A method, device and electronic equipment for liquid temperature calibration based on data analysis

Abstract

The present application discloses a liquid temperature calibration determination method, device and electronic equipment based on data analysis, which relates to the photovoltaic and semiconductor single crystal pulling technology field. The method includes: when the process for preparing single crystal silicon enters the shoulder release process, the first temperature parameter of the liquid temperature at the beginning of seeding, the second temperature parameter of the liquid temperature at the end of seeding, the average seeding pulling speed and the average seeding pulling speed in the later stage are detected; according to the first temperature parameter and the second temperature parameter, the target temperature group corresponding to the liquid temperature is determined; if the average seeding pulling speed and the average seeding pulling speed in the later stage are not in the effective seeding pulling speed interval corresponding to the target temperature group, the temperature calibration process of the liquid temperature is executed. Based on the pre-determined effective interval for judging whether the liquid temperature needs to be calibrated, the present application compares the data obtained by the current detection with the characteristics of the effective interval, and can automatically determine whether liquid temperature calibration is needed, which solves the problem of inaccurate and low efficiency of manual liquid temperature calibration.

Description

Technical Field

[0001] The present application relates to the field of photovoltaic and semiconductor single crystal pulling technology, and in particular to a liquid temperature calibration determination method, device and electronic equipment based on data analysis. Background Art

[0002] In the field of photovoltaics and semiconductors, direct-pulling single crystal silicon is a method of preparing semiconductor materials. It is a process of melting silicon into liquid and then pulling a single crystal rod from the liquid silicon. In the process of direct-pulling single crystal silicon, the temperature of the liquid silicon material is one of the important indicators. The temperature for melting the original polycrystalline silicon material from solid to liquid is relatively high, that is, the temperature needs to reach the melting point of the polycrystalline silicon material. In the growth of single crystal silicon, the temperature needs to be lowered to the crystallization point of single crystal silicon before the liquid material can be crystallized into solid single crystal silicon. However, due to the repeated use of the furnace body of the single crystal furnace, the thermal insulation of each furnace body or the long-term slow chemical reaction inside the furnace body, the temperature distribution will have uneven changes, which will affect the measurement of the liquid silicon material temperature (referred to as liquid temperature), and then affect the determination of the liquid silicon material temperature of the single crystal silicon crystallization point.

[0003] At present, whether the temperature of the liquid silicon material needs to be calibrated is mainly determined by experienced staff based on the status of the crystal seeding process. However, due to the different experiences of each person, the temperature calibration of the liquid silicon material is easily untimely and inaccurate, which has a certain impact on production costs and efficiency. Summary of the invention

[0004] In response to the above problems, the present application provides a liquid temperature calibration determination method, device and electronic equipment based on data analysis, which solves the problem of inaccurate and uncertain manual liquid temperature calibration.

[0005] In order to achieve the above objectives, this application provides the following technical solutions:

[0006] A liquid temperature calibration determination method based on data analysis, comprising:

[0007] When the process for preparing single crystal silicon enters the shoulder release step, the first temperature parameter of the liquid temperature at the start of seeding, the second temperature parameter of the liquid temperature at the end of seeding, the average seeding pulling speed and the average seeding pulling speed at the later stage are detected and obtained, and the liquid temperature represents the temperature of the liquid silicon material;

[0008] determining a target temperature group corresponding to the liquid temperature according to the first temperature parameter and the second temperature parameter;

[0009] If the average seeding pulling speed and the average seeding pulling speed in the later stage are both not in the seeding pulling speed effective range corresponding to the target temperature group, a temperature calibration process for the liquid temperature is performed.

[0010] Optionally, determining a target temperature group corresponding to the liquid temperature according to the first temperature parameter and the second temperature parameter includes:

[0011] Determine a first temperature set according to the first temperature parameter and a first temperature range of the seeding start liquid temperature, wherein the first temperature set includes a plurality of temperature groups, and each temperature group corresponds to a different temperature range of the seeding end liquid temperature;

[0012] A target temperature group is determined in the first temperature set according to the second temperature parameter.

[0013] Optionally, the performing of the temperature calibration process for the liquid temperature includes:

[0014] Calculate the temperature difference between the current liquid temperature and the standard liquid temperature;

[0015] Based on the temperature difference, the seeding end liquid temperature is calibrated.

[0016] Optionally, the method further comprises:

[0017] Obtaining a target data set in a process for preparing single crystal silicon, wherein the target data set includes a plurality of historical data, each of which represents seeding process data from the start of seeding to the end of seeding;

[0018] Determine at least one target temperature group based on the seeding start liquid temperature and the seeding end liquid temperature corresponding to each piece of historical data;

[0019] Based on the average seeding pulling speed and the average seeding pulling speed in the later stage of each historical data, the effective seeding pulling speed range corresponding to each target temperature group is determined.

[0020] Optionally, obtaining a target data set in a process for preparing single crystal silicon comprises:

[0021] Obtain at least one historical seeding process data in a process for preparing single crystal silicon;

[0022] Screening each of the historical seeding process data based on the target screening condition to obtain screened seeding process data;

[0023] Determine a target data set based on the screened seeding process data;

[0024] The target screening condition includes at least one of the following:

[0025] Delete the seeding process data corresponding to the failure of the equal diameter stage;

[0026] Delete the seeding process data in which the main power of the single crystal furnace changes during the seeding process;

[0027] Delete the seeding process data in which the temperature is reversed during the seeding process.

[0028] Optionally, determining the effective range of seeding pulling speed corresponding to each target temperature group based on the average seeding pulling speed and the average seeding pulling speed in the later stage in each piece of historical data includes:

[0029] Based on the average pulling speed range of the average seeding pulling speed and the average pulling speed in the late seeding period of the historical data corresponding to each target temperature group, the historical data corresponding to the target temperature group is divided into a plurality of intervals, wherein the average pulling speed range of each interval is different;

[0030] Determine a data proportion of each interval according to a proportion of a first number of historical data items included in each interval to a total number of historical data items corresponding to the target temperature group;

[0031] Based on the data proportion of each interval, an effective interval of the seeding pulling speed corresponding to the target temperature group is determined.

[0032] Optionally, determining the effective seeding speed interval corresponding to the target temperature group based on the data proportion of each interval includes:

[0033] Determine a first interval based on the data proportion of each interval corresponding to the target temperature group, where the first interval represents an interval with the largest data proportion among all intervals;

[0034] If the data proportion of the second interval is greater than the data proportion of the third interval, the first interval and the second interval are merged into a valid interval; wherein the second interval and the third interval are adjacent intervals of the first interval.

[0035] Optionally, the method further comprises:

[0036] At the end of seeding, based on the seeding end time, forward search is performed to record the target time point corresponding to the target length of the seeding;

[0037] The duration from the target time point to the end of seeding is determined as the late seeding period;

[0038] Based on the duration of the late seeding stage and the seeding length corresponding to the late seeding stage, an average pulling speed in the late seeding stage is determined.

[0039] A liquid temperature calibration determination device based on data analysis, comprising:

[0040] A detection unit, used for detecting and obtaining a first temperature parameter of a seeding start liquid temperature, a second temperature parameter of a seeding end liquid temperature, an average seeding pull speed, and an average seeding late pull speed when the single crystal silicon preparation process enters the shoulder release step, wherein the liquid temperature represents the temperature of the liquid silicon material;

[0041] a determination unit, configured to determine a target temperature group corresponding to the liquid temperature according to the first temperature parameter and the second temperature parameter;

[0042] The execution unit is used to execute the temperature calibration process of the liquid temperature if the average seeding pulling speed and the average seeding pulling speed in the later stage are not in the effective seeding pulling speed range corresponding to the target temperature group.

[0043] An electronic device, comprising:

[0044] Memory, used to store programs;

[0045] A processor is used to execute the program, and the program is specifically used to implement the liquid temperature calibration determination method based on data analysis as described in any one of the above.

[0046] Compared with the prior art, the present application provides a method, device and electronic device for liquid temperature calibration based on data analysis, the method comprising: when the process for preparing single crystal silicon enters the shoulder release process, the first temperature parameter of the liquid temperature at the beginning of seeding, the second temperature parameter of the liquid temperature at the end of seeding, the average seeding pulling speed and the average seeding pulling speed in the later stage are detected; according to the first temperature parameter and the second temperature parameter, the target temperature group corresponding to the liquid temperature is determined; if the average seeding pulling speed and the average seeding pulling speed in the later stage are not in the effective seeding pulling speed interval corresponding to the target temperature group, the temperature calibration process of the liquid temperature is executed. Among them, the liquid temperature represents the temperature of the liquid silicon material. The present application classifies and counts the effective features that can reflect the seeding process, determines the effective interval for judging whether liquid temperature calibration is needed, compares the data obtained by the current detection with the features of the effective interval, and can automatically determine whether liquid temperature calibration is needed, which solves the problems of inaccurate, uncertain and inefficient manual liquid temperature calibration. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.

[0048] Figure 1 A flow chart of a method for liquid temperature calibration determination based on data analysis provided in an embodiment of the present application;

[0049] Figure 2 A schematic diagram of a process flow based on a liquid temperature calibration model provided in an embodiment of the present application;

[0050] Figure 3A schematic diagram of the structure of a liquid temperature calibration and determination device based on data analysis provided in an embodiment of the present application. DETAILED DESCRIPTION

[0051] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0052] The terms "first" and "second" in this application are used to distinguish different objects, rather than to describe a specific order. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not limited to the listed steps or units, but may include steps or units that are not listed.

[0053] The embodiment of the present application provides a liquid temperature calibration determination method based on data analysis, wherein the liquid temperature refers to the temperature of the liquid silicon material in the single crystal silicon preparation process. The method is applied to the process of preparing single crystal silicon, by acquiring relevant features in the seeding process, and calculating the current data interval and comparing it with the preset effective interval. According to the comparison result, it is determined whether the liquid silicon material temperature needs to be calibrated. If calibration is required, the corresponding liquid temperature calibration process can be executed. There is no need to rely on manual experience to determine whether liquid temperature calibration is needed, which solves the problems of uncertainty and low efficiency of manual liquid temperature calibration.

[0054] In order to facilitate the description of the liquid temperature calibration determination method based on data analysis in the embodiment of the present application, the relevant data will now be described.

[0055] Direct-pulling single crystal silicon: refers to a method of preparing semiconductor materials by melting silicon into liquid and then directly pulling a single crystal rod from the liquid silicon.

[0056] Seed crystal: refers to the starting point in the growth process of single crystal silicon, which is a small crystal with the same crystal orientation as the required crystal.

[0057] Seeding: refers to bringing the seed crystal into contact with liquid silicon material, slowly lifting it up, and gradually shrinking it to eliminate local atomic arrangement distortion in the crystal structure.

[0058] Liquid temperature: refers to the temperature of liquid silicon material.

[0059] Full melting of silicon material: refers to the time point when the silicon material is heated and melted in the crucible until the molten material in the crucible is all liquid.

[0060] Shoulder release time: the time of the crystal lateral growth process after the seeding is completed.

[0061] Residual material: the weight of the remaining silicon material in the furnace body.

[0062] Stabilization time: Since the power needs to be adjusted to the seeding power before seeding begins, seeding begins after the temperature stabilizes and the seeding timing is met. The time from the power being adjusted to the seeding power to the start of seeding is called the stabilization time.

[0063] See also Figure 1 , is a flow chart of a liquid temperature calibration determination method based on data analysis provided in an embodiment of the present application, the method may include the following steps:

[0064] S101. When the process for preparing single crystal silicon enters the shoulder release step, the first temperature parameter of the liquid temperature at the start of seeding, the second temperature parameter of the liquid temperature at the end of seeding, the average seeding pulling speed and the average seeding pulling speed at the later stage are detected and obtained.

[0065] Among them, the liquid temperature characterizes the temperature of the liquid silicon material, that is, the temperature of the liquid silicon material in the process of preparing single crystal silicon. In the process of preparing single crystal silicon, the process changes can be monitored in real time, starting from the shoulder release process, and the key features of the seeding process can be obtained. Then these key features are compared with the preset effective interval to determine whether it is necessary to perform liquid silicon material temperature (ie, liquid temperature) calibration. Specifically, the key features from the beginning to the end of crystallization can be determined by analyzing historical data, and these key features can reflect the state of the entire seeding process. The temperature of the liquid silicon material at the beginning of seeding (referred to as the seeding start liquid temperature) and the temperature of the liquid silicon material at the end of seeding (referred to as the seeding end liquid temperature) can directly reflect the insulation state of the furnace body of the single crystal furnace. For example, the standard furnace body is the temperature value that reaches the silicon material crystallization start temperature seeding. After long-term use, the seeding temperature value will be low. The average pulling speed during seeding and the average pulling speed in the later stage of seeding are key indicators that indirectly reflect the liquid temperature. If seeding is carried out when the liquid temperature is low, the crystallization rate will become faster and the pulling speed will become faster. Therefore, these four key characteristics can reflect the temperature state of the liquid silicon material during the seeding process.

[0066] In an embodiment of the present application, in the actual process of preparing single crystal silicon, the value of the liquid temperature at the beginning of seeding can be monitored in real time as the first temperature parameter, such as the first temperature parameter is 1448°C, and the value of the liquid temperature at the end of seeding is the second temperature parameter, such as the second temperature parameter is 1447.8°C. The average seeding pulling speed and the average seeding pulling speed in the late stage are calculated by the crystal length and the corresponding seeding time during the seeding process. For example, the current average seeding pulling speed is 200 mm / h. The late stage of seeding refers to the period from the target moment to the end of seeding, such as the range of 50 mm in the final growth length of the crystal is the late stage of seeding.

[0067] S102: Determine a target temperature group corresponding to the liquid temperature according to the first temperature parameter and the second temperature parameter.

[0068] In order to accurately determine whether liquid temperature calibration is required, multiple temperature groups can be pre-determined according to the starting liquid temperature of seeding and the ending liquid temperature of seeding, so as to determine the target temperature group to which the current seeding state belongs according to the current first temperature parameter and the second temperature parameter.

[0069] In one embodiment of the present application, a target temperature group corresponding to the liquid temperature is determined according to a first temperature parameter and a second temperature parameter, including: determining a first temperature set according to the first temperature parameter and a first temperature range of the seeding start liquid temperature; and determining a target temperature group in the first temperature set according to the second temperature parameter. The first temperature set includes several temperature groups, and each temperature group corresponds to a different temperature range of the seeding end liquid temperature.

[0070] In order to be able to perform accurate liquid temperature calibration, in an embodiment of the present application, based on the state data of the seeding process in the historical data of single crystal silicon preparation, the seeding start liquid temperature and the seeding end liquid temperature are divided into multiple temperature groups as key features, wherein the first temperature set can be a large temperature group determined based on the seeding start liquid temperature. For example, by analyzing the historical data, the first temperature range corresponding to the seeding start liquid temperature is usually 1448°C to 1449°C, which can be divided into multiple temperature sets with a step size of 0.2°C. For example, temperature set 1 is 1448°C to 1448.2°C, temperature set 2 is 1448.3 to 1448.5, and so on. If the current first temperature parameter is 1448.1°C, it belongs to the first temperature set. In order to further improve the accuracy, after the temperature set is determined based on the seeding start liquid temperature, the corresponding temperature group is determined according to the seeding end temperature. For example, each temperature set is divided into several temperature groups according to the seeding end liquid temperature in the historical data in advance. If the seeding end liquid temperature range is 1447.5℃ to 1448℃, then according to the seeding end liquid temperature step, the temperature set is divided into several temperature groups. For example, the seeding start liquid temperature corresponding to the first temperature set is 1448℃ to 1448.2℃, and the seeding end liquid temperature is 1447.5℃ to 1447.7 degrees Celsius as the first temperature group, and so on to obtain multiple temperature groups. If the current first temperature parameter is 1448.1℃, it is determined to be the first temperature set, and the second temperature parameter is 1447.6℃, then it belongs to the first temperature group, that is, the target temperature group.

[0071] S103, if the average seeding pulling speed and the average seeding pulling speed in the later stage are both not in the seeding pulling speed effective range corresponding to the target temperature group, execute the temperature calibration process for the liquid temperature.

[0072] After obtaining the target temperature group corresponding to the current seeding process, determine the preset seeding pull speed effective area corresponding to the target temperature group. The effective interval of the seeding pull speed represents the normal speed range of crystal growth when the temperature of the liquid silicon material is normal, that is, the effective interval of the seeding pull speed represents the range of the seeding pull speed when the temperature of the liquid silicon material is normal. If the effective interval of the seeding pull speed corresponding to the current target temperature group is 215-220, and the average seeding pull speed obtained by the current detection is 213, and the average pull speed in the late stage of seeding is 214, both of which do not belong to the range of the effective interval, it means that the temperature of the liquid silicon material needs to be calibrated, and the temperature calibration process for the liquid temperature is further executed. The temperature calibration process can be determined according to the current process for preparing single crystal silicon. In one embodiment, the temperature calibration process for the temperature of the liquid silicon material includes: calculating the temperature difference between the current liquid temperature and the standard liquid temperature; based on the temperature difference, calibrating the temperature of the liquid silicon material at the end of seeding. Specifically, the end liquid temperature of seeding can be calibrated by adjusting the exposure according to the temperature difference.

[0073] In the embodiment of the present application, it is possible to perform analysis based on historical data, establish a valid interval based on key features that can reflect the temperature of the liquid silicon material, and automatically determine whether the current liquid silicon material temperature needs to be calibrated by comparing the current real-time detection data with the valid interval. This eliminates the need for manual real-time monitoring and judgment based on subjective experience, thereby improving the certainty and accuracy of the liquid temperature calibration.

[0074] The process of determining the effective range of the seeding pull speed in the embodiment of the present application is described below. In one implementation of the embodiment of the present application, the method also includes: obtaining a target data set in the process of preparing single crystal silicon, the target data set includes a number of historical data, each of which represents the seeding process data from the beginning of seeding to the end of seeding; based on the corresponding seeding start liquid temperature and seeding end liquid temperature in each historical data, determine at least one target temperature group; based on the average seeding pull speed and the average seeding pull speed in the later stage of each historical data, determine the effective range of seeding pull speed corresponding to each target temperature group.

[0075] The target data set is a historical data set recorded in the process of preparing single crystal silicon, which includes several historical data, each of which corresponds to the data of a seeding process. By analyzing these data, the effective range of seeding pulling speed that can characterize whether liquid temperature calibration is required can be determined. Extract the liquid temperature and pulling speed from the beginning to the end of each seeding and the length of the crystal after the equal diameter process in this straight-pulling single crystal silicon process from the historical data; calculate the average liquid temperature before and after the seeding start liquid temperature as the seeding start liquid temperature, and calculate the liquid temperature before and after the end of the seeding as the seeding end liquid temperature; calculate the crystal length and time from the beginning of the seeding to the end of the seeding, and divide the length by the time as the average seeding pulling speed; find and record the length and time of the end of the seeding and search forward to the moment when it is determined to be the late stage of the seeding, and use the length and time of this moment to make a difference with the length and time of the end of the seeding, and divide the calculated length difference by the time difference as the average pulling speed in the late stage of the seeding. Among them, forward retrieval refers to looking forward for the point of corresponding length at the end of the seeding.

[0076] In one implementation of the embodiments of the present application, at the end of seeding, a forward search is performed based on the end time of seeding, and the target time point corresponding to the target length is recorded; the duration from the target time point to the end time of seeding is determined as the late seeding period; based on the duration of the late seeding period and the seeding length corresponding to the late seeding period, the average pulling speed in the late seeding period is determined. For example, the seeding length is used to determine whether the seeding period has entered the late seeding period. During the seeding process, the crystal length increases, and the last 50 mm of the seeding period can be determined as the late seeding period.

[0077] In order to improve the accuracy of the judgment, it is necessary to screen the historical data to filter out unqualified data. In one embodiment of the present application, a target data set in a process for preparing a single crystal silicon is obtained, including: obtaining at least one historical seeding process data in a process for preparing a single crystal silicon; screening each of the historical seeding process data based on a target screening condition to obtain the seeding process data after screening; determining the target data set according to the seeding process data after screening; wherein the target screening condition includes at least one of the following: deleting the seeding process data corresponding to the failure of the equal diameter stage; deleting the seeding process data in which the main power of the single crystal furnace body changes during the seeding process; deleting the seeding process data in which the reverse temperature state appears during the seeding process.

[0078] It should be noted that there will be differences in historical data when preparing single crystal silicon through different equipment. Therefore, the target screening conditions can be determined according to the actual application scenarios. The target screening conditions illustrated in the above examples cannot represent all the data. Some rules can be added or deleted according to different differences to determine the target screening conditions. Specifically, the seeding process data corresponding to the failure of the equal diameter stage is deleted because the failure of the equal diameter stage belongs to the failure of crystal pulling, and the characteristics of the seeding process are not of reference significance. Delete the seeding process data in which the main power of the single crystal furnace body changes during the seeding process. In this case, there is manual intervention temperature and it cannot be used as a basis for automatic determination of liquid temperature calibration; delete the seeding process data in which the inverse temperature state occurs during the seeding process. Among them, the inverse temperature state is mainly that the liquid temperature at the end of seeding is greater than the liquid temperature at the beginning of seeding and the average pulling speed in the later stage of seeding is less than the average pulling speed of seeding.

[0079] In order to accurately set the effective interval in the embodiment of the present application, the effective interval can be determined by the data proportion. In one embodiment, based on the average seeding pulling speed and the average seeding later pulling speed in each historical data, the effective seeding pulling speed interval corresponding to each target temperature group is determined, including: based on the average pulling speed range of the seeding average pulling speed and the average pulling speed in the later stage of seeding of the historical data corresponding to each target temperature group, the historical data corresponding to the target temperature group is divided into several intervals, wherein the average pulling speed range of each interval is different; the data proportion of each interval is determined according to the proportion of the first number of historical data items included in each interval to the total number of historical data items corresponding to the target temperature group; based on the data proportion of each interval, the effective interval is determined. For example, the historical data corresponding to the target temperature group is divided into three intervals, and the average pulling speed range of each interval is different. For example, the average pulling speed of the first interval is A, the average pulling speed of the second interval is B, and the average pulling speed of the third interval is C. Then, the corresponding starting liquid temperature and ending liquid temperature of the seeding for several historical data included in the first interval both meet the liquid temperature characteristics corresponding to the target temperature group, and the average pulling speed of each historical data in the first interval is A. If the first interval includes 10 historical data, the first number of historical data included in the first interval is 10. If the total number of historical data included in these three intervals is 30, the data in the first interval accounts for 1 / 3.

[0080] First, according to the average pull speed range of the average pull speed of the seeding and the average pull speed of the late seeding of the historical data corresponding to the target temperature group, it is divided into multiple intervals, and then the effective interval is determined according to the data volume ratio of each interval. Specifically, the effective area of ​​the seeding pull speed corresponding to the target temperature group is determined based on the data ratio of each interval, including: based on the data ratio of each interval corresponding to the target temperature group, the first interval is determined, and the first interval represents the interval with the largest data ratio among all intervals; if the data ratio of the second interval is greater than the data ratio of the third interval, the first interval and the second interval are merged into a valid interval; wherein the second interval and the third interval are adjacent intervals of the first interval.

[0081] Specifically, the target temperature group is divided into multiple intervals, the data proportion of each interval is counted, and the interval with the largest data proportion is found as the first interval. In fact, the first interval is expanded according to the data proportions of the two adjacent intervals on the left and right of the first interval (i.e., the second interval and the third interval). The first interval is expanded in the direction of the interval with a larger data proportion until the data proportion of the first interval can represent the characteristics of the target temperature group.

[0082] For example, according to the screening conditions, multiple valid seeding data have been obtained, and each data has the seeding starting liquid temperature, seeding ending liquid temperature. The four characteristics of the average seeding pulling speed and the average seeding pulling speed in the later stage; for example, the starting liquid temperature is distributed between 1448-1449, and the starting liquid temperature is divided into multiple large groups with a step size of 0.1, such as 1448-1448.1 is the first group of starting liquid temperature, and this group contains multiple data. If the ending liquid temperature range in this group is 1447.5-1448, the ending liquid temperature in this group is divided into multiple small groups with another step size. Each small group has multiple data, and the valid data range of the average seeding pulling speed and the average seeding pulling speed in the later stage in this group is counted; the effective calculation method is: if the average seeding pulling speed range in this group is 200-220, divide the large range into multiple groups with a step size of 5. For a small range, calculate the data proportion of each small range, such as 200-205 data proportion is 10%, 205-210 data proportion is 50%, 210-215 data proportion is 30%, 215-220 data proportion is 10%; first calculate the interval with the largest data proportion as 205-210, and calculate the data proportion of the two adjacent intervals on the left and right of the interval, and expand the one with the larger data proportion on the left and right sides. After expansion, the interval becomes 205-215, and the data proportion is 80%. If the threshold is set to 70%, the threshold has been reached now, and it is considered that the interval 205-215 is a valid data interval for the large interval 200-220.

[0083] Furthermore, in order to facilitate the processing of liquid temperature calibration judgment and improve the judgment efficiency, the grouping information of the calculated seeding start liquid temperature, seeding end liquid temperature, seeding average pulling speed and seeding late average pulling speed can be saved in an array. When the management platform passes the feature into the liquid temperature calibration model, first determine the large group to which it belongs based on the seeding start liquid temperature, and then determine the small group to which it belongs based on the seeding end liquid temperature. After finding the group, determine whether the seeding average pulling speed and the seeding late average pulling speed are both within the corresponding valid range. If so, the liquid temperature is not calibrated, otherwise the liquid temperature needs to be calibrated.

[0084] See also Figure 2 The liquid temperature calibration model can be stored in the single crystal silicon preparation management platform. The management platform monitors the process changes in real time, obtains four features from the beginning of the shoulder release, namely, the liquid temperature at the beginning of the seeding, the liquid temperature at the end of the seeding, the average pulling speed of the seeding, and the average pulling speed at the end of the seeding, and calls the liquid temperature calibration model after obtaining them. The liquid temperature calibration model determines whether the liquid temperature needs to be calibrated. If it is sent to the host computer by the management platform, the host computer calculates the deviation between the liquid temperature and the standard liquid temperature, and calibrates the liquid temperature at the end of the seeding by adjusting the exposure.

[0085] It should be noted that there will be many abnormal situations during the crystal seeding process. The temperature and pulling speed data in abnormal situations cannot reflect the normal crystal seeding situation of the furnace body. Therefore, the liquid temperature is not calibrated for the following abnormal situations: First, the reverse temperature is not calibrated. The standard situation is that the liquid temperature at the beginning of crystal seeding is greater than the liquid temperature at the end of crystal seeding. If the liquid temperature at the end of crystal seeding is greater, it is judged as reverse temperature. At this time, the temperature cannot reflect the actual temperature performance of the furnace body; Second, if the power change during crystal seeding is not calibrated, under standard circumstances, the power does not change during crystal seeding. If the crystal seeding temperature is adjusted manually, then this time The information has no reference significance; third, the liquid mouth distance deviation from the last crystal seeding is too large, and the large liquid mouth distance deviation will lead to inaccurate temperature measurement; fourth, the seeding length exceeds the standard range of crystal seeding; fifth, the stabilization time is too long, and a long stabilization time may lead to manual intervention of power; sixth, the shoulder release time is not within the standard range, and the shoulder release time is not within the standard range, which means that the crystal growth process is not standard; seventh, the amount of remaining material is small. Under normal circumstances, the material is full. When the amount of remaining material is small, the temperature and pulling speed during the crystal seeding process cannot reflect the temperature and pulling speed of the furnace under standard production conditions.

[0086] When liquid temperature calibration was performed based on the above method, 52 tests were conducted (excluding abnormal situations), and the liquid temperature calibration was judged to be accurate 39 times at this time, with an accuracy rate of 75%, which is higher than the judgment rate of manually judging whether to perform liquid temperature calibration. In the embodiment of the present application, the key features from the beginning of seeding to the end of seeding are selected from historical data in combination with artificial experience, and the effective features are further screened out by whether abnormal situations such as manual intervention or reverse temperature occur in this seeding and whether the shoulder equal diameter after seeding survives as the standard. These effective features are classified and counted to calculate the effective interval, and whether liquid temperature calibration is needed is judged by determining whether the key information in the seeding process is within the effective interval. Solve the uncertainty of artificial liquid temperature calibration in the prior art.

[0087] In the embodiment of the present application, a liquid temperature calibration determination device based on data analysis is also provided, see Figure 3 ,include:

[0088] The detection unit 301 is used to detect and obtain a first temperature parameter of the liquid temperature at the start of seeding, a second temperature parameter of the liquid temperature at the end of seeding, an average seeding pulling speed, and an average seeding pulling speed at the end of seeding when the process for preparing single crystal silicon enters the shoulder release step, wherein the liquid temperature represents the temperature of the liquid silicon material;

[0089] A determination unit 302, configured to determine a target temperature group corresponding to the liquid temperature according to the first temperature parameter and the second temperature parameter;

[0090] The execution unit 303 is used to execute a temperature calibration process for the liquid temperature if the average seeding pulling speed and the average seeding late pulling speed are both not within the seeding pulling speed effective range corresponding to the target temperature group.

[0091] Optionally, the determining unit includes:

[0092] A first determining subunit is used to determine a first temperature set according to the first temperature parameter and a first temperature range of the seeding start liquid temperature, wherein the first temperature set includes a plurality of temperature groups, and each temperature group corresponds to a different temperature range of the seeding end liquid temperature;

[0093] The second determining subunit is configured to determine a target temperature group in the first temperature set according to the second temperature parameter.

[0094] Optionally, the execution unit includes:

[0095] A calculation subunit, used to calculate the temperature difference between the current liquid temperature and the standard liquid temperature;

[0096] The calibration subunit is used to calibrate the end liquid temperature of the seeding based on the temperature difference.

[0097] Optionally, the device further comprises:

[0098] A data acquisition unit, used to obtain a target data set in a process for preparing single crystal silicon, wherein the target data set includes a plurality of historical data, each of which represents seeding process data from the start of seeding to the end of seeding;

[0099] A temperature group determination unit, configured to determine at least one target temperature group based on a seeding start liquid temperature and a seeding end liquid temperature corresponding to each piece of historical data;

[0100] The interval determination unit is used to determine the effective interval of the seeding pulling speed corresponding to each target temperature group based on the average seeding pulling speed and the average seeding pulling speed in the later stage of each historical data.

[0101] Optionally, the data acquisition unit includes:

[0102] A first acquisition subunit is used to obtain at least one historical seeding process data in the process of preparing single crystal silicon;

[0103] A screening subunit, used for screening each of the historical seeding process data based on a target screening condition to obtain screened seeding process data;

[0104] A third determination subunit is used to determine a target data set according to the screened seeding process data;

[0105] The target screening condition includes at least one of the following:

[0106] Delete the seeding process data corresponding to the failure of the equal diameter stage;

[0107] Delete the seeding process data in which the main power of the single crystal furnace changes during the seeding process;

[0108] Delete the seeding process data in which the temperature is reversed during the seeding process.

[0109] Optionally, the interval determining unit includes:

[0110] A division subunit is used to divide the historical data corresponding to each target temperature group into a plurality of intervals based on the average seeding pulling speed and the average pulling speed range of the late seeding average pulling speed of the historical data corresponding to each target temperature group, wherein the average pulling speed range of each interval is different;

[0111] A fourth determining subunit is used to determine a data proportion of each interval according to a proportion of a first number of historical data items included in each interval to a total number of historical data items corresponding to the target temperature group;

[0112] The fifth determination subunit is used to determine the effective range of the seeding pulling speed corresponding to the target temperature group based on the data proportion of each range.

[0113] Optionally, the fifth determining subunit is configured to:

[0114] Determine a first interval based on the data proportion of each interval corresponding to the target temperature group, where the first interval represents an interval with the largest data proportion among all intervals;

[0115] If the data proportion of the second interval is greater than the data proportion of the third interval, the first interval and the second interval are merged into a valid interval; wherein the second interval and the third interval are adjacent intervals of the first interval.

[0116] Optionally, the device further comprises:

[0117] A recording subunit, used for searching forward based on the seeding end time at the end of seeding, and recording the target time point corresponding to the target length when the seeding length is the target length;

[0118] A sixth determination subunit, for determining the duration from the target time point to the end of seeding as the late seeding period;

[0119] The seventh determination subunit is used to determine the average pulling speed in the late stage of seeding based on the duration of the late stage of seeding and the seeding length corresponding to the late stage of seeding.

[0120] It should be noted that the specific implementation of each unit and sub-unit in this embodiment can refer to the corresponding content in the previous text and will not be described in detail here.

[0121] Based on the aforementioned embodiments, an embodiment of the present application provides a computer-readable storage medium, which stores one or more programs, and the one or more programs can be executed by one or more processors to implement any of the above data analysis-based liquid temperature calibration determination methods.

[0122] The present application also provides an electronic device, including:

[0123] Memory, used to store programs;

[0124] A processor is used to execute the program, wherein the program is specifically used to implement the liquid temperature calibration determination method based on data analysis as described in any one of the above.

[0125] The above is only a specific implementation method of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of changes or replacements within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same and similar parts between the various embodiments can be referenced to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part description.

[0126] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A liquid temperature calibration determination method based on data analysis, characterized in that: include: When the process for preparing single crystal silicon enters the shoulder release step, the first temperature parameter of the liquid temperature at the start of seeding, the second temperature parameter of the liquid temperature at the end of seeding, the average seeding pulling speed and the average seeding pulling speed at the later stage are detected and obtained, and the liquid temperature represents the temperature of the liquid silicon material; Determine a target temperature group corresponding to the liquid temperature according to the first temperature parameter and the second temperature parameter; wherein, determining the target temperature group corresponding to the liquid temperature according to the first temperature parameter and the second temperature parameter comprises: determining a first temperature set according to the first temperature parameter and a first temperature range of the seeding start liquid temperature, wherein the first temperature set comprises a plurality of temperature groups, and each temperature group corresponds to a different temperature range of the seeding end liquid temperature; and determining a target temperature group in the first temperature set according to the second temperature parameter; If both the average seeding pulling speed and the average seeding pulling speed in the later stage are not in the seeding pulling speed effective range corresponding to the target temperature group, a temperature calibration process for the liquid temperature is performed; The process of determining the effective range of the seeding pulling speed includes: obtaining a target data set in the process of preparing single crystal silicon, wherein the target data set includes a plurality of historical data, each of which represents the seeding process data from the start of seeding to the end of seeding; determining at least one target temperature group based on the corresponding seeding start liquid temperature and seeding end liquid temperature in each historical data; determining the effective range of the seeding pulling speed corresponding to each target temperature group based on the average seeding pulling speed and the average seeding late pulling speed in each historical data; The method of determining the effective interval of seeding pulling speed corresponding to each target temperature group based on the average seeding pulling speed and the average seeding pulling speed in the later stage in each piece of historical data includes: dividing the historical data corresponding to the target temperature group into a plurality of intervals based on the average seeding pulling speed and the average seeding pulling speed in the later stage of the historical data corresponding to each target temperature group, wherein the average pulling speed range of each interval is different; determining the proportion of data in each interval according to the proportion of a first number of historical data items included in each interval to the total number of historical data items corresponding to the target temperature group; and determining the effective interval of seeding pulling speed corresponding to the target temperature group based on the data proportion of each interval; The method of determining the effective interval of the seeding pulling speed corresponding to the target temperature group based on the data proportion of each interval includes: determining a first interval based on the data proportion of each interval corresponding to the target temperature group, the first interval representing the interval with the largest data proportion among all intervals; if the data proportion of the second interval is greater than the data proportion of the third interval, merging the first interval and the second interval into a valid interval; wherein the second interval and the third interval are adjacent intervals of the first interval.

2. The method according to claim 1, characterized in that The process of performing temperature calibration of the liquid temperature comprises: Calculate the temperature difference between the current liquid temperature and the standard liquid temperature; Based on the temperature difference, the seeding end liquid temperature is calibrated.

3. The method according to claim 1, characterized in that The method of obtaining a target data set in a process for preparing single crystal silicon comprises: Obtain at least one historical seeding process data in a process for preparing single crystal silicon; Screening each of the historical seeding process data based on the target screening condition to obtain screened seeding process data; Determine a target data set based on the screened seeding process data; The target screening condition includes at least one of the following: Delete the seeding process data corresponding to the failure of the equal diameter stage; Delete the seeding process data in which the main power of the single crystal furnace changes during the seeding process; Delete the seeding process data in which the temperature is reversed during the seeding process.

4. The method according to claim 1, characterized in that: The method further comprises: At the end of seeding, based on the seeding end time, forward search is performed to record the target time point corresponding to the target length of the seeding; The duration from the target time point to the end of seeding is determined as the late seeding period; Based on the duration of the late seeding stage and the seeding length corresponding to the late seeding stage, an average pulling speed in the late seeding stage is determined.

5. A liquid temperature calibration determination device based on data analysis, characterized in that: include: A detection unit, used for detecting and obtaining a first temperature parameter of a seeding start liquid temperature, a second temperature parameter of a seeding end liquid temperature, an average seeding pull speed, and an average seeding late pull speed when the single crystal silicon preparation process enters the shoulder release step, wherein the liquid temperature represents the temperature of the liquid silicon material; A determination unit, configured to determine a target temperature group corresponding to the liquid temperature according to the first temperature parameter and the second temperature parameter; wherein the determination unit comprises: a first determination subunit, configured to determine a first temperature set according to the first temperature parameter and a first temperature range of the seeding start liquid temperature, wherein the first temperature set comprises a plurality of temperature groups, and each temperature group corresponds to a different temperature range of the seeding end liquid temperature; and a second determination subunit, configured to determine a target temperature group in the first temperature set according to the second temperature parameter; An execution unit, configured to execute a temperature calibration process for the liquid temperature if both the average seeding pulling speed and the average seeding late pulling speed are not within the seeding pulling speed effective range corresponding to the target temperature group; The device also includes: a data acquisition unit, which is used to obtain a target data set in the process of preparing single crystal silicon, wherein the target data set includes a plurality of historical data, each of which represents the seeding process data from the start of seeding to the end of seeding; a temperature group determination unit, which is used to determine at least one target temperature group based on the seeding start liquid temperature and the seeding end liquid temperature corresponding to each historical data; an interval determination unit, which is used to determine the effective seeding pull speed interval corresponding to each target temperature group based on the seeding average pull speed and the seeding late average pull speed in each historical data; wherein the interval determination unit includes: a division subunit, which is used to divide the historical data corresponding to the target temperature group into a plurality of intervals based on the average pull speed range of the seeding average pull speed and the seeding late average pull speed of the historical data corresponding to each target temperature group, wherein the average pull speed range of each interval is different; a fourth determination subunit, which is used to determine the data proportion of each interval according to the proportion of the first number of historical data items included in each interval to the total number of historical data items corresponding to the target temperature group; a fifth determination subunit, which is used to determine the effective seeding pull speed interval corresponding to the target temperature group based on the data proportion of each interval; The fifth determination subunit is configured to: determine a first interval based on the data proportion of each interval corresponding to the target temperature group, the first interval representing the interval with the largest data proportion among all intervals; if the data proportion of the second interval is greater than the data proportion of the third interval, merge the first interval and the second interval into a valid interval; wherein the second interval and the third interval are adjacent intervals of the first interval.

6. An electronic device, characterized in that: include: Memory, used to store programs; A processor is used to execute the program, wherein the program is specifically used to implement the liquid temperature calibration determination method based on data analysis as described in any one of claims 1 to 4.

Images