Single crystal furnace pressure alarm method, device, equipment, storage medium and program product

Abstract

The embodiment of the invention provides a single crystal furnace pressure alarm method, device and equipment, a storage medium and a program product. The furnace pressure alarm method for the single crystal furnace comprises the following steps: collecting the furnace pressure of the single crystal furnace in a preset time period to obtain a furnace pressure sequence, extracting furnace pressure characteristics of the single crystal furnace from the furnace pressure sequence, determining whether the furnace pressure of the single crystal furnace is normal or abnormal according to the furnace pressure characteristics, under the condition that the furnace pressure is abnormal, alarm information is output; according to the method, the accuracy of furnace pressure alarm of the single crystal furnace is improved, and the risks of missing alarm and false alarm are reduced.

Description

Technical Field

[0001] The present application relates to the technical field of semiconductor manufacturing equipment, and in particular, to a method, device, equipment, storage medium and program product for alarming the furnace pressure of a single crystal furnace. Background Art

[0002] In the single crystal silicon growth process, maintaining stable control of the furnace pressure in the single crystal furnace is crucial. Abnormal furnace pressure not only damages the quality of crystal growth but may also pose safety hazards.

[0003] In the prior art, an alarm device is usually used to detect the furnace pressure value of the single crystal furnace. When the furnace pressure exceeds the preset pressure threshold range, it is determined that the furnace pressure is in an abnormal state and the alarm mechanism is triggered.

[0004] However, the accuracy of the alarm method using the above alarm device is relatively low, and there are relatively high risks of false alarms and missed alarms. Summary of the Invention

[0005] Embodiments of the present application provide a method, device, equipment, storage medium and program product for alarming the furnace pressure of a single crystal furnace, so as to solve the problem that the accuracy of the existing method for alarming the furnace pressure of a single crystal furnace is relatively low, and there are relatively high risks of false alarms and missed alarms.

[0006] In a first aspect, embodiments of the present application provide a method for alarming the furnace pressure of a single crystal furnace, including:

[0007] Collect the furnace pressure of the single crystal furnace within a preset time period to obtain a furnace pressure sequence;

[0008] Extract the furnace pressure characteristics of the single crystal furnace from the furnace pressure sequence, where the furnace pressure characteristics include at least one of time domain characteristics, frequency domain characteristics and statistical distribution characteristics;

[0009] Determine whether the furnace pressure of the single crystal furnace is normal or abnormal according to the furnace pressure characteristics;

[0010] Output an alarm message when the furnace pressure is abnormal.

[0011] Optionally, the determining whether the furnace pressure of the single crystal furnace is normal or abnormal according to the furnace pressure characteristics includes:

[0012] Input the furnace pressure characteristics into a pre-trained anomaly recognition model to obtain a recognition result output by the anomaly recognition model, where the recognition result is normal or abnormal furnace pressure, and the anomaly recognition model is trained using the historical furnace pressure data of the single crystal furnace.

[0013] Optionally, the furnace pressure characteristics include the time domain characteristics, and the extracting the furnace pressure characteristics of the single crystal furnace from the furnace pressure sequence includes:

[0014] Extract the average furnace pressure, the maximum furnace pressure, the minimum furnace pressure, and the rate of change of furnace pressure from the furnace pressure sequence;

[0015] Determine the average furnace pressure, the maximum furnace pressure, the minimum furnace pressure, and the rate of change of furnace pressure as the time-domain features.

[0016] Optionally, the furnace pressure features include the frequency-domain features. Extracting the furnace pressure features of the single-crystal furnace from the furnace pressure sequence includes:

[0017] Perform a Fourier transform on the furnace pressure sequence to obtain the spectrum corresponding to the furnace pressure sequence;

[0018] Determine the band energy of a preset frequency band based on the spectrum, and determine the band energy of the preset frequency band as the frequency-domain feature.

[0019] Optionally, the furnace pressure features include the statistical distribution features. Extracting the furnace pressure features of the single-crystal furnace from the furnace pressure sequence includes:

[0020] Determine the distribution of the furnace pressure sequence, and determine the standard deviation, skewness, and kurtosis of the distribution;

[0021] Determine the standard deviation, the skewness, and the kurtosis as the statistical distribution features.

[0022] Optionally, collecting the furnace pressure of the single-crystal furnace within a preset time period to obtain a furnace pressure sequence includes:

[0023] Collect the furnace pressure of the single-crystal furnace within a preset time period through multiple sensors arranged at different positions of the single-crystal furnace;

[0024] Determine the average value of the furnace pressures collected by the multiple sensors at each sampling moment to obtain an initial furnace pressure sequence;

[0025] Preprocess the initial furnace pressure sequence to obtain the furnace pressure sequence.

[0026] In a second aspect, an embodiment of the present application provides a single-crystal furnace furnace pressure alarm device, including:

[0027] A collection module for collecting the furnace pressure of the single-crystal furnace within a preset time period to obtain a furnace pressure sequence;

[0028] An extraction module for extracting the furnace pressure features of the single-crystal furnace from the furnace pressure sequence, where the furnace pressure features include at least one of time-domain features, frequency-domain features, and statistical distribution features;

[0029] A determination module for determining whether the furnace pressure of the single-crystal furnace is normal or abnormal according to the furnace pressure features.

[0030] An alarm module, configured to output an alarm message when the furnace pressure is abnormal.

[0031] In a third aspect, an embodiment of the present application provides an electronic device, including: a processor, and a memory communicatively connected to the processor;

[0032] The memory stores computer-executable instructions;

[0033] The processor executes the computer-executable instructions stored in the memory to implement the method according to any one of the first aspect.

[0034] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, which stores computer-executable instructions, and when the computer-executable instructions are executed by a processor, they are used to implement the method according to any one of the first aspect.

[0035] In a fifth aspect, an embodiment of the present application provides a computer program product, including a computer program, and when the computer program is executed by a processor, it implements the method according to any one of the first aspect.

[0036] In summary, the present application provides a single crystal furnace pressure alarm method, device, equipment, storage medium and program product, which can collect the furnace pressure data within a preset time period in the single crystal furnace in real time or at regular intervals to form a furnace pressure sequence, that is, time series furnace pressure data. Then, furnace pressure features in multiple dimensions are extracted from the furnace pressure sequence, including time domain features, frequency domain features, statistical distribution features, etc. Based on the extracted furnace pressure features, comprehensive analysis is performed to determine whether the current furnace pressure state is normal or abnormal. Finally, if it is determined that the furnace pressure is abnormal, the alarm mechanism is triggered to output an alarm message. By comprehensively analyzing the furnace pressure features, the method improves the accuracy of identifying abnormal furnace pressure, thereby improving the accuracy of furnace pressure alarm and reducing false alarms and missed alarms. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] The accompanying drawings herein are incorporated into the specification and constitute a part of this specification, showing embodiments consistent with the present application, and are used together with the specification to explain the principles of the present application.

[0038] Figure 1 Schematic flowchart of the single crystal furnace pressure alarm method provided by the present application Figure 1 ;

[0039] Figure 2 Schematic flowchart of the single crystal furnace pressure alarm method provided by the present application Figure 2 ;

[0040] Figure 3 Schematic structural diagram of the single crystal furnace pressure alarm device provided by the present application;

[0041] Figure 4Schematic structural diagram of the electronic device provided by this application.

[0042] Through the above-mentioned drawings, specific embodiments of this application have been shown, and more detailed descriptions will be provided hereinafter. These drawings and textual descriptions are not intended to limit the scope of the concept of this application in any way, but to illustrate the concept of this application to those skilled in the art by referring to specific embodiments. Detailed implementation manners

[0043] For the convenience of clearly describing the technical solutions of the embodiments of this application, in the embodiments of this application, terms such as "first" and "second" are used to distinguish identical or similar items with basically the same functions and roles. For example, the first device and the second device are only used to distinguish different devices, and do not limit their sequence. Those skilled in the art can understand that terms such as "first" and "second" do not limit the quantity and execution order, and terms such as "first" and "second" do not necessarily mean different.

[0044] It should be noted that in this application, words such as "exemplary" or "for example" are used to represent examples, illustrations or explanations. Any embodiment or design solution described as "exemplary" or "for example" in this application should not be construed as being more preferred or having more advantages than other embodiments or design solutions. Exactly speaking, using words such as "exemplary" or "for example" is intended to present relevant concepts in a specific manner.

[0045] In this application, "at least one" means one or more, and "a plurality" means two or more. "And / or" describes the association relationship of associated objects, indicating that three relationships may exist. For example, A and / or B may represent: A exists alone, A and B exist simultaneously, and B exists alone, where A and B may be singular or plural. The character " / " generally represents an "or" relationship between the associated objects before and after. "At least one (item)" or its similar expression below refers to any combination of these items, including any combination of single item (item) or plural items (items). For example, at least one (item) of a, b, or c may represent: a, b, c, a - b, a - c, b - c, or a - b - c, where a, b, and c may be single or multiple.

[0046] During the single-crystal silicon growth process, maintaining precise control of the internal pressure of the single-crystal furnace is crucial. This is because abnormal fluctuations in the furnace pressure will not only seriously affect the quality of crystal growth, resulting in product defects, but may also bring a series of safety hazards, threatening the safety of the production environment.

[0047] Currently, a simple alarm device is used to monitor the furnace pressure. Specifically, by detecting the furnace pressure value inside the single crystal furnace, when the monitored furnace pressure value exceeds the pre-set pressure threshold range, it will be determined that the current furnace pressure state is abnormal, and the alarm mechanism will be triggered to notify the relevant personnel for handling.

[0048] However, the accuracy of this method of determining whether it is abnormal only based on the threshold is relatively low, often accompanied by a high false alarm rate and missed alarm rate, that is, it may wrongly identify normal situations as abnormal conditions and issue alarms (false alarms), or fail to detect real abnormal situations in time (missed alarms).

[0049] In view of this, the present application proposes a method for alarming the furnace pressure of a single crystal furnace. It can collect the furnace pressure data within a preset time period in the single crystal furnace to form a furnace pressure sequence, that is, time series furnace pressure data. Then, extract the furnace pressure characteristics of each dimension from the furnace pressure sequence, including time domain characteristics, frequency domain characteristics, and statistical distribution characteristics, etc. Then, based on the extracted furnace pressure characteristics, perform comprehensive analysis to determine whether the current furnace pressure state is normal or abnormal. Finally, if it is determined that the furnace pressure is abnormal, trigger the alarm mechanism and output the alarm information. This method improves the accuracy of identifying furnace pressure abnormalities through comprehensive analysis of the furnace pressure characteristics within a preset time period, thereby improving the accuracy of furnace pressure alarms and reducing false alarms and missed alarms.

[0050] The following will specifically describe the technical solution of the present application and how the technical solution of the present application solves the above technical problems with specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be repeated in some embodiments. The embodiments of the present application will be described below with reference to the drawings.

[0051] Figure 1 Flow schematic of the method for alarming the furnace pressure of a single crystal furnace provided by the embodiment of the present application Figure 1 Taking the execution subject as the alarm system as an example, as Figure 1 shown, the method for alarming the furnace pressure of a single crystal furnace provided in this embodiment includes the following steps:

[0052] S101. Collect the furnace pressure of the single crystal furnace within a preset time period to obtain a furnace pressure sequence.

[0053] In the embodiment of the present application, the furnace pressure may refer to the gas pressure inside the single crystal furnace, and the preset time period may be a time range preset according to the furnace pressure control requirement, used to determine the start and end times of collecting the furnace pressure data. The furnace pressure sequence may refer to the set of furnace pressure data of the single crystal furnace collected in chronological order within the preset time period, for example, it may be stored in the form of time-pressure value pairs.

[0054] For example, the furnace pressure data can be automatically collected at a set sampling frequency within a preset time period by a pressure sensor installed on the side wall or top of the furnace body of the single crystal furnace, and the corresponding sequence of time and pressure values can be recorded in real time to obtain the furnace pressure sequence.

[0055] S102. Extract the furnace pressure characteristics of the single crystal furnace from the furnace pressure sequence. The furnace pressure characteristics include at least one of time domain characteristics, frequency domain characteristics, and statistical distribution characteristics.

[0056] In the embodiments of the present application, the time domain characteristics can be characteristics directly calculated based on the furnace pressure data in the furnace pressure sequence; the frequency domain characteristics can be obtained by converting the furnace pressure sequence to the frequency domain through Fourier transform; the statistical distribution characteristics are used to describe the probability distribution characteristics of the furnace pressure values. Each of the above three types of characteristics contains at least one characteristic value.

[0057] It can be understood that the time domain characteristics are used to reflect the variation characteristics of the furnace pressure over time, the frequency domain characteristics are used to indicate the periodicity and frequency components of the pressure variation, and the statistical distribution characteristics can be used to detect abnormal points or trends deviating from the normal range in the furnace pressure data distribution.

[0058] In an actual production scenario, the furnace pressure of the single crystal furnace is affected by various factors, which may cause the pressure to exhibit complex dynamic behaviors, including sudden changes in the time domain, the emergence of new components in the frequency domain, or shifts in the statistical distribution. By extracting different types of furnace pressure characteristics for analysis, the patterns and laws of pressure variation can be revealed from different perspectives, thereby helping to more accurately identify normal and abnormal states.

[0059] S103. Determine whether the furnace pressure of the single crystal furnace is normal or abnormal based on the furnace pressure characteristics.

[0060] In a possible implementation manner, a reference model in a normal furnace pressure state can be established based on historical furnace pressure data. Specifically, by obtaining a dataset of furnace pressure sequences collected under known normal conditions, the time domain characteristics, frequency domain characteristics, and statistical distribution characteristics are calculated from these datasets, and corresponding upper and lower limits are set as the normal range according to the distribution of the characteristic values of various calculated characteristics.

[0061] For time-domain features, compare the current time-domain features (such as mean, variance, maximum value, minimum value, etc.) with the corresponding values in the baseline model. If a certain feature value exceeds the pre-set threshold range, it indicates that there may be an abnormality in the furnace pressure; for frequency-domain features, analyze the energy change of the current frequency components. If new high-frequency noises are found, or the energy of some frequency components increases or decreases significantly, it indicates that there may be an abnormality in the furnace pressure; for statistical distribution features, analyze whether the probability distribution of the pressure values deviates from the distribution pattern in the baseline model (such as changes in skewness and kurtosis). If the proportion of the tail (i.e., the part far from the average value) in the data distribution increases, or the distribution becomes more dispersed, it indicates that there are more abnormal pressure fluctuations.

[0062] Finally, if at least one feature value in any of the time-domain features, frequency features, and statistical distribution features exceeds the threshold, it is determined that the furnace pressure is abnormal; otherwise, it is determined that the furnace pressure is normal.

[0063] In another possible implementation, the extracted furnace pressure features can be input into a pre-trained anomaly recognition model to obtain the recognition result output by the anomaly recognition model. The recognition result is that the furnace pressure is abnormal or the furnace pressure is normal.

[0064] S104. In the case of abnormal furnace pressure, output an alarm message.

[0065] In the embodiments of the present application, the alarm message may include audible and visual alarms, screen prompts, SMS or email notifications, etc. For example, sound or light signals are emitted through alarm devices such as buzzers and warning lights, specific abnormal information and recommended measures are displayed on the monitoring interface, and a notification containing detailed information is sent to the designated contacts.

[0066] The purpose of this step is to quickly respond to abnormal situations and give an alarm so that relevant personnel can handle the abnormal situations in a timely manner, thereby ensuring that the furnace pressure during the single-crystal silicon growth process is maintained within a stable range.

[0067] The single-crystal furnace furnace pressure alarm method provided by the embodiments of the present application collects the furnace pressure of the single-crystal furnace within a preset time period to obtain a furnace pressure sequence, extracts the furnace pressure features of the single-crystal furnace from the furnace pressure sequence. The furnace pressure features include at least one of time-domain features, frequency-domain features, and statistical distribution features, determines whether the furnace pressure of the single-crystal furnace is normal or abnormal according to the furnace pressure features, and outputs an alarm message in the case of abnormal furnace pressure; this method improves the accuracy of the single-crystal furnace furnace pressure alarm and reduces the risks of missed alarms and false alarms.

[0068] Figure 2 It is a flow schematic of the single-crystal furnace furnace pressure alarm method provided by the embodiments of the present application Figure 2 。This embodiment is based on Figure 1 the embodiment, and details the single-crystal furnace furnace pressure alarm method. As Figure 2As shown in the figure, the single crystal furnace pressure alarm method provided in this embodiment includes:

[0069] S201. Collect the furnace pressure of the single crystal furnace within a preset time period through multiple sensors arranged at different positions of the single crystal furnace.

[0070] In the embodiments of the present application, multiple sensors can be distributed at key positions of the single crystal furnace, such as the top, bottom, side wall, air inlet, exhaust port, etc. of the furnace cavity, so as to cover the pressure distribution of the entire furnace body. The sensors at these different positions synchronously collect furnace pressure data within a preset process time period to form a multi-dimensional time series furnace pressure data set.

[0071] It can be understood that by arranging multiple sensors at different positions, more comprehensive data can be obtained, avoiding misjudgment caused by local anomalies, and still being able to detect anomalies through other nodes when a single sensor fails, avoiding missed judgment caused by the failure of a single sensor.

[0072] It should be noted that there are multiple process stages in the single crystal silicon growth process, and the requirements for the furnace environment may be different in different process stages. Therefore, in order to accurately control the furnace pressure, in some embodiments, the setting of the preset time period can be matched with the time requirements of different process stages. For example, the preset time period corresponding to the vacuum pumping stage is 1.5 h; the preset time period corresponding to the melting stage is 3 h; the preset time period corresponding to the crystal seeding stage is 1 h.

[0073] S202. Determine the average value of the furnace pressures collected by multiple sensors at each sampling moment to obtain an initial furnace pressure sequence.

[0074] In the embodiments of the present application, the initial furnace pressure sequence can be used to reflect the change trend of the furnace pressure of the single crystal furnace within the entire preset time period. Through step S201, the furnace pressure values of each sensor at each sampling moment within the preset time period can be obtained. Then, for each sampling moment, calculate the average value of the furnace pressure values collected by all sensors, and arrange the average furnace pressure values at each sampling moment in chronological order to obtain the initial furnace pressure sequence.

[0075] It can be understood that if the original furnace pressure data of multiple sensors is directly used, due to having multiple pressure values at each moment, the data dimension will be relatively high, increasing the complexity of subsequent processing and analysis. In this way, by calculating the average value to generate the initial furnace pressure sequence, the data of multiple sensors can be integrated into a time series, simplifying the data structure; at the same time, calculating the average value of multiple sensors can effectively smooth the random errors of multiple furnace pressure values, improving the stability and reliability of the data.

[0076] S203. Perform preprocessing on the initial furnace pressure sequence to obtain a furnace pressure sequence.

[0077] Among them, in order to improve data quality and thus provide more accurate and reliable data for subsequent data analysis, it is necessary to preprocess the initial furnace pressure sequence to obtain the furnace pressure sequence. The preprocessing may include, for example: filtering, standardization, and missing value processing, etc.

[0078] Filtering processing refers to removing high-frequency noise in the initial furnace pressure sequence through filtering techniques such as low-pass filters, moving average filters, Kalman filters, etc., and retaining the basic trend of furnace pressure changes.

[0079] Standardization means standardizing the data from different sensors or different time periods to ensure that all input data is on the same scale, which is convenient for subsequent algorithm processing.

[0080] Missing value processing refers to using statistical methods such as interpolation, mean or median filling to process the missing values in the data to maintain the continuity of the data in the furnace pressure sequence.

[0081] S204. Extract the furnace pressure characteristics of the single crystal furnace from the furnace pressure sequence.

[0082] In the embodiments of the present application, by analyzing the furnace pressure sequence from multiple angles, various types of characteristics can be extracted. Using different types of furnace pressure characteristics, the actual situation and change law of the furnace pressure of the single crystal furnace can be obtained more comprehensively.

[0083] Optionally, the furnace pressure characteristics include time domain characteristics. Extracting the furnace pressure characteristics of the single crystal furnace from the furnace pressure sequence includes:

[0084] Extract the furnace pressure average value, furnace pressure maximum value, furnace pressure minimum value, and furnace pressure change rate from the furnace pressure sequence, and determine the furnace pressure average value, furnace pressure maximum value, furnace pressure minimum value, and furnace pressure change rate as time domain characteristics.

[0085] Among them, the furnace pressure average value refers to the arithmetic average of the furnace pressure values at all sampling points within the entire preset time period, which is used to reflect the average pressure level of the single crystal furnace during this time period. The furnace pressure maximum value refers to the largest furnace pressure value among all sampling points within the entire preset time period. The furnace pressure maximum value represents the pressure peak of the single crystal furnace during this period. On the contrary, the furnace pressure minimum value refers to the smallest furnace pressure value among all sampling points within the entire preset time period. The furnace pressure minimum value represents the lowest pressure point of the single crystal furnace during this period. The furnace pressure change rate refers to the absolute value of the difference between adjacent sampling points, which is used to reflect the change speed of the furnace pressure.

[0086] In this way, after extracting the furnace pressure average value, furnace pressure maximum value, furnace pressure minimum value, and furnace pressure change rate from the furnace pressure sequence, using these key characteristic values as time domain characteristics, the characteristics of the furnace pressure changing with time can be obtained.

[0087] Optionally, the furnace pressure feature includes a frequency-domain feature. Extracting the furnace pressure feature of the single-crystal furnace from the furnace pressure sequence includes:

[0088] Perform a Fourier transform on the furnace pressure sequence to obtain the spectrum corresponding to the furnace pressure sequence. Determine the band energy of a preset frequency band based on the spectrum, and determine the band energy of the preset frequency band as the frequency-domain feature.

[0089] Among them, the Fourier transform is used to convert the furnace pressure data in the time domain into furnace pressure data in the frequency domain. By calculating the amplitude of each frequency component, the spectrum corresponding to the furnace pressure sequence is obtained. After determining the frequency axis according to the sampling frequency, select a preset frequency band according to the requirements of the single-crystal silicon production process or the characteristics of the single-crystal furnace equipment, find the frequency components corresponding to the preset frequency band in the spectrum, and sum the squares of the amplitudes of these frequency components to obtain the band energy of the preset frequency band, so as to use this band energy as the frequency-domain feature for subsequent analysis.

[0090] Optionally, the furnace pressure feature includes a statistical distribution feature. Extracting the furnace pressure feature of the single-crystal furnace from the furnace pressure sequence includes:

[0091] Determine the distribution of the furnace pressure sequence, determine the standard deviation, skewness, and kurtosis of the distribution, and determine the standard deviation, skewness, and kurtosis as the statistical distribution features.

[0092] Among them, the distribution of the furnace pressure sequence describes the probability characteristics of the occurrence of pressure values. For example, the data in the furnace pressure sequence can be divided into multiple intervals, and the frequency in each interval can be calculated, and a histogram can be drawn to determine the distribution of the data to initially judge the distribution shape, and then the distribution of the furnace pressure sequence can be determined through distribution fitting tests, such as normal distribution, exponential distribution, etc.

[0093] For example, first calculate the average value of all furnace pressure data points, calculate the difference between each furnace pressure data point and the mean value, square each difference, and then calculate the average value of all squared differences to obtain the variance. Finally, take the square root of the variance to obtain the standard deviation.

[0094] Since the mean value and the standard deviation have been obtained in the standard deviation calculation, therefore, for each difference between the furnace pressure data point and the mean value, cube it, calculate the average value of all cubed differences, and divide the average cubed difference by the cube of the standard deviation to obtain the skewness. The skewness describes the symmetry of the furnace pressure data distribution.

[0095] Furthermore, on the basis of calculating the mean value and the standard deviation of the furnace pressure data, for each furnace pressure data point, calculate the difference between it and the mean value, and raise it to the fourth power, find the average value of all fourth-power difference values, and divide the average fourth-power difference by the fourth power of the standard deviation to obtain the kurtosis. The kurtosis measures the sharpness of the furnace pressure data distribution.

[0096] S205. Input the furnace pressure characteristics into a pre-trained anomaly recognition model to obtain the recognition result output by the anomaly recognition model, where the recognition result is that the furnace pressure is normal or abnormal.

[0097] Among them, the anomaly recognition model is trained using the historical furnace pressure data of the single crystal furnace. This anomaly recognition model is used to judge whether the current furnace pressure state is normal according to the input furnace pressure characteristics, such as at least one of time domain characteristics, frequency domain characteristics, and statistical distribution characteristics. The output of the model is a binary classification result of normal furnace pressure or abnormal furnace pressure.

[0098] By obtaining historical furnace pressure data and annotating the status of each group of data, that is, normal furnace pressure or abnormal furnace pressure, normal samples and abnormal samples can be obtained. Here, the normal samples are used to train the model to learn the benchmark mode of the single crystal furnace under normal furnace pressure, including time domain characteristics, frequency domain characteristics, statistical distribution characteristics, etc. under normal furnace pressure. The abnormal samples are used to help the model learn the abnormal mode. Then, the annotated data is divided into a training set and a validation set, and the training set is used to train the initial model, adjust the hyperparameters to optimize the model performance, and evaluate indicators such as accuracy and recall rate on the validation set to train the initial model to obtain the anomaly recognition model.

[0099] Input the real-time extracted furnace pressure characteristics into the already trained anomaly recognition model. This anomaly recognition model will output a classification result according to the input feature vector: if the model predicts "normal furnace pressure", it indicates that the current furnace pressure is within the safe furnace pressure range required by the process; if the model predicts "abnormal furnace pressure", it indicates that there may be abnormal fluctuations or other problems with the current furnace pressure, and further inspection or measures need to be taken.

[0100] Since the time lengths of the various process stages during the growth of the single crystal furnace are different, and the change rules of the furnace pressure are also different. For example, the furnace pressure during the vacuum pumping stage should continuously decrease until it reaches the preset value. If the furnace pressure does not decrease as expected or even shows an upward trend, it indicates that there may be an anomaly in the furnace pressure; while the furnace pressure during the melting stage needs to be highly stable, and even a small fluctuation may be abnormal. During the cooling stage, the furnace pressure slowly decreases, but if the decrease rate exceeds the threshold, there may also be an anomaly. Therefore, in some embodiments, if the preset time period is set to a duration corresponding to the time requirements of different process stages, the anomaly recognition model can also learn the normal furnace pressure modes under different process stages, so as to distinguish real anomalies from process characteristics and improve the detection accuracy.

[0101] S206. Output an alarm message in the case of abnormal furnace pressure.

[0102] Step S206 is similar to the above step S104 and will not be elaborated here.

[0103] The single crystal furnace pressure alarm method provided by the embodiment of the present application collects the furnace pressure of the single crystal furnace within a preset time period through multiple sensors arranged at different positions of the single crystal furnace, determines the average value of the furnace pressures collected by the multiple sensors at each sampling moment, obtains an initial furnace pressure sequence, preprocesses the initial furnace pressure sequence to obtain a furnace pressure sequence, extracts the furnace pressure characteristics of the single crystal furnace from the furnace pressure sequence, inputs the furnace pressure characteristics into a pre-trained anomaly recognition model, obtains the recognition result output by the anomaly recognition model, the recognition result is that the furnace pressure is normal or the furnace pressure is abnormal, and in the case of abnormal furnace pressure, an alarm message is output; this method improves the accuracy of the single crystal furnace pressure alarm and reduces the risks of missed alarms and false alarms.

[0104] In the foregoing embodiment, the single crystal furnace pressure alarm method provided by the embodiment of the present application is introduced. In order to implement each function in the method provided by the embodiment of the present application, as an execution subject, the electronic device may include a hardware structure and / or a software module, and implement the above functions in the form of a hardware structure, a software module, or a combination of a hardware structure and a software module. Whether a certain function among the above functions is executed in the form of a hardware structure, a software module, or a combination of a hardware structure and a software module depends on the specific application and design constraint conditions of the technical solution.

[0105] For example, Figure 3 is a schematic structural diagram of a single crystal furnace pressure alarm device provided by an embodiment of the present application, as Figure 3 shown, the device includes:

[0106] A collection module 301, configured to collect the furnace pressure of the single crystal furnace within a preset time period to obtain a furnace pressure sequence;

[0107] An extraction module 302, configured to extract the furnace pressure characteristics of the single crystal furnace from the furnace pressure sequence, where the furnace pressure characteristics include at least one of time domain characteristics, frequency domain characteristics, and statistical distribution characteristics;

[0108] A determination module 303, configured to determine whether the furnace pressure of the single crystal furnace is normal or abnormal according to the furnace pressure characteristics;

[0109] An alarm module 304, configured to output an alarm message in the case of abnormal furnace pressure.

[0110] Optionally, the determination module 303 is specifically configured to: input the furnace pressure characteristics into a pre-trained anomaly recognition model to obtain the recognition result output by the anomaly recognition model, the recognition result is that the furnace pressure is normal or abnormal, and the anomaly recognition model is trained using the historical furnace pressure data of the single crystal furnace.

[0111] Optionally, the furnace pressure characteristics include the time domain characteristics, and the extraction module 302 is specifically configured to:

[0112] Extract the average furnace pressure, the maximum furnace pressure, the minimum furnace pressure, and the furnace pressure change rate from the furnace pressure sequence;

[0113] Determine the average furnace pressure, the maximum furnace pressure, the minimum furnace pressure, and the furnace pressure change rate as the time-domain features.

[0114] Optionally, the furnace pressure feature includes the frequency-domain feature. The extraction module 302 is specifically configured to:

[0115] Perform a Fourier transform on the furnace pressure sequence to obtain the spectrum corresponding to the furnace pressure sequence;

[0116] Determine the band energy of a preset frequency band based on the spectrum, and determine the band energy of the preset frequency band as the frequency-domain feature.

[0117] Optionally, the furnace pressure feature includes the statistical distribution feature. The extraction module 302 is specifically configured to:

[0118] Determine the distribution of the furnace pressure sequence, and determine the standard deviation, skewness, and kurtosis of the distribution;

[0119] Determine the standard deviation, the skewness, and the kurtosis as the statistical distribution features.

[0120] Optionally, the acquisition module 301 is specifically configured to: collect the furnace pressure of the single crystal furnace within a preset time period through a plurality of sensors disposed at different positions of the single crystal furnace;

[0121] Determine the average value of the furnace pressure collected by the plurality of sensors at each sampling moment to obtain an initial furnace pressure sequence;

[0122] Preprocess the initial furnace pressure sequence to obtain the furnace pressure sequence.

[0123] It should be noted that the specific implementation principle and effect of the above single crystal furnace pressure alarm device can be referred to the relevant descriptions and effects corresponding to the above embodiments, and will not be elaborated here.

[0124] The embodiment of the present application also provides a schematic structural diagram of an electronic device, Figure 4 which is a schematic structural diagram of an electronic device provided by the embodiment of the present application. As Figure 4 shown, the electronic device may include: a processor 401 and a memory 402 communicatively connected to the processor; the memory 402 stores a computer program; the processor 401 executes the computer program stored in the memory 402, so that the processor 401 executes the method described in any one of the above embodiments.

[0125] Wherein, the memory 402 and the processor 401 may be connected through a bus 403.

[0126] The embodiments of the present application also provide a computer-readable storage medium storing computer-executable instructions, which are used to implement the method described in any of the foregoing embodiments of the present application when executed by a processor.

[0127] The embodiments of the present application also provide a chip for running instructions, which is used to execute the method described in any of the foregoing embodiments executed by an electronic device in any of the foregoing embodiments.

[0128] The embodiments of the present application also provide a computer program product, which includes a computer program that can implement the method described in any of the foregoing embodiments of the present application when executed by a processor.

[0129] In several embodiments provided by the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only illustrative. For example, the division of modules is only a logical function division. In actual implementation, there may be other division methods. For example, multiple modules or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed mutual coupling or direct coupling or communication connection can be through some interfaces. The indirect coupling or communication connection of devices or modules can be in electrical, mechanical or other forms.

[0130] The modules described as separate components may or may not be physically separated. The components displayed as modules may or may not be physical units, that is, they can be located in one place or distributed to multiple network units. Some or all of the modules can be selected according to actual needs to implement the solution of this embodiment.

[0131] In addition, the functional modules in each embodiment of the present application can be integrated in a processing unit, or each module can exist physically alone, or two or more modules can be integrated in one unit. The unit formed by the above modules can be implemented in the form of hardware or in the form of a hardware plus software functional unit.

[0132] The integrated modules implemented in the form of software functional modules can be stored in a computer-readable storage medium. The above software functional modules are stored in a storage medium, including several instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) or a processor to execute some steps of the methods described in each embodiment of the present application.

[0133] It should be understood that the above-mentioned processor may be a Central Processing Unit (CPU), or other general-purpose processors, Digital Signal Processors (DSPs), Application Specific Integrated Circuits (ASICs), etc. The general-purpose processor may be a microprocessor or any conventional processor, etc. The steps of the method disclosed in combination with the application can be directly embodied as being completed by a hardware processor, or completed by a combination of hardware and software modules in the processor.

[0134] The memory may include high-speed random access memory (RAM), and may also include non-volatile memory (NVM), such as at least one disk memory, and may also be a USB flash drive, a mobile hard disk, a read-only memory, a magnetic disk, or an optical disc, etc.

[0135] The bus may be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, an Extended Industry Standard Architecture (EISA) bus, etc. The bus can be divided into an address bus, a data bus, a control bus, etc. For the sake of convenience in representation, the buses in the drawings of this application are not limited to only one bus or one type of bus.

[0136] The above storage medium can 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 (EPROM), Programmable Read-Only Memory (PROM), Read-Only Memory (ROM), magnetic memory, flash memory, a magnetic disk or an optical disk. The storage medium can be any available medium accessible by a general-purpose or special-purpose computer.

[0137] An exemplary storage medium is coupled to the processor, enabling the processor to read information from the storage medium and write information to the storage medium. Of course, the storage medium can also be a component of the processor. The processor and the storage medium can be located in an Application Specific Integrated Circuits (ASIC). Of course, the processor and the storage medium can also exist as discrete components in an electronic device or a master device.

[0138] It should be noted that for the foregoing method embodiments, for the sake of simple description, they are all expressed as a series of action combinations. However, those skilled in the art should know that this application is not limited by the described action sequence, because according to this application, certain steps can be performed in other sequences or simultaneously. Secondly, those skilled in the art should also know that the embodiments described in the specification are all optional embodiments, and the actions and modules involved are not necessarily essential to this application.

[0139] Furthermore, it should be noted that although the steps in the flowchart are displayed sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless there is a clear description in this article, the execution of these steps has no strict order limitation, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowchart can include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily executed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed alternately or in turn with at least a part of other steps or sub-steps or stages of other steps.

[0140] In the above embodiments, the descriptions of the respective embodiments have their own emphases. For parts not detailed in a certain embodiment, reference may be made to the relevant descriptions of other embodiments. The technical features of the above embodiments can be combined arbitrarily. For the sake of brevity of description, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered as falling within the scope described in this specification.

[0141] Those skilled in the art will readily conceive of other embodiments of the present application after considering the specification and practicing the invention disclosed herein. The present application is intended to cover any variations, uses, or adaptations of the present application, which follow the general principles of the present application and include known common general knowledge or conventional technical means in the technical field not disclosed in the present application. The specification and the embodiments are only regarded as exemplary, and the true scope and spirit of the present application are pointed out by the claims.

[0142] As described above, the above are only the specific implementation manners of the embodiments of the present application, but the protection scope of the embodiments of the present application is not limited thereto. Any changes or substitutions within the technical scope disclosed in the embodiments of the present application should be covered within the protection scope of the embodiments of the present application. Therefore, the protection scope of the embodiments of the present application shall be subject to the protection scope of the claims.

Claims

1. A method for alarming the furnace pressure of a single crystal furnace, characterized in that, Comprising: Collect the furnace pressure of the single crystal furnace within a preset time period to obtain a furnace pressure sequence; Extract the furnace pressure characteristics of the single crystal furnace from the furnace pressure sequence, where the furnace pressure characteristics include at least one of time domain characteristics, frequency domain characteristics, and statistical distribution characteristics; Determine whether the furnace pressure of the single crystal furnace is normal or abnormal according to the furnace pressure characteristics; Output an alarm message in the case of abnormal furnace pressure.

2. The method according to claim 1, wherein The determining whether the furnace pressure of the single crystal furnace is normal or abnormal according to the furnace pressure characteristics includes: Input the furnace pressure characteristics into a pre-trained anomaly recognition model to obtain an identification result output by the anomaly recognition model, where the identification result is normal furnace pressure or abnormal furnace pressure, and the anomaly recognition model is trained using historical furnace pressure data of the single crystal furnace.

3. The method according to claim 1 or 2, characterized in that, The furnace pressure characteristics include the time domain characteristics. The extracting the furnace pressure characteristics of the single crystal furnace from the furnace pressure sequence includes: Extract the average furnace pressure, maximum furnace pressure, minimum furnace pressure, and furnace pressure change rate from the furnace pressure sequence; Determine the average furnace pressure, maximum furnace pressure, minimum furnace pressure, and furnace pressure change rate as the time domain characteristics.

4. The method according to claim 1 or 2, characterized in that, The furnace pressure characteristics include the frequency domain characteristics. The extracting the furnace pressure characteristics of the single crystal furnace from the furnace pressure sequence includes: Perform a Fourier transform on the furnace pressure sequence to obtain the spectrum corresponding to the furnace pressure sequence; Determine the band energy of a preset frequency band based on the spectrum, and determine the band energy of the preset frequency band as the frequency domain characteristics.

5. The method according to claim 1 or 2, characterized in that, The furnace pressure characteristics include the statistical distribution characteristics. The extracting the furnace pressure characteristics of the single crystal furnace from the furnace pressure sequence includes: Determine the distribution of the furnace pressure sequence, and determine the standard deviation, skewness, and kurtosis of the distribution; Determine the standard deviation, skewness, and kurtosis as the statistical distribution characteristics.

6. The method according to claim 1 or 2, characterized in that, The collecting the furnace pressure of the single crystal furnace within a preset time period to obtain a furnace pressure sequence includes: Collect the furnace pressure of the single crystal furnace within a preset time period through a plurality of sensors arranged at different positions of the single crystal furnace; Determine the average value of the furnace pressures collected by the plurality of sensors at each sampling moment to obtain an initial furnace pressure sequence; Perform preprocessing on the initial furnace pressure sequence to obtain the furnace pressure sequence.

7. A pressure alarm device for a single crystal furnace, characterized in that, Comprising: A collecting module for collecting the furnace pressure of the single crystal furnace within a preset time period to obtain a furnace pressure sequence; An extracting module for extracting the furnace pressure characteristics of the single crystal furnace from the furnace pressure sequence, where the furnace pressure characteristics include at least one of time domain characteristics, frequency domain characteristics, and statistical distribution characteristics; A determining module for determining whether the furnace pressure of the single crystal furnace is normal or abnormal according to the furnace pressure characteristics; An alarm module for outputting an alarm message in the case of abnormal furnace pressure.

8. An electronic device, characterized in that, Comprising: A memory, a processor; The memory stores computer execution instructions; The processor executes the computer execution instructions stored in the memory, so that the processor executes the single crystal furnace furnace pressure alarm method according to any one of claims 1-6.

9. A computer-readable storage medium, characterized in that, Computer execution instructions are stored in the computer-readable storage medium, and when the computer execution instructions are executed by a processor, they are used to implement the single crystal furnace furnace pressure alarm method according to any one of claims 1-6.

10. A computer program product, characterized in that, Comprising a computer program, when the computer program is executed by a processor, it implements the single crystal furnace pressure alarm method according to any one of the above claims 1-6.

Images