As a further aspect of the present invention, the oxygen-carbon operating condition time series dataset includes an oxygen content sequence, a carbon content sequence, a growth stage identifier sequence, a furnace temperature sequence, and a furnace pressure sequence. The stage-specific oxygen-carbon feature set includes feature point coordinates, cluster density values, feature point change amplitude, and change rate data. The crystal-to-crate rotation parameter set includes crystal rotation direction, crucible rotation direction, and rotation start-stop sequence. The dynamic speed difference parameter includes radial / axial gradient correction values ​​and target speeds for crystal and crucible rotation. The oxygen-carbon control adaptation coefficient includes an oxygen / carbon content uniformity coefficient and a control parameter adaptation value.

[0013] As a further aspect of the present invention, the oxygen and carbon time-series data acquisition module includes:

[0014] Data acquisition submodule: Collects real-time data on oxygen and carbon content in the single crystal growth furnace, synchronously collects growth stage identification parameters, classifies and records the three types of data collected, and generates a basic data set;

[0015] Parameter association submodule: Based on the basic data set, it collects furnace temperature and pressure parameters, matches the temperature and pressure parameters with the basic data set according to the same time nodes, and generates an associated parameter set;

[0016] The time series integration submodule calls the associated parameter set, sorts and integrates all data in the set according to the time node order, verifies the consistency of the corresponding data nodes, and generates an oxygen and carbon operating condition time series dataset.

[0017] As a further aspect of the present invention, the oxygen and carbon feature extraction module includes:

[0018] Stage Determination Submodule: Based on the oxygen and carbon operating condition time series dataset, determine the single crystal growth stage, call the corresponding stage time window threshold, extract the complete oxygen and carbon time series data within the stage, associate and bind the time window threshold with the stage amplitude threshold and stage rate threshold, verify the consistency of the parameters, and generate the stage threshold parameter set.

[0019] Feature filtering submodule: Calls the stage threshold parameter set, extracts the oxygen and carbon time series data of the corresponding stage, calculates the change amplitude and change rate of adjacent oxygen and carbon data point by point, compares the change amplitude with the stage amplitude threshold, and compares the change rate with the stage rate threshold, filters out feature points that exceed both thresholds at the same time, records the coordinates and numerical information of the feature points, and generates a set of feature points that exceed the standard.

[0020] Cluster Integration Submodule: Based on the set of feature points exceeding the standard, it calls the stage feature density parameter to perform clustering operation on the feature points, calculates the density value of each cluster, compares the cluster density value with the set density threshold, retains the feature clusters that reach the set density threshold, classifies and organizes the cluster data according to the stage, and generates a stage-specific oxygen and carbon feature set.

[0021] As a further aspect of the present invention, the steering coordination control module includes:

[0022] Change rate calculation submodule: Calls the stage-specific oxygen and carbon feature set, extracts oxygen and carbon content data from the feature set, calculates the change rate of oxygen and carbon content according to adjacent data nodes, retrieves preset positive and negative critical values, associates and matches the change rate with the two types of critical values, records the matching correspondence, and generates a change rate critical comparison table.

[0023] Trend determination submodule: Based on the critical comparison table of change rate, the change rate of oxygen and carbon content is compared with the positive and negative critical values ​​respectively to determine the upward or downward trend of oxygen and carbon content, mark the critical value range corresponding to the trend, and generate a trend determination result set.

[0024] Steering parameter generation submodule: Based on the trend judgment result set, adjust the steering direction of the crystal rotation mechanism and the crucible rotation mechanism according to the judgment trend, adjust the steering start and stop timing parameters, integrate the steering direction parameters and timing parameters, classify and organize them according to parameter categories, and generate the crystal rotation and crucible rotation steering parameter group.

[0025] As a further aspect of the present invention, the speed parameter correction module includes:

[0026] Gradient calculation submodule: Based on the stage-specific oxygen and carbon feature set, extract the radial and axial distribution data of oxygen and carbon content in the feature set, calculate the difference of oxygen and carbon content between each sampling point in the radial direction to obtain the radial gradient value, calculate the difference of oxygen and carbon content between each sampling point in the axial direction to obtain the axial gradient value, and perform a weighted summation operation on the radial gradient value and the axial gradient value to generate a comprehensive gradient value.

[0027] Gradient Comparison Submodule: Calls the gradient synthesis value, retrieves the preset gradient threshold range, compares the gradient synthesis value with the gradient threshold range, determines the threshold range to which the gradient synthesis value belongs, retrieves the speed difference benchmark value of the corresponding range, and generates a benchmark value matching table.

[0028] Parameter correction submodule: Based on the benchmark value matching table, calculate the degree of deviation between the gradient comprehensive value and the corresponding interval median value, perform linear correction on the speed difference benchmark value according to the percentage of deviation, and allocate the corrected speed difference to the crystal rotation speed parameter and the crucible rotation speed parameter to generate dynamic speed difference parameters.

[0029] As a further aspect of the present invention, the regulation effect verification module includes:

[0030] Data acquisition submodule: calls the crystal rotating crucible rotation parameter group and dynamic speed difference parameter, collects real-time oxygen content data and real-time carbon content data under the corresponding working conditions of the two sets of parameters, integrates the two sets of parameters and oxygen and carbon data according to the time nodes, and generates the adjusted oxygen and carbon dataset.

[0031] Uniformity calculation submodule: Based on the oxygen and carbon dataset after regulation, extract oxygen content distribution data and carbon content distribution data, calculate the dispersion of oxygen content data and carbon content data, obtain the oxygen and carbon content uniformity through the dispersion of distribution data, and generate oxygen and carbon uniformity values.

[0032] The adaptation coefficient generation submodule retrieves the oxygen and carbon uniformity value, retrieves the preset uniformity threshold, compares the oxygen and carbon uniformity value with the uniformity threshold, calculates the parameter adaptation ratio based on the comparison result, and generates the oxygen and carbon regulation adaptation coefficient.

[0033] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0034] This invention integrates oxygen and carbon content, growth stage indicators, and furnace temperature and pressure parameters, forming a complete data chain according to time nodes. After stage-by-stage judgment, corresponding thresholds are matched, oxygen and carbon data change characteristics are calculated and clustered, and the turning and timing are adjusted based on the change trend. The rotation speed difference is corrected according to the gradient distribution, and the control effect is verified by uniformity comparison. This achieves deep synergy between oxygen and carbon feedback and crystal-cruise rotation control, improves the targeting and accuracy of control, ensures the stability of the single crystal growth environment, strengthens the linkage and adaptability between parameters, and helps to cultivate single crystal materials with regular crystal structure, meeting the needs of high-tech fields for core basic materials. Attached Figure Description

[0035] Figure 1 This is a system flowchart of the present invention.

[0036] Figure 2This is a flowchart of the control system of the present invention. Detailed Implementation

[0037] The technical solution of the present invention will now be described with reference to the accompanying drawings.

[0038] In embodiments of the present invention, words such as "exemplarily," "for example," etc., are used to indicate that something is an example, illustration, or description. Any embodiment or design described as "exemplary" in the present invention should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of the word "exemplary" is intended to present the concept in a concrete manner. Furthermore, in embodiments of the present invention, the meaning expressed by "and / or" can be both, or either one.

[0039] In the embodiments of this invention, the terms "image" and "picture" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, their intended meanings are consistent. Similarly, the terms "of," "corresponding (relevant)," and "corresponding" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, their intended meanings are consistent.

[0040] In this embodiment of the invention, sometimes a subscript such as W1 may be written in a non-subscript form such as W1. When the difference is not emphasized, the meaning they express is the same.

[0041] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0042] Please see Figure 1 This invention provides a technical solution: a synergistic control system for crystal rotation and crucible rotation based on oxygen and carbon content feedback, the system comprising:

[0043] Oxygen and carbon time series data acquisition module: Collects real-time data of oxygen and carbon content in single crystal growth furnace, synchronously collects growth stage identification parameters such as crystal pulling and shoulder formation, associates furnace temperature and pressure parameters, integrates them one by one according to time nodes, and generates oxygen and carbon condition time series dataset.

[0044] Oxygen and carbon feature extraction module: Based on the oxygen and carbon working condition time series dataset, determine the current single crystal growth stage, call the corresponding stage time window threshold, calculate the change amplitude and change rate of adjacent oxygen and carbon data point by point, compare the change amplitude with the stage amplitude threshold, and compare the change rate with the stage rate threshold, filter out feature points that exceed both thresholds at the same time, call the stage feature density parameter to perform clustering operation on feature points, retain feature clusters with cluster density reaching the set value, and generate stage-specific oxygen and carbon feature set;

[0045] Steering Coordination and Control Module: Calls the stage-specific oxygen and carbon feature set, calculates the rate of change of oxygen and carbon content in adjacent periods, compares the rate of change with the forward and reverse critical values, determines the trend and rate of change, controls the steering of the crystal rotation mechanism and the crucible rotation mechanism, and adjusts the steering start and stop sequence in a coordinated manner to obtain the crystal rotation and crucible rotation steering parameter set.

[0046] Speed ​​parameter correction module: Based on the stage-specific oxygen and carbon feature set, calculate the radial and axial gradient values ​​of oxygen and carbon, calculate the weighted sum to obtain the gradient comprehensive value, compare it with the preset gradient threshold range, call the corresponding speed difference benchmark value, correct the benchmark value according to the percentage of deviation, and allocate it to the crystal rotation and crucible rotation speed parameters to obtain the dynamic speed difference parameter.

[0047] The regulation effect verification module collects real-time oxygen and carbon data corresponding to the crystal rotating crucible rotation parameter group and dynamic speed difference parameter, calculates the oxygen and carbon content uniformity, compares it with the preset uniformity threshold, and generates an oxygen and carbon regulation adaptation coefficient.

[0048] The oxygen and carbon operating condition time series dataset includes real-time oxygen content sequence, real-time carbon content sequence, growth stage identifier sequence, furnace temperature sequence, and furnace pressure sequence. The stage-specific oxygen and carbon feature set includes stage feature point coordinates, feature cluster density values, feature point change amplitude data, and feature point change rate data. The crystal-to-crate-to-rotation parameter set includes crystal-to-rotation direction, crucible-to-rotation direction, and rotation start-stop timing parameters. The dynamic speed difference parameters include radial gradient correction value, axial gradient correction value, crystal-to-rotation target speed, and crucible-to-rotation target speed. The oxygen and carbon regulation adaptation coefficients include oxygen content uniformity coefficient, carbon content uniformity coefficient, and regulation parameter adaptation value.

[0049] Please see Figure 1 The oxygen and carbon time-series data acquisition module includes:

[0050] Data acquisition submodule: Collects real-time data on oxygen and carbon content in the single crystal growth furnace, synchronously collects growth stage identification parameters, classifies and records the three types of data collected, and generates a basic data set;

[0051] Data acquisition submodule: Collects real-time data on oxygen and carbon content in the single crystal growth furnace. High-frequency sensors are used to collect data at a frequency of 1 time / second. The oxygen content collection range is 0-5ppm and the carbon content is 0-2ppm. Simultaneously, growth stage identification parameters are collected through furnace body sensors. The stages of crystal introduction (1), shoulder formation (2), equal diameter (3), and tailing (4) are distinguished by digital codes. The three types of data are classified and recorded in the format of "timestamp-parameter type-value". For example, the timestamp 1690000001s corresponds to an oxygen content of 2.1ppm, a carbon content of 0.8ppm, and a growth stage identification of 1. After recording 100 sets of data in sequence, the classification and organization are completed to generate a basic data set.

[0052] Parameter association submodule: Based on the basic data set, it collects furnace temperature and pressure parameters, matches the temperature and pressure parameters with the basic data set according to the same time nodes, and generates an associated parameter set;

[0053] Parameter Association Submodule: Based on the basic dataset, temperature data in the range of 800-1200℃ is collected by the furnace temperature sensor, and pressure data in the range of 0.1-0.5MPa is collected by the pressure sensor. The collection frequency is consistent with the basic data. The temperature and pressure parameters are matched point by point with the basic dataset according to the millisecond-level timestamp. For example, timestamp 1690000001s corresponds to a temperature of 1050℃ and a pressure of 0.3MPa. After verifying that there is no timestamp deviation, the matching is completed and the associated parameter set is generated.

[0054] The time series integration submodule calls the associated parameter set, sorts and integrates all data in the set according to the time node order, verifies the consistency of the corresponding data nodes, and generates an oxygen and carbon operating condition time series dataset.

[0055] The time series integration submodule calls the associated parameter set, sorts all data in the set in ascending order of timestamp, checks the continuity of timestamps point by point, removes duplicate timestamp data (retains the first collection value), and supplements missing timestamp data (filling with the average of two adjacent points). The example integrated data is 1690000001s: oxygen 2.1ppm, carbon 0.8ppm, stage 1, temperature 1050℃, pressure 0.3MPa. The entire data is then processed in sequence to generate an oxygen and carbon operating condition time series dataset.

[0056] Please see Figure 1 and Figure 2 The oxygen and carbon feature extraction module includes:

[0057] Stage Determination Submodule: Based on the oxygen and carbon operating condition time series dataset, determine the single crystal growth stage, call the corresponding stage time window threshold, extract the complete oxygen and carbon time series data within the stage, associate and bind the time window threshold with the stage amplitude threshold and stage rate threshold, verify the consistency of the parameters, and generate the stage threshold parameter set.

[0058] Stage Determination Submodule: Based on the oxygen and carbon operating condition time series dataset, the growth stage identifier parameters are extracted from the data. Combined with the duration threshold, the current stage is determined (crystal introduction stage duration ≤ 10 min, shoulder formation 10-30 min, isodiameter 30-120 min, and finishing stage > 120 min). The corresponding stage time window threshold is called (determined by 3 sets of single crystal growth experiments: crystal introduction 5 s, shoulder formation 10 s, isodiameter 15 s, and finishing stage 8 s). The complete oxygen and carbon time series data within this stage is extracted. The time window threshold is bound to the stage amplitude threshold (crystal introduction 0.2 ppm, shoulder formation 0.3 ppm, isodiameter 0.4 ppm, and finishing stage 0.2 ppm) and the stage rate threshold (crystal introduction 0.05 ppm / s, shoulder formation 0.08 ppm / s, isodiameter 0.1 ppm / s, and finishing stage 0.04 ppm / s). The parameters are checked to ensure there is no deviation, and a stage threshold parameter set is generated.

[0059] Feature filtering submodule: Calls the stage threshold parameter set, extracts the oxygen and carbon time series data of the corresponding stage, calculates the change amplitude and change rate of adjacent oxygen and carbon data point by point, compares the change amplitude with the stage amplitude threshold, and compares the change rate with the stage rate threshold, filters out feature points that exceed both thresholds at the same time, records the coordinates and numerical information of the feature points, and generates a set of feature points that exceed the standard.

[0060] Feature Filtering Submodule: Calls the stage threshold parameter set, extracts oxygen and carbon time series data of the shoulder release stage (example time window 10s, containing 10 data points), calculates the change amplitude of adjacent data points point by point (the absolute value of the subsequent value minus the previous value, such as oxygen 2.4ppm at point 2 minus 2.1ppm at point 1 to get 0.3ppm) and change rate (amplitude divided by the time interval 1s to get 0.3ppm / s), compares the change amplitude with the 0.3ppm amplitude threshold of the shoulder release stage, and compares the rate with the 0.08ppm / s rate threshold, filters out feature points with amplitude ≥0.3ppm and rate ≥0.08ppm / s, records the timestamp and numerical information of the feature points, and generates a set of feature points exceeding the standard.

[0061] Cluster Integration Submodule: Based on the set of feature points exceeding the standard, it calls the stage feature density parameter to perform clustering operation on the feature points, calculates the density value of each cluster, compares the cluster density value with the set density threshold, retains the feature clusters that reach the set density threshold, classifies and organizes the cluster data according to the stage, and generates a stage-specific oxygen and carbon feature set.

[0062] Cluster Integration Submodule: Based on the set of feature points exceeding the standard, it calls the feature density parameters of the shoulder release stage (which have been experimentally verified to be set to 3 points / 10s window), performs clustering operations on the feature points according to the 10s time window, counts the number of feature points in each window (e.g., there are 4 feature points in the 1690000010-1690000020s window), compares the cluster density value with the density threshold set by 3 points / 10s, retains the feature clusters with density values ​​≥ the set value, classifies and organizes the cluster data according to the shoulder release stage, and generates a stage-specific oxygen and carbon feature set.

[0063] Please see Figure 1 and Figure 2 The steering coordination and control module includes:

[0064] Change rate calculation submodule: Calls the stage-specific oxygen and carbon feature set, extracts oxygen and carbon content data from the feature set, calculates the change rate of oxygen and carbon content according to adjacent data nodes, retrieves preset positive and negative critical values, associates and matches the change rate with the two types of critical values, records the matching correspondence, and generates a change rate critical comparison table.

[0065] The rate of change calculation submodule calls the stage-specific oxygen and carbon feature set to extract oxygen and carbon content data during the shoulder release stage (example: 10 consecutive data points, 1 second time interval). It calculates the rate of change of oxygen and carbon content based on adjacent data nodes, retrieves the preset forward critical value of 0.08 ppm / s and the reverse critical value of -0.08 ppm / s (determined through 5 repeated experiments, with a deviation ≤ ±0.01 ppm / s), and associates and matches each rate of change with the two types of critical values ​​point by point. It records the matching results (e.g., a rate of change of 0.09 ppm / s corresponds to the forward critical value) and generates a rate of change critical comparison table.

[0066] Trend determination submodule: Based on the critical comparison table of change rate, the change rate of oxygen and carbon content is compared with the positive and negative critical values ​​respectively to determine the upward or downward trend of oxygen and carbon content, mark the critical value range corresponding to the trend, and generate a trend determination result set.

[0067] Trend determination submodule: Based on the critical comparison table of change rate, the change rate of oxygen and carbon content is compared point by point with the positive and negative critical values. A change rate > 0.08 ppm / s is determined to be an upward trend, < -0.08 ppm / s is determined to be a downward trend, and a change rate between the two is determined to be a stable trend. The critical value interval corresponding to the trend is marked (e.g., an upward trend corresponds to the interval > 0.08 ppm / s), the duration of each trend segment is recorded (e.g., an upward trend lasts for 3 seconds), and a trend determination result set is generated.

[0068] Steering parameter generation submodule: Based on the trend judgment result set, adjust the steering direction of the crystal rotation mechanism and the crucible rotation mechanism according to the judgment trend, adjust the steering start and stop timing parameters, integrate the steering direction parameters and timing parameters, classify and organize them according to parameter categories, and generate the crystal rotation and crucible rotation steering parameter group.

[0069] Steering parameter generation submodule: Based on the trend determination result set, the crystal rotation mechanism is adjusted to turn clockwise when there is an upward trend, and counterclockwise when there is a downward trend. The current rotation is maintained when there is a stable trend. The steering start and stop timing parameters are adjusted (0.5s delay to start steering when there is an upward trend, and immediate start when there is a downward trend). The steering direction parameters and timing parameters are integrated and sorted into categories such as "steering direction - start and stop time - duration" to generate the crystal rotation crucible steering parameter group.

[0070] Please see Figure 1 and Figure 2 The speed parameter correction module includes:

[0071] Gradient calculation submodule: Based on the stage-specific oxygen and carbon feature set, extract the radial and axial distribution data of oxygen and carbon content in the feature set, calculate the difference of oxygen and carbon content between each sampling point in the radial direction to obtain the radial gradient value, calculate the difference of oxygen and carbon content between each sampling point in the axial direction to obtain the axial gradient value, and perform a weighted summation operation on the radial gradient value and the axial gradient value to generate a comprehensive gradient value.

[0072] The gradient calculation submodule extracts radial distribution data (3 collection points: center, middle, and edge, spaced 5cm) and axial distribution data (2 collection points: upper and lower, spaced 8cm) of oxygen and carbon content in the isodiameter stage based on the stage-specific oxygen and carbon feature set. It calculates the radial gradient value (difference between adjacent collection points, e.g., 2.5ppm at the center minus 2.3ppm at the middle equals 0.2ppm) and the axial gradient value (2.4ppm at the upper part minus 2.2ppm at the lower part equals 0.2ppm). A radial weight of 0.6 and an axial weight of 0.4 are set (experimentally verified to accurately reflect the gradient distribution). The weighted sum is then used to obtain the gradient composite value, calculated as: Gradient composite value = Radial gradient value × 0.6 + Axial gradient value × 0.4. Substituting the values, we get 0.2 × 0.6 + 0.2 × 0.4 = 0.2ppm.

[0073] Gradient Comparison Submodule: Calls the gradient synthesis value, retrieves the preset gradient threshold range, compares the gradient synthesis value with the gradient threshold range, determines the threshold range to which the gradient synthesis value belongs, retrieves the speed difference benchmark value of the corresponding range, and generates a benchmark value matching table.

[0074] Gradient Comparison Submodule: Calls the gradient composite value of 0.17ppm, retrieves the preset gradient threshold range (low range 0-0.2ppm, middle range 0.2-0.4ppm, high range >0.4ppm, experimentally verified to cover all operating conditions), compares the gradient composite value with the threshold range, determines that it belongs to the low range, retrieves the corresponding speed difference benchmark value of 5rpm (experimentally determined that low gradient corresponds to low speed difference), records the range attributes and benchmark value, and generates a benchmark value matching table.

[0075] Parameter correction submodule: Based on the benchmark value matching table, calculate the degree of deviation between the gradient comprehensive value and the corresponding interval median value, perform linear correction on the speed difference benchmark value according to the percentage of deviation, and allocate the corrected speed difference to the crystal rotation speed parameter and the crucible rotation speed parameter to generate dynamic speed difference parameters.

[0076] Based on the benchmark matching table, the median value in the lower range is 0.1 ppm. Calculate the degree of deviation between the gradient composite value and the median value.

[0077] The 5 rpm speed difference baseline value is linearly corrected according to a 70% deviation. After correction, the speed difference is 5 × (1 + 70%) = 8.5 rpm. The 8.5 rpm is then distributed to the crystal rotation speed (30 rpm) and the crucible rotation speed (21.5 rpm) to generate dynamic speed difference parameters.

[0078] Please see Figure 1 and Figure 2 The regulation effect verification module includes:

[0079] Data acquisition submodule: calls the crystal rotating crucible rotation parameter group and dynamic speed difference parameter, collects real-time oxygen content data and real-time carbon content data under the corresponding working conditions of the two sets of parameters, integrates the two sets of parameters and oxygen and carbon data according to the time nodes, and generates the adjusted oxygen and carbon dataset.

[0080] Data acquisition submodule: Calls the crystal rotating crucible rotation parameter group (clockwise rotation, start-stop timing 0.5s) and dynamic speed difference parameter (8.5rpm), collects real-time oxygen and carbon content data under the corresponding working conditions of the two sets of parameters, with a collection duration of 20s and a frequency of 1 time / second. The two sets of parameters are integrated with the oxygen and carbon data point by point according to the timestamp. For example, timestamp 1690000100s corresponds to clockwise rotation, speed difference of 8.5rpm, oxygen 2.3ppm, and carbon 0.7ppm. After integration, the regulated oxygen and carbon dataset is generated.

[0081] Uniformity calculation submodule: Based on the oxygen and carbon dataset after regulation, extract oxygen content distribution data and carbon content distribution data, calculate the dispersion of oxygen content data and carbon content data, obtain the oxygen and carbon content uniformity through the dispersion of distribution data, and generate oxygen and carbon uniformity values.

[0082] Uniformity Calculation Submodule: Based on the post-regulation oxygen and carbon dataset, 20 sets of oxygen and carbon content distribution data are extracted. The dispersion is calculated by first obtaining the mean of the data, then calculating the sum of squares of the differences between each data point and the mean, dividing by the total data, and taking the square root to obtain the dispersion value. Substituting the data, the mean oxygen content is calculated to be 2.3 ppm with a dispersion value of 0.08 ppm, and the mean carbon content is 0.7 ppm with a dispersion value of 0.03 ppm. The uniformity of oxygen and carbon content is obtained through dispersion calculation. Subtracting the result of the dispersion value divided by the mean from 1, the oxygen uniformity is 0.96 and the carbon uniformity is 0.96, generating the oxygen and carbon uniformity values.

[0083] The adaptation coefficient generation submodule retrieves the oxygen and carbon uniformity value, retrieves the preset uniformity threshold, compares the oxygen and carbon uniformity value with the uniformity threshold, calculates the parameter adaptation ratio based on the comparison result, and generates the oxygen and carbon regulation adaptation coefficient.

[0084] The adaptation coefficient generation submodule retrieves the oxygen and carbon uniformity values ​​(oxygen 0.96, carbon 0.96), retrieves the preset uniformity threshold of 0.9 (which has been experimentally verified to correspond to a qualified control effect), compares the uniformity values ​​with the threshold, calculates the parameter adaptation ratio, divides the uniformity values ​​by the threshold, the oxygen adaptation ratio is 1.07, the carbon adaptation ratio is 1.07, and the average value is taken as the final result, generating an oxygen and carbon control adaptation coefficient of 1.07.

[0085] The above are merely specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A crystal and crucible transfer co-regulation system based on oxygen and carbon content feedback, characterized in that, The system includes: Oxygen and carbon time series data acquisition module: Collects real-time data on oxygen and carbon content in the single crystal growth furnace, synchronously collects growth stage identification parameters, associates furnace temperature and pressure operating parameters, integrates them according to time nodes, and generates an oxygen and carbon operating condition time series dataset. Oxygen and carbon feature extraction module: Based on the oxygen and carbon operating condition time series dataset, determine the single crystal growth stage, call the corresponding time window threshold, calculate the change amplitude and rate of adjacent oxygen and carbon data, compare them with the stage threshold respectively, filter out the feature points that exceed the standard, retain the clusters that meet the standard after clustering operation, and generate a stage-specific oxygen and carbon feature set. Steering Coordination and Control Module: Calls the stage-specific oxygen and carbon feature set, calculates the rate of change of oxygen and carbon content and compares it with the critical value, determines the trend of change, and controls the crystal rotation and crucible rotation steering and start-stop sequence to obtain the crystal rotation and crucible rotation steering parameter set. Speed ​​parameter correction module: Based on the stage-specific oxygen and carbon feature set, calculate the radial and axial gradient values ​​of oxygen and carbon and sum them with weights. After comparing with the gradient threshold, correct the speed difference benchmark value and allocate it to the corresponding parameter to obtain the dynamic speed difference parameter. Control effect verification module: Collect oxygen and carbon data corresponding to the crystal rotating crucible rotation parameter group and dynamic speed difference parameter, calculate the uniformity and compare it with the threshold, and generate oxygen and carbon control adaptation coefficient.

2. The crystal and crucible transfer coordinated control system based on oxygen and carbon content feedback according to claim 1, characterized in that: The oxygen-carbon operating condition time series dataset includes oxygen content sequence, carbon content sequence, growth stage identifier sequence, furnace temperature sequence, and furnace pressure sequence. The stage-specific oxygen-carbon feature set includes feature point coordinates, cluster density value, feature point change amplitude, and change rate data. The crystal-to-crate rotation parameter set includes crystal rotation direction, crucible rotation direction, and rotation start-stop sequence. The dynamic speed difference parameter includes radial / axial gradient correction value and target speeds for crystal and crucible rotation. The oxygen-carbon control adaptation coefficient includes oxygen / carbon content uniformity coefficient and control parameter adaptation value.

3. The crystal rotation and crucible rotation coordinated control system based on oxygen and carbon content feedback according to claim 1, characterized in that, The oxygen and carbon time-series data acquisition module includes: Data acquisition submodule: Collects real-time data on oxygen and carbon content in the single crystal growth furnace, synchronously collects growth stage identification parameters, classifies and records the three types of data collected, and generates a basic data set; Parameter association submodule: Based on the basic data set, it collects furnace temperature and pressure parameters, matches the temperature and pressure parameters with the basic data set according to the same time nodes, and generates an associated parameter set; The time series integration submodule calls the associated parameter set, sorts and integrates all data in the set according to the time node order, verifies the consistency of the corresponding data nodes, and generates an oxygen and carbon operating condition time series dataset.

4. The crystal rotation and crucible rotation coordinated control system based on oxygen and carbon content feedback according to claim 3, characterized in that, The oxygen and carbon feature extraction module includes: Stage Determination Submodule: Based on the oxygen and carbon operating condition time series dataset, determine the single crystal growth stage, call the corresponding stage time window threshold, extract the complete oxygen and carbon time series data within the stage, associate and bind the time window threshold with the stage amplitude threshold and stage rate threshold, verify the consistency of the parameters, and generate the stage threshold parameter set. Feature filtering submodule: Calls the stage threshold parameter set, extracts the oxygen and carbon time series data of the corresponding stage, calculates the change amplitude and change rate of adjacent oxygen and carbon data point by point, compares the change amplitude with the stage amplitude threshold, and compares the change rate with the stage rate threshold, filters out feature points that exceed both thresholds at the same time, records the coordinates and numerical information of the feature points, and generates a set of feature points that exceed the standard. Cluster Integration Submodule: Based on the set of feature points exceeding the standard, it calls the stage feature density parameter to perform clustering operation on the feature points, calculates the density value of each cluster, compares the cluster density value with the set density threshold, retains the feature clusters that reach the set density threshold, classifies and organizes the cluster data according to the stage, and generates a stage-specific oxygen and carbon feature set.

5. The crystal rotation and crucible rotation coordinated control system based on oxygen and carbon content feedback according to claim 4, characterized in that, The steering coordination control module includes: Change rate calculation submodule: Calls the stage-specific oxygen and carbon feature set, extracts oxygen and carbon content data from the feature set, calculates the change rate of oxygen and carbon content according to adjacent data nodes, retrieves preset positive and negative critical values, associates and matches the change rate with the two types of critical values, records the matching correspondence, and generates a change rate critical comparison table. Trend determination submodule: Based on the critical comparison table of change rate, the change rate of oxygen and carbon content is compared with the positive and negative critical values ​​respectively to determine the upward or downward trend of oxygen and carbon content, mark the critical value range corresponding to the trend, and generate a trend determination result set. Steering parameter generation submodule: Based on the trend judgment result set, adjust the steering direction of the crystal rotation mechanism and the crucible rotation mechanism according to the judgment trend, adjust the steering start and stop timing parameters, integrate the steering direction parameters and timing parameters, classify and organize them according to parameter categories, and generate the crystal rotation and crucible rotation steering parameter group.

6. The crystal rotation and crucible rotation coordinated control system based on oxygen and carbon content feedback according to claim 1, characterized in that, The speed parameter correction module includes: Gradient calculation submodule: Based on the stage-specific oxygen and carbon feature set, extract the radial and axial distribution data of oxygen and carbon content in the feature set, calculate the difference of oxygen and carbon content between each sampling point in the radial direction to obtain the radial gradient value, calculate the difference of oxygen and carbon content between each sampling point in the axial direction to obtain the axial gradient value, and perform a weighted summation operation on the radial gradient value and the axial gradient value to generate a comprehensive gradient value. Gradient Comparison Submodule: Calls the gradient synthesis value, retrieves the preset gradient threshold range, compares the gradient synthesis value with the gradient threshold range, determines the threshold range to which the gradient synthesis value belongs, retrieves the speed difference benchmark value of the corresponding range, and generates a benchmark value matching table. Parameter correction submodule: Based on the benchmark value matching table, calculate the degree of deviation between the gradient comprehensive value and the corresponding interval median value, perform linear correction on the speed difference benchmark value according to the percentage of deviation, and allocate the corrected speed difference to the crystal rotation speed parameter and the crucible rotation speed parameter to generate dynamic speed difference parameters.

7. The crystal rotation and crucible rotation coordinated control system based on oxygen and carbon content feedback according to claim 6, characterized in that, The regulation effect verification module includes: Data acquisition submodule: calls the crystal rotating crucible rotation parameter group and dynamic speed difference parameter, collects real-time oxygen content data and real-time carbon content data under the corresponding working conditions of the two sets of parameters, integrates the two sets of parameters and oxygen and carbon data according to the time nodes, and generates the adjusted oxygen and carbon dataset. Uniformity calculation submodule: Based on the oxygen and carbon dataset after regulation, extract oxygen content distribution data and carbon content distribution data, calculate the dispersion of oxygen content data and carbon content data, obtain the oxygen and carbon content uniformity through the dispersion of distribution data, and generate oxygen and carbon uniformity values. The adaptation coefficient generation submodule retrieves the oxygen and carbon uniformity value, retrieves the preset uniformity threshold, compares the oxygen and carbon uniformity value with the uniformity threshold, calculates the parameter adaptation ratio based on the comparison result, and generates the oxygen and carbon regulation adaptation coefficient.

Images