WO2023221576A1 - 氧含量控制方法、装置、电子设备及存储介质 - Google Patents

氧含量控制方法、装置、电子设备及存储介质 Download PDF

Info

Publication number
WO2023221576A1
WO2023221576A1 PCT/CN2023/076448 CN2023076448W WO2023221576A1 WO 2023221576 A1 WO2023221576 A1 WO 2023221576A1 CN 2023076448 W CN2023076448 W CN 2023076448W WO 2023221576 A1 WO2023221576 A1 WO 2023221576A1
Authority
WO
WIPO (PCT)
Prior art keywords
oxygen content
characteristic data
data
threshold
preset
Prior art date
Application number
PCT/CN2023/076448
Other languages
English (en)
French (fr)
Inventor
李广砥
张伟建
王莎莎
王正远
郭力
杨正华
Original Assignee
隆基绿能科技股份有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 隆基绿能科技股份有限公司 filed Critical 隆基绿能科技股份有限公司
Publication of WO2023221576A1 publication Critical patent/WO2023221576A1/zh

Links

Classifications

    • C—CHEMISTRY; METALLURGY
    • C30—CRYSTAL GROWTH
    • C30B—SINGLE-CRYSTAL GROWTH; UNIDIRECTIONAL SOLIDIFICATION OF EUTECTIC MATERIAL OR UNIDIRECTIONAL DEMIXING OF EUTECTOID MATERIAL; REFINING BY ZONE-MELTING OF MATERIAL; PRODUCTION OF A HOMOGENEOUS POLYCRYSTALLINE MATERIAL WITH DEFINED STRUCTURE; SINGLE CRYSTALS OR HOMOGENEOUS POLYCRYSTALLINE MATERIAL WITH DEFINED STRUCTURE; AFTER-TREATMENT OF SINGLE CRYSTALS OR A HOMOGENEOUS POLYCRYSTALLINE MATERIAL WITH DEFINED STRUCTURE; APPARATUS THEREFOR
    • C30B15/00—Single-crystal growth by pulling from a melt, e.g. Czochralski method
    • C—CHEMISTRY; METALLURGY
    • C30—CRYSTAL GROWTH
    • C30B—SINGLE-CRYSTAL GROWTH; UNIDIRECTIONAL SOLIDIFICATION OF EUTECTIC MATERIAL OR UNIDIRECTIONAL DEMIXING OF EUTECTOID MATERIAL; REFINING BY ZONE-MELTING OF MATERIAL; PRODUCTION OF A HOMOGENEOUS POLYCRYSTALLINE MATERIAL WITH DEFINED STRUCTURE; SINGLE CRYSTALS OR HOMOGENEOUS POLYCRYSTALLINE MATERIAL WITH DEFINED STRUCTURE; AFTER-TREATMENT OF SINGLE CRYSTALS OR A HOMOGENEOUS POLYCRYSTALLINE MATERIAL WITH DEFINED STRUCTURE; APPARATUS THEREFOR
    • C30B15/00—Single-crystal growth by pulling from a melt, e.g. Czochralski method
    • C30B15/20—Controlling or regulating
    • C—CHEMISTRY; METALLURGY
    • C30—CRYSTAL GROWTH
    • C30B—SINGLE-CRYSTAL GROWTH; UNIDIRECTIONAL SOLIDIFICATION OF EUTECTIC MATERIAL OR UNIDIRECTIONAL DEMIXING OF EUTECTOID MATERIAL; REFINING BY ZONE-MELTING OF MATERIAL; PRODUCTION OF A HOMOGENEOUS POLYCRYSTALLINE MATERIAL WITH DEFINED STRUCTURE; SINGLE CRYSTALS OR HOMOGENEOUS POLYCRYSTALLINE MATERIAL WITH DEFINED STRUCTURE; AFTER-TREATMENT OF SINGLE CRYSTALS OR A HOMOGENEOUS POLYCRYSTALLINE MATERIAL WITH DEFINED STRUCTURE; APPARATUS THEREFOR
    • C30B29/00—Single crystals or homogeneous polycrystalline material with defined structure characterised by the material or by their shape
    • C30B29/02—Elements
    • C30B29/06—Silicon

Definitions

  • the present invention relates to the technical field of crystal preparation, and in particular to an oxygen content control method, an oxygen content control device, an electronic device and a storage medium.
  • the preparation process of single crystal silicon materials is mainly Czochralski process/CZ, which uses the Czochralski process to refine polycrystalline silicon raw materials into single crystal silicon.
  • the process of generating rod-shaped single crystal silicon crystals during the Czochralski single crystal process is divided into steps such as charging, melting, feeding, temperature adjustment, seeding, shoulder placement, shoulder rotation, equal diameter, and finishing.
  • silicon reacts with the quartz crucible containing silicon, namely: SiO 2 +Si ⁇ 2SiO ⁇ ,
  • the reaction results in an increase in the oxygen content in the molten silicon, which is the main source of oxygen in the silicon.
  • the concentration of oxygen incorporated into the crystal depends on the three diffusion boundary layers and the three interface areas.
  • the thickness of the boundary layer depends on the melt heat convection, while the interface area depends on the charging amount and the size and shape of the crucible, as well as the crystal diameter.
  • the ratio of the crucible and melt interface area to the free surface area of the melt is an important factor in determining the oxygen content in the crystal.
  • the atmosphere in the furnace chamber also has an important impact on the oxygen content in the growing crystals.
  • the weight continues to increase, while the weight of the melt in the crucible decreases, and the melt level drops.
  • the interface area between the crucible and the melt gradually decreases, while the free surface of the melt remains constant, that is, the amount of oxygen evaporated from the melt through the free surface remains constant. Therefore, the distribution of oxygen in the crystal is uneven.
  • the content in the head of the crystal is high and the content in the tail is low. The content is high in the center of the crystal and low in the edge.
  • the adsorption and device technology require that the single crystal oxygen content provided by the material manufacturer is uniformly distributed, especially the radial distribution uniformity.
  • the dissolution rate of quartz crucible is mainly related to factors such as temperature, furnace chamber pressure, quartz crucible surface state, and boundary layer thickness at the crucible/melt interface. The higher the temperature, the lower the pressure, the greater the surface roughness, the thinner the thickness of the boundary layer, and the faster the dissolution rate of the quartz crucible.
  • the vapor pressure of SiO can reach 12mbar. Under the current commonly used vacuum crystal pulling process conditions, SiO volatilization will not be hindered, so it is easy to volatilize from the free surface of the melt.
  • Oxygen is the impurity with the highest content and the most complex behavior in Czochralski single crystal silicon. In the past 40 years, the control and behavior of oxygen content in silicon has been one of the important research topics in the field of silicon materials. Oxygen in crystals is both beneficial and harmful.
  • oxygen in the crystal is both beneficial and harmful. Therefore, the oxygen content needs to be controlled during the process of pulling silicon single crystal.
  • the experimental process of the current method of controlling oxygen content is complex, the cycle is long, and the effect is not good enough.
  • embodiments of the present invention are proposed to provide a parameter adjustment method that overcomes the above problems or at least partially solves the above problems, so as to solve the problem that the experimental process of the method of controlling oxygen content is complicated, the cycle is long, and the effect is not good enough.
  • the problem is proposed to provide a parameter adjustment method that overcomes the above problems or at least partially solves the above problems, so as to solve the problem that the experimental process of the method of controlling oxygen content is complicated, the cycle is long, and the effect is not good enough. The problem.
  • embodiments of the present invention also provide an oxygen content prediction device, an electronic device and a storage medium to ensure the implementation and application of the above method.
  • an oxygen content prediction method which includes:
  • For each type of characteristic data determine the oxygen content difference information between the characteristic data whose oxygen content is higher than the preset high oxygen content threshold and the characteristic data whose oxygen content is lower than the preset low oxygen content threshold;
  • the oxygen content difference information corresponding to the characteristic data determine the order of importance of various characteristic data to the oxygen content
  • the process parameters corresponding to the various characteristic data are adjusted to control the oxygen content.
  • determining the oxygen content difference information between the characteristic data whose oxygen content is higher than the preset high oxygen content threshold and the characteristic data whose oxygen content is lower than the preset low oxygen content threshold includes:
  • the characteristic data whose oxygen content is higher than the preset high oxygen content threshold is classified into a high oxygen content data set, and the characteristic data whose oxygen content is lower than the preset low oxygen content threshold is classified into a low oxygen content data set.
  • content data set For each type of characteristic data, the characteristic data whose oxygen content is higher than the preset high oxygen content threshold is classified into a high oxygen content data set, and the characteristic data whose oxygen content is lower than the preset low oxygen content threshold is classified into a low oxygen content data set.
  • the oxygen content difference information between the two data sets is determined based on the high oxygen content data set and the low oxygen content data set.
  • determining the oxygen content difference information between the two data sets according to the high oxygen content data set and the low oxygen content data set includes:
  • the difference between the median oxygen content of the high oxygen content data set and the median oxygen content of the low oxygen content data set is calculated as the difference between the two data sets.
  • determining the order of importance of various characteristic data to oxygen content based on the oxygen content difference information corresponding to the characteristic data includes:
  • the oxygen content difference information corresponding to the characteristic data calculate the first characteristic weight of various characteristic data on the oxygen content
  • the importance ranking of various feature data to the oxygen content is determined.
  • the method also includes:
  • the various characteristic data and the corresponding oxygen content are input into the oxygen content prediction model; wherein the oxygen content prediction model passes the characteristic data samples before the head stage of equal-diameter growth, and the corresponding marked head stages of equal-diameter growth.
  • the sample oxygen content training in the external phase is obtained;
  • For each type of feature data perform a weighted average of the first feature weight and the second feature weight, and use the average as the total feature weight corresponding to the feature data;
  • the importance ranking of various feature data to the oxygen content is determined.
  • the method before sorting the importance of the oxygen content according to the various characteristic data and adjusting the process parameters corresponding to the various characteristic data, the method further includes:
  • the characteristic data whose oxygen content is between a preset high oxygen content threshold and a preset low oxygen content threshold is divided into a medium oxygen content data set;
  • the data distribution diagrams of the high oxygen content data set, the medium oxygen content data set, the low oxygen content data set and all characteristic data are respectively displayed; wherein, in the data distribution diagram Includes oxygen content difference information corresponding to the characteristic data;
  • Sorting the importance of oxygen content according to the various characteristic data, and adjusting the process parameters corresponding to the various characteristic data includes:
  • the process parameters corresponding to the various characteristic data are adjusted.
  • the method before determining, for each type of characteristic data, the oxygen content difference information between the characteristic data whose oxygen content is higher than the preset high oxygen content threshold and the characteristic data whose oxygen content is lower than the preset low oxygen content threshold. , the method also includes:
  • the preset high oxygen content threshold and the preset low oxygen content threshold are determined based on the density map of the characteristic data with respect to the oxygen content, so that the oxygen content is between the preset high oxygen content threshold and the preset low oxygen content threshold.
  • the quantity proportion of the characteristic data between the preset low oxygen content thresholds is higher than the preset proportion, and the preset high oxygen content threshold is lower than the preset maximum oxygen content.
  • An embodiment of the present invention also discloses an oxygen content prediction device, which includes:
  • the data acquisition module is used to acquire various characteristic data during the multiple Czochralski single crystal processes, and the corresponding oxygen content in the head stage of equal-diameter growth;
  • a difference determination module for each type of characteristic data, determine the oxygen content difference information between the characteristic data whose oxygen content is higher than the preset high oxygen content threshold and the characteristic data whose oxygen content is lower than the preset low oxygen content threshold;
  • a ranking determination module configured to determine the importance ranking of various characteristic data to the oxygen content based on the oxygen content difference information corresponding to the characteristic data
  • the oxygen content control module is used to rank the importance of the oxygen content according to the various characteristic data, and adjust the process parameters corresponding to the various characteristic data to control the oxygen content.
  • the difference determination module includes:
  • the data set division submodule is used to divide the characteristic data whose oxygen content is higher than the preset high oxygen content threshold into a high oxygen content data set for each type of characteristic data, and divide the characteristic data whose oxygen content is lower than the preset low oxygen content into a high oxygen content data set.
  • the threshold feature data is divided into low oxygen content data sets;
  • the difference determination sub-module is used to determine, for each type of characteristic data, the oxygen content difference information between the two data sets according to the high oxygen content data set and the low oxygen content data set.
  • the difference determination sub-module includes:
  • a difference calculation unit used to calculate the oxygen content of the high oxygen content data set for each characteristic data.
  • the difference between the median of the oxygen content and the median of the oxygen content of the low oxygen content data set is used as the oxygen content difference information between the two data sets.
  • the ranking determination module includes:
  • the first determination calculation submodule is used to calculate the first characteristic weight of various characteristic data on the oxygen content according to the oxygen content difference information corresponding to the characteristic data;
  • the ranking determination sub-module is used to determine the importance ranking of various characteristic data to the oxygen content according to the first characteristic weight.
  • the device also includes:
  • a data input module for inputting the various characteristic data and the corresponding oxygen content into an oxygen content prediction model; wherein the oxygen content prediction model passes the characteristic data samples before the head stage of equal-diameter growth, and the corresponding markers The sample oxygen content of the head stage of isometric growth is obtained by training;
  • a model training module used to train the oxygen content prediction model according to the multiple characteristic data and the corresponding oxygen content
  • a weight output module is used to output the second feature weight corresponding to various feature data in the trained oxygen content prediction model
  • the ordering determination sub-module includes:
  • a total weight determination unit configured to perform a weighted average of the first feature weight and the second feature weight for each type of feature data, and use the average value as the total feature weight corresponding to the feature data;
  • a ranking determination unit is configured to determine the importance ranking of various feature data to the oxygen content according to the total feature weight.
  • the device also includes:
  • a data set dividing module configured to sort the oxygen content according to the importance of the various characteristic data and adjust the process parameters corresponding to each characteristic data, for each characteristic data, divide the oxygen content into Characteristic data between the preset high oxygen content threshold and the preset low oxygen content threshold are divided into medium oxygen content data sets;
  • a distribution diagram display module configured to display the data distribution diagrams of the high oxygen content data set, the medium oxygen content data set, the low oxygen content data set and all characteristic data respectively for each type of characteristic data; wherein, the data distribution diagram Includes oxygen content difference information corresponding to the characteristic data;
  • the oxygen content control module includes:
  • Parameter adjustment submodule used to rank the importance of oxygen content according to the various characteristic data and Data distribution chart, adjust the process parameters corresponding to each characteristic data.
  • the device also includes:
  • a threshold determination module configured to determine, for each type of characteristic data, an oxygen content difference between the characteristic data whose oxygen content is higher than a preset high oxygen content threshold and the characteristic data whose oxygen content is lower than the preset low oxygen content threshold.
  • An embodiment of the present invention also discloses an electronic device, which is characterized in that it includes a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory complete communication with each other through the communication bus;
  • Memory used to store computer programs
  • the processor is used to implement the above method steps when executing the program stored in the memory.
  • An embodiment of the present invention also discloses a readable storage medium.
  • the instructions in the storage medium are executed by a processor of an electronic device, the electronic device can perform one or more of the oxygen content described in the embodiment of the present invention. method of prediction.
  • the embodiment of the present invention by obtaining a variety of characteristic data in the process of multiple Czochralski single crystals and the corresponding oxygen content in the head stage of equal diameter growth, for each characteristic data, it is determined that the oxygen content is higher than the preset high Oxygen content difference information between the characteristic data of the oxygen content threshold and the characteristic data whose oxygen content is lower than the preset low oxygen content threshold.
  • the oxygen content difference information corresponding to the characteristic data determine the impact of various characteristic data on the oxygen content. Sort the importance of oxygen content according to the importance of various characteristic data to oxygen content, and adjust the process parameters corresponding to each characteristic data, so that the characteristic data that affects oxygen content can be analyzed using the data accumulated in long-term production operations. Sorting, so that the oxygen content can be automatically controlled more accurately, thereby improving the quality of the product.
  • Figure 1 is a step flow chart of an embodiment of an oxygen content control method of the present invention
  • Figure 2 is a step flow chart of another embodiment of the oxygen content control method of the present invention.
  • Figure 3 is a schematic diagram of an example of oxygen content prediction
  • FIG. 4 is a structural block diagram of an embodiment of an oxygen content control device of the present invention.
  • FIG. 5 is a structural block diagram of a computing device for oxygen content prediction according to an exemplary embodiment.
  • FIG. 1 a step flow chart of an embodiment of an oxygen content control method of the present invention is shown. Specifically, it may include the following steps:
  • Step 101 Obtain various characteristic data in the process of multiple Czochralski single crystals and the corresponding oxygen content in the head stage of equal-diameter growth.
  • the Czochralski single crystal process is a process of refining raw materials into single crystals using the Czochralski method, for example, the process of Czochralski single crystal silicon.
  • the process of Czochralski single crystal can be divided into melting stage, feeding stage, temperature adjustment stage, seeding stage, shoulder placing stage, shoulder turning stage, equal diameter growth stage, etc.
  • the beginning part of the equal diameter growth stage is called the head stage of equal diameter growth (for example, the stage from the beginning of equal diameter to the crystal length increasing by 50 mm).
  • various characteristic data are obtained.
  • the core characteristic data includes: maxsetmainheaterpowertime (maximum main power when melting), bodysumtime (equal diameter time), body0remainweight (remaining material amount at zero time of equal diameter) , bodyavgdiameter (average pulling speed of equal diameter), crystallength (crystal length), bottomheater (bottom heater power), bodystdseedrotation (standard deviation of rotation speed of equal diameter crystals), bodyavgseedrotation (average crystal rotation speed of equal diameters), bodyendargonflow (argon gas flow at the end of equal diameters), temp_t (temperature adjustment duration), neck_t (shoulder length), head_o (head oxygen content value) etc.
  • it may include any applicable feature data, and the embodiment of the present invention does not limit this.
  • a single crystal is produced during each Czochralski single crystal process, and the oxygen content in the head stage of equal-diameter growth is detected through equipment for detecting oxygen content, so as to obtain the oxygen content.
  • equipment for detecting oxygen content so as to obtain the oxygen content.
  • a detection device using the principle of spectral imaging can accurately detect the oxygen content.
  • Step 102 For each type of characteristic data, determine the oxygen content difference information between the characteristic data whose oxygen content is higher than the preset high oxygen content threshold and the characteristic data whose oxygen content is lower than the preset low oxygen content threshold.
  • the preset high oxygen content threshold and the preset low oxygen content threshold are set according to business logic, and can be specifically set according to the actual situation.
  • the embodiment of the present invention does not limit this. If it is higher than the preset high oxygen content threshold, it indicates that the oxygen content is high, and if it is lower than the preset low oxygen content threshold, it indicates that the oxygen content is low.
  • the oxygen content difference information includes one or more of the difference in the median of the oxygen content, the difference in the mean of the oxygen content, the difference in the variance of the oxygen content, or any other suitable representation of oxygen. Information on content differences is not limited in the embodiments of the present invention.
  • the oxygen content difference information between the characteristic data whose oxygen content is higher than the preset high oxygen content threshold and the characteristic data whose oxygen content is lower than the preset low oxygen content threshold may include multiple methods. For example, take the median of the oxygen content corresponding to the characteristic data whose oxygen content is higher than the preset high oxygen content threshold, and the median of the oxygen content corresponding to the characteristic data whose oxygen content is lower than the preset low oxygen content threshold, and calculate the two The difference between the medians.
  • Step 103 Determine various characteristic data based on the oxygen content difference information corresponding to the characteristic data. Rank the importance of oxygen content.
  • the greater the value of the oxygen content difference information corresponding to a certain characteristic data the greater the impact of the characteristic data on the oxygen content, and the greater the value of the oxygen content difference information corresponding to a certain characteristic data.
  • the oxygen content difference information corresponding to the characteristic data the importance ranking of various characteristic data on the oxygen content can be determined, so that the importance of the characteristic data on the oxygen content is ranked from high to low.
  • Step 104 Sort the oxygen content according to the importance of the various characteristic data, and adjust the process parameters corresponding to the various characteristic data to control the oxygen content.
  • the process parameters correspond to the characteristic data and are process-related parameters set for the equipment, including body0remainweight (remaining material amount at zero time of equal diameter), bodyendargonflow (argon flow rate at the end of equal diameter), bodyavgseedrotation (average equal diameter Crystal rotation speed value), etc., or any other applicable process parameters, the embodiments of the present invention do not limit this.
  • the embodiment of the present invention by obtaining a variety of characteristic data in the process of multiple Czochralski single crystals and the corresponding oxygen content in the head stage of equal diameter growth, for each characteristic data, it is determined that the oxygen content is higher than the preset high Oxygen content difference information between the characteristic data of the oxygen content threshold and the characteristic data whose oxygen content is lower than the preset low oxygen content threshold.
  • the oxygen content difference information corresponding to the characteristic data determine the impact of various characteristic data on the oxygen content. Sort the importance of oxygen content according to the importance of various characteristic data to oxygen content, and adjust the process parameters corresponding to each characteristic data, so that the characteristic data that affects oxygen content can be analyzed using the data accumulated in long-term production operations. Sorting, so that the oxygen content can be automatically controlled more accurately, thereby improving the quality of the product.
  • each characteristic data it is determined that the characteristic data whose oxygen content is higher than the preset high oxygen content threshold and the characteristic data whose oxygen content is lower than the preset low oxygen content threshold are determined.
  • the oxygen content difference information between the two characteristic data it also includes: for each characteristic data, based on the density map of the characteristic data for the oxygen content, determining the preset high oxygen content threshold and the preset low oxygen content threshold, so that the oxygen content The quantity proportion of the characteristic data whose content is between the preset high oxygen content threshold and the preset low oxygen content threshold is higher than the preset proportion, and the preset high oxygen content threshold is lower than the preset maximum oxygen content .
  • the preset high oxygen content threshold and the preset low oxygen content threshold are determined.
  • the determined preset high oxygen content threshold and the preset low oxygen content threshold satisfy two conditions at the same time, that is, the proportion of the number of characteristic data with oxygen content between the preset high oxygen content threshold and the preset low oxygen content threshold is greater than
  • the preset proportion and the preset high oxygen content threshold are lower than the preset maximum oxygen content.
  • the preset proportion can be set according to actual needs, and the embodiment of the present invention does not limit this.
  • the preset maximum oxygen content can be set according to actual needs, and this is not limited in the embodiment of the present invention.
  • the preset high oxygen content threshold is set to 8.5
  • the preset low oxygen content threshold is set to 6.5.
  • FIG. 2 there is shown a step flow chart of another embodiment of the oxygen content control method of the present invention, which may specifically include the following steps:
  • Step 201 Obtain various characteristic data in the process of multiple Czochralski single crystals and the corresponding oxygen content in the head stage of equal-diameter growth.
  • Step 202 For each type of characteristic data, classify the characteristic data whose oxygen content is higher than the preset high oxygen content threshold into a high oxygen content data set, and classify the characteristic data whose oxygen content is lower than the preset low oxygen content threshold. to the low oxygen content data set.
  • each type of characteristic data all characteristic data is divided into multiple data sets according to oxygen content. Among them, the characteristic data whose oxygen content is higher than the preset high oxygen content threshold is classified into a high oxygen content data set, and the characteristic data whose oxygen content is lower than the preset low oxygen content threshold is classified into a low oxygen content data set.
  • Step 203 For each characteristic data, determine the oxygen content difference information between the two data sets based on the high oxygen content data set and the low oxygen content data set.
  • the oxygen content difference information between the high oxygen content data set and the low oxygen content data set there are multiple methods for determining the oxygen content difference information between the high oxygen content data set and the low oxygen content data set. For example, calculating high oxygen content data The difference between the median oxygen content of the set and the median oxygen content of the low oxygen content data set is used as the oxygen content difference information between the two data sets. For another example, the difference between the mean value of the oxygen content of the high oxygen content data set and the mean value of the oxygen content of the low oxygen content data set is calculated as the oxygen content difference information between the two data sets. For another example, the difference between each oxygen content of the high oxygen content data set and each oxygen content of the low oxygen content data set is calculated as the oxygen content difference information between the two data sets.
  • a method of determining the oxygen content difference information between the two data sets is determined. In an implementation manner, it may include: for each characteristic data, calculating the difference between the median of the oxygen content of the high oxygen content data set and the median of the oxygen content of the low oxygen content data set, As the oxygen content difference information between the two data sets.
  • the oxygen content corresponding to the 50th or 51st feature data is the median oxygen content of the high oxygen content data set.
  • the oxygen content corresponding to the 26th feature data is the median oxygen content of the low oxygen content data set.
  • the difference between the two medians is calculated as the oxygen content difference information between the two data sets.
  • Step 204 Based on the oxygen content difference information corresponding to the characteristic data, determine the order of importance of various characteristic data to the oxygen content.
  • a specific implementation of determining the order of importance of various characteristic data to oxygen content based on the oxygen content difference information corresponding to the characteristic data may also include:
  • the oxygen content difference information corresponding to the characteristic data is used to calculate the first characteristic weight of various characteristic data to the oxygen content, and the order of importance of various characteristic data to the oxygen content is determined based on the first characteristic weight.
  • the first feature weight represents the importance of the feature data to the oxygen content. The greater the weight of the first feature, the greater the impact of this feature data on the oxygen content. The smaller the weight of the first feature, the smaller the impact of this feature data on the oxygen content. The higher the weight of the first feature, the higher the importance of feature data to oxygen content.
  • the oxygen content difference information corresponding to the feature data there are many ways to calculate the first feature weight of various feature data on the oxygen content.
  • the difference in medians can be directly used as the first feature weight, or the value obtained by multiplying the difference in medians by a preset coefficient can be used as the first feature weight.
  • a feature weight The embodiment of the present invention does not limit this.
  • it may also include: inputting the various characteristic data and the corresponding oxygen content into an oxygen content prediction model; wherein the oxygen content prediction model is configured by The characteristic data samples before the stage and the oxygen content of the samples corresponding to the marked equal-diameter growth head stage are trained; the oxygen content prediction model is trained according to the various characteristic data and the corresponding oxygen content; and the trained output Second feature weights corresponding to various feature data in the oxygen content prediction model.
  • a specific implementation manner of determining the importance ranking of various characteristic data to the oxygen content according to the first characteristic weight may include: for each characteristic data, performing an evaluation on the first characteristic weight and the second characteristic weight. A weighted average is used, and the average is used as the total feature weight corresponding to the feature data; based on the total feature weight, the importance ranking of various feature data to the oxygen content is determined.
  • the correlation between the machine learning feature data and the oxygen content in the head stage of equal-diameter growth can be used to obtain an oxygen content prediction model that can predict the oxygen content.
  • Oxygen content prediction models include: Naive Bayes model, extreme gradient boosting model, or category gradient boosting model.
  • Naive Bayes models are a simple way to build a classifier that assigns class labels represented by feature values to problem instances, drawn from a finite set.
  • the extreme gradient boosting (eXtreme Gradient Boosting, XGboost) model is a model that uses boosting trees for prediction in large-scale parallelism.
  • the Categorical Gradient Boosting (CATboost) model is a model of the gradient boosting algorithm that can handle categorical features well.
  • the Light Gradient Boosting Machine (lightGBM) model is a fast, distributed, high-performance gradient boosting framework based on the decision tree algorithm. Can be used for sorting, classification, regression and many other machine learning tasks.
  • the data used to train the model is divided into a training set and a test set, of which the test set generally accounts for 20%.
  • Use integrated learning models such as XGboost, lightGBM, and CATboost to build a learner, use Bayesian optimization methods to tune model parameters, and finally select the overall optimal CATboost model as the optimal model.
  • the R2 (R-Square, coefficient of determination) index of the model is 0.51.
  • the training process of the oxygen content prediction model is the process of determining the feature weights corresponding to various feature data based on the training data. After the oxygen content prediction model is trained, the feature weights are determined. From the trained oxygen content prediction model, obtain the second feature weight corresponding to various feature data. Among them, the greater the weight of the second feature, the greater the impact of the second feature data on the oxygen content. The smaller the weight, the smaller the impact of this second characteristic data on the oxygen content.
  • the first feature weight and the second feature weight are weighted and averaged, and the average is used as the total feature weight corresponding to the feature data.
  • the weights of the two can be the same, then just find the average of the first feature weight and the second feature weight directly.
  • the weights of the two can also be different, and the weights of the two can be adjusted according to the actual effect, so that the obtained total feature weight can more accurately reflect the importance of various feature data to the oxygen content.
  • the method of determining the importance ranking of various feature data to the oxygen content based on the total feature weight can be found in determining the importance ranking of various feature data to the oxygen content based on the first feature weight.
  • Step 205 Sort the oxygen content according to the importance of the various characteristic data, and adjust the process parameters corresponding to the various characteristic data to control the oxygen content.
  • the process parameters corresponding to the various characteristic data before adjusting the process parameters corresponding to the various characteristic data according to the importance of the oxygen content according to the various characteristic data, it may also include: data, classify the characteristic data whose oxygen content is between the preset high oxygen content threshold and the preset low oxygen content threshold into a medium oxygen content data set, and display the high oxygen content data set, medium oxygen content data set, and medium oxygen content data set respectively for each type of characteristic data.
  • a specific implementation of sorting the importance of the oxygen content according to the various characteristic data and adjusting the process parameters corresponding to the various characteristic data may include: ranking the oxygen content according to the various characteristic data.
  • the content importance ranking and data distribution chart are used to adjust the process parameters corresponding to each characteristic data.
  • the characteristic data whose oxygen content is between the preset high oxygen content threshold and the preset low oxygen content threshold is divided into a medium oxygen content data set.
  • the feature data in the medium oxygen content data set is the feature data that makes the oxygen content moderate.
  • a data distribution chart is a graph that reflects the distribution of data.
  • Data distribution diagrams include box plots, density diagrams, histograms, etc., or any other applicable distribution diagrams, which are not limited in the embodiment of the present invention.
  • the distribution shows the data distribution diagrams of the high oxygen content data set, the medium oxygen content data set, the low oxygen content data set and all the characteristic data, that is, a total of four data distribution diagrams.
  • the data distribution chart also displays the oxygen content difference information corresponding to the characteristic data for reference.
  • the process parameters corresponding to each characteristic data are adjusted.
  • the data distribution chart can display the parameter areas of various characteristic data time, as well as key values such as median, mean, oxygen content difference information, characteristic data of oxygen content at the preset high oxygen content threshold, characteristic data of oxygen content at the preset low oxygen content threshold. Based on the data distribution diagram, when adjusting the process parameters corresponding to each characteristic data, the adjusted values of the process parameters can be determined more conveniently and accurately.
  • Figure 3 shows a schematic diagram of a box plot of characteristic data versus oxygen content.
  • the high oxygen content data set corresponds to the box plot labeled more
  • the low oxygen content data set corresponds to the box plot labeled less
  • the medium oxygen content data set corresponds to the box plot labeled between
  • all feature data corresponds to Box plot labeled all.
  • the horizontal line in each box represents the median of the data set. Since between data accounts for the vast majority, the graph corresponding to all data is visually almost the same as the graph corresponding to between data.
  • the oxygen content is higher than the predetermined value.
  • the characteristic data with a high oxygen content threshold is divided into a high oxygen content data set
  • the characteristic data with an oxygen content lower than the preset low oxygen content threshold is divided into a low oxygen content data set.
  • the oxygen content data set and the low oxygen content data set determine the oxygen content difference information between the two data sets, and determine the importance of various characteristic data to the oxygen content based on the oxygen content difference information corresponding to the characteristic data.
  • Sexual ranking according to the importance of various characteristic data to oxygen content, adjust the process parameters corresponding to each characteristic data, so that the data accumulated in long-term production operations can be used to analyze the ranking of characteristic data that affects oxygen content. , so that the oxygen content can be automatically controlled more accurately, thereby improving the quality of the product.
  • FIG. 4 a structural block diagram of an embodiment of an oxygen content control device of the present invention is shown, which may specifically include the following modules:
  • the data acquisition module 301 is used to acquire various characteristic data in the process of multiple Czochralski single crystals, and the corresponding oxygen content in the head stage of equal-diameter growth;
  • the difference determination module 302 is configured to determine, for each type of characteristic data, the oxygen content difference information between the characteristic data whose oxygen content is higher than the preset high oxygen content threshold and the characteristic data whose oxygen content is lower than the preset low oxygen content threshold. ;
  • the ranking determination module 303 is used to determine the importance ranking of various characteristic data to the oxygen content based on the oxygen content difference information corresponding to the characteristic data;
  • the oxygen content control module 304 is used to rank the importance of the oxygen content according to the various characteristic data, and adjust the process parameters corresponding to the various characteristic data to control the oxygen content.
  • the difference determination module includes:
  • the data set division submodule is used to divide the characteristic data whose oxygen content is higher than the preset high oxygen content threshold into a high oxygen content data set for each type of characteristic data, and divide the characteristic data whose oxygen content is lower than the preset low oxygen content into a high oxygen content data set.
  • the threshold feature data is divided into low oxygen content data sets;
  • the difference determination sub-module is used to determine, for each type of characteristic data, the oxygen content difference information between the two data sets according to the high oxygen content data set and the low oxygen content data set.
  • the difference determination sub-module includes:
  • a difference calculation unit configured to calculate, for each characteristic data, a difference between the median of the oxygen content of the high oxygen content data set and the median of the oxygen content of the low oxygen content data set, as Oxygen content difference information between the two data sets.
  • the ranking determination module includes:
  • the first determination calculation submodule is used to calculate the first characteristic weight of various characteristic data on the oxygen content according to the oxygen content difference information corresponding to the characteristic data;
  • the ranking determination sub-module is used to determine the importance ranking of various characteristic data to the oxygen content according to the first characteristic weight.
  • the device also includes:
  • a data input module for inputting the various characteristic data and the corresponding oxygen content into an oxygen content prediction model; wherein the oxygen content prediction model passes the characteristic data samples before the head stage of equal-diameter growth, and the corresponding markers The sample oxygen content of the head stage of isometric growth is obtained by training;
  • a model training module used to train the oxygen content prediction model according to the multiple characteristic data and the corresponding oxygen content
  • a weight output module is used to output the second feature weight corresponding to various feature data in the trained oxygen content prediction model
  • the ordering determination sub-module includes:
  • a total weight determination unit configured to perform a weighted average of the first feature weight and the second feature weight for each type of feature data, and use the average value as the total feature weight corresponding to the feature data;
  • a ranking determination unit is configured to determine the importance ranking of various feature data to the oxygen content according to the total feature weight.
  • the device also includes:
  • a data set dividing module configured to sort the oxygen content according to the importance of the various characteristic data and adjust the process parameters corresponding to each characteristic data, for each characteristic data, divide the oxygen content into Characteristic data between the preset high oxygen content threshold and the preset low oxygen content threshold are divided into medium oxygen content data sets;
  • a distribution diagram display module configured to display the data distribution diagrams of the high oxygen content data set, the medium oxygen content data set, the low oxygen content data set and all characteristic data respectively for each type of characteristic data; wherein, the data distribution diagram Includes oxygen content difference information corresponding to the characteristic data;
  • the oxygen content control module includes:
  • the parameter adjustment sub-module is used to adjust the process parameters corresponding to each of the characteristic data according to the importance ranking of the oxygen content and the data distribution diagram of the various characteristic data.
  • the device also includes:
  • a threshold determination module configured to determine, for each type of characteristic data, an oxygen content difference between the characteristic data whose oxygen content is higher than a preset high oxygen content threshold and the characteristic data whose oxygen content is lower than the preset low oxygen content threshold.
  • the embodiment of the present invention by obtaining a variety of characteristic data in the process of multiple Czochralski single crystals and the corresponding oxygen content in the head stage of equal diameter growth, for each characteristic data, it is determined that the oxygen content is higher than the preset high Oxygen content difference information between the characteristic data of the oxygen content threshold and the characteristic data whose oxygen content is lower than the preset low oxygen content threshold. According to the oxygen content difference information corresponding to the characteristic data, determine the impact of various characteristic data on the oxygen content. Sort the importance of oxygen content according to the importance of various characteristic data to oxygen content, and adjust the process parameters corresponding to each characteristic data, so that the characteristic data that affects oxygen content can be analyzed using the data accumulated in long-term production operations. sorting, so that it can automatically sort Oxygen content can be more precisely controlled, thereby improving product quality.
  • the description is relatively simple. For relevant details, please refer to the partial description of the method embodiment.
  • FIG. 5 is a structural block diagram of an electronic device 400 for oxygen content prediction according to an exemplary embodiment.
  • the electronic device 400 may be a mobile phone, a computer, a digital broadcast terminal, a messaging device, a game console, a tablet device, a medical device, a fitness device, a personal digital assistant, or the like.
  • the electronic device 400 may include one or more of the following components: a processing component 402 , a memory 404 , a power supply component 406 , a multimedia component 408 , an audio component 410 , an input/output (I/O) interface 412 , and a sensor component 414 , and communication component 416.
  • Processing component 402 generally controls the overall operations of electronic device 400, such as operations associated with display, phone calls, data communications, camera operations, and recording operations.
  • the processing component 402 may include one or more processors 420 to execute instructions to complete all or part of the steps of the oxygen content prediction method described above.
  • processing component 402 may include one or more modules that facilitate interaction between processing component 402 and other components.
  • processing component 402 may include a multimedia module to facilitate interaction between multimedia component 408 and processing component 402.
  • Memory 404 is configured to store various types of data to support operations at device 400 . Examples of such data include instructions for any application or method operating on the electronic device 400, contact data, phonebook data, messages, pictures, videos, etc.
  • Memory 404 may be implemented by any type of volatile or non-volatile storage device, or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EEPROM), Programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic or optical disk.
  • SRAM static random access memory
  • EEPROM electrically erasable programmable read-only memory
  • EEPROM erasable programmable read-only memory
  • EPROM Programmable read-only memory
  • PROM programmable read-only memory
  • ROM read-only memory
  • magnetic memory flash memory, magnetic or optical disk.
  • Power component 404 provides power to various components of electronic device 400 .
  • Power components 404 may include a power management system, one or more power supplies, and other components associated with generating, managing, and distributing power to electronic device 400 .
  • Multimedia component 408 includes a screen that provides an output interface between the electronic device 400 and the user.
  • the screen may include a liquid crystal display (LCD) and a touch panel (TP). If the screen includes a touch panel, the screen may be implemented as a touch screen to receive custom input signal from the user.
  • the touch panel includes one or more touch sensors to sense touches, swipes, and gestures on the touch panel. The touch sensor may not only sense the boundary of a touch or slide action, but also detect the duration and pressure associated with the touch or slide action.
  • multimedia component 408 includes a front-facing camera and/or a rear-facing camera.
  • the front camera and/or the rear camera may receive external multimedia data.
  • Each front-facing camera and rear-facing camera can be a fixed optical lens system or have a focal length and optical zoom capabilities.
  • Audio component 410 is configured to output and/or input audio signals.
  • audio component 410 includes a microphone (MIC) configured to receive external audio signals when electronic device 400 is in operating modes, such as call mode, recording mode, and voice recognition mode. The received audio signals may be further stored in memory 404 or sent via communication component 416 .
  • audio component 410 also includes a speaker for outputting audio signals.
  • the I/O interface 412 provides an interface between the processing component 402 and a peripheral interface module, which may be a keyboard, a click wheel, a button, etc. These buttons may include, but are not limited to: Home button, Volume buttons, Start button, and Lock button.
  • Sensor component 414 includes one or more sensors for providing various aspects of status assessment for electronic device 400 .
  • the sensor component 414 can detect the open/closed state of the device 400, the relative positioning of components, such as the display and keypad of the electronic device 400, the sensor component 414 can also detect the electronic device 400 or a component of the electronic device 400. position changes, the presence or absence of user contact with the electronic device 400 , the orientation or acceleration/deceleration of the electronic device 400 and temperature changes of the electronic device 400 .
  • Sensor assembly 414 may include a proximity sensor configured to detect the presence of nearby objects without any physical contact.
  • Sensor assembly 414 may also include a light sensor, such as a CMOS or CCD image sensor, for use in imaging applications.
  • the sensor component 414 may also include an acceleration sensor, a gyroscope sensor, a magnetic sensor, a pressure sensor, or a temperature sensor.
  • Communication component 416 is configured to facilitate wired or wireless communication between electronic device 400 and other devices.
  • the electronic device 400 can access a wireless network based on a communication standard, such as WiFi, 2G or 3G, or a combination thereof.
  • the communication component 414 receives broadcast signals or broadcast related information from an external broadcast management system via a broadcast channel.
  • the communication component 414 also includes a near field communication (NFC) module to facilitate short-range communications.
  • NFC near field communication
  • the NFC module can be implemented based on radio frequency identification (RFID) technology, infrared data association (IrDA) technology, ultra-wideband (UWB) technology, Bluetooth (BT) technology and other technologies.
  • RFID radio frequency identification
  • IrDA infrared data association
  • UWB ultra-wideband
  • Bluetooth Bluetooth
  • electronic device 400 may be configured by one or more application specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable Programming gate array (FPGA), controller, microcontroller, microprocessor or other electronic components are implemented for executing the above oxygen content prediction method.
  • ASICs application specific integrated circuits
  • DSPs digital signal processors
  • DSPDs digital signal processing devices
  • PLDs programmable logic devices
  • FPGA field programmable Programming gate array
  • controller microcontroller, microprocessor or other electronic components are implemented for executing the above oxygen content prediction method.
  • a non-transitory computer-readable storage medium including instructions such as a memory 404 including instructions, which can be executed by the processor 420 of the electronic device 400 to complete the above oxygen content prediction method is also provided.
  • the non-transitory computer-readable storage medium may be ROM, random access memory (RAM), CD-ROM, magnetic tape, floppy disk, optical data storage device, etc.
  • a non-transitory computer-readable storage medium that, when instructions in the storage medium are executed by a processor of a terminal, enables the terminal to perform an oxygen content control method, the method further comprising:
  • For each type of characteristic data determine the oxygen content difference information between the characteristic data whose oxygen content is higher than the preset high oxygen content threshold and the characteristic data whose oxygen content is lower than the preset low oxygen content threshold;
  • the oxygen content difference information corresponding to the characteristic data determine the order of importance of various characteristic data to the oxygen content
  • the process parameters corresponding to the various characteristic data are adjusted to control the oxygen content.
  • determining the oxygen content difference information between the characteristic data whose oxygen content is higher than the preset high oxygen content threshold and the characteristic data whose oxygen content is lower than the preset low oxygen content threshold includes:
  • the characteristic data whose oxygen content is higher than the preset high oxygen content threshold is classified into a high oxygen content data set, and the characteristic data whose oxygen content is lower than the preset low oxygen content threshold is classified into a low oxygen content data set.
  • content data set For each type of characteristic data, the characteristic data whose oxygen content is higher than the preset high oxygen content threshold is classified into a high oxygen content data set, and the characteristic data whose oxygen content is lower than the preset low oxygen content threshold is classified into a low oxygen content data set.
  • the oxygen content difference information between the two data sets is determined based on the high oxygen content data set and the low oxygen content data set.
  • the information to determine the oxygen content difference between the two data sets includes:
  • the difference between the median oxygen content of the high oxygen content data set and the median oxygen content of the low oxygen content data set is calculated as the difference between the two data sets.
  • determining the order of importance of various characteristic data to oxygen content based on the oxygen content difference information corresponding to the characteristic data includes:
  • the oxygen content difference information corresponding to the characteristic data calculate the first characteristic weight of various characteristic data on the oxygen content
  • the importance ranking of various feature data to the oxygen content is determined.
  • the method also includes:
  • the various characteristic data and the corresponding oxygen content are input into the oxygen content prediction model; wherein the oxygen content prediction model passes the characteristic data samples before the head stage of equal-diameter growth, and the corresponding marked head stages of equal-diameter growth.
  • the sample oxygen content training in the external phase is obtained;
  • For each type of feature data perform a weighted average of the first feature weight and the second feature weight, and use the average as the total feature weight corresponding to the feature data;
  • the importance ranking of various feature data to the oxygen content is determined.
  • the method before sorting the importance of the oxygen content according to the various characteristic data and adjusting the process parameters corresponding to the various characteristic data, the method further includes:
  • the characteristic data whose oxygen content is between a preset high oxygen content threshold and a preset low oxygen content threshold is divided into a medium oxygen content data set;
  • the data distribution diagram of the high oxygen content data set, the medium oxygen content data set, the low oxygen content data set and all the characteristic data are respectively displayed; wherein, the data distribution diagram includes the oxygen content corresponding to the characteristic data.
  • Content difference information For each type of characteristic data, the data distribution diagram of the high oxygen content data set, the medium oxygen content data set, the low oxygen content data set and all the characteristic data are respectively displayed; wherein, the data distribution diagram includes the oxygen content corresponding to the characteristic data.
  • Sorting the importance of oxygen content according to the various characteristic data, and adjusting the process parameters corresponding to the various characteristic data includes:
  • the process parameters corresponding to the various characteristic data are adjusted.
  • the method before determining, for each type of characteristic data, the oxygen content difference information between the characteristic data whose oxygen content is higher than the preset high oxygen content threshold and the characteristic data whose oxygen content is lower than the preset low oxygen content threshold. , the method also includes:
  • the preset high oxygen content threshold and the preset low oxygen content threshold are determined based on the density map of the characteristic data with respect to the oxygen content, so that the oxygen content is between the preset high oxygen content threshold and the preset low oxygen content threshold.
  • the quantity proportion of the characteristic data between the preset low oxygen content thresholds is higher than the preset proportion, and the preset high oxygen content threshold is lower than the preset maximum oxygen content.
  • embodiments of the present invention may be provided as methods, devices, or computer program products.
  • embodiments of the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment that combines software and hardware aspects.
  • embodiments of the invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) having computer-usable program code embodied therein.
  • Embodiments of the invention are described with reference to flowcharts and/or block diagrams of methods, terminal devices (systems), and computer program products according to embodiments of the invention. It will be understood that each process and/or block in the flowchart illustrations and/or block diagrams, and combinations of processes and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing terminal device to produce a machine such that the instructions are executed by the processor of the computer or other programmable data processing terminal device. Means are generated for implementing the functions specified in the process or processes of the flowchart diagrams and/or the block or blocks of the block diagrams.
  • These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing terminal equipment to operate in a predictive manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including the instruction means, the The instruction means implements the functions specified in a process or processes of the flowchart and/or a block or blocks of the block diagram.
  • These computer program instructions can also be loaded onto a computer or other programmable data processing terminal equipment, so that a series of operating steps are performed on the computer or other programmable terminal equipment to produce computer-implemented processing, thereby causing the computer or other programmable terminal equipment to perform a computer-implemented process.
  • the instructions executed on provide steps for implementing the functions specified in a process or processes of the flow diagrams and/or a block or blocks of the block diagrams.
  • the device embodiments described above are only illustrative.
  • the units described as separate components may or may not be physically separated.
  • the components shown as units may or may not be physical units, that is, they may be located in One location, or it can be distributed across multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. Persons of ordinary skill in the art can understand and implement the method without any creative effort.

Landscapes

  • Chemical & Material Sciences (AREA)
  • Engineering & Computer Science (AREA)
  • Crystallography & Structural Chemistry (AREA)
  • Materials Engineering (AREA)
  • Metallurgy (AREA)
  • Organic Chemistry (AREA)
  • Crystals, And After-Treatments Of Crystals (AREA)

Abstract

提供一种氧含量控制方法、装置、设备及介质。该方法包括：获取多次直拉单晶过程中的多种特征数据，和对应的等径生长的头部阶段的氧含量，针对每种特征数据，确定氧含量高于预设高氧含量阈值的特征数据和氧含量低于预设低氧含量阈值的特征数据之间的氧含量差值信息，根据所述特征数据对应的氧含量差值信息，确定各种特征数据对氧含量的重要性排序，根据所述各种特征数据对氧含量的重要性排序，对所述各个特征数据对应的工艺参数进行调整，使得利用长期生产运行中积累的数据分析出影响氧含量的特征数据的排序，从而据此能够自动对氧含量进行更加精确的控制，从而提高产品的品质。

Description

氧含量控制方法、装置、电子设备及存储介质
本申请要求在2022年5月20日提交中国专利局、申请号为202210550299.2、名称为“氧含量控制方法、装置、电子设备及存储介质”的专利申请的优先权，其全部内容通过引用结合在本申请中。
技术领域
本发明涉及晶体制备技术领域，特别是涉及一种氧含量控制方法、一种氧含量控制装置、一种电子设备以及一种存储介质。
背景技术
单晶硅材料的制备工艺以直拉法(Czochralski process/CZ)为主，利用直拉法将多晶硅原料提炼成单晶硅。在直拉单晶过程中生成棒状单晶硅晶体的过程分为装料、熔料、加料、调温、引晶、放肩、转肩、等径、收尾等步骤。
在单晶硅生长过程中硅与盛硅的石英坩埚发生反应，即：
SiO2+Si→2SiO↑，
反应结果造成熔硅中氧含量增加，这是硅中氧的主要来源。
氧掺入晶体的浓度，取决于三个扩散边界层和三个界面面积。边界层厚度取决于熔体热对流，而界面面积取决于装料量和坩埚尺寸与形状以及晶体直径等。其中，坩埚和熔体界面面积与熔体自由表面面积之比决定进入晶体中氧含量的重要因素。除此之外，炉室内的气氛对生长晶体中的氧含量也有着重要影响。
在单晶生长过程中，随着晶体生长，重量不断增加，而坩埚内的熔体重量随之减少，熔体液面下降。这意味着坩埚与熔体间界面面积逐渐减小，而熔体自由表面仍然保持恒定，即氧从熔体通过自由表面的挥发量保持不变。故氧在晶体中的分布是不均匀的，一般为晶体头部含量高，尾部含量低。晶体中心部位含量高，边缘部位含量低。而吸除和器件工艺要求材料厂家提供的单晶氧含量分布均匀，特别是径向分布均匀性要好。
石英坩埚的溶解速度主要与温度，炉室内压力，石英坩埚表面状态，坩埚/熔体界面上的边界层厚度等有因素关。温度越高，压力越低、表面粗糙度越大，边界层厚度越薄，石英坩埚的溶解速度越快。
在1420摄氏度时，SiO的蒸汽压可达12mbar。在目前普遍采用减压拉晶工艺条件下，SiO挥发不会受到阻碍，因此它是很容易从熔体自由表面挥发的。
氧是直拉单晶硅中含量最高，行为最复杂的一种杂质。近40多年来，对硅中氧含量控制和行为，一直是硅材料领域中的重要研究课题之一。晶体中的氧既有益又有害。
如前所述，晶体中的氧既有益又有害，因此，在拉制硅单晶过程中需要对氧含量进行控制。目前控制氧含量的方法的实验过程复杂，周期长，而且效果不够好。
发明内容
鉴于上述问题，提出了本发明实施例以便提供一种克服上述问题或者至少部分地解决上述问题的一种参数调整方法，以解决控制氧含量的方法的实验过程复杂，周期长，而且效果不够好的问题。
相应的，本发明实施例还提供了一种氧含量预测装置、一种电子设备以及一种存储介质，用以保证上述方法的实现及应用。
为了解决上述问题，本发明实施例公开了一种氧含量预测方法，包括：
获取多次直拉单晶过程中的多种特征数据，和对应的等径生长的头部阶段的氧含量；
针对每种特征数据，确定氧含量高于预设高氧含量阈值的特征数据和氧含量低于预设低氧含量阈值的特征数据之间的氧含量差值信息；
根据所述特征数据对应的氧含量差值信息，确定各种特征数据对氧含量的重要性排序；
根据所述各种特征数据对氧含量的重要性排序，对所述各个特征数据对应的工艺参数进行调整，以控制氧含量。
可选地，所述针对每种特征数据，确定氧含量高于预设高氧含量阈值的特征数据和氧含量低于预设低氧含量阈值的特征数据之间的氧含量差值信息包括：
针对每种特征数据，将所述氧含量高于预设高氧含量阈值的特征数据划分到高氧含量数据集，将所述氧含量低于预设低氧含量阈值的特征数据划分到低氧含量数据集；
针对每种特征数据，根据所述高氧含量数据集和所述低氧含量数据集，确定两个数据集之间的氧含量差值信息。
可选地，所述针对每种特征数据，根据所述高氧含量数据集和所述低氧含量数据集，确定两个数据集之间的氧含量差值信息包括：
针对每种特征数据，计算所述高氧含量数据集的氧含量的中位数和所述低氧含量数据集的氧含量的中位数之间的差值，作为所述两个数据集之间的氧含量差值信息。
可选地，所述根据所述特征数据对应的氧含量差值信息，确定各种特征数据对氧含量的重要性排序包括：
根据所述特征数据对应的氧含量差值信息，计算各种特征数据对氧含量的第一特征权重；
根据所述第一特征权重，确定各种特征数据对氧含量的重要性排序。
可选地，所述方法还包括：
将所述多种特征数据和对应的氧含量输入氧含量预测模型；其中，所述氧含量预测模型通过在等径生长的头部阶段之前的特征数据样本，以及对应标记的等径生长的头部阶段的样本氧含量训练得到；
根据所述多种特征数据和对应的氧含量，训练所述氧含量预测模型；
输出训练好的所述氧含量预测模型中各种特征数据对应的第二特征权重；
所述根据所述第一特征权重，确定各种特征数据对氧含量的重要性排序包括：
针对每种特征数据，对所述第一特征权重和第二特征权重进行加权求均值，将所述均值作为特征数据对应的总特征权重；
根据所述总特征权重，确定各种特征数据对氧含量的重要性排序。
可选地，在所述根据所述各种特征数据对氧含量的重要性排序，对所述各个特征数据对应的工艺参数进行调整之前，所述方法还包括：
针对每种特征数据，将所述氧含量介于预设高氧含量阈值和预设低氧含量阈值的特征数据划分到中氧含量数据集；
针对每种特征数据，分别显示所述高氧含量数据集、中氧含量数据集、低氧含量数据集以及全部特征数据的数据分布图；其中，所述数据分布图中 包括特征数据对应的氧含量差值信息；
所述根据所述各种特征数据对氧含量的重要性排序，对所述各个特征数据对应的工艺参数进行调整包括：
根据所述各种特征数据对氧含量的重要性排序和数据分布图，对所述各个特征数据对应的工艺参数进行调整。
可选地，在所述针对每种特征数据，确定氧含量高于预设高氧含量阈值的特征数据和氧含量低于预设低氧含量阈值的特征数据之间的氧含量差值信息之前，所述方法还包括：
针对每种特征数据，基于所述特征数据对于氧含量的密度图，确定所述预设高氧含量阈值和预设低氧含量阈值，以使氧含量介于所述预设高氧含量阈值和预设低氧含量阈值之间的特征数据的数量占比高于预设占比，且所述预设高氧含量阈值低于预设最高氧含量。
本发明实施例还公开了一种氧含量预测装置，包括：
数据获取模块，用于获取多次直拉单晶过程中的多种特征数据，和对应的等径生长的头部阶段的氧含量；
差值确定模块，用于针对每种特征数据，确定氧含量高于预设高氧含量阈值的特征数据和氧含量低于预设低氧含量阈值的特征数据之间的氧含量差值信息；
排序确定模块，用于根据所述特征数据对应的氧含量差值信息，确定各种特征数据对氧含量的重要性排序；
氧含量控制模块，用于根据所述各种特征数据对氧含量的重要性排序，对所述各个特征数据对应的工艺参数进行调整，以控制氧含量。
可选地，所述差值确定模块包括：
数据集划分子模块，用于针对每种特征数据，将所述氧含量高于预设高氧含量阈值的特征数据划分到高氧含量数据集，将所述氧含量低于预设低氧含量阈值的特征数据划分到低氧含量数据集；
差值确定子模块，用于针对每种特征数据，根据所述高氧含量数据集和所述低氧含量数据集，确定两个数据集之间的氧含量差值信息。
可选地，所述差值确定子模块包括：
差值计算单元，用于针对每种特征数据，计算所述高氧含量数据集的氧 含量的中位数和所述低氧含量数据集的氧含量的中位数之间的差值，作为所述两个数据集之间的氧含量差值信息。
可选地，所述排序确定模块包括：
第一确定计算子模块，用于根据所述特征数据对应的氧含量差值信息，计算各种特征数据对氧含量的第一特征权重；
排序确定子模块，用于根据所述第一特征权重，确定各种特征数据对氧含量的重要性排序。
可选地，所述装置还包括：
数据输入模块，用于将所述多种特征数据和对应的氧含量输入氧含量预测模型；其中，所述氧含量预测模型通过在等径生长的头部阶段之前的特征数据样本，以及对应标记的等径生长的头部阶段的样本氧含量训练得到；
模型训练模块，用于根据所述多种特征数据和对应的氧含量，训练所述氧含量预测模型；
权重输出模块，用于输出训练好的所述氧含量预测模型中各种特征数据对应的第二特征权重；
所述排序确定子模块包括：
总权重确定单元，用于针对每种特征数据，对所述第一特征权重和第二特征权重进行加权求均值，将所述均值作为特征数据对应的总特征权重；
排序确定单元，用于根据所述总特征权重，确定各种特征数据对氧含量的重要性排序。
可选地，所述装置还包括：
数据集划分模块，用于在所述根据所述各种特征数据对氧含量的重要性排序，对所述各个特征数据对应的工艺参数进行调整之前，针对每种特征数据，将所述氧含量介于预设高氧含量阈值和预设低氧含量阈值的特征数据划分到中氧含量数据集；
分布图显示模块，用于针对每种特征数据，分别显示所述高氧含量数据集、中氧含量数据集、低氧含量数据集以及全部特征数据的数据分布图；其中，所述数据分布图中包括特征数据对应的氧含量差值信息；
所述氧含量控制模块包括：
参数调整子模块，用于根据所述各种特征数据对氧含量的重要性排序和 数据分布图，对所述各个特征数据对应的工艺参数进行调整。
可选地，所述装置还包括：
阈值确定模块，用于在所述针对每种特征数据，确定氧含量高于预设高氧含量阈值的特征数据和氧含量低于预设低氧含量阈值的特征数据之间的氧含量差值信息之前，针对每种特征数据，基于所述特征数据对于氧含量的密度图，确定所述预设高氧含量阈值和预设低氧含量阈值，以使氧含量介于所述预设高氧含量阈值和预设低氧含量阈值之间的特征数据的数量占比高于预设占比，且所述预设高氧含量阈值低于预设最高氧含量。
本发明实施例还公开了一种电子设备，其特征在于，包括处理器、通信接口、存储器和通信总线，其中，处理器，通信接口，存储器通过通信总线完成相互间的通信；
存储器，用于存放计算机程序；
处理器，用于执行存储器上所存放的程序时，实现如上所述的方法步骤。
本发明实施例还公开了一种可读存储介质，当所述存储介质中的指令由电子设备的处理器执行时，使得电子设备能够执行本发明实施例中一个或多个所述的氧含量预测方法。
本发明实施例包括以下优点：
依据本发明实施例，通过获取多次直拉单晶过程中的多种特征数据，和对应的等径生长的头部阶段的氧含量，针对每种特征数据，确定氧含量高于预设高氧含量阈值的特征数据和氧含量低于预设低氧含量阈值的特征数据之间的氧含量差值信息，根据所述特征数据对应的氧含量差值信息，确定各种特征数据对氧含量的重要性排序，根据所述各种特征数据对氧含量的重要性排序，对所述各个特征数据对应的工艺参数进行调整，使得利用长期生产运行中积累的数据分析出影响氧含量的特征数据的排序，从而据此能够自动对氧含量进行更加精确的控制，从而提高产品的品质。
上述说明仅是本申请技术方案的概述，为了能够更清楚了解本申请的技术手段，而可依照说明书的内容予以实施，并且为了让本申请的上述和其它目的、特征和优点能够更明显易懂，以下特举本申请的具体实施方式。
附图说明
为了更清楚地说明本申请实施例或现有技术中的技术方案，下面将对实 施例或现有技术描述中所需要使用的附图作一简单地介绍，显而易见地，下面描述中的附图是本申请的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动的前提下，还可以根据这些附图获得其他的附图。
图1是本发明的一种氧含量控制方法实施例的步骤流程图；
图2是本发明的另一种氧含量控制方法实施例的步骤流程图；
图3是氧含量预测的一种示例的示意图；
图4是本发明的一种氧含量控制装置实施例的结构框图；
图5是根据一示例性实施例示出的一种用于氧含量预测的计算设备的结构框图。
具体实施例
下面详细描述本申请的实施例，所述实施例的示例在附图中示出，其中自始至终相同或类似的标号表示相同或类似的元件或具有相同或类似功能的元件。下面通过参考附图描述的实施例是示例性的，旨在用于解释本申请，而不能理解为对本申请的限制。
为使本发明的上述目的、特征和优点能够更加明显易懂，下面结合附图和具体实施方式对本发明作进一步详细的说明。
参照图1，示出了本发明的一种氧含量控制方法实施例的步骤流程图，具体可以包括如下步骤：
步骤101，获取多次直拉单晶过程中的多种特征数据，和对应的等径生长的头部阶段的氧含量。
在本发明实施例中，直拉单晶过程是利用直拉法将原料提炼成单晶的过程，例如，直拉单晶硅的过程。直拉单晶过程可以划分为熔料阶段、加料阶段、调温阶段、引晶阶段、放肩阶段、转肩阶段、等径生长阶段等。其中，将等径生长阶段刚开始的部分，称为等径生长的头部阶段(例如，从等径开始到晶体长度增加50毫米的阶段)。
在本发明实施例中，在直拉单晶过程中，获取多种特征数据。例如，获取工业传感器及相关控制设备采集的特征数据，其中核心的特征数据包括：maxsetmainheaterpowertime(熔料时主加功率最大值)、bodysumtime(等径时长)、body0remainweight(等径零时刻剩料量)、bodyavgdiameter(等径平均拉速)、crystallength(晶体长度)、bottomheater(底加热器功率)、 bodystdseedrotation(等径晶体转速标准差)、bodyavgseedrotation(等径平均晶体转速值)、bodyendargonflow(等径结束时氩气流量)、temp_t(调温时长)、neck_t(放肩时长)、head_o(头氧含量值)等。具体可以包括任意适用的特征数据，本发明实施例对此不做限制。
在本发明实施例中，每次直拉单晶过程都会产出单晶，通过检测氧含量的设备对等径生长的头部阶段的氧含量进行检测，以便获取到氧含量。例如，对单晶硅棒进行切片，采用光谱成像原理的检测设备可以准确检测出氧含量。
在本发明实施例中，对于获取的多种特征数据，还可以删除文本编码特征、冗余特征、使用箱图法对异常数据进行剔除操作，使用相关性分析进行初步特征筛选等，最后采用剩余的特征数据。
步骤102，针对每种特征数据，确定氧含量高于预设高氧含量阈值的特征数据和氧含量低于预设低氧含量阈值的特征数据之间的氧含量差值信息。
在本发明实施例中，预设高氧含量阈值和预设低氧含量阈值是根据业务逻辑设置的，具体可以根据实际情况设置，本发明实施例对此不做限制。高于预设高氧含量阈值，则表明氧含量偏高，低于预设低氧含量阈值，则表明氧含量偏低。
在本发明实施例中，氧含量差值信息包括氧含量的中位数的差，氧含量的均值的差，氧含量的方差的差等中一种或多种，或者其他任意适用的表征氧含量的差距的信息，本发明实施例对此不做限制。
在本发明实施例中，针对每种特征数据，氧含量高于预设高氧含量阈值的特征数据有多个，氧含量低于预设低氧含量阈值的特征数据也有多个，因此，确定氧含量高于预设高氧含量阈值的特征数据和氧含量低于预设低氧含量阈值的特征数据之间的氧含量差值信息的方式可以包括多种。例如，取氧含量高于预设高氧含量阈值的特征数据对应的氧含量的中位数，和氧含量低于预设低氧含量阈值的特征数据对应的氧含量的中位数，计算两个中位数之间的差值。又例如，计算氧含量高于预设高氧含量阈值的特征数据对应的氧含量的均值，和氧含量低于预设低氧含量阈值的特征数据对应的氧含量的均值，计算两个均值之间的差值。
步骤103，根据所述特征数据对应的氧含量差值信息，确定各种特征数据 对氧含量的重要性排序。
在本发明实施例中，某种特征数据对应的氧含量差值信息的数值越大，表明该种特征数据对氧含量的影响越大，某种特征数据对应的氧含量差值信息的数值越小，表明该种特征数据对氧含量的影响越小。因此，根据特征数据对应的氧含量差值信息，可以确定各种特征数据对氧含量的重要性排序，从而得到对氧含量的影响来说，特征数据的重要性从高到低排序。
步骤104，根据所述各种特征数据对氧含量的重要性排序，对所述各个特征数据对应的工艺参数进行调整，以控制氧含量。
在本发明实施例中，为了控制氧含量，根据各种特征数据对氧含量的重要性排序，对所述各个特征数据对应的工艺参数进行调整。由于改变相关工艺参数对氧含量进行控制时，氧含量的变化有延后性，所以需要在等径生长的头部阶段之前，对氧含量进行控制。其中，工艺参数与特征数据相对应，是给设备设定的与工艺有关的参数，包括body0remainweight(等径零时刻剩料量)、bodyendargonflow(等径结束时氩气流量)、bodyavgseedrotation(等径平均晶体转速值)等，或者其他任意适用的工艺参数，本发明实施例对此不做限制。
在本发明实施例中，对工艺参数的调整时，除了综合考虑其他因素外，优先对重要性排序靠前的一种或多种特征数据对应的工艺参数进行调整，以便在对氧含量控制的同时，调整更少的工艺参数，以及更小的调整量。
依据本发明实施例，通过获取多次直拉单晶过程中的多种特征数据，和对应的等径生长的头部阶段的氧含量，针对每种特征数据，确定氧含量高于预设高氧含量阈值的特征数据和氧含量低于预设低氧含量阈值的特征数据之间的氧含量差值信息，根据所述特征数据对应的氧含量差值信息，确定各种特征数据对氧含量的重要性排序，根据所述各种特征数据对氧含量的重要性排序，对所述各个特征数据对应的工艺参数进行调整，使得利用长期生产运行中积累的数据分析出影响氧含量的特征数据的排序，从而据此能够自动对氧含量进行更加精确的控制，从而提高产品的品质。
在本发明的一种可选实施例中，在针对每种特征数据，确定氧含量高于预设高氧含量阈值的特征数据和氧含量低于预设低氧含量阈值的特征数据之 间的氧含量差值信息之前，还包括：针对每种特征数据，基于所述特征数据对于氧含量的密度图，确定所述预设高氧含量阈值和预设低氧含量阈值，以使氧含量介于所述预设高氧含量阈值和预设低氧含量阈值之间的特征数据的数量占比高于预设占比，且所述预设高氧含量阈值低于预设最高氧含量。
通常氧含量控制得越低越好，但由于设备的限制，氧含量控制得太低，所需要的付出的成本也将超出可以接受的范围。因此，针对每种特征数据，生成特征数据对于氧含量的密度图。根据密度图，确定出预设高氧含量阈值和预设低氧含量阈值。确定的预设高氧含量阈值和预设低氧含量阈值同时满足两个条件，即氧含量介于预设高氧含量阈值和预设低氧含量阈值之间的特征数据的数量占比高于预设占比，以及预设高氧含量阈值低于预设最高氧含量。其中，预设占比可以根据实际需要进行设置，本发明实施例对此不做限制。一般来说，预设占比越高，所需要的付出的成本也越低。预设最高氧含量可以根据实际需要进行设置，本发明实施例对此不做限制。例如，预设高氧含量阈值设置为8.5，预设低氧含量阈值设置为6.5。
参照图2，示出了本发明的另一种氧含量控制方法实施例的步骤流程图，具体可以包括如下步骤：
步骤201，获取多次直拉单晶过程中的多种特征数据，和对应的等径生长的头部阶段的氧含量。
步骤202，针对每种特征数据，将所述氧含量高于预设高氧含量阈值的特征数据划分到高氧含量数据集，将所述氧含量低于预设低氧含量阈值的特征数据划分到低氧含量数据集。
在本发明实施例中，针对每种特征数据，按照氧含量，将所有的特征数据划分为多个数据集。其中，将氧含量高于预设高氧含量阈值的特征数据划分到高氧含量数据集，将氧含量低于预设低氧含量阈值的特征数据划分到低氧含量数据集。
步骤203，针对每种特征数据，根据所述高氧含量数据集和所述低氧含量数据集，确定两个数据集之间的氧含量差值信息。
在本发明实施例中，针对每种特征数据，确定高氧含量数据集和低氧含量数据集之间的氧含量差值信息的方法包括多种。例如，计算高氧含量数据 集的氧含量的中位数和低氧含量数据集的氧含量的中位数之间的差值，作为两个数据集之间的氧含量差值信息。又例如，计算高氧含量数据集的氧含量的均值和低氧含量数据集的氧含量的均值之间的差值，作为两个数据集之间的氧含量差值信息。又例如，计算高氧含量数据集的每个氧含量和低氧含量数据集的每个氧含量之间的差值，作为两个数据集之间的氧含量差值信息。
在本发明的一种可选实施例中，针对每种特征数据，根据所述高氧含量数据集和所述低氧含量数据集，确定两个数据集之间的氧含量差值信息的一种实现方式中，可以包括：针对每种特征数据，计算所述高氧含量数据集的氧含量的中位数和所述低氧含量数据集的氧含量的中位数之间的差值，作为所述两个数据集之间的氧含量差值信息。
例如，高氧含量数据集中有100个特征数据，则按照特征数据从大到小的顺序，第50或51个特征数据对应的氧含量为高氧含量数据集的氧含量的中位数。低氧含量数据集中有51个特征数据，则按照特征数据从大到小的顺序，第26个特征数据对应的氧含量为低氧含量数据集的氧含量的中位数。计算两个中位数之间的差值，作为两个数据集之间的氧含量差值信息。
步骤204，根据所述特征数据对应的氧含量差值信息，确定各种特征数据对氧含量的重要性排序。
在本发明的一种可选实施例中，根据所述特征数据对应的氧含量差值信息，确定各种特征数据对氧含量的重要性排序的一种具体实现方式中，还可以包括：根据所述特征数据对应的氧含量差值信息，计算各种特征数据对氧含量的第一特征权重，根据所述第一特征权重，确定各种特征数据对氧含量的重要性排序。
第一特征权重表征特征数据对氧含量的重要程度。第一特征权重越大，该种特征数据对氧含量的影响也越大，第一特征权重越小，该种特征数据对氧含量的影响也越小。第一特征权重越高，特征数据对氧含量的重要性排序越靠前。
根据特征数据对应的氧含量差值信息，计算各种特征数据对氧含量的第一特征权重的方式包括多种。例如，当氧含量差值信息是中位数的差时，可以直接将中位数的差作为第一特征权重，或者也可以将中位数的差乘以预设系数后得到的值作为第一特征权重。本发明实施例对此不做限制。
在本发明的一种可选实施例中，还可以包括：将所述多种特征数据和对应的氧含量输入氧含量预测模型；其中，所述氧含量预测模型通过在等径生长的头部阶段之前的特征数据样本，以及对应标记的等径生长的头部阶段的样本氧含量训练得到；根据所述多种特征数据和对应的氧含量，训练所述氧含量预测模型；输出训练好的所述氧含量预测模型中各种特征数据对应的第二特征权重。根据所述第一特征权重，确定各种特征数据对氧含量的重要性排序的一种具体实现方式中，可以包括：针对每种特征数据，对所述第一特征权重和第二特征权重进行加权求均值，将所述均值作为特征数据对应的总特征权重；根据所述总特征权重，确定各种特征数据对氧含量的重要性排序。
预测等径生长的头部阶段的氧含量可以采用机器学习特征数据和等径生长的头部阶段的氧含量之间的相关关系，得到一个可以预测氧含量的氧含量预测模型。为了训练氧含量预测模型，需要准备样本数据以及对应的标签数据，即获取的多种特征数据，以及对应标记的等径生长的头部阶段的氧含量。
氧含量预测模型包括：朴素贝叶斯模型、或极端梯度提升模型、或类别梯度提升模型。朴素贝叶斯模型是一种构建分类器的简单方法，该分类模型会给问题实例分配用特征值表示的类标签，类标签取自有限集合。极端梯度提升(eXtreme Gradient Boosting，XGboost)模型是大规模并行利用提升树进行预测的一种模型。类别梯度提升(Categorical Gradient Boosting，CATboost)模型是一种能够很好地处理类别型特征的梯度提升算法的模型。光梯度提升机(Light Gradient Boosting Machine，lightGBM)模型是是个快速的，分布式的，高性能的基于决策树算法的梯度提升框架。可用于排序，分类，回归以及很多其他的机器学习任务中。
例如，用于训练模型的数据划分为训练集和测试集，其中，测试集一般占比20％。使用XGboost、lightGBM、CATboost等集成学习模型构建学习器，使用贝叶斯优化方法进行模型参数调优、最终选择整体最优的CATboost模型作为最优模型。模型的R2(R-Square，决定系数)指标为0.51。
氧含量预测模型的训练过程，就是根据训练数据，确定各种特征数据对应的特征权重的过程。在氧含量预测模型训练好后，特征权重确定下来。从训练好的氧含量预测模型中，获取各种特征数据对应的第二特征权重。其中，第二特征权重越大，该种第二特征数据对氧含量的影响也越大，第二特征权 重越小，该种第二特征数据对氧含量的影响也越小。
针对每种特征数据，对第一特征权重和第二特征权重进行加权求均值，将所述均值作为特征数据对应的总特征权重。两者的权重可以一样，则直接求第一特征权重和第二特征权重的均值即可。两者的权重也可以不一样，两者的权重可以根据实际效果进行调整，以便求得的总特征权重能够更加准确的反应各种特征数据对氧含量的重要性。
根据总特征权重，确定各种特征数据对氧含量的重要性排序的方式可以参见根据第一特征权重，确定各种特征数据对氧含量的重要性排序。
步骤205，根据所述各种特征数据对氧含量的重要性排序，对所述各个特征数据对应的工艺参数进行调整，以控制氧含量。
在本发明的一种可选实施例中，在根据所述各种特征数据对氧含量的重要性排序，对所述各个特征数据对应的工艺参数进行调整之前，还可以包括：针对每种特征数据，将所述氧含量介于预设高氧含量阈值和预设低氧含量阈值的特征数据划分到中氧含量数据集，针对每种特征数据，分别显示所述高氧含量数据集、中氧含量数据集、低氧含量数据集以及全部特征数据的数据分布图；其中，所述数据分布图中包括特征数据对应的氧含量差值信息。
相应的，根据所述各种特征数据对氧含量的重要性排序，对所述各个特征数据对应的工艺参数进行调整的一种具体实现方式中，可以包括：根据所述各种特征数据对氧含量的重要性排序和数据分布图，对所述各个特征数据对应的工艺参数进行调整。
针对每种特征数据，将氧含量介于预设高氧含量阈值和预设低氧含量阈值的特征数据划分到中氧含量数据集。中氧含量数据集中的特征数据是使氧含量适中的特征数据。
数据分布图是反应数据分布情况的图形。数据分布图包括箱线图，密度图，直方图等，或者其他任意适用的分布图，本发明实施例对此不做限制。针对每个特征数据，分布显示高氧含量数据集、中氧含量数据集、低氧含量数据集以及全部特征数据的数据分布图，即一共四个数据分布图。其中，数据分布图中还显示有特征数据对应的氧含量差值信息以供参考。
根据各种特征数据对氧含量的重要性排序和数据分布图，对各个特征数据对应的工艺参数进行调整。数据分布图可以显示出各种特征数据的参数区 间，以及中位数、均值、氧含量差值信息、氧含量在预设高氧含量阈值时的特征数据、氧含量在预设低氧含量阈值时的特征数据等关键数值。基于数据分布图，对各个特征数据对应的工艺参数进行调整时，能够更加方便和准确的确定工艺参数调整后的值。
例如，如图3所示的特征数据对于氧含量的箱线图的示意图。高氧含量数据集对应于标识为more的箱线图，低氧含量数据集对应于标识为less的箱线图，中氧含量数据集对应于标识为between的箱线图，所有特征数据对应于标识为all的箱线图。其中，每个箱体中的横线表示该数据集的中位数。由于between数据占绝大多数，所以all数据对应的图，跟between数据对应的图，在视觉上几乎一样。
依据本发明实施例，通过获取多次直拉单晶过程中的多种特征数据，和对应的等径生长的头部阶段的氧含量，针对每种特征数据，将所述氧含量高于预设高氧含量阈值的特征数据划分到高氧含量数据集，将所述氧含量低于预设低氧含量阈值的特征数据划分到低氧含量数据集，针对每种特征数据，根据所述高氧含量数据集和所述低氧含量数据集，确定两个数据集之间的氧含量差值信息，根据所述特征数据对应的氧含量差值信息，确定各种特征数据对氧含量的重要性排序，根据所述各种特征数据对氧含量的重要性排序，对所述各个特征数据对应的工艺参数进行调整，使得利用长期生产运行中积累的数据分析出影响氧含量的特征数据的排序，从而据此能够自动对氧含量进行更加精确的控制，从而提高产品的品质。
需要说明的是，对于方法实施例，为了简单描述，故将其都表述为一系列的动作组合，但是本领域技术人员应该知悉，本发明实施例并不受所描述的动作顺序的限制，因为依据本发明实施例，某些步骤可以采用其他顺序或者同时进行。其次，本领域技术人员也应该知悉，说明书中所描述的实施例均属于优选实施例，所涉及的动作并不一定是本发明实施例所必须的。
参照图4，示出了本发明的一种氧含量控制装置实施例的结构框图，具体可以包括如下模块：
数据获取模块301，用于获取多次直拉单晶过程中的多种特征数据，和对应的等径生长的头部阶段的氧含量；
差值确定模块302，用于针对每种特征数据，确定氧含量高于预设高氧含量阈值的特征数据和氧含量低于预设低氧含量阈值的特征数据之间的氧含量差值信息；
排序确定模块303，用于根据所述特征数据对应的氧含量差值信息，确定各种特征数据对氧含量的重要性排序；
氧含量控制模块304，用于根据所述各种特征数据对氧含量的重要性排序，对所述各个特征数据对应的工艺参数进行调整，以控制氧含量。
可选地，所述差值确定模块包括：
数据集划分子模块，用于针对每种特征数据，将所述氧含量高于预设高氧含量阈值的特征数据划分到高氧含量数据集，将所述氧含量低于预设低氧含量阈值的特征数据划分到低氧含量数据集；
差值确定子模块，用于针对每种特征数据，根据所述高氧含量数据集和所述低氧含量数据集，确定两个数据集之间的氧含量差值信息。
可选地，所述差值确定子模块包括：
差值计算单元，用于针对每种特征数据，计算所述高氧含量数据集的氧含量的中位数和所述低氧含量数据集的氧含量的中位数之间的差值，作为所述两个数据集之间的氧含量差值信息。
可选地，所述排序确定模块包括：
第一确定计算子模块，用于根据所述特征数据对应的氧含量差值信息，计算各种特征数据对氧含量的第一特征权重；
排序确定子模块，用于根据所述第一特征权重，确定各种特征数据对氧含量的重要性排序。
可选地，所述装置还包括：
数据输入模块，用于将所述多种特征数据和对应的氧含量输入氧含量预测模型；其中，所述氧含量预测模型通过在等径生长的头部阶段之前的特征数据样本，以及对应标记的等径生长的头部阶段的样本氧含量训练得到；
模型训练模块，用于根据所述多种特征数据和对应的氧含量，训练所述氧含量预测模型；
权重输出模块，用于输出训练好的所述氧含量预测模型中各种特征数据对应的第二特征权重；
所述排序确定子模块包括：
总权重确定单元，用于针对每种特征数据，对所述第一特征权重和第二特征权重进行加权求均值，将所述均值作为特征数据对应的总特征权重；
排序确定单元，用于根据所述总特征权重，确定各种特征数据对氧含量的重要性排序。
可选地，所述装置还包括：
数据集划分模块，用于在所述根据所述各种特征数据对氧含量的重要性排序，对所述各个特征数据对应的工艺参数进行调整之前，针对每种特征数据，将所述氧含量介于预设高氧含量阈值和预设低氧含量阈值的特征数据划分到中氧含量数据集；
分布图显示模块，用于针对每种特征数据，分别显示所述高氧含量数据集、中氧含量数据集、低氧含量数据集以及全部特征数据的数据分布图；其中，所述数据分布图中包括特征数据对应的氧含量差值信息；
所述氧含量控制模块包括：
参数调整子模块，用于根据所述各种特征数据对氧含量的重要性排序和数据分布图，对所述各个特征数据对应的工艺参数进行调整。
可选地，所述装置还包括：
阈值确定模块，用于在所述针对每种特征数据，确定氧含量高于预设高氧含量阈值的特征数据和氧含量低于预设低氧含量阈值的特征数据之间的氧含量差值信息之前，针对每种特征数据，基于所述特征数据对于氧含量的密度图，确定所述预设高氧含量阈值和预设低氧含量阈值，以使氧含量介于所述预设高氧含量阈值和预设低氧含量阈值之间的特征数据的数量占比高于预设占比，且所述预设高氧含量阈值低于预设最高氧含量。
依据本发明实施例，通过获取多次直拉单晶过程中的多种特征数据，和对应的等径生长的头部阶段的氧含量，针对每种特征数据，确定氧含量高于预设高氧含量阈值的特征数据和氧含量低于预设低氧含量阈值的特征数据之间的氧含量差值信息，根据所述特征数据对应的氧含量差值信息，确定各种特征数据对氧含量的重要性排序，根据所述各种特征数据对氧含量的重要性排序，对所述各个特征数据对应的工艺参数进行调整，使得利用长期生产运行中积累的数据分析出影响氧含量的特征数据的排序，从而据此能够自动对 氧含量进行更加精确的控制，从而提高产品的品质。
对于装置实施例而言，由于其与方法实施例基本相似，所以描述的比较简单，相关之处参见方法实施例的部分说明即可。
图5是根据一示例性实施例示出的一种用于氧含量预测的电子设备400的结构框图。例如，电子设备400可以是移动电话，计算机，数字广播终端，消息收发设备，游戏控制台，平板设备，医疗设备，健身设备，个人数字助理等。
参照图5，电子设备400可以包括以下一个或多个组件：处理组件402，存储器404，电源组件406，多媒体组件408，音频组件410，输入/输出(I/O)的接口412，传感器组件414，以及通信组件416。
处理组件402通常控制电子设备400的整体操作，诸如与显示，电话呼叫，数据通信，相机操作和记录操作相关联的操作。处理组件402可以包括一个或多个处理器420来执行指令，以完成上述的氧含量预测方法的全部或部分步骤。此外，处理组件402可以包括一个或多个模块，便于处理组件402和其他组件之间的交互。例如，处理部件402可以包括多媒体模块，以方便多媒体组件408和处理组件402之间的交互。
存储器404被配置为存储各种类型的数据以支持在设备400的操作。这些数据的示例包括用于在电子设备400上操作的任何应用程序或方法的指令，联系人数据，电话簿数据，消息，图片，视频等。存储器404可以由任何类型的易失性或非易失性存储设备或者它们的组合实现，如静态随机存取存储器(SRAM)，电可擦除可编程只读存储器(EEPROM)，可擦除可编程只读存储器(EPROM)，可编程只读存储器(PROM)，只读存储器(ROM)，磁存储器，快闪存储器，磁盘或光盘。
电力组件404为电子设备400的各种组件提供电力。电力组件404可以包括电源管理系统，一个或多个电源，及其他与为电子设备400生成、管理和分配电力相关联的组件。
多媒体组件408包括在所述电子设备400和用户之间的提供一个输出接口的屏幕。在一些实施例中，屏幕可以包括液晶显示器(LCD)和触摸面板(TP)。如果屏幕包括触摸面板，屏幕可以被实现为触摸屏，以接收来自用 户的输入信号。触摸面板包括一个或多个触摸传感器以感测触摸、滑动和触摸面板上的手势。所述触摸传感器可以不仅感测触摸或滑动动作的边界，而且还检测与所述触摸或滑动操作相关的持续时间和压力。在一些实施例中，多媒体组件408包括一个前置摄像头和/或后置摄像头。当电子设备400处于操作模式，如拍摄模式或视频模式时，前置摄像头和/或后置摄像头可以接收外部的多媒体数据。每个前置摄像头和后置摄像头可以是一个固定的光学透镜系统或具有焦距和光学变焦能力。
音频组件410被配置为输出和/或输入音频信号。例如，音频组件410包括一个麦克风(MIC)，当电子设备400处于操作模式，如呼叫模式、记录模式和语音识别模式时，麦克风被配置为接收外部音频信号。所接收的音频信号可以被进一步存储在存储器404或经由通信组件416发送。在一些实施例中，音频组件410还包括一个扬声器，用于输出音频信号。
I/O接口412为处理组件402和外围接口模块之间提供接口，上述外围接口模块可以是键盘，点击轮，按钮等。这些按钮可包括但不限于：主页按钮、音量按钮、启动按钮和锁定按钮。
传感器组件414包括一个或多个传感器，用于为电子设备400提供各个方面的状态评估。例如，传感器组件414可以检测到设备400的打开/关闭状态，组件的相对定位，例如所述组件为电子设备400的显示器和小键盘，传感器组件414还可以检测电子设备400或电子设备400一个组件的位置改变，用户与电子设备400接触的存在或不存在，电子设备400方位或加速/减速和电子设备400的温度变化。传感器组件414可以包括接近传感器，被配置用来在没有任何的物理接触时检测附近物体的存在。传感器组件414还可以包括光传感器，如CMOS或CCD图像传感器，用于在成像应用中使用。在一些实施例中，该传感器组件414还可以包括加速度传感器，陀螺仪传感器，磁传感器，压力传感器或温度传感器。
通信组件416被配置为便于电子设备400和其他设备之间有线或无线方式的通信。电子设备400可以接入基于通信标准的无线网络，如WiFi，2G或3G，或它们的组合。在一个示例性实施例中，通信部件414经由广播信道接收来自外部广播管理系统的广播信号或广播相关信息。在一个示例性实施例中，所述通信部件414还包括近场通信(NFC)模块，以促进短程通信。例 如，在NFC模块可基于射频识别(RFID)技术，红外数据协会(IrDA)技术，超宽带(UWB)技术，蓝牙(BT)技术和其他技术来实现。
在示例性实施例中，电子设备400可以被一个或多个应用专用集成电路(ASIC)、数字信号处理器(DSP)、数字信号处理设备(DSPD)、可编程逻辑器件(PLD)、现场可编程门阵列(FPGA)、控制器、微控制器、微处理器或其他电子元件实现，用于执行上述氧含量预测方法。
在示例性实施例中，还提供了一种包括指令的非临时性计算机可读存储介质，例如包括指令的存储器404，上述指令可由电子设备400的处理器420执行以完成上述氧含量预测方法。例如，所述非临时性计算机可读存储介质可以是ROM、随机存取存储器(RAM)、CD-ROM、磁带、软盘和光数据存储设备等。
一种非临时性计算机可读存储介质，当所述存储介质中的指令由终端的处理器执行时，使得终端能够执行一种氧含量控制方法，所述方法还包括：
获取多次直拉单晶过程中的多种特征数据，和对应的等径生长的头部阶段的氧含量；
针对每种特征数据，确定氧含量高于预设高氧含量阈值的特征数据和氧含量低于预设低氧含量阈值的特征数据之间的氧含量差值信息；
根据所述特征数据对应的氧含量差值信息，确定各种特征数据对氧含量的重要性排序；
根据所述各种特征数据对氧含量的重要性排序，对所述各个特征数据对应的工艺参数进行调整，以控制氧含量。
可选地，所述针对每种特征数据，确定氧含量高于预设高氧含量阈值的特征数据和氧含量低于预设低氧含量阈值的特征数据之间的氧含量差值信息包括：
针对每种特征数据，将所述氧含量高于预设高氧含量阈值的特征数据划分到高氧含量数据集，将所述氧含量低于预设低氧含量阈值的特征数据划分到低氧含量数据集；
针对每种特征数据，根据所述高氧含量数据集和所述低氧含量数据集，确定两个数据集之间的氧含量差值信息。
可选地，所述针对每种特征数据，根据所述高氧含量数据集和所述低氧 含量数据集，确定两个数据集之间的氧含量差值信息包括：
针对每种特征数据，计算所述高氧含量数据集的氧含量的中位数和所述低氧含量数据集的氧含量的中位数之间的差值，作为所述两个数据集之间的氧含量差值信息。
可选地，所述根据所述特征数据对应的氧含量差值信息，确定各种特征数据对氧含量的重要性排序包括：
根据所述特征数据对应的氧含量差值信息，计算各种特征数据对氧含量的第一特征权重；
根据所述第一特征权重，确定各种特征数据对氧含量的重要性排序。
可选地，所述方法还包括：
将所述多种特征数据和对应的氧含量输入氧含量预测模型；其中，所述氧含量预测模型通过在等径生长的头部阶段之前的特征数据样本，以及对应标记的等径生长的头部阶段的样本氧含量训练得到；
根据所述多种特征数据和对应的氧含量，训练所述氧含量预测模型；
输出训练好的所述氧含量预测模型中各种特征数据对应的第二特征权重；
所述根据所述第一特征权重，确定各种特征数据对氧含量的重要性排序包括：
针对每种特征数据，对所述第一特征权重和第二特征权重进行加权求均值，将所述均值作为特征数据对应的总特征权重；
根据所述总特征权重，确定各种特征数据对氧含量的重要性排序。
可选地，在所述根据所述各种特征数据对氧含量的重要性排序，对所述各个特征数据对应的工艺参数进行调整之前，所述方法还包括：
针对每种特征数据，将所述氧含量介于预设高氧含量阈值和预设低氧含量阈值的特征数据划分到中氧含量数据集；
针对每种特征数据，分别显示所述高氧含量数据集、中氧含量数据集、低氧含量数据集以及全部特征数据的数据分布图；其中，所述数据分布图中包括特征数据对应的氧含量差值信息；
所述根据所述各种特征数据对氧含量的重要性排序，对所述各个特征数据对应的工艺参数进行调整包括：
根据所述各种特征数据对氧含量的重要性排序和数据分布图，对所述各个特征数据对应的工艺参数进行调整。
可选地，在所述针对每种特征数据，确定氧含量高于预设高氧含量阈值的特征数据和氧含量低于预设低氧含量阈值的特征数据之间的氧含量差值信息之前，所述方法还包括：
针对每种特征数据，基于所述特征数据对于氧含量的密度图，确定所述预设高氧含量阈值和预设低氧含量阈值，以使氧含量介于所述预设高氧含量阈值和预设低氧含量阈值之间的特征数据的数量占比高于预设占比，且所述预设高氧含量阈值低于预设最高氧含量。
本说明书中的各个实施例均采用递进的方式描述，每个实施例重点说明的都是与其他实施例的不同之处，各个实施例之间相同相似的部分互相参见即可。
本领域内的技术人员应明白，本发明实施例的实施例可提供为方法、装置、或计算机程序产品。因此，本发明实施例可采用完全硬件实施例、完全软件实施例、或结合软件和硬件方面的实施例的形式。而且，本发明实施例可采用在一个或多个其中包含有计算机可用程序代码的计算机可用存储介质(包括但不限于磁盘存储器、CD-ROM、光学存储器等)上实施的计算机程序产品的形式。
本发明实施例是参照根据本发明实施例的方法、终端设备(系统)、和计算机程序产品的流程图和/或方框图来描述的。应理解可由计算机程序指令实现流程图和/或方框图中的每一流程和/或方框、以及流程图和/或方框图中的流程和/或方框的结合。可提供这些计算机程序指令到通用计算机、专用计算机、嵌入式处理机或其他可编程数据处理终端设备的处理器以产生一个机器，使得通过计算机或其他可编程数据处理终端设备的处理器执行的指令产生用于实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能的装置。
这些计算机程序指令也可存储在能引导计算机或其他可编程数据处理终端设备以预测方式工作的计算机可读存储器中，使得存储在该计算机可读存储器中的指令产生包括指令装置的制造品，该指令装置实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能。
这些计算机程序指令也可装载到计算机或其他可编程数据处理终端设备上，使得在计算机或其他可编程终端设备上执行一系列操作步骤以产生计算机实现的处理，从而在计算机或其他可编程终端设备上执行的指令提供用于实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能的步骤。
尽管已描述了本发明实施例的优选实施例，但本领域内的技术人员一旦得知了基本创造性概念，则可对这些实施例做出另外的变更和修改。所以，所附权利要求意欲解释为包括优选实施例以及落入本发明实施例范围的所有变更和修改。
最后，还需要说明的是，在本文中，诸如第一和第二等之类的关系术语仅仅用来将一个实体或者操作与另一个实体或操作区分开来，而不一定要求或者暗示这些实体或操作之间存在任何这种实际的关系或者顺序。而且，术语“包括”、“包含”或者其任何其他变体意在涵盖非排他性的包含，从而使得包括一系列要素的过程、方法、物品或者终端设备不仅包括那些要素，而且还包括没有明确列出的其他要素，或者是还包括为这种过程、方法、物品或者终端设备所固有的要素。在没有更多限制的情况下，由语句“包括一个……”限定的要素，并不排除在包括所述要素的过程、方法、物品或者终端设备中还存在另外的相同要素。
以上对本发明所提供的一种氧含量控制方法和装置、一种电子设备以及一种储存介质，进行了详细介绍，本文中应用了具体个例对本发明的原理及实施方式进行了阐述，以上实施例的说明只是用于帮助理解本发明的方法及其核心思想；同时，对于本领域的一般技术人员，依据本发明的思想，在具体实施方式及应用范围上均会有改变之处，综上所述，本说明书内容不应理解为对本发明的限制。
以上所描述的装置实施例仅仅是示意性的，其中所述作为分离部件说明的单元可以是或者也可以不是物理上分开的，作为单元显示的部件可以是或者也可以不是物理单元，即可以位于一个地方，或者也可以分布到多个网络单元上。可以根据实际的需要选择其中的部分或者全部模块来实现本实施例方案的目的。本领域普通技术人员在不付出创造性的劳动的情况下，即可以理解并实施。
本文中所称的“一个实施例”、“实施例”或者“一个或者多个实施例”意味着，结合实施例描述的特定特征、结构或者特性包括在本申请的至少一个实施例中。此外，请注意，这里“在一个实施例中”的词语例子不一定全指同一个实施例。
在此处所提供的说明书中，说明了大量具体细节。然而，能够理解，本申请的实施例可以在没有这些具体细节的情况下被实践。在一些实例中，并未详细示出公知的方法、结构和技术，以便不模糊对本说明书的理解。
最后应说明的是：以上实施例仅用以说明本申请的技术方案，而非对其限制；尽管参照前述实施例对本申请进行了详细的说明，本领域的普通技术人员应当理解：其依然可以对前述各实施例所记载的技术方案进行修改，或者对其中部分技术特征进行等同替换；而这些修改或者替换，并不使相应技术方案的本质脱离本申请各实施例技术方案的精神和范围。

Claims (10)

  1. 一种氧含量控制方法，其特征在于，包括：
    获取多次直拉单晶过程中的多种特征数据，和对应的等径生长的头部阶段的氧含量；
    针对每种特征数据，确定氧含量高于预设高氧含量阈值的特征数据和氧含量低于预设低氧含量阈值的特征数据之间的氧含量差值信息；
    根据所述特征数据对应的氧含量差值信息，确定各种特征数据对氧含量的重要性排序；
    根据所述各种特征数据对氧含量的重要性排序，对所述各个特征数据对应的工艺参数进行调整，以控制氧含量。
  2. 根据权利要求1所述的方法，其特征在于，所述针对每种特征数据，确定氧含量高于预设高氧含量阈值的特征数据和氧含量低于预设低氧含量阈值的特征数据之间的氧含量差值信息包括：
    针对每种特征数据，将所述氧含量高于预设高氧含量阈值的特征数据划分到高氧含量数据集，将所述氧含量低于预设低氧含量阈值的特征数据划分到低氧含量数据集；
    针对每种特征数据，根据所述高氧含量数据集和所述低氧含量数据集，确定两个数据集之间的氧含量差值信息。
  3. 根据权利要求2所述的方法，其特征在于，所述针对每种特征数据，根据所述高氧含量数据集和所述低氧含量数据集，确定两个数据集之间的氧含量差值信息包括：
    针对每种特征数据，计算所述高氧含量数据集的氧含量的中位数和所述低氧含量数据集的氧含量的中位数之间的差值，作为所述两个数据集之间的氧含量差值信息。
  4. 根据权利要求1所述的方法，其特征在于，所述根据所述特征数据对应的氧含量差值信息，确定各种特征数据对氧含量的重要性排序包括：
    根据所述特征数据对应的氧含量差值信息，计算各种特征数据对氧含量的第一特征权重；
    根据所述第一特征权重，确定各种特征数据对氧含量的重要性排序。
  5. 根据权利要求4述的方法，其特征在于，所述方法还包括：
    将所述多种特征数据和对应的氧含量输入氧含量预测模型；其中，所述氧含量预测模型通过在等径生长的头部阶段之前的特征数据样本，以及对应标记的等径生长的头部阶段的样本氧含量训练得到；
    根据所述多种特征数据和对应的氧含量，训练所述氧含量预测模型；
    输出训练好的所述氧含量预测模型中各种特征数据对应的第二特征权重；
    所述根据所述第一特征权重，确定各种特征数据对氧含量的重要性排序包括：
    针对每种特征数据，对所述第一特征权重和第二特征权重进行加权求均值，将所述均值作为特征数据对应的总特征权重；
    根据所述总特征权重，确定各种特征数据对氧含量的重要性排序。
  6. 根据权利要求2所述的方法，其特征在于，在所述根据所述各种特征数据对氧含量的重要性排序，对所述各个特征数据对应的工艺参数进行调整之前，所述方法还包括：
    针对每种特征数据，将所述氧含量介于预设高氧含量阈值和预设低氧含量阈值的特征数据划分到中氧含量数据集；
    针对每种特征数据，分别显示所述高氧含量数据集、中氧含量数据集、低氧含量数据集以及全部特征数据的数据分布图；其中，所述数据分布图中包括特征数据对应的氧含量差值信息；
    所述根据所述各种特征数据对氧含量的重要性排序，对所述各个特征数据对应的工艺参数进行调整包括：
    根据所述各种特征数据对氧含量的重要性排序和数据分布图，对所述各个特征数据对应的工艺参数进行调整。
  7. 根据权利要求1所述的方法，其特征在于，在所述针对每种特征数据，确定氧含量高于预设高氧含量阈值的特征数据和氧含量低于预设低氧含量阈值的特征数据之间的氧含量差值信息之前，所述方法还包括：
    针对每种特征数据，基于所述特征数据对于氧含量的密度图，确定所述预设高氧含量阈值和预设低氧含量阈值，以使氧含量介于所述预设高氧含量阈值和预设低氧含量阈值之间的特征数据的数量占比高于预设占比，且所述预设高氧含量阈值低于预设最高氧含量。
  8. 一种氧含量控制装置，其特征在于，包括：
    数据获取模块，用于获取多次直拉单晶过程中的多种特征数据，和对应的等径生长的头部阶段的氧含量；
    差值确定模块，用于针对每种特征数据，确定氧含量高于预设高氧含量阈值的特征数据和氧含量低于预设低氧含量阈值的特征数据之间的氧含量差值信息；
    排序确定模块，用于根据所述特征数据对应的氧含量差值信息，确定各种特征数据对氧含量的重要性排序；
    氧含量控制模块，用于根据所述各种特征数据对氧含量的重要性排序，对所述各个特征数据对应的工艺参数进行调整，以控制氧含量。
  9. 一种电子设备，其特征在于，包括处理器、通信接口、存储器和通信总线，其中，处理器，通信接口，存储器通过通信总线完成相互间的通信；
    存储器，用于存放计算机程序；
    处理器，用于执行存储器上所存放的程序时，实现权利要求1-7任一所述的方法步骤。
  10. 一种可读存储介质，其特征在于，当所述存储介质中的指令由电子设备的处理器执行时，使得电子设备能够执行如方法权利要求1-7中一个或多个所述的氧含量控制方法。
PCT/CN2023/076448 2022-05-20 2023-02-16 氧含量控制方法、装置、电子设备及存储介质 WO2023221576A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202210550299.2 2022-05-20
CN202210550299.2A CN117127252A (zh) 2022-05-20 2022-05-20 氧含量控制方法、装置、电子设备及存储介质

Publications (1)

Publication Number Publication Date
WO2023221576A1 true WO2023221576A1 (zh) 2023-11-23

Family

ID=88834524

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2023/076448 WO2023221576A1 (zh) 2022-05-20 2023-02-16 氧含量控制方法、装置、电子设备及存储介质

Country Status (2)

Country Link
CN (1) CN117127252A (zh)
WO (1) WO2023221576A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117364231A (zh) * 2023-12-08 2024-01-09 苏州晨晖智能设备有限公司 基于多参数协同控制的硅棒含氧量调控方法及系统

Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US4511428A (en) * 1982-07-09 1985-04-16 International Business Machines Corporation Method of controlling oxygen content and distribution in grown silicon crystals
JPH06172081A (ja) * 1992-12-03 1994-06-21 Komatsu Denshi Kinzoku Kk 半導体単結晶の酸素濃度制御方法
JPH0959084A (ja) * 1995-08-23 1997-03-04 Sumitomo Metal Ind Ltd 結晶中酸素濃度の制御方法
JPH11322485A (ja) * 1998-05-11 1999-11-24 Sumitomo Metal Ind Ltd 結晶中酸素濃度の制御方法
US20030051658A1 (en) * 2001-07-27 2003-03-20 Shigemasa Nakagawa Method and apparatus for controlling the oxygen concentration of a silicon single crystal, and method and apparatus for providing guidance for controlling the oxygen concentration
US20060283378A1 (en) * 2005-06-15 2006-12-21 Siltronic Ag Silicone single crystal production process
CN105019017A (zh) * 2015-06-30 2015-11-04 内蒙古中环光伏材料有限公司 一种降低直拉单晶硅中氧含量的方法
CN109154103A (zh) * 2015-12-04 2019-01-04 环球晶圆股份有限公司 用于生产低氧含量硅的系统及方法
CN112863620A (zh) * 2020-12-31 2021-05-28 杭州富加镓业科技有限公司 一种基于深度学习和提拉法的导电型氧化镓的质量预测方法、制备方法及系统
CN113417003A (zh) * 2021-06-22 2021-09-21 宁夏中欣晶圆半导体科技有限公司 能够降低头部氧含量的大直径单晶硅生产方法及装置

Patent Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US4511428A (en) * 1982-07-09 1985-04-16 International Business Machines Corporation Method of controlling oxygen content and distribution in grown silicon crystals
JPH06172081A (ja) * 1992-12-03 1994-06-21 Komatsu Denshi Kinzoku Kk 半導体単結晶の酸素濃度制御方法
JPH0959084A (ja) * 1995-08-23 1997-03-04 Sumitomo Metal Ind Ltd 結晶中酸素濃度の制御方法
JPH11322485A (ja) * 1998-05-11 1999-11-24 Sumitomo Metal Ind Ltd 結晶中酸素濃度の制御方法
US20030051658A1 (en) * 2001-07-27 2003-03-20 Shigemasa Nakagawa Method and apparatus for controlling the oxygen concentration of a silicon single crystal, and method and apparatus for providing guidance for controlling the oxygen concentration
US20060283378A1 (en) * 2005-06-15 2006-12-21 Siltronic Ag Silicone single crystal production process
CN105019017A (zh) * 2015-06-30 2015-11-04 内蒙古中环光伏材料有限公司 一种降低直拉单晶硅中氧含量的方法
CN109154103A (zh) * 2015-12-04 2019-01-04 环球晶圆股份有限公司 用于生产低氧含量硅的系统及方法
CN112863620A (zh) * 2020-12-31 2021-05-28 杭州富加镓业科技有限公司 一种基于深度学习和提拉法的导电型氧化镓的质量预测方法、制备方法及系统
CN113417003A (zh) * 2021-06-22 2021-09-21 宁夏中欣晶圆半导体科技有限公司 能够降低头部氧含量的大直径单晶硅生产方法及装置

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117364231A (zh) * 2023-12-08 2024-01-09 苏州晨晖智能设备有限公司 基于多参数协同控制的硅棒含氧量调控方法及系统
CN117364231B (zh) * 2023-12-08 2024-04-12 苏州晨晖智能设备有限公司 基于多参数协同控制的硅棒含氧量调控方法及系统

Also Published As

Publication number Publication date
CN117127252A (zh) 2023-11-28

Similar Documents

Publication Publication Date Title
WO2023221576A1 (zh) 氧含量控制方法、装置、电子设备及存储介质
US20220413798A1 (en) Playlist configuration and preview
WO2020000961A1 (zh) 图像标签识别方法、装置及服务器
US20110016150A1 (en) System and method for tagging multiple digital images
US10254928B1 (en) Contextual card generation and delivery
WO2020108023A1 (zh) 视频动作分类的方法、装置、计算机设备和存储介质
WO2017096773A1 (zh) 智能设备控制方法及装置
WO2020088069A1 (zh) 手势关键点检测方法、装置、电子设备及存储介质
CN104021148B (zh) 调节音效的方法和装置
CN110362711A (zh) 歌曲推荐方法及装置
CN111210844B (zh) 语音情感识别模型的确定方法、装置、设备及存储介质
EP3364285A1 (en) Method, apparatus, computer program and storage medium for updating information
CN105912190A (zh) 界面操作方法和移动终端
US20180217719A1 (en) Menu modification based on controller manipulation data
CN109819288A (zh) 广告投放视频的确定方法、装置、电子设备及存储介质
WO2016026293A1 (zh) 消息发送方法及装置
CN106652981A (zh) Bpm检测方法及装置
CN117166042A (zh) 断线控制方法、装置、电子设备及存储介质
WO2018018912A1 (zh) 一种搜索方法、装置及电子设备
US20150134661A1 (en) Multi-Source Media Aggregation
CN108573706A (zh) 一种语音识别方法、装置及设备
US20130242200A1 (en) Measuring device and method for calculating response time of electronic device
CN108256542A (zh) 一种通信标识的特征确定方法、装置及设备
CN107436896A (zh) 一种输入推荐方法、装置及电子设备
CN116024649A (zh) 拉速控制方法、装置、电子设备及存储介质

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 23806538

Country of ref document: EP

Kind code of ref document: A1